https://ntp.niehs.nih.gov/ntp/about_ntp/trpanel/2018/march/tr596peerdraft.pdf
ON THE TOXICOLOGY AND CARCINOGENESI STUDIES IN B6C3F1/N MICE EXPOSED TO WHOLE-BODY RADIO FREQUENCY RADIATION AT A FREQUENCY (1,900 MHz) AND MODULATIONS (GSM AND CDMA) USED BY CELL PHONES
Scheduled Peer Review Date: March 26-28, 2018
proof that WiFi and radio waves are causing cancer and DNA damage and whole genome damage, the cell towers that you see all over, are a weaponized system to cause cancers, and a whole spectrum of diseases in humans, and biological systems, and they know full well that it does, and have been silently killing us all,
National Institutes of Health
Public Health Service
U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES
FOREWORD
The National Toxicology Program (NTP) is an interagency program within the Public Health Service (PHS) of the
Department of Health and Human Services (HHS) and is headquartered at the National Institute of Environmental
Health Sciences of the National Institutes of Health (NIEHS/NIH). Three agencies contribute resources to the
program: NIEHS/NIH, the National Institute for Occupational Safety and Health of the Centers for Disease Control
and Prevention (NIOSH/CDC), and the National Center for Toxicological Research of the Food and Drug
Administration (NCTR/FDA). Established in 1978, the NTP is charged with coordinating toxicological testing
activities, strengthening the science base in toxicology, developing and validating improved testing methods, and
providing information about potentially toxic substances to health regulatory and research agencies, scientific and
medical communities, and the public.
The Technical Report series began in 1976 with carcinogenesis studies conducted by the National Cancer Institute.
In 1981, this bioassay program was transferred to the NTP. The studies described in the Technical Report series are
designed and conducted to characterize and evaluate the toxicologic potential, including carcinogenic activity, of
selected substances in laboratory animals (usually two species, rats and mice). Substances selected for NTP toxicity
and carcinogenicity studies are chosen primarily on the basis of human exposure, level of production, and chemical
structure. The interpretive conclusions presented in NTP Technical Reports are based only on the results of these
NTP studies. Extrapolation of these results to other species, including characterization of hazards and risks to
humans, requires analyses beyond the intent of these reports. Selection per se is not an indicator of a substance’s
carcinogenic potential.
The NTP conducts its studies in compliance with its laboratory health and safety guidelines and FDA Good
Laboratory Practice Regulations and must meet or exceed all applicable federal, state, and local health and safety
regulations. Animal care and use are in accordance with the Public Health Service Policy on Humane Care and Use
of Animals. Studies are subjected to retrospective quality assurance audits before being presented for public review.
NTP Technical Reports are indexed in the NIH/NLM PubMed database and are available free of charge
electronically on the NTP website (http://ntp.niehs.nih.gov). Additional information regarding this study may be
requested through Central Data Management (CDM) at cdm@niehs.nih.gov. Toxicity data are available through
NTP’s Chemical Effects in Biological Systems (CEBS) database: https://www.niehs.nih.gov/research/resources/
databases/cebs/index.cfm.
NTP TECHNICAL REPORT
ON THE
TOXICOLOGY AND CARCINOGENESIS
STUDIES IN B6C3F1/N MICE EXPOSED
TO WHOLE-BODY RADIO FREQUENCY RADIATION
AT A FREQUENCY (1,900 MHz) AND MODULATIONS
(GSM AND CDMA) USED BY CELL PHONES
Scheduled Peer Review Date: March 26-28, 2018
NOTICE
This DRAFT Technical Report is distributed solely for the purpose of predissemination peer review under the applicable information
quality guidelines. It has not been formally disseminated by the NTP. It does not represent and should not be construed to represent
NTP determination or policy.
NTP TR 596
National Toxicology Program
National Institutes of Health
Public Health Service
U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES
2
NOT FOR ATTRIBUTION Peer Review Draft
CONTRIBUTORS
National Toxicology Program
Evaluated and interpreted results and reported findings
M.E. Wyde, Ph.D., Study Scientist
A.E. Brix, D.V.M., Ph.D., Study Pathologist
Experimental Pathology Laboratories, Inc.
C.R. Blystone, Ph.D.
J.R. Bucher, Ph.D.
M.F. Cesta, D.V.M., Ph.D.
M.C. Cora, D.V.M.
S.A. Elmore, D.V.M., M.S.
P.M. Foster, Ph.D.
M.J. Hooth, Ph.D.
A.P. King-Herbert, D.V.M.
G.E. Kissling, Ph.D.
D.E. Malarkey, D.V.M., Ph.D.
G.K. Roberts, Ph.D.
K.R. Shockley, Ph.D.
R.C. Sills, D.V.M., Ph.D.
S.L. Smith-Roe, Ph.D.
M.D. Stout, Ph.D.
N.J. Walker, Ph.D.
K.L. Witt, M.S.
IIT Research Institute
Conducted studies and evaluated pathology findings
D.L. McCormick, Ph.D., Principal Investigator
T.L. Horn, Ph.D., Study Director
J.R. Gauger, B.S., Engineer
L.H. Brennecke, D.V.M.
Charles River Laboratories, Inc.
R.M. Kovatch, D.V.M.
Charles River Laboratories, Inc.
Integrated Laboratory Systems, Inc.
Provided pathology review
E.T. Adams, D.V.M., Ph.D., Principal Investigator
G.D. Hill, D.V.M., Ph.D.
R.R. Moore, D.V.M.
RTI International
Provided SCVCE analysis
R.W. Tyl, Ph.D., Principal Investigator
F.T. Les, M.S. (sperm motility)
Charles River Laboratories, Inc.
M.C. Marr, B.A.
C.S. Sloan, M.S.
IT’IS Foundation
Constructed and maintained exposure system
N. Kuster, Ph.D.
M. Capstick, Ph.D.
CSS, Inc.
Prepared quality assessment audits
S. Brecher, Ph.D., Principal Investigator
S. Iyer, B.S.
V.S. Tharakan, D.V.M.
NTP Pathology Working Group
Evaluated slides and contributed to pathology reports
on male mice (May 2, 2017) or female mice (May 1, 2017)
exposed to GSM- or CDMA-modulated cell phone RFR for 2 Years
R.R. Moore, D.V.M., Coordinator (male mice)
ILS, Inc.
G.D. Hill, D.V.M., Ph.D., Coordinator (female mice)
ILS, Inc.
A.E. Brix, D.V.M., Ph.D.
Experimental Pathology Laboratories, Inc.
M.F. Cesta, D.V.M., Ph.D.
National Toxicology Program
S.A. Elmore, D.V.M., M.S.
National Toxicology Program
K.S. Frazier, D.V.M., Ph.D.
GlaxoSmithKline
M.P. Jokinen, D.V.M.
ILS, Inc.
D.E. Malarkey, D.V.M., Ph.D.
National Toxicology Program
A.R. Pandiri, B.V.Sc. & A.H., Ph.D.
National Toxicology Program
K.S. Regan, D.V.M.
Regan Pathology/Toxicology Services, Inc.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 3
Peer Review Draft NOT FOR ATTRIBUTION
Social & Scientific Systems, Inc.
Provided statistical analyses
M.V. Smith, Ph.D., Principal Investigator
L.J. Betz, M.S.
S.F. Harris, B.S.
J.D. Krause, Ph.D.
C.G. Leach, M.S.
Biotechnical Services, Inc.
Prepared Technical Report
S.R. Gunnels, M.A., Principal Investigator
K.K. Coker, Ph.D.
P.A. Gideon, B.A.
L.M. Harper, B.S.
P.C. Nader, B.S.E.
J.I. Powers, M.A.P.
D.C. Serbus, Ph.D.
4
NOT FOR ATTRIBUTION Peer Review Draft
CONTENTS
ABSTRACT ................................................................................................................................................................. 7
EXPLANATION OF LEVELS OF EVIDENCE OF CARCINOGENIC ACTIVITY........................................14
PEER REVIEW PANEL ..........................................................................................................................................15
SUMMARY OF PEER REVIEW PANEL COMMENTS.....................................................................................16
INTRODUCTION .....................................................................................................................................................17
MATERIALS AND METHODS..............................................................................................................................39
RESULTS...................................................................................................................................................................67
DISCUSSION AND CONCLUSIONS.....................................................................................................................99
REFERENCES ........................................................................................................................................................105
APPENDIX A Summary of Lesions in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years .....................................................A-1
APPENDIX B Summary of Lesions in Female Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years .....................................................B-1
APPENDIX C Summary of Lesions in Male Mice
Exposed to CDMA-Modulated Cell Phone RFR for 2 Years..................................................C-1
APPENDIX D Summary of Lesions in Female Mice
Exposed to CDMA-Modulated Cell Phone RFR for 2 Years..................................................D-1
APPENDIX E Genetic Toxicology......................................................................................................................E-1
APPENDIX F Hematology Results.....................................................................................................................F-1
APPENDIX G Organ Weights and Organ-Weight-to-Body-Weight Ratios.................................................. G-1
APPENDIX H Reproductive Tissue Evaluations and Estrous Cycle Characterization................................ H-1
APPENDIX I GSM- and CDMA-Modulated Cell Phone RFR Exposure Data..............................................I-1
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 5
Peer Review Draft NOT FOR ATTRIBUTION
APPENDIX J Ingredients, Nutrient Composition, and Contaminant Levels
in NTP-2000 Rat and Mouse Ration...........................................................................................J-1
APPENDIX K Sentinel Animal Program.......................................................................................................... K-1
6 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
7
Peer Review Draft NOT FOR ATTRIBUTION
ABSTRACT
GSM- AND CDMA-MODULATED CELL PHONE RADIO FREQUENCY RADIATION
Synonyms: Cell phone radio frequency radiation; mobile phone radio frequency radiation
The predominant source of human exposure to radio frequency radiation (RFR) occurs through usage of cellular
phone handsets. The Food and Drug Administration nominated cell phone RFR emission for toxicology and
carcinogenicity testing in 1999. At that time, animal experiments were deemed crucial because meaningful human
exposure data from epidemiological studies were not available. Male and female B6C3F1/N mice were exposed to
time-averaged whole-body specific absorption rates of 0 (sham control), 5, 10, or 15 W/kg Global System for
Mobile Communications (GSM)- or Code Division Multiple Access (CDMA)-modulated cell phone RFR at
1,900 MHz for 28 days or 0, 2.5, 5, or 10 W/kg GSM- or CDMA-modulated cell phone RFR for up to 2 years.
Genetic toxicology studies were conducted in mouse peripheral blood erythrocytes and leukocytes, brain cells, and
liver cells.
GSM
28-DAY STUDY
Groups of 10 male and 10 female core study mice and groups of 20 male and 20 female special study mice were
housed in specially designed reverberation chambers and received whole-body exposures to GSM-modulated cell
phone RFR at power levels of 0 (sham control), 5, 10, or 15 W/kg, for up to 18 hours and 20 minutes per day, 5 or
7 (last week of study) days per week for at least 28 days with continuous cycling of 10 minutes on and 10 minutes
off during the exposure periods. The sham control animals were housed in reverberation chambers identical to those
8 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
used for the exposed groups, but were not exposed to cell phone RFR; a shared group of unexposed mice of each sex
served as sham controls for both cell phone RFR modulations. All mice survived to the end of the study. Mean
body weights of exposed groups of males and females were similar to controls. There were no exposure-related
clinical signs, differences in organ weights, or histopathologic findings. Differences in body temperatures between
the exposed groups and the control group were not considered to be related to cell phone RFR exposure.
2-YEAR STUDY
Groups of 105 male and 105 female mice were housed in reverberation chambers and received whole-body
exposures to GSM-modulated cell phone RFR at power levels of 0 (sham control), 2.5, 5, or 10 W/kg, 9 hours and
10 minutes per day, 7 days per week for 106 (males) or 108 (females) weeks with continuous cycling of 10 minutes
on and 10 minutes off during a period of 18 hours and 20 minutes each day. The sham control animals were housed
in reverberation chambers identical to those used for the exposed groups, but were not exposed to cell phone RFR;
shared groups of unexposed mice of each sex served as sham controls for both cell phone RFR modulations. Fifteen
mice per group were randomly selected from the core group after 10 weeks of study; ten of those 15 mice per group
were used for interim evaluation at 14 weeks, and five mice per group were used for genetic toxicity testing at
14 weeks. The remaining 90 animals per group were exposed up to 2 years.
At the 14-week interim evaluation in the 2-year study, mean body weights of exposed groups of males and females
were similar to those of the sham controls. There were no changes to the hematology variables attributable to GSM
cell phone RFR exposure. Differences in organ weights were not associated with histopathologic findings and were
not considered related to exposure. In males, there were no exposure-related effects on reproductive organ weights,
testis spermatid concentrations, caudal epididymal sperm concentrations, or sperm motility. In females, there were
no exposure related effects on estrous cycle length, number of cycling females, or relative amount of time spent in
the estrous stages. The only histopathologic finding at the 14-week interim evaluation was an increased incidence of
minimal focal inflammation in the liver of the 5 W/kg males.
In the 2-year study, percent survival was significantly higher for the 5 W/kg males than the sham control group.
Survival of the other exposed groups of males and females was generally similar to that of the sham controls. Mean
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 9
Peer Review Draft NOT FOR ATTRIBUTION
body weights of exposed groups of males and females were similar to those of the sham controls throughout the
study.
The combined incidences of fibrosarcoma, sarcoma, or malignant fibrous histiocytoma of the skin were increased in
5 and 10 W/kg males, although not significantly or in an exposure concentration-related manner; however, the
incidences exceeded the overall historical control ranges for malignant fibrous histiocytoma. In the lung, there was
a significant positive trend in the incidences of alveolar/bronchiolar adenoma or carcinoma (combined) in males.
Compared to the sham controls, all exposed groups of females had increased incidences of malignant lymphoma and
the incidences in the 2.5 and 5 W/kg groups were significantly increased. The sham control group had a low
incidence of malignant lymphoma compared to the range seen in historical controls.
There were no nonneoplastic lesions that were considered related to exposure to GSM-modulated cell phone RFR.
CDMA
28-DAY STUDY
Groups of 10 male and 10 female core study mice and groups of 20 male and 20 female special study mice were
housed in reverberation chambers and received whole-body exposures to CDMA-modulated cell phone RFR at
power levels of 0 (sham control), 5, 10, or 15 W/kg, for up to 18 hours and 20 minutes per day, 5 or 7 (last week of
study) days per week for at least 28 days with continuous cycling of 10 minutes on and 10 minutes off during the
exposure periods. The sham control animals were housed in reverberation chambers identical to those used for the
exposed groups, but were not exposed to cell phone RFR; a shared group of unexposed mice of each sex served as
sham controls for both cell phone RFR modulations. All mice survived to the end of the study. Mean body weights
of exposed groups of males and females were similar to controls. There were no exposure-related clinical signs,
differences in organ weights, or histopathologic findings. Differences in body temperatures between the exposed
groups and the control group were not considered to be related to cell phone RFR exposure.
10 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
2-YEAR STUDY
Groups of 105 male and 105 female mice were housed in reverberation chambers and received whole-body
exposures to CDMA-modulated cell phone RFR at power levels of 0 (sham control), 2.5, 5, or 10 W/kg, 9 hours and
10 minutes per day, 7 days per week for 106 (males) or 108 (females) weeks with continuous cycling of 10 minutes
on and 10 minutes off during a period of 18 hours and 20 minutes each day. The sham control animals were housed
in reverberation chambers identical to those used for the exposed groups, but were not exposed to cell phone RFR;
shared groups of unexposed mice of each sex served as sham controls for both cell phone RFR modulations. Fifteen
mice per group were randomly selected from the core group after 10 weeks of study; ten of those 15 mice per group
were used for interim evaluation at 14 weeks, and five mice per group were used for genetic toxicity testing at
14 weeks. The remaining 90 animals per group were exposed up to 2 years.
At the 14-week interim evaluation of the 2-year study, mean body weights of exposed groups of males and females
were similar to those of the sham controls. There were no changes to the hematology variables attributable to
CDMA cell phone RFR exposure. Differences in organ weights in male mice were not associated with
histopathologic findings and were not considered related to exposure; there were no significant changes in organ
weights in females. In males, there were no exposure-related effects on reproductive organ weights, testis spermatid
concentrations, caudal epididymal sperm concentrations, or sperm motility. In females, there were no exposure
related effects on estrous cyclicity. Compared to the sham controls, there were statistically significant differences
for extended estrous in the 2.5 W/kg group and extended diestrus in the 5 W/kg group; however, these changes were
considered sporadic due to the lack of an exposure-related response. In the kidney of 10 W/kg females, there was a
significantly increased incidence of minimal to mild interstitial lymphocytic cellular infiltration.
Percent survival was significantly higher in 2.5 W/kg males compared to that in the sham controls in the 2-year
study. Survival of males and females in all other exposed groups was generally similar to that of the sham controls.
Mean body weights of exposed groups of males and females were similar to those of the sham controls throughout
the study.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 11
Peer Review Draft NOT FOR ATTRIBUTION
There was a significantly increased incidence of hepatoblastoma in 5 W/kg males. Compared to the sham controls,
the incidences of malignant lymphoma were increased in all exposed groups of females, and the increase was
significant in the 2.5 W/kg group. As noted for the GSM study, the shared sham control group had a low incidence
of malignant lymphoma compared to the range observed in historical controls.
There were no nonneoplastic lesions that were considered related to exposure to CDMA-modulated cell phone RFR.
GENETIC TOXICOLOGY
Comet Assay
As part of the 14-week interim evaluation, samples of frontal cortex, hippocampus, cerebellum, liver, and blood
leukocytes were evaluated for DNA damage using the comet assay (two sexes, two cell phone RFR modulations,
and five tissues per animal). Samples of peripheral blood were also evaluated for chromosome damage in the
micronucleus assay. Results are based on the 100-cell scoring approach that was standard at the time of the study;
data obtained using a second 150-cell scoring approach, recommended in a recently adopted international guideline
for the in vivo comet assay, are noted for the few instances where results differed between the two methods.
Significant increases in DNA damage were observed in cells of the frontal cortex of male mice exposed to both
modulations, GSM and CDMA. No other tissues showed evidence of a treatment-related effect in male mice. In
female mice exposed to the CDMA modulation, significant increases in DNA damage were seen in blood leukocytes
at all three exposure levels using both scoring approaches. No statistically significant increases in percent comet tail
DNA were observed in any of the samples from female mice exposed to the GSM modulation with the 100-cell
scoring method. Scoring 150 cells resulted in an equivocal response in liver of female mice exposed to CDMA; a
similar pattern of response was seen with the 100-cell scoring method, but none of the increases were significant.
Micronucleus Assay
No significant increases in micronucleated red blood cells or changes in the percentage of immature erythrocytes
among total erythrocytes were observed in the peripheral blood of mice of either sex exposed to either modulation of
cell phone RFR.
12 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
CONCLUSIONS
Under the conditions of these 2-year studies, there was equivocal evidence of carcinogenic activity of
GSM-modulated cell phone RFR at 1,900 MHz in male B6C3F1/N mice based on the combined incidences of
fibrosarcoma, sarcoma, or malignant fibrous histiocytoma in the skin and the incidences of alveolar/bronchiolar
adenoma or carcinoma (combined) in the lung. There was equivocal evidence of carcinogenic activity of
GSM-modulated cell phone RFR at 1,900 MHz in female B6C3F1/N mice based on the incidences of malignant
lymphoma (all organs). There was equivocal evidence of carcinogenic activity of CDMA-modulated cell phone
RFR at 1,900 MHz in male B6C3F1/N mice based on the incidences of hepatoblastoma of the liver. There was
equivocal evidence of carcinogenic activity of CDMA-modulated cell phone RFR at 1,900 MHz in female
B6C3F1/N mice based on the incidences of malignant lymphoma (all organs).
Exposure to GSM- or CDMA-modulated cell phone RFR at 1,900 MHz did not increase the incidence of any
nonneoplastic lesions in male or female B6C3F1/N mice.
* Explanation of Levels of Evidence of Carcinogenic Activity is on page 14.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 13
Peer Review Draft NOT FOR ATTRIBUTION
Summary of the 2-Year Carcinogenesis and Genetic Toxicology Studies
of GSM- and CDMA-Modulated Cell Phone RFR Exposure in Mice
GSM-Modulated
Cell Phone RFR
Male Mice
GSM-Modulated
Cell Phone RFR
Female Mice
CDMA-Modulated
Cell Phone RFR
Male Mice
CDMA-Modulated
Cell Phone RFR
Female Mice
Whole-body GSM- or
CDMA-modulated cell
phone RFR exposure
0, 2.5, 5, or 10 W/kg 0, 2.5, 5, or 10 W/kg 0, 2.5, 5, or 10 W/kg 0, 2.5, 5, or 10 W/kg
Survival rates 66/90, 63/90, 80/90,
72/90
67/90, 74/90, 70/90,
73/90
66/90, 83/91, 71/90,
71/90
67/90, 75/89, 70/90,
72/90
Body weights Exposed groups similar
to the sham control group
Exposed groups similar
to the sham control group
Exposed groups similar
to the sham control group
Exposed groups similar
to the sham control group
Nonneoplastic effects None None None None
Neoplastic effects None None None None
Equivocal findings Skin: fibrosarcoma,
sarcoma, or malignant
fibrous histiocytoma
(1/90, 1/89, 5/90, 4/90)
Lung:
alveolar/bronchiolar
adenoma or carcinoma
(23/90, 24/89, 32/90,
34/90)
All organs: malignant
lymphoma (2/90, 13/90,
9/90, 6/90)
Liver: hepatoblastoma
(6/90, 6/89, 16/90, 7/90)
All organs: malignant
lymphoma (2/90, 9/89,
6/90, 7/90)
Level of evidence of
carcinogenic activity Equivocal evidence Equivocal evidence Equivocal evidence Equivocal evidence
Genetic toxicology
DNA damage:
GSM-modulated
CDMA-modulated
Positive in frontal cortex (males); negative in frontal cortex (females); negative in
hippocampus, cerebellum, liver, and leukocytes (males and females)
Positive in frontal cortex (males) and leukocytes (females); negative in
hippocampus, cerebellum, and liver (males and females); negative in leukocytes
(males) and frontal cortex (females)
Micronucleated erythrocytes in peripheral blood in vivo:
GSM-modulated
CDMA-modulated
Negative in males and females
Negative in males and females
14 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
EXPLANATION OF LEVELS OF EVIDENCE OF CARCINOGENIC ACTIVITY
The National Toxicology Program describes the results of individual experiments on a test agent and notes the strength of the evidence for
conclusions regarding each study. Negative results, in which the study animals do not have a greater incidence of neoplasia than control animals,
do not necessarily mean that a test agent is not a carcinogen, inasmuch as the experiments are conducted under a limited set of conditions.
Positive results demonstrate that a test agent is carcinogenic for laboratory animals under the conditions of the study and indicate that exposure to
the test agent has the potential for hazard to humans. Other organizations, such as the International Agency for Research on Cancer, assign a
strength of evidence for conclusions based on an examination of all available evidence, including animal studies such as those conducted by the
NTP, epidemiologic studies, and estimates of exposure. Thus, the actual determination of risk to humans from test agents found to be
carcinogenic in laboratory animals requires a wider analysis that extends beyond the purview of these studies.
Five categories of evidence of carcinogenic activity are used in the Technical Report series to summarize the strength of evidence observed in
each experiment: two categories for positive results (clear evidence and some evidence); one category for uncertain findings (equivocal
evidence); one category for no observable effects (no evidence); and one category for experiments that cannot be evaluated because of major
flaws (inadequate study). These categories of interpretative conclusions were first adopted in June 1983 and then revised on March 1986 for use
in the Technical Report series to incorporate more specifically the concept of actual weight of evidence of carcinogenic activity. For each
separate experiment (male rats, female rats, male mice, female mice), one of the following five categories is selected to describe the findings.
These categories refer to the strength of the experimental evidence and not to potency or mechanism.
• Clear evidence of carcinogenic activity is demonstrated by studies that are interpreted as showing a dose-related (i) increase of
malignant neoplasms, (ii) increase of a combination of malignant and benign neoplasms, or (iii) marked increase of benign neoplasms
if there is an indication from this or other studies of the ability of such tumors to progress to malignancy.
• Some evidence of carcinogenic activity is demonstrated by studies that are interpreted as showing a test agent-related increased
incidence of neoplasms (malignant, benign, or combined) in which the strength of the response is less than that required for clear
evidence.
• Equivocal evidence of carcinogenic activity is demonstrated by studies that are interpreted as showing a marginal increase of
neoplasms that may be test agent related.
• No evidence of carcinogenic activity is demonstrated by studies that are interpreted as showing no test agent-related increases in
malignant or benign neoplasms
• Inadequate study of carcinogenic activity is demonstrated by studies that, because of major qualitative or quantitative limitations,
cannot be interpreted as valid for showing either the presence or absence of carcinogenic activity.
For studies showing multiple test agent-related neoplastic effects that if considered individually would be assigned to different levels of evidence
categories, the following convention has been adopted to convey completely the study results. In a study with clear evidence of carcinogenic
activity at some tissue sites, other responses that alone might be deemed some evidence are indicated as “were also related” to test agent
exposure. In studies with clear or some evidence of carcinogenic activity, other responses that alone might be termed equivocal evidence are
indicated as “may have been” related to test agent exposure.
When a conclusion statement for a particular experiment is selected, consideration must be given to key factors that would extend the actual
boundary of an individual category of evidence. Such consideration should allow for incorporation of scientific experience and current
understanding of long-term carcinogenesis studies in laboratory animals, especially for those evaluations that may be on the borderline between
two adjacent levels. These considerations should include:
• adequacy of the experimental design and conduct;
• occurrence of common versus uncommon neoplasia;
• progression (or lack thereof) from benign to malignant neoplasia as well as from preneoplastic to neoplastic lesions;
• some benign neoplasms have the capacity to regress but others (of the same morphologic type) progress. At present, it is impossible
to identify the difference. Therefore, where progression is known to be a possibility, the most prudent course is to assume that benign
neoplasms of those types have the potential to become malignant;
• combining benign and malignant tumor incidence known or thought to represent stages of progression in the same organ or tissue;
• latency in tumor induction;
• multiplicity in site-specific neoplasia;
• metastases;
• supporting information from proliferative lesions (hyperplasia) in the same site of neoplasia or other experiments (same lesion in
another sex or species);
• presence or absence of dose relationships;
• statistical significance of the observed tumor increase;
• concurrent control tumor incidence as well as the historical control rate and variability for a specific neoplasm;
• survival-adjusted analyses and false positive or false negative concerns;
• structure-activity correlations; and
• in some cases, genetic toxicology.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 15
Peer Review Draft NOT FOR ATTRIBUTION
NATIONAL TOXICOLOGY PROGRAM TECHNICAL REPORTS
PEER REVIEW PANEL
The members of the Peer Review Panel who evaluated the draft NTP Technical Report on GSM- and CDMA-modulated cell phone RFR in mice
on March 26-28, 2018, are listed below. Panel members serve as independent scientists, not as representatives of any institution, company, or
governmental agency. In this capacity, panel members have five major responsibilities in reviewing the NTP studies:
• to ascertain that all relevant literature data have been adequately cited and interpreted,
• to determine if the design and conditions of the NTP studies were appropriate,
• to ensure that the Technical Report presents the experimental results and conclusions fully and clearly,
• to judge the significance of the experimental results by scientific criteria, and
• to assess the evaluation of the evidence of carcinogenic activity and other observed toxic responses.
16 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
SUMMARY OF PEER REVIEW PANEL COMMENTS
NOTE: A summary of the Peer Review Panel’s remarks will appear in a future draft of this report.
17
Peer Review Draft NOT FOR ATTRIBUTION
INTRODUCTION
GSM- AND CDMA-MODULATED CELL PHONE RADIO FREQUENCY RADIATION
Synonyms: Cell phone radio frequency radiation; mobile phone radio frequency radiation
OVERVIEW
All consumer cell phone devices function through the transmission of radio waves on a cellular network. The
cellular network itself is composed of a collection of individual “cells” that include a fixed-location transceiver (a
device that transmits and receives radio signals), also referred to as a cell tower. The collection of adjacent smaller
“cells” in the cellular network enables cell phones and towers to use low-power transmitters, thereby allowing for
the same frequencies to be reused in non-adjacent cells without interference. Together the individual “cells”
comprise the cellular network that provides coverage over a large geographical area. In the United States, there are
two major nation-wide cellular networks: CDMA (Code Division Multiple Access) and GSM (Global System for
Mobile Communications). With technologies rapidly evolving to meet consumers’ increased demand for better
coverage, increased call quality, faster data transfer rates, and increased accessibility, the terms CDMA and GSM
tend to group together multiple, sometimes successive, technologies that are implemented by the service providers
that maintain the two networks. In the United States, Sprint® and Verizon® use and maintain the CDMA network;
AT&T® and T-Mobile® use and maintain the GSM network.
For both the GSM and CDMA networks, transmissions occur at specific radio wave frequencies, which are allocated
and regulated by the Federal Communications Commission (FCC). While the transmission of radio signals occurs at
the same frequencies for both networks, the networks differ in the method by which their signal is modulated. In
telecommunications, modulation is a process of conveying a signal, like a cell phone user’s voice during a call,
18 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
inside another signal that can be physically transmitted. This process involves modulation of the signal prior to
transmission at one end, and then demodulation at the other end. Since this process requires different technologies
for CDMA and GSM, many cell phones are not interchangeable between the two networks and will only function on
one or the other of the networks, not both.
The constantly evolving cellular technologies are commonly referred to by their successive generations (G). The
first generation (1G) devices were analogue phones, as opposed to the digital phones of today. Digital voice
systems of the second generation (2G) replaced the analogue system of 1G. At the time that these studies were
being designed, 2G technology was the primary technology in use and 3G technologies were emerging. Therefore,
the current studies were conducted using modulated signals that replicated the 2G and 3G technology in use at the
time. Over the course of the studies, however, more advanced 4G technologies were developed. Currently, all of
these technologies (2G, 3G, and 4G) are still actively in use for mobile communication applications. 2G and 3G are
still the basis for voice calling applications, while 3G and 4G technologies were primarily developed to offer faster
access to the internet. Some of the 3G technology is based on 2G technology. While 2G technology is being phased
out in the United States, this technology will remain in use in other places throughout the world. More advanced
and efficient technologies that are currently in development, such as 5G, will utilize higher frequencies than existing
technologies.
RADIO FREQUENCY RADIATION (RFR)
In the context of this report, radio frequency (RF) radiation refers to the broad range of electromagnetic fields from
3 kilohertz (3 kHz) to 300 gigahertz (300 GHz). Different applications utilize different frequency bands within the
RF portion of the electromagnetic spectrum. The range of frequencies for radio and television are in the 145 kHz to
850 MHz range. These include long, medium, shortwave, and very high frequency (VHF) radio transmissions and
VHF and ultra-high frequency (UHF) over-the-air television transmissions. Wireless communications and
networking typically utilize frequencies between 800 MHz and 6 GHz. Cell phone networks (2G, 3G, and 4G)
utilize frequencies in the range of 600 MHz to 5.7 GHz. In the United States, wireless telecommunications networks
and devices operate in bands at frequencies of nominally 800 MHz, 850 MHz, or 1,900 MHz for 2G; 850 MHz,
1,700 MHz, 1,900 MHz, or 2,100 MHz for 3G; and 600 MHz, 700 MHz, 800 MHz, 850MHz, 1,700 MHz,
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 19
Peer Review Draft NOT FOR ATTRIBUTION
1,900 MHz, 2,100 MHz, 2,300 MHz, 2,500 MHz, 5,200 MHz, or 5,700 MHz for 4G. The next generation, i.e., the
5th generations of wireless communications, will also utilize the RFR spectrum above 6 GHz. Other terms are also
used in the literature for part of the RFR spectrum, e.g., microwaves for frequencies above 1 GHz, millimeter waves
for frequencies above 30 GHz.
CELL PHONES AND RFR
Cell phones and other commonly used wireless communication devices transmit their signals via RFR to enable
voice calls and data transfer, including communication through the internet. Wireless phones are two-way radios
that contain both a receiver and a transmitter. When a user makes a call, voice sound is converted into digital
information. The information is imposed on to RFR and transmitted to the nearest base station. Base stations,
commonly referred to as cell towers, have antennas placed on towers that are free standing or mounted on existing
structures such as trees, water tanks, or tall buildings and contain electronic equipment and antennas that receive and
transmit RF signals and form a bridge to the rest of the communications infrastructure. The base station receives
and transmits radio signals in its area or “cell.” As the user moves around, the radio signal can be relayed within the
communications network from one “cell” of coverage to another, maintaining call connection. The call is routed
through the communications network either through a land line phone or another wireless phone again using radio
signals. To conserve energy and minimize interference, mobile phones automatically regulate the RFR signal
strength, and hence the emitted field, to the lowest power level possible for a connection to be made. However, in a
poor transmission environment (caused by, e.g., a distant base station, presence of obstacles between the base station
and the mobile phone, or interferences from adjacent calls) there is a higher output power and emission from the
mobile phone in order to make a connection. Therefore, the better the connection, the lower the power output of the
wireless device.
20 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
PROPERTIES OF CELL PHONE RFR
Cell phone RFR is a form of nonionizing electromagnetic energy that consists of propagating electromagnetic waves
of oscillating electric (E-) and magnetic (H-) fields that move together through space at the speed of light. As
opposed to ionizing radiation, which contains enough energy when passing through matter to break chemical bonds
or remove an electron from an atom or molecule to produce charged ions, nonionizing radiation refers to
electromagnetic energy that at most only has sufficient energy for excitation of an electron to a higher energy state.
Nonionizing radiation includes a broad range of the electromagnetic spectrum from extremely low frequency (ELF)
radiation to radio and microwaves, infrared, visible light, and near ultraviolet radiation. It has a lower frequency and
longer wavelength than ionizing radiation (Figure 1).
FIGURE 1
Electromagnetic Spectrum (OET, 1999)
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 21
Peer Review Draft NOT FOR ATTRIBUTION
Cell phone RFR fields transport large amounts of data at a very fast rate over long distances. RF waves are
characterized by their wavelength (the distance covered by one complete cycle of the electromagnetic wave) and
their frequency (the number of electromagnetic waves passing a given point in 1 second). The frequency of an RF
signal is expressed in terms of Hertz (Hz), where one Hz is equivalent to one cycle per second. The RF segment of
the electromagnetic spectrum is generally defined as the frequencies between approximately 3 kHz to 300 GHz.
The intensity of an RF field can be expressed by its electric and magnetic components and is measured in volts per
meter (V/m) for electric fields and amperes per meter (A/m) for magnetic fields. Another measure of RFR is the
power density, which is defined as the power per unit area and is expressed in watts per square meter (W/m2
) in the
far-field of sources. The quantity used to describe the amount of RFR energy absorbed by the body is referred to as
the specific absorption rate (SAR), which is expressed in watts per kilogram (W/kg). SAR is a function of the
geometry and the dielectric properties of biological tissues absorbing the energy (which result from the interaction
of electromagnetic radiation with constituents at the cellular and molecular level), the square of the strength of the
induced E-field, and the mass density of the exposed tissue. The SAR value is derived by averaging the absorbed
energy over a specific volume (typically 1 gram, 10 grams, or the whole body for regulatory purposes).
Cell Phone RFR Signal Modulation
In wireless telecommunications, modulation is the process of conveying digital or analog signals or information (the
message) by varying one or more parameters of another signal (the carrier), typically at a much higher frequency,
that can be transmitted over a distance. The modulated carrier contains complete information about the message
signal and the original message can be recovered by suitable signal processing of the signal when received at a
remote location (base station). One of the main goals of the modulation used in mass wireless communication
systems is to transfer as much data as possible in the least amount of spectrum. Over the years, multiple modulation
techniques have emerged to achieve and improve spectral efficiency, either when considering a single user in
isolation or multiple users simultaneously using the same spectrum.
Cell phone technology is typically referred to in “generations.” The first generation (1G) of wireless technology
was an analog system that used analog frequency modulation for voice calls. The 1G networks were introduced in
the 1980s and continued until they were replaced by networks of the second-generation (2G) networks. These
22 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
networks differed from the 1G networks in that they were digital, provided encryption, were significantly more
efficient, and introduced data services [i.e., text messages, picture messages, and Multimedia Message Service
(MMS)] in addition to voice calls. The 2G networks became commercially available in 1992 and used three
common multiple access technologies for accommodating multiple simultaneous users:
• Frequency Division Multiple Access (FDMA): the available spectrum is split into a number of distinct
parts (channels) each large enough to accommodate a single user or call without overlap; all users utilize
their channel 100% of the time for the duration of the call or message. The channels are normally of equal
bandwidth;
• Time Division Multiple Access (TDMA): the available spectrum is allocated to a single channel; each user
or call is assigned a certain portion of time;
• Code Division Multiple Access (CDMA): the available spectrum is allocated to a single channel; each user
or call is assigned a unique sequence code to spread the message over the available spectrum. All users use
the whole of the spectrum all of the time. At the receiver, the same unique sequence code is used to
recover the desired signal from the sum of all the user calls.
2G systems used a combination of FDMA/TDMA or CDMA for, for example, GSM and cdmaOne (IS-95),
respectively. While the 2G technology continues to operate, subsequent third and fourth generations of network
technologies were introduced in 1998 (3G), 2006 (4G), and 2011 (4G-LTE). These technologies were developed to
support increased data needs for multimedia access with increased bandwidth and transfer rates to accommodate
internet-based broadband applications, including video conferencing, streaming video, sending and receiving faxes,
and instantly downloading e-mail messages with attachments. With the introduction of 3G technology,
“smartphones” were developed. With these devices, the newer technologies were overlaid with 2G to support
multiple access modes (2G, 3G, and 4G) (Buddhikot et al., 2009). Although the 2G technologies will be phased out
over time and replaced by newer technologies, the current wireless communication networks continue to utilize 2G
for voice and text.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 23
Peer Review Draft NOT FOR ATTRIBUTION
All 3G systems utilize CDMA/WCDMA technology and fall into two groups complying with the 3rd Generation
Partnership Project (3GPP) or 3GPP2 family of standards. Universal Mobile Telecommunications Service (UMTS),
Wideband Code Division Multiple Access (WCDMA), and Time Division-Synchronous Code Division Multiple
Access (TD-SCDMA) are 3GPP variants; CDMA2000 (which is based on 2G cdmaOne) is 3GPP2. 4G systems use
Orthogonal Frequency Division Multiplexing (OFDM) within the E-UTRAS (LTE-Advanced) or Worldwide
Interoperability for Microwave Access (WiMAX) standards.
Modulation Schemes (GSM and CDMA)
The Global System for Mobile Communications (originally Groupe Spécial Mobile; GSM) was developed to
establish a digital standard for compatibility throughout Europe. GSM is a circuit-switched system that uses both
FDMA and TDMA technologies. The frequency division mechanism divides the GSM band into 200 kHz-wide
channels. The time division mechanism enables up to eight time slots (voice channels) per frequency channel
wherein a single cell phone transmits in only one out of eight available time slots during a voice communication.
This introduces a pulsed signal shape with a pulse repetition rate of 217 Hz. Such a TDMA frame has a length of
4.6 milliseconds (ms), and 26 TDMA frames make up a multiframe with a 120 ms duration. During a multiframe, a
mobile phone transmits in 25 out of 26 possible time slots. This TDMA frame structure causes significant low
frequency amplitude modulation components to be superimposed on the RF carrier at 8.3 and 217 Hz.
With GSM, the duplexing between uplink (when the handset transmits to the base station) and downlink (when
the base station transmits to the handset) is implemented in the frequency and time domain. Constant frequency
spacing is maintained between up and downlink frequencies: in the United States, the uplink is 1,850 to 1,910 MHz,
and the downlink is 1,930 to 1,990 MHz. The uplink and downlink frequencies are chosen according to the cell
(area that is covered by a base station) into which the mobile is registered. In order to minimize interference
between neighboring cells, a frequency reuse policy is applied. In this approach, when a mobile phone moves from
one cell into an adjacent cell, frequencies used for data uplink and downlink change in association with this
movement (i.e., transmission frequencies change at handover from one cell to another).
24 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
GSM technology implements a power control in order to increase the battery life of mobile handsets. The power
control has a dynamic range of 30 decibels (dB) subdivided into 2 dB power-level steps. The power control is
typically implemented using the Slow Associated Control Channel (SACCH), which facilitates a power control
update rate no faster than every four multiframes (480 ms). Once a target power level is received, the mobile station
is able to regulate its power in 2 dB steps every 60 ms. This means that power regulation over 15 steps (full
dynamic range) takes 900 ms. GSM base stations typically average the received signal strength from a mobile
phone over 1 second, such that the actual power regulation usually takes place after multiples of 480 ms.
The GSM supports data transfer speeds up to 9.6 kilobits/second, allowing the transmission of basic data services
such as Short Message Service (SMS), but not large packets of data such as internet access and streaming video.
CDMA technology uses a form of coded transmission known as Direct Sequence Spread Spectrum (DSSS) in which
data multiplies by a much faster pseudo random code before being modulated on to the carrier. The effect of the
multiplication is to spread the message across all frequency bands available for use at any time but with very
specific characteristics. CDMA signal access technology is based on code division separation of mobile stations as
well as base stations. This implies strong differences of the signal structure compared to GSM. For example, in
Interim Standard 95 (IS-95), in the forwardlink (downlink), a set of 64 Walsh codes (which are deterministic and
orthogonal) are applied to spread/separate the individual channels in the downlink of a cell. After the orthogonal
spreading, a short (16-bit) Pseudo Noise (PN) code is applied to further spread the signal and identify the cell.
Hence, a separation of neighboring cells in the frequency domain is no longer necessary. Eventually, there is no
need for the mobile station to change its transmission frequency during the transition from one cell into another. As
with GSM systems, the duplexing between the forward and reverse links is implemented in the frequency domain.
In CDMA systems, an efficient power control is crucial. Because all mobile stations transmit and interfere in the
same frequency channel, each mobile device decreases the signal to noise ratio of all the other mobile devices.
Hence, the output power of a mobile phone should be kept at a minimum that guarantees good transmission quality.
On the other hand, when moving around, the mobile device is subject to slow and fast fading, shadowing, external
interference, etc. In order to keep the signal received at the base station constant and compensate for effects on the
communication channel, a fast power control is necessary. Therefore, when a CDMA mobile station is active
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 25
Peer Review Draft NOT FOR ATTRIBUTION
(communicating), a closed-loop power control is applied. The base station monitors the signal quality in the reverse
link and inserts power-control bits in the communication channel. For example, in IS-95, the power control over a
dynamic range of 48 dB in 1 dB steps with an update rate of 800 MHz is implemented. The power control is
implemented by sending a binary value of “1” to regulate the transmit power 1 dB down, and “0” to regulate the
transmit power 1 dB up. A quasistatic power level is therefore implemented by an alternating 0101 power-control
pattern.
IS-95 (also known as cdmaOne) was developed by Qualcomm (San Diego, CA) as the first 2G CDMA-based digital
cellular technology. The term IS-95 generally applies to a protocol revision (P_REV=1) that was adopted as a
standard (TIA-EIA-95) by the Telecommunications Industry Association (TIA) in 1995. Over time, subsequent
iterations of the IS-95 protocol such as IS-95A, TSB-74, and IS-95B were developed, each with incremental
improvements over the previous protocols. Later, more advanced versions of the CDMA technology evolved to
include IS-2000, which incorporated much higher transfer rates than the previous 2G versions.
SOURCES, USE, AND HUMAN EXPOSURE
The predominant source of RFR for the majority of the population is in telecommunications and mobile internet
access applications for wireless devices. Aside from telecommunications, there are other man-made applications of
RFR, which include microwave ovens, radar, industrial heating and sealing, medical diagnostics [Magnetic
Resonance Imaging (MRI)] and therapy (surgical diathermy and ablation), and remote tracking or detection of
objects [anti-theft, Radio Frequency Identification (RFID)]. However, there are also natural sources of RFR such as
atmospheric electrical discharges (lightning) and solar and cosmic radiation. RFR exposures from natural sources
are much smaller and tend to be spread over a much wider range of frequencies compared to man-made fields
(IARC, 2013).
Highest human exposure to cell phone RFR primarily occurs through the use of cellular phone handsets and other
wireless devices held in closest proximity to the human body such as tablets and laptop computers. The use of cell
phones has become widespread over the last two decades amongst adults and children, thereby increasing the level
of RFR the population is exposed to. Concern has been expressed regarding the potential health risks associated
26 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
with use of cell phones. Particularly, there has been a great deal of focus on the possibility of increased risk of brain
cancer because traditionally, these devices were used in close proximity (0 to 2 cm) to the head, yet the advent of
smart phones has altered dramatically the usage scenarios for such devices away from a simple phone call. The
RFR exposure of a person is defined in terms of SAR, the power absorbed in the body, because the body has
complex geometry and tissue distributions, and even exposure to uniform RFR electromagnetic fields (EMF) will
result in nonuniform SAR distributions. In general (apart from the case when very close to the antenna), the level of
RFR exposure by a cell phone is inversely proportional to the square of the distance of the body from the device’s
antenna, and the highest SAR levels occur in the parts of the body nearest to the antenna. Accordingly, there is a
very nonuniform exposure to cell phone RFR across the whole body of cell phone users and even of bystanders.
Accurate and detailed estimation of cell phone RFR exposure in humans is difficult to obtain because the output
power of wireless devices constantly varies depending on several factors. Overall, the network carrier adjusts the
output power of each connected device to the lowest level that is still compatible with a good quality signal. This
adaptive power control occurs continuously and is achieved by a logarithmic downscaling of the time-averaged
power from the maximum of 0.125 and 0.25 W to a level as low as 1 mW. When a device is in use, the output
power (and subsequent exposure to cell phone RFR) from the device is increased compared to the output from that
same device in “standby” mode. Therefore, levels of exposures are related to the amount of active time a user
spends on the device. The output power of a device changes based on the signal received at the base station.
Decreases in signal strength result in higher output powers. Therefore, there are increases in the output power as the
distance between the device and the base station increases, if there are physical obstacles between the device and the
base station, multipath reflections, and during handovers in the case of GSM (handover is the passing of a call from
one base station to another when the user moves across the borders of cells or by network request to optimize
communication traffic). The proximity of the device to the body and the type, number, and position of antennas in
the device are other important factors affecting the amount of exposure to cell phone RFR.
Potential exposure to cell phone RFR also occurs from the cell phone towers (or base stations) that form the
network. While modern towers emit substantially more power than devices, exposures from base station antennas
are considerably lower in users than from the hand-held device. Typically, base station antennas are placed at
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 27
Peer Review Draft NOT FOR ATTRIBUTION
heights of 50 to 200 feet, in order to adequately cover an area (or cell). The antennas direct RF energy toward the
horizon, with some downward tilt. As with all forms of radiation (ionizing and nonionizing), the RF energy level
decreases rapidly as the distance from the antenna increases. As a result, the level of exposure to cell phone RFR at
ground level is very low compared to the level close to the antenna. Overall, the exposure level from base stations is
very small compared to exposure from the handheld devices.
Some base station antennas are installed on rooftops and at the top of lamp poles that are in close proximity or
adjacent to office space and residential buildings. Levels of exposure from these sources can approach or exceed
Federal Communications Commission (FCC) safety guidelines. Occupational exposure occurs during maintenance
on base stations. As a result, the FCC established guidelines for occupational exposures. Safety guidelines and
regulatory compliance are discussed below.
The levels of cell phone RFR inside buildings with base station antennas mounted on the roof or on the side of the
building are typically much lower than the levels outside, depending on the construction materials of the building.
Wood or cement block reduces the exposure to cell phone RFR by a factor of about 10. Due to the directional
nature of the signals, the energy level behind an antenna is orders of magnitude lower than in front of the antenna.
According to a Pew Research poll (Pew, 2017), approximately 95% of adult Americans own a cell phone. As of
December 2015, the number of active wireless subscriber connections was 377.9 million, which exceeded the
population of the United States (CTIA, 2017). According to the same survey, 49.3% of households in the United
States utilize only a wireless phone, and not a landline.
Safety Guidelines for Exposure
The Federal Communications Commission (FCC) and U.S. Food and Drug Administration (FDA) are jointly
responsible for the regulation of wireless communication devices.
28 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
Federal Communications Commission
The FCC is required by its responsibilities under the National Environmental Policy Act of 1969 to evaluate the
impact of emissions from FCC-regulated transmitters on the quality of the human environment (42 USC §4321
et seq.). As a result, the FCC regulates both the wireless devices as well as the base stations that form the cells of
the network. Since 1996, the FCC has required that all wireless communications devices (transmitting in the
100 kHz to 6 GHz frequency range) sold in the United States comply with its minimum guidelines for safety and
maximum RFR absorption standards based on SAR. The FCC requires a formal approval process for all devices
sold in the United States. FCC approval is contingent on the demonstration that the device does not exceed the
maximum allowable SAR level when the device is operating at its maximum power. The SAR limit adopted by the
FCC for exposure in the general population is 0.08 W/kg, as averaged over the whole body, and a peak
spatial-average SAR of 1.6 W/kg, averaged over any 1 gram of tissue (47 CFR §1.1310) when averaged over
6 minutes. Exceptions are made for the extremities (hands, wrists, feet, ankles, and pinnae), where the peak
spatial-average SAR limit is 4 W/kg, averaged over any 10 grams of tissue for an exposure period of no longer than
30 minutes. For occupational exposures, the whole-body SAR limit is 0.4 W/kg, and the limit for the peak
spatial-average SAR is 8 W/kg, averaged over any 1 gram of tissue. For the hands, wrists, feet, ankles, and pinnae,
the peak spatial-average SAR limit for occupational exposure is 20 W/kg, averaged over any 10 grams of tissue for
an exposure period not to exceed 6 minutes.
The FCC rules and guidelines for cell phone RFR exposure are based upon standards initially developed by the
Institute of Electrical and Electronics Engineers (IEEE) and the National Council on Radiation Protection and
Measurements (NCRP). These standards for RF exposure in workers and the general population are based on
protection against adverse effects that might occur due to increases in tissue or body temperature in excess of 1° C
(wbSAR, approximately 4 W/kg) or less (after applying safety factors). Because RF-energy absorption and any
induced effects are dependent on the frequency of incident-field parameters and the composition of exposed tissues,
it has been suggested that quantifying SARs in small averaging regions is more relevant for evaluations of human
health effects.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 29
Peer Review Draft NOT FOR ATTRIBUTION
Food and Drug Administration
The FDA does not currently regulate the use of wireless communications devices or the devices themselves. The
FDA also does not require safety evaluations for radiation-emitting wireless communication devices. It does
maintain the authority to take regulatory action if it is demonstrated that exposure to the emitted cell phone RFR
from these devices is hazardous to the user.
ABSORPTION OF CELL PHONE RFR
RFR interacts with the human body via inductive or capacitive coupling or a combination of both. The absorption
of the coupled RFR is dependent on the frequency of the signal and the dielectric properties of the exposed tissue. It
generates oscillating currents in the tissue, which in turn give rise to induced E-fields. The energy is transferred into
molecular motion of polar molecules like water, a strongly dipolar molecule and major component of biological
tissues. Resonant oscillations in polar subgroups of cellular macromolecules are damped by collisions with
surrounding water molecules that disperse the energy of the RF signal into random molecular motion. Tissue
heating occurs as the energy is transferred to the surrounding aqueous environment as heat (IARC, 2013).
The SAR (W/kg) is a measure of the absorption of RF energy by biological tissues. It is a function of several main
factors: the electrical conductivity (Siemens/meter) of the tissue, the square of the strength (Volts/meter) of the
induced E-field, and the geometry and mass density (kg/meter3) of the tissue absorbing the energy. The SAR is
calculated as the average of the absorbed power over a specific volume of tissue (typically 1 or 10 gram volume of
tissue or the whole body).
TOXICITY
A comprehensive review of the toxicity of cell phone RFR in in vitro models, laboratory animals, and humans was
recently conducted and published in the IARC Monograph series (IARC, 2013).
30 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
Thermal Effects
Given the ability of cell phone RFR to heat tissues, the toxic effects of cell phone RFR are often classified as
thermal or nonthermal effects, based on whether the observed effect was a result of a significant temperature change
(thermal effects) or independent of any change in temperature considered in excess of thermal noise (nonthermal
effects). The most well-established and biologically plausible mechanism for cell phone RFR-induced effects in
biological systems is through tissue heating resulting in damage. It has been well established that excessive heating
causes significant damage to cells, tissues, and organs. At high enough levels of cell phone RFR exposure, the
absorption of energy could lead to increased heating to the point that it overwhelms an organism’s ability to
thermoregulate and maintain an acceptable body temperature. Because human exposures to cell phone RFR occur at
intensities that are not expected to cause significant thermal effects, the nonthermal effects are more appropriate to
the evaluation of potential effects in humans.
Nonthermal effects refer to biological changes that occur with body temperature increases that are below 1° C.
Changes of temperature up to 1° C are considered in the range of thermal noise (IARC, 2013). There is an ongoing
debate regarding whether nonthermal biological effects can occur as a result of exposures to low-intensity cell phone
RFR. It has been suggested that there is no plausible nonthermal mechanism by which exposure to low-intensity
RFR could induce significant biological effects (Adair, 2003; Prohofsky, 2004; Sheppard et al., 2008). However,
there are numerous reports of specific biological effects associated with cell phone RFR exposures at levels
considered below those expected to result in a measurable amount of tissue heating. Other than tissue heating, the
mechanisms of interaction between cell phone RFR and biological systems have not been well characterized, but
several mechanisms have been proposed for these nonthermal effects in biological systems, including the generation
of reactive oxygen species, induction of ferromagnetic resonance, demodulation of pulsed RF signals, and the
alteration of ligand binding to hydrophobic sites in receptor proteins (IARC, 2013). Additionally, low levels of
exposure to cell phone RFR may result in small temperature changes in localized areas of exposed tissues that cause
conformational changes in temperature-sensitive proteins and induce the expression of heat-shock proteins.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 31
Peer Review Draft NOT FOR ATTRIBUTION
Experimental Animals
Toxic effects have been reported in various types of studies in cell phone RFR-exposed laboratory animals and
in vitro systems. Most studies investigating the potential toxicity of cell phone RFR have focused primarily on
genotoxicity and related effects; these findings are summarized in the Genetic Toxicity section. However, several
studies have been conducted to evaluate other aspects of toxicity, including specific studies on gene and protein
expression, immunotoxicity, and permeability of the blood-brain barrier. The results of these studies have been
mixed. It is important to note that these studies were conducted with cell phone RFR of differing parameters
(frequency, power density, continuous wave versus amplitude-modulated signals, etc.). Because there may be
differences in cell phone RFR-induced responses depending on the frequency, modulation, and power density, it is
not surprising that the results reported in the literature can be somewhat inconsistent.
Several effects on the humoral and cell-mediated responses of the immune system have been reported at various
frequencies of cell phone RFR in rats and mice. These include effects on the activity of NK cells, plaque-forming
cell response to sheep erythrocytes, production of tumor necrosis factor (TNF) in peritoneal macrophages and
splenic T-cells, mitogenic response in T lymphocytes, phagocytic activity of neutrophils, leukocyte profile, and
thymic and splenic cellularity (Smialowicz et al., 1983; Guy et al., 1985; Veyret et al., 1991; Novoselova et al.,
1999; Lushnikov et al., 2001; Kolomytseva et al., 2002). However, many of these effects were observed in studies
conducted with cell phone RFR at frequencies greater than 10 GHz. Other studies have demonstrated no exposurerelated
effects on the immune system (Elekes et al., 1996; Chagnaud and Veyret, 1999; Lushnikov et al., 2001;
Gatta et al., 2003; Nasta et al., 2006).
A few studies have investigated the impact of cell phone RFR at frequencies between 800 and 1,900 MHz on gene
and protein expression. Several studies have demonstrated that cell phone RFR can alter the expression of certain
genes in the brain (Fritze et al., 1997; Belyaev et al., 2006; Nittby et al., 2008), while others have failed to associate
cell phone RFR exposure with changes in gene expression (Stagg et al., 2001; Paparini et al., 2008). The expression
of various proteins has also been investigated in rats and mice. These studies have primarily yielded negative results
for the specific proteins being evaluated in the rat brain (Fritze et al., 1997; Belyaev et al., 2006; Ammari et al.,
2008, 2010; Dasdag et al., 2009). Similarly, no effects of cell phone RFR on protein expression have been reported
32 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
in the testis (Lee et al., 2010) or in the skin (Masuda et al., 2006; Sanchez et al., 2006, 2008). Changes in the
expression of bone morphogenic protein and bone morphogenic protein receptors have been reported in the kidney
of newborn rats (Pyrpasopoulou et al., 2004). A study by EÅŸmekaya et al. (2010) also demonstrated increased
expression and activity for caspase 3 and caspase 9 in the thyroid gland of Wistar rats.
Exposure to cell phone RFR induces changes in markers for oxidative stress in multiple tissues, including the brain
(Ilhan et al., 2004; Meral et al., 2007; Ammari et al., 2008; Sokolovic et al., 2008; Imge et al., 2010), heart
(Ozguner et al., 2005a), kidney (Oktem et al., 2005; Ozguner et al., 2005b), eye (Ozguner et al., 2006), liver (Ozgur
et al., 2010; Tomruk et al., 2010), endometrium (Oral et al., 2006; Guney et al., 2007), and testis and epididymis
(Mailankot et al., 2009). A few studies have also demonstrated cell phone RFR-mediated effects on differentiation
and apoptosis in the endometrium (Oral et al., 2006; Guney et al., 2007) and brain (Dasdag et al., 2009; Sonmez
et al., 2010). Changes have also been noted in the permeability of the blood-brain barrier in some studies (Eberhardt
et al., 2008; Nittby et al., 2009, 2011). However, other studies conducted under similar experimental conditions
failed to demonstrate any effect of cell phone RFR exposure on the permeability of the blood-brain barrier
(Grafström et al., 2008; de Gannes et al., 2009; McQuade et al., 2009; Masuda et al., 2009).
Humans
Numerous epidemiology studies have been conducted to investigate the association between exposure to cell phone
RFR and health effects in humans. However, many of these studies were conducted in small groups exposed to cell
phone RFR signals with different characteristics (frequencies, modulations, intensities, etc.) than the specific
frequency bands and modulated cell phone RFR signals used in wireless communication. Many of these studies
evaluate microwaves, ELF fields, and radar, which are all different forms of RFR. While these studies may provide
additional data for the evaluation of the toxicity of RFR in general, a smaller subset of these studies, which
specifically evaluated cell phone RFR at the frequencies and modulations used in wireless communications is more
critical to evaluating the potential toxicity of cell phone RFR from mobile communication devices.
There is a very limited set of research investigating the general toxicity of cell phone RFR in humans because most
of the focus for research has been on the potential for carcinogenic effects. Studies in humans have failed to
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 33
Peer Review Draft NOT FOR ATTRIBUTION
demonstrate any consistent adverse health effects in cell phone RFR-exposed populations. There are reports of
some exposed individuals that describe acute, subjective effects following exposure to cell phone RFR, including
headaches, fatigue, skin itching, and sensations of heat (Frey, 1998; Chia et al., 2000; Hocking and Westerman,
2000; Sandström et al., 2001; Santini et al., 2002a,b). However, these have primarily been reported in people that
consider themselves electrosensitive, and not in the general population. It has been suggested that there are likely
other causes, not cell phone RFR, for these subjective symptoms (Kwon and Hämäläinen, 2011). In fact, the
validity of electrosensitivity as an actual phenomenon has been questioned and debated. Variable results have been
observed in the electroencephalogram (EEG) of volunteers exposed to RFR during sleep. Some studies indicate that
exposure to cell phone RFR induces changes in sleep latency and sleep EEG (Mann and Röschke, 1996; Wagner
et al., 1998, 2000; Borbély et al.,1999; Huber et al., 2000, 2002, 2003; Loughran et al., 2005; Hung et al., 2007;
Regel et al., 2007; Lowden et al., 2011). Glucose metabolism in the brain, a marker for brain activity, is increased
in the region of the brain closest to the antenna (Volkow et al., 2011). While these results demonstrate
exposure-related effects, the toxicologic significance of these findings is unclear.
No effects of cell phone RFR on the neuroendocrine system, auditory and vestibular systems, or consistent effects
on cognitive performance have been reported in humans. There is also no clear evidence of effects on heart rate or
blood pressure.
CARCINOGENICITY
The carcinogenic potential of cell phone RFR in animals and humans is widely debated. A comprehensive review of
the carcinogenicity of cell phone RFR in laboratory animals and humans was recently conducted and published in
the International Agency for Research on Cancer (IARC) Monograph series (IARC, 2013).
Experimental Animals
Studies published to date have not demonstrated consistently increased incidences of tumors at any site associated
with exposure to cell phone RFR in rats or mice. No increases in tumor incidences were observed in B6C3F1 mice
exposed to GSM-modulated cell phone RFR for 24 months (Tillmann et al., 2007), F344 rats exposed to
CDMA-modulated cell phone RFR for 24 months (La Regina et al., 2003), or Wistar rats exposed to
34 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
GSM-modulated cell phone RFR for 24 months (Smith et al., 2007). In studies conducted in transgenic and
tumor-prone mouse strains, exposure to cell phone RFR has not been consistently associated with an increased
incidence of tumors at any site (Utteridge et al., 2002; Sommer et al., 2004, 2007; Oberto et al., 2007; Lee et al.,
2011). While these studies have advanced the knowledge of the potential toxicity of cell phone RFR, critical
limitations in the design of many of these studies severely limit the utility of the information to adequately evaluate
the carcinogenicity of cell phone RFR. These limitations include studies with very short daily exposure durations
(≤ 2 hours per day) in restrained animals or with levels of cell phone RFR exposures too low to adequately assess
carcinogenic potential. The focus of many of these studies conducted in genetically-altered and tumor-susceptible
mice was not to evaluate the overall carcinogenicity of cell phone RFR, but to investigate the effects in the specific
predisposed tissues in that model.
Based on the constraints in the designs of the existing studies, it is difficult to definitively conclude that these
negative results clearly indicate that cell phone RFR is not carcinogenic. To adequately evaluate the potential
chronic toxicity and carcinogenicity of cell phone RFR, further studies with enhanced study designs and improved
exposure paradigms were needed.
Humans
As a result of the IARC review conducted in 2011, RF electromagnetic fields were classified as possibly
carcinogenic to humans (Group 2B). This classification was based on limited evidence of carcinogenicity in humans
based on positive associations between exposure to RFR from wireless phones and increased risk for gliomas and
acoustic neuromas, specifically in users with the greatest amount of cell phone usage. The IARC Working Group
acknowledged that the findings were affected by potential selection and information bias, weakness of associations,
and inconsistencies between study results (IARC, 2011).
While several other studies were considered, the IARC evaluation was based primarily on reports from the
INTERPHONE Study, the largest research effort conducted to date examining the potential association between
exposure to cell phone RFR and cancer in humans. INTERPHONE was an IARC-coordinated research effort that
included a series of studies conducted with a common core protocol at 16 study centers in 13 countries: Australia,
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 35
Peer Review Draft NOT FOR ATTRIBUTION
Canada, Denmark, Finland, France, Germany, Israel, Italy, Japan, New Zealand, Norway, Sweden, and the
United Kingdom (Cardis et al., 2007). The studies were specifically designed to investigate the association between
cell phone RFR and tumors of the brain (glioma and meningioma), acoustic nerve (schwannoma), and parotid gland.
The final report for the INTERPHONE studies was published in 2011 (IARC, 2011).
The results of these studies seemingly demonstrated an elevated risk of glioma and acoustic neuroma in the group in
the highest decile for exposure (cumulative phone call time). However, the INTERPHONE study group concluded
that recall and selection biases and implausible values for usage reported by the participants in the study may
explain the increased risk (INTERPHONE Study Group, 2010, 2011). Further, the INTERPHONE studies and other
published epidemiological studies may have been concluded prior to the potential lag time (the interval between the
time of the onset of exposure and the subsequent development of a tumor) for the development of slow-growing
brain tumors. Overall, the authors of these studies concluded that there was no significant increase in risk of glioma,
meningioma, or acoustic neuroma associated with the use of cell phones.
Other studies have compared time trends in cell phone usage and the incidences of different types of cancers to
investigate indirect evidence of an association between cell phone RFR and cancer. These studies were conducted
across several different countries (Saika and Katanoda, 2011), and in a group of European countries (Lönn et al.,
2004; Nelson et al., 2006; Röösli et al., 2007; Deltour et al., 2009; de Vocht et al., 2011), the United States (Muscat
et al., 2006; Propp et al., 2006; Inskip et al., 2010), Japan (Nomura et al., 2011), New Zealand (Cook et al., 2003),
and Israel (Czerninski et al., 2011). Overall, the evaluations suggest that there were no significant changes in the
trends of cancer incidences. Any minor increases in cancer rates that were observed in these studies were attributed
to enhanced detection capabilities for cancer that were the result of advances in diagnostic medical equipment, like
computerized tomography (CT) scans and MRI.
Several cohort studies have been conducted, but also failed to establish a clear association between cell phone RFR
and the development of any of the investigated cancer types (Johansen et al., 2001; Schüz et al., 2006, 2011).
Additional studies have demonstrated that there was no association between cell phone usage and pituitary gland
tumors (Takebayashi et al., 2008; Schoemaker and Swerdlow, 2009), testicular tumors (Schüz et al., 2006; Hardell
36 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
et al., 2007), parotid gland tumors (Hardell et al., 2004; Lönn et al., 2006), uveal melanoma in the eye (Schüz et al.,
2006; Stang et al., 2009), and cutaneous melanoma (Hardell et al., 2011). Some studies have demonstrated that
there was no association between cell phone usage and leukemia (Johansen et al., 2001; Schüz et al., 2006) and
non-Hodgkin’s lymphoma (Hardell et al., 2005), whereas others have reported increased risk of non-Hodgkin’s
lymphoma (Linet et al., 2006) and leukemia (Kaufman et al., 2009).
Many of the epidemiological studies that have been published are limited in their ability to definitively establish a
causal association between cell phone usage and increased cancer incidences due to recall and selection bias,
confounding factors, and low study participation.
As mentioned previously, the utility of human studies with regard to evaluation of the carcinogenic potential of cell
phone RFR is dependent upon the length of time the subjects in the studies were exposed to cell phone RFR. Given
the long latency period between the initiation of exposures and the development of tumors, a sufficient duration of
exposure must be reached in order to evaluate the association between exposure and cancer outcome. Because
widespread usage did not occur until the 1990s in some countries, these populations may not have been exposed
long enough to expect any changes in cancer incidences compared to studies in populations where widespread use
occurred five or more years earlier in the late 1980s.
GENETIC TOXICITY
Extensive reviews of the literature on the genotoxicity of various frequencies and modulations of cell phone RFR,
covering experimental systems ranging broadly from cell-free DNA preparations to cells of exposed animals and
humans, have concluded that evidence for cell phone RFR-associated genotoxicity is inconsistent and weak (Brusick
et al., 1998; Verschaeve et al., 2010; Repacholi et al., 2012; Vijayalaxmi and Prihoda, 2012). Interpretations of the
genotoxicity studies and the ability to draw definitive conclusions based on weight-of-evidence from the large
number of studies that have been reported have been hampered by inadequacies in experimental design, especially
related to exposure standards and radiation-measuring procedures (Brusick et al., 1998). Although the majority of
studies report a lack of effect, the several reports of a positive response are concentrated among experiments
assessing chromosomal or DNA damage in mammalian cell systems in vitro and in vivo. Some key studies
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 37
Peer Review Draft NOT FOR ATTRIBUTION
reporting cell phone RFR-associated genotoxicity in human cell lines, including DNA damage and chromosomal
effects, could not be replicated (Speit et al., 2007, 2013). A critical complicating factor in the study of the genotoxic
effects of cell phone RFR is that under certain conditions, cell phone RFR is sufficiently energetic to heat cells and
tissues, and not all studies have considered this factor in their design. Exposure to heat in vivo and in vitro has
produced positive results in tests for genotoxicity, such as the comet assay and micronucleus assay (Asanami and
Shimono, 1997; Komae et al., 1999; Speit and Schütz, 2013). The mode of action whereby heat induces these
effects appears to be through induction of protein denaturation and aggregation, which can interfere with chromatin
structure and slow the kinetics of DNA repair (Kampinga and Dikomey, 2001; Hunt et al., 2007). Thus, heatinduced
increases in DNA migration seen in the comet assay may reflect slowed repair of endogenous lesions, and
similarly, activity in the micronucleus assay may be due to aneugenic rather than clastogenic events (Asanami and
Shimono, 1997; Komae et al., 1999; Speit and Schütz, 2013). Therefore, it is important to distinguish between
nonthermal and thermal conditions when studying measures of genotoxicity following exposure to cell phone RFR.
STUDY RATIONALE
The FDA nominated cell phone RFR emissions of wireless communication devices for toxicology and
carcinogenicity testing. Current exposure guidelines are based on protection from acute injury from thermal effects
and little is known about the potential for health effects from long-term exposure to RFR below the thermal hazard
threshold. Epidemiology studies that have been conducted to date have not demonstrated a causal link between cell
phone RFR and any health problems in humans, however the results of these studies are complicated by
confounding factors and potential biases. Additionally, exposures in the general population may not have occurred
for a long enough period to account for the long latency period of some types of cancers in humans. Similar to the
challenges faced in epidemiological studies, studies in laboratory animals have been complicated by limitations that
researchers have faced in conducting robust studies designed to characterize the toxicity and carcinogenicity of cell
phone RFR.
For years, the primary concern regarding the potential health risk of chronic exposure to cell phone RFR was brain
cancer based on the proximity of wireless devices near the head during use. While the brain is an organ of concern,
understanding the potential toxicity and carcinogenicity of whole-body exposure is critical. Cell phone RFR is
38 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
constantly emitted from wireless devices to communicate with base stations, regardless of whether the user is on a
call or not. As the public has become more aware of the uncertainty regarding the potential effects of cell phone
RFR on the brain, more emphasis has been placed on the use of wired or wireless headsets (like Bluetooth), which
minimize cell phone RFR exposure to the head. In recent years, the density of cell towers has increased to cope with
the increasing demand for capacity, resulting in installations closer to residential neighborhoods and schools.
Additional cell phone RFR technologies, like SmartMeters used by power companies, transmit data in real time
using cell phone-type RFR. These existing and emerging technologies may potentially increase the levels of
exposures in human populations. These and other additional sources also expose different parts of the body, not
only the head.
In 2013, cell phone RFR was classified by the IARC as possibly carcinogenic to humans based on limited evidence
of an association between exposure to cell phone RFR from heavy wireless phone use and glioma and vestibular
schwannoma (acoustic neuroma) in human epidemiology studies and limited evidence for the carcinogenicity of cell
phone RFR in experimental animals (IARC, 2013). While ionizing radiation is a well-accepted human carcinogen,
theoretical arguments have been raised against the possibility that nonionizing radiation could induce tumors
(discussed in IARC, 2013). Given the extremely large number of people who use wireless communication devices,
even a very small increase in the incidence of disease resulting from exposure to cell phone RFR generated by those
devices would translate to a large number of affected individuals, which would have broad implications for public
health.
39
Peer Review Draft NOT FOR ATTRIBUTION
MATERIALS AND METHODS
OVERVIEW
The establishment of the National Toxicology Program (NTP) research program on radio frequency radiation (RFR)
has required the coordination of expertise from multiple scientific and engineering disciplines. At the initiation of
the RFR research program, a collaboration was established with technical experts from the Radio-Frequency Fields
Group in the Radio Frequency (RF) Technology Division, which is part of the Communications Technology
Laboratory (CTL) at the National Institute of Standards and Technology (NIST, Boulder, CO). NIST evaluated the
existing exposure systems and identified the types of improvements that would be required to provide a system of
sufficient size and power to conduct robust toxicology and carcinogenicity studies with uniform RFR exposures in
unrestrained, individually housed animals for a minimum of 6 hours a day at frequencies and modulations that
reflected those in use at the time. The design of the chambers and toxicology studies required special consideration
of logistical, financial, and engineering limitations.
NIST tested the feasibility of a reverberation chamber-type exposure system by conducting a series of studies on
field strengths, field uniformity, and power requirements under various conditions of RFR exposure in such
chambers. These studies provided critical information for the design of experimental studies with respect to the
number of cages that could be placed in specific size chambers, the arrangement of cages within each chamber, and
the input power requirements.
NTP also worked with the Foundation for Research on Information Technologies in Society (IT’IS, Zurich,
Switzerland), which conducted studies using computational models that simulated RFR dosimetry to provide
estimates of whole-body and organ-specific internal field strengths and specific absorption rates (SARs) during
exposure. Based on information and parameters obtained during the NIST feasibility studies, IT’IS built a prototype
reverberation chamber as the basis for an exposure system to study health effects of long-term exposure of
40 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
laboratory animals. Following completion, NIST evaluated the prototype exposure chamber to determine if it met
the requirements specified by the NTP.
After prototype-testing by IT’IS Foundation and NIST, the IT’IS Foundation built the reverberation chambers
required for the NTP RFR exposure facility. Chambers were installed at the Illinois Institute of Technology (IIT)
Research Institute (IITRI, Chicago, IL). Following the installation and initial testing of the exposure system by
IT’IS and IITRI, technical experts from NIST conducted an independent validation of the system. NIST confirmed
that the probe readings in the system were consistent, that field uniformity was within expected specifications, and
that the signal quality was acceptable. NIST performed additional evaluations prior to initiation of the 2-year
studies and after completion of the studies to determine if any changes occurred in the signal quality, field
uniformity, or consistency of in-chamber field measurements. All studies were conducted at IITRI with real-time
monitoring of the system performance at IT’IS Foundation.
Institution Role
National Institute of Standards and Technology (NIST)
(Boulder, CO)
Suggested reverberation chamber exposure system
Conducted feasibility studies for reverberation chambers
Established various technical parameters for chambers
Evaluated the prototype chamber built by IT’IS Foundation
Validated the system prior to the conduct of studies at IITRI
Reevaluated RFR exposures prior to and after 2-year studies
IT’IS Foundation
(Zurich, Switzerland)
Constructed and tested prototype chamber
Refined technical parameters
Built the chambers for the NTP exposure facility
Installed chambers at IITRI
Monitored system performance throughout all phases of the studies
Conducted maintenance on exposure system hardware and software
IIT Research Institute (IITRI)
(Chicago, IL)
Tested exposure system after installation
Conducted maintenance of exposure system hardware
Conducted all toxicology and carcinogenicity studies
Conducted day-to-day operations
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 41
Peer Review Draft NOT FOR ATTRIBUTION
REVERBERATION CHAMBER METHOD OF EXPOSURE
The use of the reverberation exposure chamber as a method for exposing rats and mice to cell phone RFR was
conceptualized by NIST and further designed and tested by NIST and the IT’IS Foundation. A reverberation
chamber is a resonant box where the resonances and field structure are continuously modified under the influence of
metallic stirrers, introduced to change the effective geometry, such that when averaged over time, the field strength
is uniform over the entire exposure volume. A reverberation chamber exposure system was selected by the NTP for
the primary benefit that controlled exposures can be achieved in unrestrained animals (rats and mice) with extended
daily RFR exposure periods compared to other methods of exposure for up to 2 years.
Preliminary studies were first conducted at NIST to test the concept of reverberation chambers. In these studies,
field strengths and field uniformity were measured under various conditions of cell phone RFR exposure, including
an empty chamber and a chamber loaded with water bottles (simulating animals) at different locations in the
chamber. Power requirements were evaluated to achieve desired SAR levels. The effects of proximity between
water bottles were also investigated to avoid electromagnetic coupling. These studies provided critical information
for the design of experimental studies with respect to the number of cages that could be placed in specific size
chambers, the arrangement of cages within each chamber, and the input power requirements. The results of these
investigations demonstrated that while variations occurred over time and space the average cell phone RFR field
was uniform over the large volume of the chamber. These studies also demonstrated that cell phone RFR field
exposure occurred from all directions and all polarizations, and that there was uniformity of SAR in reverberation
chambers. Based on the information and parameters obtained during the NIST feasibility studies, a custom-built
prototype reverberation chamber was constructed and tested by the IT’IS Foundation. The development of the
prototype chamber involved the design of amplifiers and antennas for signal generation, the design of vertical and
horizontal stirrers to improve the homogeneity of experimentally generated RF fields, the development of both
hardware and software for the control and monitoring of experimentally generated RF signals, and testing of
chamber performance. During the design of the prototype exposure chamber, engineering studies were performed to
optimize the following prior to construction:
42 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
• The uniform field volume within each chamber to minimize spatial variability in the characteristics of
generated RF fields within a chamber such that all animals housed within the chamber space were exposed
to comparable RF field strengths
• The design and placement of stirrers in each chamber in order to maximize homogeneity of
experimentally-generated RF fields
• The design and location of RF antennas in each chamber
• The location of cage racks within the exposure chamber in order to provide appropriate separation of
individual animal cages and cage racks from all reflective surfaces (chamber walls, chamber floor and
ceiling, antennas, and stirrers) in the reverberation chamber
• Chamber volume to provide adequate space for staff to observe animals, collect data, and perform routine
animal husbandry operations, while minimizing overall chamber volume to minimize the chamber
size/footprint and the RF power required to maintain target SARs
The final reverberation chamber design for use in these studies was a fully-shielded room constructed of stainless
steel, equipped with a shielded room door to eliminate leakage of RFR signals, two rotating stirrers (one horizontal
and one vertical), ventilation structures, and RFR excitation antennas. A detailed rationale for the selection of
reverberation chambers for exposure to RFR and a full description of the exposure system are provided in Capstick
et al. (2017) and Gong et al. (2017).
As part of the validation of the reverberation chamber exposure system design, a team of engineers from NIST
conducted an independent evaluation of chamber design and exposure system operation in order to evaluate the
suitability of the reverberation chamber model for use in the program. NIST engineers evaluated the design and
operation of the prototype chamber and performed an extensive series of RF measurements to support an evaluation
of system performance.
CELL PHONE RFR EXPOSURE FACILITY
The exposure facility was specifically designed to expose mice in reverberation chambers to three different power
levels of modulated cell phone RFR [Global System for Mobile Communications (GSM) or Code Division Multiple
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 43
Peer Review Draft NOT FOR ATTRIBUTION
Access (CDMA)] at 1,900 MHz for up to 2 years to evaluate toxicity and carcinogenicity. The completed exposure
facility consisted of a total of 21 RFR reverberation exposure chambers (seven designated for mice); the RFR signal
generation, amplification, and monitoring systems; software for chamber operation; and hardware and software for
monitoring of environmental and exposure conditions within each chamber. All system hardware and software were
installed by the IT’IS Foundation.
During exposures, modulated (GSM or CDMA) cell phone RFR signals were generated by a signal generator,
amplifiers amplified the signals, and the signals were delivered by antennas in the reverberation chambers. RFR
field strengths were monitored in real time and were adjusted throughout the studies to achieve specific exposure
levels [based on SARs quantitated in watts (W) per kg body weight]. Environmental conditions were also
monitored and controlled in real time throughout the study. RFR exposures and environmental conditions were
monitored and controlled by a computer in a control room at the study laboratory at IITRI; the IT’IS Foundation was
also capable of remote system monitoring and control.
Facility Design and Reverberation Chambers
Each reverberation chamber was permanently programmed for a specified modulation (GSM or CDMA) of the
1,900 MHz cell phone RFR specified for the mouse studies. SARs for each chamber were adjustable and selected
prior to exposures. The field strength required to achieve a given target SAR (W/kg) exposure level is a function of
animal body weight (kg); however, separate chambers were not required for male and female B6C3F1/N mice
because their body weights and growth curves are sufficiently similar to yield similar SARs. To conduct robust
toxicology studies with three exposure groups (low, medium, and high), three chambers were required for different
levels of exposures for GSM modulation and three for CDMA modulation. A sham exposure chamber without any
cell phone RFR signal provided shared control groups for the parallel studies of the two modulations. As per these
requirements, the cell phone RFR exposure facility consisted of seven reverberation chambers for exposures in mice
including:
• Three power levels for mice exposed to GSM-modulated cell phone RFR at 1,900 MHz
• Three power levels for mice exposed to CDMA-modulated cell phone RFR at 1,900 MHz
• One sham control chamber for mice with no RFR exposure.
44 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
The chamber size was designed to accommodate the RF field stirring paddles (described below), approximately
220 individually housed mice, and a minimum distance (3/4 of a wavelength) between the cages and the walls, floor,
ceiling and stirrers, respectively. The interior of the chamber was suitable for cleaning using high-pressure water
(after the RF antennas were protected). The internal dimensions of the chambers were 2.2 m (width) × 3.7 m
(length) × 2.6 m (height); the exterior dimensions were 2.3 m (width) × 3.8 m (length) × 2.85 m (height). A
floorplan for the exposure facility and images of the interior and exterior of the chambers are presented in Figures 2
and 3.
Each chamber contained two motor-controlled stirring paddles (one vertical and one horizontal) with adjustable
speed control (1 to 50 rpm) and large asymmetrical reflecting surfaces. Stirring paddles were placed off center in
the chamber for maximum scattering of the RFR fields to generate a statistically homogeneous field distribution
when averaged over time. The horizontal stirrer was mounted on the ceiling of the chamber. The vertical stirrer
was at the rear of the chamber, and was protected by rack guides that prevented contact with the animal cage racks.
Cage Racks and Watering System
Cages, cage racks, and watering systems for standard laboratory use contain elements that have the ability to alter
the exposure of the animals or introduce potential confounding factors. Because cage racks and the drinking water
delivery system were contained inside the chambers during exposure periods, it was required that these components
be constructed of durable materials that had essentially no impact on the RF fields generated in the chamber.
Metallic cage rack components, cage lids, feed dispensers, and cage grommets all needed to be eliminated. Hence,
custom engineering was required to overcome the challenges regarding potential RFR exposure-altering aspects of
the caging and cage racks used to house the animals during the studies. The safe provision of drinking water
provided the largest challenge for the studies.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 45
Peer Review Draft NOT FOR ATTRIBUTION
FIGURE 2
Exposure Facility Floor Plan for the Cell Phone RFR Studies (Not shown are the Ethernet connections to computers in the control room.)
Mouse chamber designations: low GSM=14; medium GSM=12; high GSM=11; low CDMA=3; medium CDMA=2; high CDMA=1;
sham control=13. The 14 other chambers (including 12 for cell phone RFR exposure and two for sham control) were designated for concurrent rat studies.
46 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
FIGURE 3
Exterior view of chambers, empty chamber showing the vertical and horizontal stirrers,
and chamber with cage racks in place
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 47
Peer Review Draft NOT FOR ATTRIBUTION
The absorption of RFR energy by water, if supplied by nonmetallic sipper tubes and distribution systems or bottles,
could lead to dose-dependent elevated water temperatures. As the same time, the potential for enhanced exposure
fields by metallic sipper tubes or lixits precluded the use of water bottles or a standard automatic watering system in
the reverberation chambers. The absorption of RFR energy by water could result in significant heating of the
drinking water, thereby decreasing water palatability and increasing the required RFR power to achieve the desired
exposure field strength, potentially to the extent that the exposure levels could not be met. To overcome these
challenges, adaptations were made to an automatic watering system so that the delivery of drinking water to the
animals would not interfere with cell phone RFR dosimetry. The water system was constructed from stainless steel
ensuring no dose-dependent energy absorption in the water (avoiding exposure-dependent water temperature) and in
structures around the lixits to ensure no enhanced fields that could lead to excessive SAR in the animals while
drinking.
Customized, nonmetallic animal cage racks for the reverberation chambers were designed by IITRI to minimize
any absorption of RFR or disruption of RF field homogeneity. Cage racks were constructed primarily of box beam
fiberglass (with some angle beam fiberglass used in nonweight-bearing areas of the rack). The shelves/cage lids
were constructed of a clear polycarbonate sheet with slots for increased airflow. The potential impact of the racks
on RF fields was evaluated in the prototype reverberation chamber by the IT’IS Foundation. Cage racks were
designed to accommodate the automatic watering system and position the perimeter of each animal cage at least
one-half wavelength from any reflecting surface. The specific considerations for design and further details of the
custom-designed cage racks and adapted automated watering system are provided in Capstick et al. (2017).
Cell Phone RFR Exposure System Control
The hardware and chambers designated for mice (using an exposure frequency of 1,900 MHz) were connected to a
dedicated computer control system using an Ethernet protocol. The computerized control system managed and
monitored the cell phone RFR exposures and environmental conditions in the chambers. A more detailed
description of the computer control of cell phone RFR exposure is provided in Capstick et al. (2017).
48 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
The control computer managed the exposure schedule, stirrer rotation speeds, exposure signal and level, and
monitored air flow, temperature, humidity, light, and the electric and magnetic fields (E- and H-fields, respectively)
in each chamber. The hardware for the exposure system consisted of the control computer and a rack containing
communications interfaces and instrumentation for signal generation, data acquisition, signal monitoring, signal
amplifiers, and the chamber hardware (which included the stirrer motors and environmental and RFR sensors). The
instrumentation rack contained the equipment that generated the cell phone RFR signal, acquired cell phone RFR
field strengths and environmental data, and provided an interface between the components and the control computer.
The mouse system hardware included an Ethernet to general purpose interface bus, a cell phone RFR signal
generator, three data acquisition units, four RF field measurement units, a power supply unit, and an Ethernet hub.
The amplifier array housed signal amplifiers, an amplifier cooling system, and two real-time digital control units
that directly controlled the six amplifiers in the mouse system. Each amplifier produced 400 W peak power and in
excess of 200 W average power. A closed-circuit cooling system ran cool water through the amplifiers to keep them
from overheating. The real-time digital control units controlled which chamber the amplifier output was routed to
and the level of amplifier output power while it was routed to that particular chamber.
CELL PHONE RFR SIGNAL GENERATION
GSM-modulated and CDMA-modulated cell phone RFR signals were generated experimentally via a SMIQ02B
vector signal generator with options SMIQB11 and SMIQB20 and software options 100421 – 100423 (Rohde and
Schwarz, Munich, Germany). Signals were amplified using six LSE™ amplifiers (LSE, Spanga, Sweden) in the
exposure system. The outputs of each individual amplifier were set by real-time controllers on a slot-by-slot basis
for GSM or CDMA modulation to control the E-field strength in each chamber. Each chamber contained at least
one standard gain antenna (two half-wave dipoles) that was mounted a quarter of a wavelength in front of a reflector
plate. Antennas were directed towards one of the two stirrers to maximize scattering and obtain acceptable E-field
homogeneity within the chamber space. The computerized control system managed the exposure schedule, stirrer
rotation speeds, and exposure signal type and level.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 49
Peer Review Draft NOT FOR ATTRIBUTION
The RFR power introduced into a given chamber was adjusted to achieve target field strengths; to maintain constant
exposure levels (W/kg) in a given chamber, the field strengths [measured in volts (V) per meter] were regularly
adjusted to reflect changes in the average mass of the exposed animals. The relationship between animal mass, field
strength, and SAR was determined from numerical dosimetry and programmed into the control software, hence the
required exposure field strength was computed from the average animal weights entered for each exposure group.
The interval at which animal weights were updated was determined on how rapidly the animals were growing, at the
start of the exposure period this was once per week, and as long as up to every 4 weeks later in the study.
VERIFICATION OF CELL PHONE RFR EXPOSURE
Prior to initiation of the animal studies, the RF Fields Group in the Communications Technology Laboratory at the
NIST performed an independent, detailed evaluation of 18 of the reverberation chambers (excluding the three sham
control chambers; Figure 2) to verify the cell phone RFR exposure fields, chamber characteristics (field uniformity),
and signal quality to determine the accuracy of field values reported by the developers of the exposure system (IT’IS
Foundation). Full reports detailing the procedures for measurements and calculations are available from the NTP.
All E-field measurements agreed within the estimated uncertainty bounds, indicating that the chamber fields
measured by the NIST agreed with the measurements provided by the IT’IS Foundation probes. During validation,
it was determined that the H-field probes at higher signal levels in the mid- and high-power GSM chambers reported
higher fields than indicated by other measurements, potentially leading to a modest overestimation of chamber field
strengths. In these chambers, H-field probes were replaced with E-field probes, which provided more accurate
measurements of the RF fields. The magnitude of field variation throughout the volume of a fully loaded chamber
was consistent with earlier values reported for the prototype chamber. However, it was determined that there may
have been up to ± 2.5 dB of variation in the exposure field depending on location in the cage racks. To mitigate this
positional variation, cages were routinely rotated to various locations within and between the cage racks. The
quality of the modulated signals was found to be acceptable with regard to distortion and harmonic content.
Overall, the NIST confirmed that the cell phone RFR reverberation chamber exposure system was operating
correctly and cell phone RFR exposures were within specifications.
50 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
CELL PHONE RFR EXPOSURE MONITORING
During all exposure periods, experimentally generated cell phone RFR was continuously monitored by the control
system via two RF sensors (E-field and/or H-field probes) in each exposure chamber that measured real-time signal
strengths. The use of two probes provided two independent measurements of RF field strengths and ensured that
appropriate quantitation of experimentally generated RF fields continued even in the unlikely event that one probe
failed. The E-field sensor measured electric field strength (V/m). The H-field sensor measured magnetic field
strength [measured in amperes (A) per meter]. All chambers were instrumented with one E-field sensor (ER3DV6)
and one H-field sensor (H3DV6) [both from Schmid and Partner Engineering AG (SPEAG), Zurich, Switzerland],
except for the medium and high power GSM chambers. These chambers were instrumented with two E-field probes
because H-field probes saturated at high field strengths. This change in hardware did not result in the loss of
monitoring capability. The measured E- and H-fields were communicated to the control computer in order to
maintain exposure to selected levels of RFR. During daily shutdown periods when RFR exposures were not active,
RF sensors monitored ambient RF fields in the exposure chambers. RF sensors were calibrated twice by the
manufacturer (SPEAG); once prior to initiation of any of the animal studies and once prior to initiation of the 2-year
studies. All E-field probes were calibrated in air from 100 MHz to 3.0 GHz, and had an absolute accuracy of
± 6.0% (k=2) with a spherical isotropy of better than ± 0.4 dB. All H-field probes were calibrated in air from
200 MHz to 3.0 GHz and had an absolute accuracy of ± 6.0% (k=2) with a spherical isotropy of better than ± 0.2 dB.
Data collected by the RF sensors were transmitted to the exposure and monitoring system on a real-time basis and
were recorded throughout the study. Chamber field strengths are reported as V/m and animal exposure levels (SAR
values) are reported as W/kg. The chamber field strength is the average effective E-field strength from both probes.
E-field and H-field strengths are related by the impedance of free space which is ~377 Ohms. Where an H-field
probe was used, the value in A/m was multiplied by 377 to calculate the equivalent E-field strength in V/m; it is this
effective E-field value that was used to report the chamber field strength. Field strength data reported for each day
of exposure included mean ± standard deviation, minimum field strength, maximum field strength, total number of
readings in range/total number of readings for the period, and percentage of readings in range. After each exposure
day, cell phone RFR exposure data were downloaded onto DVDs for long-term archival. Summaries of the 2-year
cell phone RFR exposure data from the studies are presented in Appendix I. The SAR and chamber-fields in the
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 51
Peer Review Draft NOT FOR ATTRIBUTION
exposure chambers were within the target ranges (defined as ± 2 dB) for >99.85% of recorded measurements over
the course of the 2-year study; ≥99.70% of recorded E-field and H-field measurements were within the target ranges
for all but one chamber (97.35% within range). All recorded SAR and field measurements were within the target
ranges for the sham control chamber. In the 28-day studies, the performance of the sham control and exposure
chambers was similar for SAR and field measurements as in the 2-year studies (data not shown).
As previously stated, the performance of the cell phone RFR exposure and monitoring system was independently
validated by engineers from the NIST prior to the initiation of the animal studies.
MONITORING AND MAINTENANCE OF ENVIRONMENTAL CONDITIONS
Environmental conditions including temperature, humidity, and airflow in all exposure chambers, as well as in other
areas of the IITRI cell phone RFR exposure facility, were maintained by a computer-controlled environmental
management system (Siemens Industries, Inc.). Monitoring instrumentation for each chamber was located in the air
exhaust duct. Each chamber was fitted by the IT’IS Foundation with a sensor box that contained sensors for
temperature and humidity (Type EE06; E + E Elektronik GmbH, Engerwitzdorf, Austria), oxygen level (Pewatron
Type FCX-MC25; Zurich, Switzerland), air speed (model EE65A; E + E Elektronik GmbH), light
(light-dependent resistor), noise (design based on WL-93 microphone; Shure Brothers, Inc., Evanston, IL), and RFR.
Outputs from the sensor box were monitored using Agilent data acquisition units, with the exception of the RF
sensor. The RF sensor was directly wired to a warning light as a safety precaution to indicate active RFR exposures
and not intended to quantitatively measure RFR field strengths.
Exposure chambers were equipped with incandescent lights located on light bars in each corner of the chamber. All
connections were RF-filtered. Chamber lighting was controlled using an adjustable daily cycle of 12 hours on,
12 hours off. In order to minimize the heat load generated by the incandescent lights, low wattage bulbs were used
that maintained chamber lighting within a range that was sufficient to support normal in vivo operations, while
minimally affecting chamber temperature. Given the expected effect of RFR field exposures on chamber
temperature during the toxicology studies, it was considered prudent to attempt to minimize heat load generated by
local sources in each chamber.
52 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
Differences in noise levels in the exposure chambers resulting from the heating, ventilation, and air conditioning
system were equalized by the installation of sound baffles in various ducts within the system. An audible signal
generated by the high intensity GSM signal was detected and equalized in all chambers by the introduction of a
“pink noise” masking sound; this masking noise equalized sound levels in all chambers. As a result of the
combination of these efforts, noise levels in all chambers were essentially equivalent and met the NC-35 noise
specification. [The noise criterion (NC) is a widely accepted numerical index commonly used to define the
maximum allowable noise. It primarily applies to the noise produced by ventilation systems, but is applied to other
noise sources, as well. Standards organizations, such as the American National Standards Institute (ANSI),
Acoustical Society of America (ASA), and International Standards Organization, provide definitions of various NCs
for ambient noise in enclosed spaces. The ANSI/ASA standard (S12.2-2008) recommends NCs for various types of
rooms, including private residences (NC 25-40), schools (NC 25-35), offices (NC 25-40), libraries (NC 30-35), and
restaurants (NC 40-45)].
ANIMAL SOURCE
Male and female B6C3F1/N mice were obtained from the NTP colony maintained at Taconic Farms, Inc.
(Germantown, NY), for the 28-day and 2-year studies.
ANIMAL WELFARE
Animal care and use are in accordance with the Public Health Service Policy on Humane Care and Use of Animals.
All animal studies were conducted in an animal facility accredited by the Association for the Assessment and
Accreditation of Laboratory Animal Care International. Studies were approved by the IITRI Animal Care and Use
Committee and conducted in accordance with all relevant NIH and NTP animal care and use policies and applicable
federal, state, and local regulations and guidelines.
28-DAY STUDIES
The 28-day studies were conducted to evaluate the cumulative effects of repeated GSM- or CDMA-modulated cell
phone RFR exposure and to determine the appropriate cell phone RFR power levels to be used in the 2-year studies.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 53
Peer Review Draft NOT FOR ATTRIBUTION
Groups of 10 male and 10 female core study mice and groups of 20 male and 20 female special study mice were
housed in reverberation chambers and received whole-body exposures to GSM- or CDMA-modulated cell phone
RFR at power levels of 0 (sham control), 5, 10, or 15 W/kg, for 9 hours and 10 minutes per day for 5 or 7 (last week
of study) days per week for at least 28 days with continuous cycling of 10 minutes on and 10 minutes off during a
period of 18 hours and 20 minutes each day. The sham control animals were housed in a reverberation chamber
identical to those used for the exposed groups, but were not exposed to cell phone RFR; a shared group of
unexposed mice of each sex served as sham controls for both cell phone RFR modulations.
Animals were observed twice daily and were weighed once during quarantine, initially, and weekly thereafter.
Clinical signs were recorded once during quarantine and then weekly. In core study mice, subcutaneously implanted
temperature microchips and monitoring equipment (Bio Medic Data Systems, Seaford, DE) were used to monitor
individual animal body temperatures. Body temperature measurements were taken prior to initial exposure at the
beginning of the study, on days 7 and 14 during inactive shutdown periods with no exposure, and on days 2, 4, 17,
20, and 27 within 5 minutes of exposure pauses at the end of the second to the last “on” cycle at the same time each
day.
Mice were quarantined for 9 or 3 days (first and second shipment, respectively) before the beginning of the studies.
Ten mice (two males and eight females) that were not assigned during randomization were selected for parasite
evaluation and gross observation of disease. Mice were approximately 5 to 6 weeks old at the beginning of the
studies. The health of the animals was monitored during the studies according to the protocols of the NTP Sentinel
Animal Program (Appendix K). All test results were negative.
Mice were housed individually. Feed and water were available ad libitum. To avoid interference with cell phone
RFR dosimetry, feed was provided in ceramic (nonmetallic) bowls and water was delivered in an adapted automatic
watering system (Capstick et al., 2017). Cages were changed weekly and rotated within the racks weekly; racks
were changed biweekly. Further details of animal maintenance are given in Table 1. Information on feed
composition and contaminants is provided in Appendix J.
54 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
Necropsies were performed on all core study mice on day 29 or 30. Organs weighed were the right adrenal gland,
brain, heart, right kidney, liver, lung, right testis, and thymus. Tissues for microscopic examination were fixed and
preserved in 10% neutral buffered formalin (except eyes, testis with epididymis, and vaginal tunics were first fixed
in Davidson’s solution or modified Davidson’s solution), processed and trimmed, embedded in paraffin, sectioned to
a thickness of 4 to 6 µm, and stained with hematoxylin and eosin. Complete histopathologic examinations were
performed by the study laboratory pathologist on all 0 (sham control) and 15 W/kg GSM- and 15 W/kg
CDMA-modulated cell phone RFR core study mice. Table 1 lists the tissues and organs routinely examined.
After a review of the laboratory reports and selected histopathology slides by a quality assessment (QA) pathologist,
the findings and differences of opinions between the study pathologist (SP) and the QA pathologist were reviewed
by the NTP pathologist. Slides containing representative lesions of exposure-related lesions or differences of
opinions between pathologists were brought to a Pathology Peer Review (PPR). Final diagnoses for reviewed
lesions represent a consensus of the PPR or a consensus between the study laboratory pathologist, NTP pathologist,
and the QA pathologist(s). Details of these review procedures have been described, in part, by Maronpot and
Boorman (1982) and Boorman et al. (1985).
2-YEAR STUDIES
Study Design
Groups of 105 male and 105 female mice were housed in reverberation chambers and received whole-body
exposures to GSM- or CDMA-modulated cell phone RFR at power levels of 0 (sham control), 2.5, 5, or 10 W/kg,
9 hours and 10 minutes per day, 7 days per week for 106 (males) or 108 (females) weeks with continuous cycling of
10 minutes on and 10 minutes off during a period of 18 hours and 20 minutes each day. The sham control animals
were housed in reverberation chambers identical to those used for the exposed groups, but were not exposed to cell
phone RFR; shared groups of unexposed mice of each sex served as sham controls for both cell phone RFR
modulations. Fifteen mice per group were randomly selected from the core group after 10 weeks of study; ten mice
per group were randomly selected for interim evaluation at 14 weeks, and five mice per group were used for genetic
toxicity testing at 14 weeks.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 55
Peer Review Draft NOT FOR ATTRIBUTION
Mice were quarantined for 9 days before the beginning of the studies. An additional five male and five female mice
not assigned during randomization were selected for parasite evaluation and gross observation of disease. Mice
were approximately 5 to 6 weeks old at the beginning of the studies. The health of the animals was monitored
during the studies according to the protocols of the NTP Sentinel Animal Program (Appendix K). All test results
were negative.
Mice were housed individually. Feed and water were available ad libitum. To avoid interference with cell phone
RFR dosimetry, feed was provided in ceramic (nonmetallic) bowls and water was delivered in an adapted automatic
watering system (Capstick et al., 2017). Cages were changed weekly and rotated within the racks biweekly; racks
were changed biweekly. Further details of animal maintenance are given in Table 1. Information on feed
composition and contaminants is provided in Appendix J.
Clinical Examinations and Pathology
Animals were observed twice daily and were weighed initially, weekly for the first 14 weeks, at 4-week intervals
during weeks 14 to 86, and then every 2 weeks from week 90 until the end of the studies. Clinical signs were
recorded once during quarantine and at least every 4 weeks during the studies.
Hematology evaluations were performed on 10 male and 10 female interim evaluation mice from each group at
14 weeks. Mice were anesthetized with 70% CO2/30% O2 and blood was collected from the retroorbital sinus and
placed into tubes containing EDTA as an anticoagulant. Hematology parameters were determined on an ADVIA™
120 automated hematology analyzer (Bayer Diagnostic Division, Tarrytown, NY). The parameters measured are
listed in Table 1. Wright Giemsa stained peripheral blood smears were prepared and evaluated for any blood cell
abnormalities. Blood was collected from the remaining five male and five female interim evaluation mice per
exposure group at 14 weeks for use in the comet and micronucleus assays; methods for these assays are presented in
Appendix E.
At 14 weeks, samples were collected for sperm motility and count and vaginal cytology evaluations on 10 male and
10 female interim evaluation mice from each group. The parameters evaluated are listed in Table 1. For 15 or
56 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
16 consecutive days prior to scheduled euthanasia, the vaginal vaults of the females were moistened with saline, if
necessary, and samples of vaginal fluid and cells were stained. Relative numbers of leukocytes, nucleated epithelial
cells, and large squamous epithelial cells were determined and used to ascertain estrous cycle stage (i.e., diestrus,
proestrus, estrus, and metestrus). Male animals were evaluated for sperm count and motility. The left testis and left
epididymis were isolated and weighed. The tail of the epididymis (cauda epididymis) was then removed from the
epididymal body (corpus epididymis) and weighed. Modified Tyrode’s buffer was applied to slides and a small
incision was made at the distal border of the cauda epididymis. The sperm effluxing from the incision were
dispersed in the buffer on the slides, and the numbers of motile and nonmotile spermatozoa were counted for five
fields per slide by two observers. Following completion of sperm motility estimates, each left cauda epididymis was
placed in buffered saline solution. Caudae were finely minced, and the tissue was incubated in the saline solution
and then heat fixed at 65° C. Sperm density was then determined microscopically with the aid of a hemacytometer.
To quantify spermatogenesis, the testicular spermatid head count was determined by removing the tunica albuginea
and homogenizing the left testis in phosphate-buffered saline containing 10% dimethyl sulfoxide. Homogenizationresistant
spermatid nuclei were counted with a hemacytometer.
All mice were necropsied. The cerebrum, frontal cortex, hippocampus, and liver were collected from five male and
five female interim sacrifice animals per exposure group at 14 weeks for use in the comet assay; methods for this
assay are presented in Appendix E. Microscopic examinations were performed on 10 male and 10 female interim
evaluation mice in each group at 14 weeks and all core study mice, including those found dead or euthanized
moribund. At the interim evaluation, the brain, right and left epididymides, heart, right and left kidneys, liver, lung,
right and left ovaries, right and left testes, and thymus were weighed. At necropsy, all organs and tissues were
examined for grossly visible lesions, and all major tissues were fixed and preserved in 10% neutral buffered
formalin (except eyes were first fixed in Davidson’s solution, and testes, vaginal tunics, and epididymides were first
fixed in modified Davidson’s solution), processed and trimmed, embedded in paraffin, sectioned to a thickness of 4
to 6 µm, and stained with hematoxylin and eosin for microscopic examination. For all paired organs (e.g., adrenal
gland, kidney, ovary), samples from each organ were examined. Tissues examined microscopically are listed in
Table 1.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 57
Peer Review Draft NOT FOR ATTRIBUTION
Microscopic evaluations were completed by the study laboratory pathologist, and the pathology data were entered
into the Toxicology Data Management System. The report, slides, paraffin blocks, residual wet tissues, and
pathology data were sent to the NTP Archives for inventory, slide/block match, wet tissue audit, and storage. The
slides, individual animal data records, and pathology tables were evaluated by an independent quality assurance
(QA) laboratory. The individual animal records and tables were compared for accuracy, the slide and tissue counts
were verified, and the histotechnique was evaluated. For the 2-year studies, a QA pathologist evaluated slides from
all tumors and all potential target organs, which included the brain, spinal cord, heart, and kidney. In addition, the
liver, large intestine (cecum and colon), small intestine (duodenum, jejunum, and ileum), lung, testis, urinary
bladder, and Harderian gland were reviewed from all male mice for specific lesions; and the bronchial and
mesenteric lymph nodes, spleen, ovary, urinary bladder, Harderian gland, and thyroid gland were reviewed from all
female mice for specific lesions.
The QA report and the reviewed slides were submitted to the NTP pathologist, who reviewed and addressed any
inconsistencies in the diagnoses made by the laboratory and QA pathologists. The QA pathologist, who served as
the coordinator of the Pathology Working Group (PWG) presented representative histopathology slides containing
examples of lesions related to test agent administration, examples of disagreements in diagnoses between the
laboratory and QA pathologists, or lesions of general interest to the PWG for review. The PWG consisted of the
NTP pathologist and other pathologists experienced in rodent toxicologic pathology. This group examined the
tissues without any knowledge of exposure groups. When the PWG consensus differed from the opinion of the
laboratory pathologist, the diagnosis was changed. Final diagnoses for reviewed lesions represent a consensus
between the laboratory pathologist, QA pathologist(s), and the PWG. Details of these review procedures have been
described, in part, by Maronpot and Boorman (1982) and Boorman et al. (1985). For subsequent analyses of the
pathology data, the decision of whether to evaluate the diagnosed lesions for each tissue type separately or combined
was generally based on the guidelines of Brix et al. (2010).
58 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
TABLE 1
Experimental Design and Materials and Methods in the Whole-Body Exposure Studies
of GSM- and CDMA-Modulated Cell Phone RFR
28-Day Studies 2-Year Studies
Study Laboratory
IIT Research Institute (Chicago, IL) IIT Research Institute (Chicago, IL)
Strain and Species
B6C3F1/N mice B6C3F1/N mice
Animal Source
Taconic Farms, Inc. (Germantown, NY) Taconic Farms, Inc. (Germantown, NY)
Time Held Before Studies
9 and 3 days (first and second shipment, respectively) 9 days
Average Age When Studies Began
Approximately 5 to 6 weeks 5 to 6 weeks
Date of First Exposure
September 6, 2010 June 18, 2012
Duration of Exposure
9 hours and 10 minutes per day over an 18 hour and 20 minute period
as exposures cycled between modulations every 10 minutes, 5 or
7 (last week of study) days per week for at least 28 days.
9 hours and 10 minutes per day over and 18 hour and 20 minute
period as exposures cycled between modulations every 10 minutes,
7 days per week for 14 weeks (interim evaluation) or 106 (males) or
108 (females) weeks (2-year studies).
Date of Last Exposure
October 3 or 4, 2010 Males: June 26, 2014
Females: July 9, 2014
Necropsy Dates
October 4 or 5, 2010 Males: June 16 to 26, 2014
Females: June 26 to July 9, 2014
Age at Necropsy
Approximately 9 to 10 weeks Males: 110 to 112 weeks
Females: 111 to 114 weeks
Size of Study Groups
10 males and 10 females Core study: 90 males and 90 females
Interim evaluation: 10 male and 10 females
Genetic toxicity: Five male and five females
Method of Distribution
Animals were distributed randomly into groups of approximately
equal initial mean body weights.
Animals were distributed randomly into groups of approximately
equal initial mean body weights.
Animals per Cage
1 1
Method of Animal Identification
Tail tattoo Tail tattoo
Diet
Certified, irradiated NTP-2000 rodent diet wafer (Zeigler Brothers,
Inc., Gardners, PA), available ad libitum, ceramic feed bowls
changed weekly
Same as 28-day studies
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 59
Peer Review Draft NOT FOR ATTRIBUTION
TABLE 1
Experimental Design and Materials and Methods in the Whole-Body Exposure Studies
of GSM- and CDMA-Modulated Cell Phone RFR
28-Day Studies 2-Year Studies
Water
Tap water (Chicago municipal supply) via an adapted automatic
watering system (SE Lab Group, Cincinnati, OH), available
ad libitum
Same as 28-day studies
Cages
Polycarbonate, solid bottom “shoebox” cages (Allentown Caging,
Allentown, NJ), changed and rotated within the rack weekly
Same as 28-day studies
Bedding
Certified, irradiated hardwood bedding (P.J. Murphy Forest Products
Corp., Montville, NJ), changed weekly
Same as 28-day studies
Racks
Custom-designed fiberglass cage racks (Ultra, Inc., Milwaukee, WI),
changed biweekly
Same as 28-day studies
Reverberation Chambers
Fully-shielded, stainless steel room equipped with a stainless steel
door to eliminate leakage of cell phone RFR signals, cell phone RFR
excitation antennas, and two rotating stirrers; chambers were cleaned
at least once weekly.
Same as 28-day studies
Reverberation Chamber Environment
Temperature: 72° ± 3° F
Relative humidity: 50% ± 15%
Room incandescent light: 12 hours/day
Chamber air changes: at least 10/hour
Temperature: 72° ± 3° F
Relative humidity: 50% ± 15%
Room incandescent light: 12 hours/day
Chamber air changes: at least 10/hour
Exposure Concentrations
Time-averaged whole-body SARs of 0 (sham control), 5, 10, and
15 W/kg GSM- or CDMA-modulated cell phone RFR
Time-averaged whole-body SARs of 0 (sham control), 2.5, 5, and
10 W/kg GSM- or CDMA-modulated cell phone RFR
Type and Frequency of Observation
Observed twice daily; animals were weighed once during quarantine,
initially, and weekly thereafter. Clinical signs were recorded once
during quarantine and then weekly.
Body temperature measurements were taken on core study mice prior
to initial exposure at the beginning of the study, on days 7 and 14
during inactive exposures, and on days 2, 4, 17, 20, and 27 within
5 minutes of exposure pauses at the end of the second to the last “on”
cycle.
Observed twice daily; animals were weighed initially, weekly for the
first 14 weeks, at 4-week intervals during weeks 14 to 86, and then
every 2 weeks from week 90 until the end of the studies. Clinical
signs were recorded once during quarantine and at least once every
4 weeks during the studies.
Method of Euthanasia
Carbon dioxide asphyxiation Carbon dioxide asphyxiation
Necropsy
Necropsies were performed on all core study mice on day 29 or 30.
Organs weighed were the right adrenal gland, brain, heart, right
kidney, liver, lung, right testis, and thymus.
Necropsies were performed on all mice. Organs weighed in 10 mice
per exposure group at 14 weeks were the brain, heart, kidneys (right
and left), liver, lung, ovaries (right and left), testes (right and left)
with epididymides (right and left), and thymus.
60 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
TABLE 1
Experimental Design and Materials and Methods in the Whole-Body Exposure Studies
of GSM- and CDMA-Modulated Cell Phone RFR
28-Day Studies 2-Year Studies
Clinical Pathology
None Blood was collected from the retroorbital sinus of 10 mice per group
at 14 weeks for hematology.
Hematology: hematocrit (auto and manual); hemoglobin
concentration; erythrocyte, reticulocyte, and platelet counts;
erythrocyte, leukocyte, and platelet morphology; mean cell volume;
mean cell hemoglobin; mean cell hemoglobin concentration; and
leukocyte count and differentials
Histopathology
Complete histopathology was performed on all 0 (sham control) and
15 W/kg groups. In addition to gross lesions and tissue masses, the
following tissues were examined: adrenal gland, aorta, bone with
marrow, brain, clitoral gland, esophagus, eyes, gallbladder,
Harderian gland, heart, large intestine (cecum, colon, rectum), small
intestine (duodenum, jejunum, ileum), kidney, larynx, liver, lung,
lymph nodes (mandibular and mesenteric), mammary gland, muscle,
nerve (sciatic), nose, oral cavity, ovary, pancreas, pharynx, pituitary
gland, preputial gland, prostate gland, salivary gland, seminal
vesicle, skin, spinal cord, spleen, stomach (forestomach and
glandular), testis with epididymis, thymus, thyroid gland, tongue,
trachea, urinary bladder, uterus, vagina, and Zymbal’s gland.
Complete histopathology was performed on 10 mice from each group
at 14 weeks, on all mice that died early, and on all mice surviving to
the end of the studies. In addition to gross lesions and tissue masses,
the following tissues were examined: adrenal gland, aorta, bone with
marrow, brain, clitoral gland, esophagus, eyes, gallbladder,
Harderian gland, heart, large intestine (cecum, colon, rectum), small
intestine (duodenum, jejunum, ileum), kidney, liver, lung with
bronchi, lymph nodes (mandibular and mesenteric), mammary gland,
muscle, nerve (sciatic, trigeminal, and ganglion), nose, ovary,
pancreas, pituitary gland, preputial gland, prostate gland, salivary
gland, seminal vesicle, skin, spinal cord, spleen, stomach
(forestomach and glandular), testis with epididymis, thymus, thyroid
gland, trachea, urinary bladder, and uterus.
Sperm Motility and Count and Vaginal Cytology
None Spermatid and sperm samples were collected from 10 male mice in
each group at 14 weeks. The following parameters were evaluated:
spermatid heads per testis and per gram testis, sperm motility, and
sperm per cauda epididymis and per gram cauda epididymis. The
left cauda, left epididymis, and left testis were weighed. Vaginal
samples were collected from 10 females in each group for 15 or
16 days prior to the 14-week interim evaluation.
STATISTICAL METHODS
For all analyses, P values less than 0.05 were considered statistically significant.
Survival Analyses
The probability of survival was estimated by the product-limit procedure of Kaplan and Meier (1958) and is
presented in the form of graphs. Animals found dead of other than natural causes or missing were censored; animals
dying from natural causes were not censored. Statistical analyses for possible dose-related effects on survival used
Cox’s (1972) method for testing two groups for equality and Tarone’s (1975) life table test to identify dose-related
trends. All reported P values for the survival analyses are two sided.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 61
Peer Review Draft NOT FOR ATTRIBUTION
Calculation of Incidence
The incidences of neoplasms or nonneoplastic lesions are presented in Tables A1, A4, B1, B4, C1, C4, D1, and D4
as the numbers of animals bearing such lesions at a specific anatomic site and the numbers of animals with that site
examined microscopically. For calculation of statistical significance, the incidences of most neoplasms (Tables A2,
B2, C2, and D2) and all nonneoplastic lesions are given as the numbers of animals affected at each site examined
microscopically. However, when macroscopic examination was required to detect neoplasms in certain tissues (e.g.,
mesentery, pleura, peripheral nerve, skeletal muscle, tongue, tooth, and Zymbal’s gland) before microscopic
evaluation, the denominators consist of the number of animals that had a gross abnormality. When neoplasms had
multiple potential sites of occurrence (e.g., leukemia or lymphoma), the denominators consist of the number of
animals on which a necropsy was performed. Tables A2, B2, C2, and D2 also give the survival-adjusted neoplasm
rate for each group and each site-specific neoplasm. This survival-adjusted rate (based on the Poly-3 method
described below) accounts for differential mortality by assigning a reduced risk of neoplasm, proportional to the
third power of the fraction of time on study, only to site-specific, lesion-free animals that do not reach terminal
euthanasia.
Analysis of Neoplasm and Nonneoplastic Lesion Incidences
The Poly-k test (Bailer and Portier, 1988; Portier and Bailer, 1989; Piegorsch and Bailer, 1997) was used to assess
neoplasm and nonneoplastic lesion prevalence. This test is a survival-adjusted quantal-response procedure that
modifies the Cochran-Armitage linear trend test to take survival differences into account. More specifically, this
method modifies the denominator in the quantal estimate of lesion incidence to approximate more closely the total
number of animal years at risk. For analysis of a given site, each animal is assigned a risk weight. This value is one
if the animal had a lesion at that site or if it survived until terminal euthanasia; if the animal died prior to terminal
euthanasia and did not have a lesion at that site, its risk weight is the fraction of the entire study time that it survived,
raised to the kth power.
This method yields a lesion prevalence rate that depends only upon the choice of a shape parameter for a Weibull
hazard function describing cumulative lesion incidence over time (Bailer and Portier, 1988). Unless otherwise
specified, a value of k=3 was used in the analysis of site-specific lesions. This value was recommended by Bailer
62 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
and Portier (1988) following an evaluation of neoplasm onset time distributions for a variety of site-specific
neoplasms in control F344 rats and B6C3F1 mice (Portier et al., 1986). Bailer and Portier (1988) showed that the
Poly-3 test gave valid results if the true value of k was anywhere in the range from 1 to 5. A further advantage of
the Poly-3 method is that it does not require lesion lethality assumptions. Variation introduced by the use of risk
weights, which reflect differential mortality, was accommodated by adjusting the variance of the Poly-3 statistic as
recommended by Bieler and Williams (1993).
Tests of significance included pairwise comparisons of each dosed group with controls and a test for an overall
dose-related trend. Continuity-corrected Poly-3 tests were used in the analysis of lesion incidence, and reported
P values are one sided. The significance of lower incidences or decreasing trends in lesions is represented as 1–P
with the letter N added (e.g., P=0.99 is presented as P=0.01N). For neoplasms and nonneoplastic lesions detected at
the interim evaluation, the Fisher exact test (Gart et al., 1979), a procedure based on the overall proportion of
affected animals, was used.
Analysis of Continuous Variables
Two approaches were employed to assess the significance of pairwise comparisons between dosed and control
groups in the analysis of continuous variables. Organ and body weight data, which historically have approximately
normal distributions and body temperatures, were analyzed with the parametric multiple comparison procedures of
Dunnett (1955) and Williams (1971, 1972). Hematology, spermatid, and epididymal spermatozoal data, which have
typically skewed distributions, were analyzed using the nonparametric multiple comparison methods of Shirley
(1977) (as modified by Williams, 1986) and Dunn (1964). Jonckheere’s test (Jonckheere, 1954) was used to assess
the significance of the dose-related trends and to determine whether a trend-sensitive test (Williams’ or Shirley’s
test) was more appropriate for pairwise comparisons than a test that does not assume a monotonic dose-related trend
(Dunnett’s or Dunn’s test). Prior to statistical analysis, extreme values identified by the outlier test of Dixon and
Massey (1957) were examined by NTP personnel, and implausible values were eliminated from the analysis. Tests
for extended periods of estrus, diestrus, metestrus, and proestrus, as well as skipped estrus and skipped diestrus,
were constructed based on a Markov chain model proposed by Girard and Sager (1987). For each dose group, a
transition probability matrix was estimated for transitions among the proestrus, estrus, metestrus, and diestrus stages,
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 63
Peer Review Draft NOT FOR ATTRIBUTION
with provision for extended stays within each stage as well as for skipping estrus or diestrus within a cycle. Equality
of transition matrices among dose groups and between the control group and each dosed group was tested using
chi-square statistics. P values for these analyses are two-sided.
Historical Control Data
The concurrent control group represents the most valid comparison to the treated groups and is the only control
group analyzed statistically in NTP bioassays. However, historical control data are often helpful in interpreting
potential treatment-related effects, particularly for uncommon or rare neoplasm types. For meaningful comparisons,
the conditions for studies in the historical control database must be generally similar. Significant factors affecting
the background incidences of neoplasms at a variety of sites are diet, sex, strain/stock, and route of exposure. The
NTP historical control database contains all 2-year studies for each species, sex, and strain/stock with histopathology
findings in control animals completed within the most recent 5-year period (Haseman, 1992, 1995; Haseman and
Rao, 1992). In general, the historical control database for a given study includes studies using the same route of
administration, and the overall incidences of neoplasms in controls for all routes of administration are included for
comparison. Because the two mouse studies presented in this report are the only two using this whole-body
exposure method, only the overall incidences for all routes are included.
QUALITY ASSURANCE METHODS
The 28-day and 2-year studies were conducted in compliance with Food and Drug Administration Good Laboratory
Practice Regulations (21 CFR, Part 58). In addition, the 28-day and 2-year study reports were audited
retrospectively by an independent QA contractor against study records submitted to the NTP Archives. Separate
audits covered completeness and accuracy of the pathology data, pathology specimens, final pathology tables, and a
draft of this NTP Technical Report. Audit procedures and findings are presented in the reports and are on file at
NIEHS. The audit findings were reviewed and assessed by NTP staff, and all comments were resolved or otherwise
addressed during the preparation of this Technical Report.
64 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
GENETIC TOXICOLOGY
The genetic toxicity of GSM- and CDMA-modulated cell phone RFR was assessed by measuring the frequency of
micronucleated erythrocytes in peripheral blood and DNA damage in five different tissues of male and female mice
following 14 weeks of exposure. Micronuclei (literally “small nuclei” or Howell-Jolly bodies) are biomarkers of
induced structural or numerical chromosomal alterations and are formed when acentric fragments or whole
chromosomes fail to incorporate into either of two daughter nuclei during cell division (Schmid, 1975; Heddle et al.,
1983). The alkaline (pH>13) comet assay (OECD, 2014) (also known as the single cell gel electrophoresis assay)
detects DNA damage in any of a variety of eukaryotic cell types (Tice et al., 2000; Collins, 2004; Brendler-Schwaab
et al., 2005; Burlinson et al., 2007); cell division is not required. The type of DNA damage detected includes nicks,
adducts, strand breaks, and abasic sites that are converted to DNA strand breaks after treatment of cells in an
alkaline (pH>13) solution. Transient DNA strand breaks generated by the process of DNA excision repair may also
be detected. DNA damage caused by crosslinking agents has been detected as a reduction of DNA migration
(Pfuhler and Wolf, 1996; Hartmann et al., 2003). The fate of the DNA damage detected by the comet assay is
varied; most of the damage is rapidly repaired resulting in no sustained impact on the tissue but some may result in
cell death or may be incorrectly processed by the repair proteins and result in a fixed mutation or chromosomal
alteration. The detailed protocols for these studies and the results are given in Appendix E.
The genetic toxicity studies have grown out of an earlier effort by the NTP to develop a comprehensive database
permitting a critical anticipation of a test article’s carcinogenicity in experimental animals based on the results
from a number of in vitro and in vivo short-term tests measuring functionally distinct genotoxicity endpoints. The
short-term tests were originally developed to clarify proposed mechanisms of chemical-induced DNA damage based
on the relationship between electrophilicity and mutagenicity (Miller and Miller, 1977) and the somatic mutation
theory of cancer (Straus, 1981; Crawford, 1985). However, it should be noted that not all cancers arise through
genotoxic mechanisms, and in these studies, the test article is not a chemical. Many studies have established the
genotoxicity of some forms of radiation including, for example, UV light radiation and X-ray radiation, which are
both forms of ionizing radiation. Because exposure to cell phone RFR requires specialized and highly technical
exposure protocols, only in vivo biomarkers associated with genotoxicity could be investigated.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 65
Peer Review Draft NOT FOR ATTRIBUTION
Clearly positive results in long-term peripheral blood micronucleus tests have high predictivity for rodent
carcinogenicity; a weak response in one sex only or negative results in both sexes in this assay do not correlate well
with either negative or positive results in rodent carcinogenicity studies (Witt et al., 2000). The relationship
between comet assay results and rodent carcinogenicity was investigated previously and a close association was
observed (Sasaki et al., 2000); however, this assay is best employed as a hazard identification assay. Because of the
theoretical and observed associations between induced genetic damage and adverse effects in somatic and germ
cells, the determination of in vivo genetic effects is important to the overall understanding of the risks associated
with exposure to a particular test article.
66 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
67
Peer Review Draft NOT FOR ATTRIBUTION
RESULTS
GSM
28-DAY STUDY
All mice survived to the end of the study (Table 2). Weekly mean body weights of exposed groups of males and
females were similar to those of the sham controls at all time points (Table 2 and Figure 4). There were no clinical
signs related to exposure to GSM cell phone RFR.
TABLE 2
Mean Body Weights and Survival of Mice Exposed to GSM-Modulated Cell Phone RFR for 28 Days
Sham Control 5 W/kg 10 W/kg 15 W/kg
Day
Av. Wt.
(g)
No. of
Survivors
Av. Wt.
(g)
Wt. (% of
Controls)
No. of
Survivors
Av. Wt.
(g)
Wt. (% of
Controls)
No. of
Survivors
Av. Wt.
(g)
Wt. (% of
Controls)
No. of
Survivors
Male
1 20.2 10 20.0 98.9 10 20.4 100.8 10 21.1 104.7 10
8 21.8 10 22.2 101.5 10 21.8 99.8 10 22.6 103.3 10
15 22.8 10 23.1 101.4 10 22.7 99.4 10 23.2 101.7 10
22 24.0 10 24.2 101.0 10 23.8 99.4 10 24.1 100.5 10
29 24.9 10 25.2 101.2 10 24.7 99.5 10 25.0 100.5 10
Female
1 18.1 10 17.8 98.3 10 17.4 96.1 10 17.9 98.9 10
8 18.9 10 19.0 100.7 10 18.4 97.3 10 18.5 98.0 10
15 20.1 10 20.1 100.0 10 19.5 97.0 10 19.6 97.3 10
22 21.0 10 21.1 100.4 10 20.4 97.1 10 20.3 96.8 10
30 21.7 10 21.9 100.9 10 21.2 97.5 10 21.0 96.6 10
68 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
FIGURE 4
Growth Curves for Mice Exposed to GSM-Modulated Cell Phone RFR for 28 Days
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 69
Peer Review Draft NOT FOR ATTRIBUTION
Body temperatures were significantly higher in RFR-exposed male mice at several time points (Table 3). In female
mice, there were a few occurrences of significantly lower body temperatures in the exposed groups, but no
significantly higher body temperatures. These changes in body temperature were inconsistent and not SAR-related.
TABLE 3
Mean Body Temperatures of Mice Exposed to GSM-Modulated Cell Phone RFR for 28 Daysa
Sham Control 5 W/kg 10 W/kg 15 W/kg
Day Temperature
(° C)
No.
Measured
Temperature
(° C)
No.
Measured
Temperature
(° C)
No.
Measured
Temperature
(° C)
No.
Measured
Male
0 37.0 ± 0.2 10 38.5 ± 0.2 10 37.3 ± 0.3 10 36.2 ± 0.2* 10
2 35.7 ± 0.1 10 37.2 ± 0.3** 10 37.1 ± 0.3** 10 37.0 ± 0.3** 10
4 36.2 ± 0.2 10 37.0 ± 0.2 10 37.1 ± 0.3* 10 37.1 ± 0.2* 10
7b 36.6 ± 0.2 9 37.4 ± 0.2 10 37.7 ± 0.4* 10 36.8 ± 0.1 10
14b 35.5 ± 0.3 10 36.0 ± 0.1 10 36.1 ± 0.4 10 35.9 ± 0.1 10
17 36.0 ± 0.3 10 37.2 ± 0.3* 10 36.7 ± 0.3 10 36.8 ± 0.4 10
20 36.5 ± 0.3 10 37.0 ± 0.3 10 37.6 ± 0.3* 10 37.2 ± 0.2 10
27 35.8 ± 0.4 9 37.6 ± 0.3** 10 37.4 ± 0.2** 10 37.2 ± 0.3** 10
2-27c 36.0 ± 0.2 10 37.1 ± 0.1** 10 37.1 ± 0.2** 10 36.9 ± 0.2** 10
Female
0 38.1 ± 0.1 10 37.9 ± 0.1 9 37.2 ± 0.3** 9 37.2 ± 0.1** 10
2 37.5 ± 0.2 10 37.4 ± 0.2 9 37.3 ± 0.3 9 37.5 ± 0.1 10
4 37.0 ± 0.2 10 37.5 ± 0.2 10 37.1 ± 0.5 10 37.6 ± 0.1 10
7b 38.6 ± 0.1 10 38.1 ± 0.2 10 37.8 ± 0.5 10 38.5 ± 0.1 10
14b 36.9 ± 0.1 10 36.4 ± 0.1 10 36.6 ± 0.2 9 37.0 ± 0.2 10
17 37.9 ± 0.1 10 37.3 ± 0.2** 10 37.7 ± 0.1 9 37.6 ± 0.1 10
20 37.7 ± 0.2 10 37.6 ± 0.2 10 37.6 ± 0.1 9 37.8 ± 0.1 10
27 37.8 ± 0.1 10 38.2 ± 0.1 10 37.2 ± 0.2* 9 37.5 ± 0.2 10
2-27 37.6 ± 0.1 10 37.5 ± 0.1 10 37.3 ± 0.2 10 37.6 ± 0.1 10
* Significantly different (P≤0.05) from the sham control group by Williams’ or Dunnett’s test
** P≤0.01 a Temperatures are given as mean ± standard error. b All temperatures were recorded within 5 minutes of the exposure cessation, except for the measurements on days 7 and 14, which were
recorded at least 1 hour after exposure. c Average of days 2 to 27, excluding days 7 and 14
70 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
There were no exposure-related effects on the organ weights of males exposed to GSM cell phone RFR (Table G1).
The absolute heart weight of 15 W/kg females was significantly less than that of the sham controls, and there were
negative trends in the absolute weights of the brain, right kidney, and liver, all of which were considered to be due to
minor reductions in body weight. There were no significantly lower relative organ weights and no associated
histopathologic findings, therefore, these organ weight changes were considered sporadic and not related to GSM
cell phone RFR exposure.
There were no histopathologic lesions related to the effects of exposure to GSM cell phone RFR.
Exposure Level Selection Rationale: In male and female mice exposed for 5 days to cell phone RFR up to 12 W/kg,
only sporadic increases were observed in body temperature, regardless of the sex or age of the animals (Wyde et al.,
2018). Because no significant effects of cell phone RFR were observed in body temperature at 12 W/kg, a higher
upper exposure level was selected for the 28-day studies. Due to limits on the maximum capacity of the exposure
system to generate high RF fields, the maximum achievable exposure level capacity was 15 W/kg, which was
selected as the highest exposure level for the 28-day studies. Selection of the highest exposure level for the 2-year
studies was also limited by the power capacity of the exposure system to generate maximum RF fields. Based on
the technical limitations and increased body temperature at various time points that were similarly observed at 10
and 15 W/kg in the 28-day studies, the exposure levels selected for the 2-year studies were 2.5, 5, and 10 W/kg.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 71
Peer Review Draft NOT FOR ATTRIBUTION
2-YEAR STUDY
Survival
Estimates of 2-year survival probabilities for male and female mice are shown in Table 4 and in the Kaplan-Meier
survival curves (Figure 5). Survival was significantly higher for the 5 W/kg males than the sham control group.
Survival of the rest of the exposed groups of males and females was generally similar to that of the sham controls.
TABLE 4
Survival of Mice Exposed to GSM-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
Male
Animals initially in study 105 105 105 105
14-week interim evaluationa 15 15 15 15
Accidental deathb 0 1 0 0
Missingb 0 1 0 0
Moribund 8 6 2 6
Natural deaths 16 19 8 12
Animals surviving to study termination 66 63 80f 72g
Percent probability of survival at end of studyc 73 72 89 80
Mean survival (days)d 687 693 717 707
Survival analysise P=0.135N P=0.959 P=0.013N P=0.360N
Female
Animals initially in study 105 105 105 105
14-week interim evaluationa 15 15 15 15
Moribund 9 9 9 6
Natural deaths 14 7 11 11
Animals surviving to study termination 67f 74h 70i 73j
Percent probability of survival at end of study 74 80 77 80
Mean survival (days) 704 715 711 712
Survival analysis P=0.476N P=0.420 P=0.709N P=0.405N
a Excluded from survival analysis b Censored in the survival analysis c Kaplan-Meier determinations d Mean of all deaths (uncensored, censored, and terminal euthanasia) e The result of the life table trend test (Tarone, 1975) is in the sham control column, and the results of the life table pairwise comparisons (Cox,
1972) with the sham controls are in the exposed group columns. A negative trend or lower mortality in an exposure group is indicated by N. f Includes one animal that died during the last week of the study g Includes four animals that died during the last week of the study h Includes four animals that died during the last week of the study; two of these were censored in the survival analysis i Includes two animals that died during the last week of the study; one of these was censored in the survival analysis
j Includes one animal that died during the last week of the study; this animal was censored in the survival analysis
72 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
FIGURE 5
Kaplan-Meier Survival Curves for Mice Exposed to GSM-Modulated Cell Phone RFR for 2 Years
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 73
Peer Review Draft NOT FOR ATTRIBUTION
Body Weights and Clinical Observations
Mean body weights of exposed groups of males and females were similar to those of the sham controls throughout
the study (Tables 5 and 6; Figure 6). Clinical signs included more occurrences of thin and ruffled fur in 10 W/kg
males and thin, ruffled fur, and mass-torso/ventral in 5 and 10 W/kg females. These findings were not correlated
with differences in body weights or incidences of neoplasms in exposed animals.
74 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
TABLE 5
Mean Body Weights and Survival of Male Mice Exposed to GSM-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
Day
Av. Wt.
(g)
No. of
Survivors
Av. Wt.
(g)
Wt. (% of
Controls)
No. of
Survivors
Av. Wt.
(g)
Wt. (% of
Controls)
No. of
Survivors
Av. Wt.
(g)
Wt. (% of
Controls)
No. of
Survivors
0 20.6 105 20.4 99.1 105 20.4 99.4 105 20.4 99.0 105
8 21.9 104 22.0 100.4 104 21.8 99.5 105 21.9 100.2 105
15 22.9 104 23.1 100.7 104 22.8 99.4 105 23.1 100.6 105
22 24.1 104 24.4 101.4 104 24.0 99.7 105 23.9 99.1 105
29 25.1 104 25.4 101.2 104 25.0 99.6 105 24.8 98.5 105
36 26.3 104 26.2 99.9 104 26.1 99.5 105 25.7 98.0 105
43 27.3 104 27.1 99.3 104 27.2 99.5 105 26.4 96.7 105
50 28.1 104 27.8 99.1 104 28.1 100.2 105 27.5 98.1 105
57 29.3 104 28.9 98.5 104 29.3 99.9 105 28.7 97.8 105
64 30.5 104 29.6 97.2 104 30.3 99.4 105 29.7 97.2 105
71 31.7 104 30.8 97.3 104 31.7 100.1 105 30.9 97.5 105
79 32.9 104 31.8 96.6 104 32.9 100.1 105 32.1 97.5 105
86 33.6 104 33.0 98.3 104 33.8 100.5 105 32.7 97.2 105
93 34.3 94 34.0 99.0 94 34.7 101.2 95 33.6 98.0 95
121 38.6 89 38.5 99.5 89 39.4 101.9 90 38.5 99.7 90
149 42.0 89 42.2 100.6 89 42.1 100.2 90 42.8 101.9 90
177 44.3 89 44.4 100.3 89 44.7 100.9 90 44.8 101.0 90
205 46.0 89 46.3 100.7 89 46.4 100.8 90 46.1 100.3 90
233 47.3 89 47.1 99.7 89 47.3 100.1 90 46.7 98.8 90
261 47.6 89 47.9 100.8 89 47.9 100.6 90 47.1 99.0 90
289 48.2 88 48.6 100.7 89 48.5 100.6 90 47.7 98.9 90
317 48.8 88 49.1 100.7 89 49.1 100.6 90 48.2 98.8 90
345 49.4 88 49.7 100.7 89 49.7 100.6 90 48.6 98.4 90
373 50.0 87 50.3 100.6 88 50.2 100.4 90 49.2 98.3 89
401 50.4 86 51.0 101.0 88 51.0 101.0 90 49.7 98.4 89
429 50.7 85 51.2 100.9 88 51.5 101.4 90 50.1 98.8 89
457 51.1 84 51.7 101.1 88 51.9 101.5 90 50.3 98.4 89
485 51.5 84 52.1 101.3 88 52.3 101.5 90 51.1 99.3 88
513 50.5 83 51.4 101.8 88 52.0 103.0 89 50.6 100.2 87
541 49.7 83 50.7 102.0 86 51.2 103.0 89 50.6 101.7 85
569 50.4 82 51.1 101.5 85 51.8 102.8 88 51.2 101.8 85
597 50.9 81 51.8 101.7 83 52.1 102.3 88 51.5 101.3 85
625 50.7 78 51.6 101.7 79 52.1 102.7 86 51.5 101.5 82
639 49.8 78 51.5 103.6 78 51.9 104.3 85 51.1 102.7 82
653 49.0 78 51.1 104.3 75 51.9 105.8 84 50.8 103.6 82
667 49.0 76 51.0 104.0 71 52.0 106.1 82 50.5 103.0 81
681 49.2 74 51.1 103.9 69 51.7 105.0 82 49.8 101.2 77
695 48.7 71 50.6 104.0 68 51.1 104.9 81 50.0 102.7 74
709 48.5 69 50.3 103.6 67 50.6 104.3 80 49.1 101.3 74
723 48.4 67 50.6 104.6 65 50.0 103.4 80 48.9 101.1 73
Mean for Weeks
1-13 27.3 27.0 99.2 27.2 99.8 26.8 98.3
14-52 44.7 44.8 100.3 45.0 100.8 44.4 99.5
53-105 49.9 51.1 102.4 51.5 103.1 50.4 100.9
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 75
Peer Review Draft NOT FOR ATTRIBUTION
TABLE 6
Mean Body Weights and Survival of Female Mice Exposed to GSM-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
Day
Av. Wt.
(g)
No. of
Survivors
Av. Wt.
(g)
Wt. (% of
Controls)
No. of
Survivors
Av. Wt.
(g)
Wt. (% of
Controls)
No. of
Survivors
Av. Wt.
(g)
Wt. (% of
Controls)
No. of
Survivors
0 17.4 105 17.2 99.1 105 17.5a 100.3 104 17.3 99.6 105
8 18.4 105 18.3 99.5 105 18.5 100.9 105 18.4 100.3 105
15 19.4 105 19.4 99.6 105 19.4 99.6 105 19.3 99.3 105
22 20.2 105 20.3 100.4 105 20.2 99.8 105 20.2 99.8 105
29 20.8 105 20.9 100.5 105 20.9 100.4 105 20.8 99.8 105
36 21.5 105 21.5 99.9 105 21.7 100.9 105 21.5 99.9 105
43 22.0 105 21.9 99.4 105 21.9 99.8 105 21.8 99.4 105
50 22.5 105 22.3 98.8 105 22.5 99.9 105 22.5 100.1 105
57 22.8 105 22.6 99.0 105 22.9 100.6 105 22.7 99.8 105
64 23.3 105 23.3 99.7 105 23.7 101.4 105 23.5 100.7 105
71 23.4 105 23.6 101.0 105 24.1 102.9 105 24.1 102.9 105
79 23.9 105 24.2 101.0 105 24.4 102.1 105 24.6 102.9 105
86 24.0 105 24.3 101.4 105 24.5 101.9 104 24.7 102.8 105
93 24.3 95 24.4 100.3 95 25.2 103.6 94 25.0 103.0 95
121 26.3 90 26.8 101.7 90 28.2 107.2 89 27.3 103.6 89
149 28.8 90 29.3 101.6 90 30.6 106.2 89 30.4 105.5 89
177 30.8 90 31.7 103.1 90 33.6 109.3 89 33.4 108.5 89
205 33.4 90 34.9 104.6 90 36.7 109.9 89 36.1 108.2 89
233 36.8 90 37.3 101.5 90 39.7 108.0 89 39.0 106.0 89
261 38.4 90 38.9 101.3 90 41.5 107.9 89 41.6 108.2 89
289 40.3 90 40.3 100.1 90 43.2 107.1 89 42.7 106.0 89
317 42.3 90 42.8 101.4 90 45.4 107.6 89 45.3 107.2 89
345 45.0 90 45.3 100.7 90 47.7 106.0 88 47.2 104.8 89
373 47.6 90 47.7 100.1 90 49.6 104.2 88 49.1 103.0 89
401 49.9 90 49.3 98.7 90 51.9 103.9 88 51.2 102.5 88
429 51.4 90 51.3 99.9 89 53.4 103.9 88 52.4 102.1 88
457 53.3 89 52.6 98.7 88 54.5 102.3 88 53.7 100.8 87
485 55.0 89 53.9 98.1 87 55.6 101.1 88 55.4 100.7 87
513 54.5 87 53.3 97.8 87 54.1 99.3 88 54.1 99.2 87
541 51.9 87 51.4 99.1 86 52.6 101.4 87 52.2 100.7 86
569 52.2 83 52.0 99.7 84 53.6 102.7 87 53.0 101.5 85
597 55.3 80 54.4 98.4 84 55.0 99.5 87 54.8 99.2 84
625 56.3 76 55.0 97.8 83 55.5 98.6 85 56.0 99.6 83
639 54.8 75 54.0 98.6 82 54.6 99.6 83 54.5 99.6 83
653 54.5 71 53.4 97.9 80 54.9 100.8 80 53.8 98.7 83
667 55.1 70 53.2 96.6 79 54.8 99.6 80 53.5 97.2 81
681 54.6 70 52.7 96.5 78 54.2 99.2 76 53.4 97.8 77
695 54.0 69 52.1 96.3 77 53.1 98.2 76 52.8 97.7 76
709 53.7 68 52.1 96.9 76 52.0 96.8 74 51.8 96.5 75
723 53.0 68 51.7 97.6 74 51.8 97.7 71 52.0 98.1 74
737 52.2 67 51.2 98.1 72 51.1 97.8 69 51.6 98.9 72
Mean for Weeks
1-13 22.4 22.5 100.1 21.7 101.0 22.6 100.8
14-52 34.6 35.2 101.6 37.2 107.3 36.8 106.1
53-107 53.3 52.3 98.2 53.5 100.4 53.1 99.7
a
One animal not weighed
76 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
FIGURE 6
Growth Curves for Mice Exposed to GSM-Modulated Cell Phone RFR for 2 Years
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 77
Peer Review Draft NOT FOR ATTRIBUTION
14-Week Interim Evaluation
There were no changes to the hematology variables attributable to GSM cell phone RFR exposure (Table F1).
At the 14-week interim evaluation, mean body weights of exposed groups of males and females were similar to
those of the sham controls (Table G2). In males, the absolute right kidney weights were significantly lower (7%) in
the 5 and 10 W/kg groups compared to the sham controls, and the absolute left kidney weight was significantly
lower (12%) in the 10 W/kg group (Table G2). The absolute liver weights of 5 and 10 W/kg males were
significantly lower (10%) and the relative liver weight was significantly lower in 5 W/kg males. These organ weight
changes were considered small changes and were not accompanied by exposure-related histopathologic lesions. In
10 W/kg females, there were significantly lower relative weights in the brain and right kidney (Table G2); these
changes were not accompanied by significant changes in absolute weights and were not considered toxicologically
important. The absolute thymus weight of 10 W/kg females was 20% higher compared to the sham controls, but this
was not correlated with any histopathologic lesions in the thymus.
In males, there were no exposure-related effects on reproductive organ weights, testis spermatid concentrations,
caudal epididymal sperm concentrations, or sperm motility (Table H1). In females, there were no exposure-related
effects on estrous cycle length, number of cycling females, or relative amount of time spent in the estrous stages
(Tables H2 and H3; Figure H1).
In the liver, a significantly higher incidence of focal inflammation occurred in 5 W/kg males (sham control, 0/10;
2.5 W/kg, 2/10; 5 W/kg, 4/10; 10 W/kg, 0/10; Table A4). Focal inflammation is commonly seen in B6C3F1/N
mice, and consisted of small clusters of mixed inflammatory cells, predominantly lymphocytes with fewer
macrophages and an occasional neutrophil. There was no zonal pattern to this finding and the inflammation was
randomly scattered within the hepatic parenchyma. All of the lesions were of minimal severity that typically
consisted of one to three small areas of inflammation, and they were not considered biologically relevant.
78 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
Pathology and Statistical Analyses
This section describes the statistically significant or potentially biologically noteworthy changes in the incidences of
malignant lymphoma and neoplasms and/or nonneoplastic lesions of the skin, lung, mediastinum, and ovary in the
2-year study. Summaries of the incidences of neoplasms and nonneoplastic lesions, statistical analyses of primary
neoplasms that occurred with an incidence of at least 5% in at least one animal group, and historical incidences for
the neoplasms mentioned in this section are presented in Appendix A for male mice and Appendix B for female
mice.
Skin (Subcutaneous Tissue): The incidences of malignant fibrous histiocytoma were higher in 5 and 10 W/kg males,
although not significantly or in an exposure concentration-related manner (Tables 7, A1, and A2); however, the
incidences exceeded the overall historical control ranges for malignant fibrous histiocytoma (Tables 7 and A3a).
The combined incidences of fibrosarcoma, sarcoma, or malignant fibrous histiocytoma were also increased in the 5
and 10 W/kg males, although still not in a statistically significant or exposure concentration-dependent manner. In
males, all but one of the malignant fibrous histiocytomas occurred on the tail; the remaining neoplasm (in a 5 W/kg
animal) was located on the pinna of the ear. Malignant fibrous histiocytomas can have a variable appearance. In
general, all the malignant fibrous histiocytomas had a portion of the neoplasm that was composed of spindle-shaped
cells arranged in interlacing or irregular bundles or whorls amongst a background of varying amounts of collagen
and a sizable population of cells resembling histiocytes – large cells with abundant eosinophilic cytoplasm and small
basophilic nuclei. Multinucleated cells were present in most of the tumors, but were more abundant in the neoplasm
on the ear. Several of the neoplasms on the tail had areas of pigment in the section – possibly from the tail tattoo.
The single malignant fibrous histiocytoma that occurred in a sham control male metastasized throughout the
abdominal cavity, involving the liver, stomach, mesentery, adrenal gland, and seminal vesicle, as well as being
found in the mesenteric lymph nodes and skeletal muscle. None of the other neoplasms had distant metastases.
The single occurrences of sarcoma in a 2.5 W/kg male (sham control, 0/90; 2.5 W/kg, 1/89; 5 W/kg, 0/90; 10 W/kg,
0/90) and fibrosarcoma in a 10 W/kg male (0/90, 0/89, 0/90, 1/90) were histologically much different from the
malignant fibrous histiocytomas (Table A1). They were much larger neoplasms, with large areas of necrosis. They
were poorly circumscribed and consisted of interlacing bundles of elongated cells in a background of varying
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 79
Peer Review Draft NOT FOR ATTRIBUTION
TABLE 7
Incidences of Neoplasms of the Skin (Subcutaneous Tissue) in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
Number Examined Microscopically 90 89 90 90
Malignant Fibrous Histiocytoma, Multiplea 0 0 1 (1%) 0
Malignant Fibrous Histiocytoma (includes multiple)b
Overall ratec 1/90 (1%) 0/89 (0%) 5/90 (6%) 3/90 (3%)
Adjusted rated 1.2% 0.0% 5.8% 3.6%
Terminal ratee 0/66 (0%) 0/63 (0%) 4/80 (5%) 3/72 (4%)
First incidence (days) 674 —g 654 729 (T)
Poly-3 testf P=0.127 P=0.499N P=0.124 P=0.321
Fibrosarcoma, Sarcoma, or Malignant Fibrous Histiocytomah
Overall rate 1/90 (1%) 1/89 (1%) 5/90 (6%) 4/90 (4%)
Adjusted rate 1.2% 1.2% 5.8% 4.7%
Terminal rate 0/66 (0%) 0/63 (0%) 4/80 (5%) 3/72 (4%)
First incidence (days) 674 523 654 488
Poly-3 test P=0.093 P=0.758N P=0.124 P=0.197
(T) Terminal euthanasia a Number of animals with neoplasm b Historical control incidence for 2-year studies (all routes) (mean ± standard deviation): 2/589 (0.3% ± 0.7%), range 0%-2% c Number of animals with neoplasm per number of animals necropsied d Poly-3 estimated neoplasm incidence after adjustment for intercurrent mortality e Observed incidence at terminal euthanasia f Beneath the sham control incidence is the P value associated with the trend test. Beneath the exposed group incidence are the P values
corresponding to pairwise comparisons between the sham controls and that exposed group. The Poly-3 test accounts for differential mortality
in animals that do not reach terminal euthanasia. A lower incidence in an exposure group is indicated by N. g Not applicable; no neoplasms in animal group h Historical control incidence: 5/589 (0.8% ± 1.0%), range 0%-2%
amounts of collagen or myxomatous material. Nuclei were long and oval and typically vesicular, in comparison to
the small, often round, densely basophilic nuclei found in the malignant fibrous histiocytomas. There was no
population of histiocyte-like cells in the sarcoma or the fibrosarcoma. Neither of these neoplasms occurred on the
tail. Fibrosarcoma, sarcoma, and malignant fibrous histiocytomas are all neoplasms of mesenchymal origin.
Lung: There was a significant positive trend in the incidences of alveolar/bronchiolar adenoma or carcinoma
(combined) in males (Tables 8 and A2). The incidences of focal alveolar epithelial hyperplasia were similar in all
groups of males (6/90, 8/89, 8/90, 7/90; Table A4). Alveolar/bronchiolar adenomas were discrete, expansile
proliferations of cuboidal to columnar cells supported by a fine fibrovascular stroma arranged in solid nests or
papillary fronds projecting into alveolar spaces and causing compression of the surrounding parenchyma.
80 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
TABLE 8
Incidences of Alveolar/bronchiolar Neoplasms of the Lung in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
Number Examined Microscopically 90 89 90 90
Alveolar/bronchiolar Adenoma, Multiplea 2 0 2 1
Alveolar/bronchiolar Adenoma (includes multiple)b
Overall ratec 13/90 (14%) 13/89 (15%) 18/90 (20%) 16/90 (18%)
Adjusted rated 16.0% 16.0% 20.7% 19.0%
Terminal ratee 9/66 (14%) 10/63 (16%) 16/80 (20%) 14/72 (19%)
First incidence (days) 488 663 604 658
Poly-3 testf P=0.297 P=0.583 P=0.279 P=0.380
Alveolar/bronchiolar Carcinoma, Multiple 2 0 1 1
Alveolar/bronchiolar Carcinoma (includes multiple)g
Overall rate 13/90 (14%) 12/89 (13%) 16/90 (18%) 18/90 (20%)
Adjusted rate 16.1% 14.7% 18.5% 21.2%
Terminal rate 12/66 (18%) 8/63 (13%) 16/80 (20%) 14/72 (19%)
First incidence (days) 568 594 729 (T) 614
Poly-3 test P=0.165 P=0.488N P=0.418 P=0.259
Alveolar/bronchiolar Adenoma or Carcinomah
Overall rate 23/90 (26%) 24/89 (27%) 32/90 (36%) 34/90 (38%)
Adjusted rate 28.1% 29.2% 36.8% 39.9%
Terminal rate 18/66 (27%) 17/63 (27%) 30/80 (38%) 28/72 (39%)
First incidence (days) 488 594 604 614
Poly-3 test P=0.040 P=0.506 P=0.149 P=0.074
(T) Terminal euthanasia a Number of animals with neoplasm b Historical control incidence for 2-year studies (all routes) (mean ± standard deviation): 84/589 (14.3% ± 5.4%), range 8%-24% c Number of animals with neoplasm per number of animals with lung examined microscopically d Poly-3 estimated neoplasm incidence after adjustment for intercurrent mortality e Observed incidence at terminal euthanasia f Beneath the sham control incidence is the P value associated with the trend test. Beneath the exposed group incidence are the P values
corresponding to pairwise comparisons between the sham controls and that exposed group. The Poly-3 test accounts for differential mortality
in animals that do not reach terminal euthanasia. A lower incidence in an exposure group is indicated by N. g Historical control incidence: 66/589 (11.0% ± 4.4%), range 4%-20% h Historical control incidence: 142/589 (24.0% ± 5.3%), range 16%-34%
Alveolar/bronchiolar carcinomas were usually larger than adenomas and tended to be poorly demarcated and locally
invasive. They were composed of cuboidal to columnar epithelial cells that displayed moderate to marked
pleomorphism and lacked a normal orderly arrangement, with multiple layers and piling up of cells. The neoplastic
cells were arranged in papillary arrangements or solid sheets of cells; most carcinomas contained both growth
patterns. Occasional mitoses were present.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 81
Peer Review Draft NOT FOR ATTRIBUTION
Malignant Lymphoma: Compared to the sham controls, all exposed groups of females had higher incidences of
malignant lymphoma and the incidences in the 2.5 and 5 W/kg groups were significantly higher (Tables 9, B1, and
B2). The sham control group had a low incidence of malignant lymphoma compared to the range seen in historical
controls (Tables 9 and B3). All of the incidences in the exposed groups fell within the overall historical control
range. Malignant lymphoma involved many organs, most frequently the spleen, lymph nodes, thymus, lung, kidney,
liver, and bone marrow, and was characterized by the effacement of normal architecture by a monomorphic
population of neoplastic lymphocytes, which tended to be larger than normal lymphocytes. In spleens with
malignant lymphoma, there was a loss of individual follicles and periarteriolar lymphoid sheaths, as the enlarged
white pulp became one solid sheet of neoplastic cells sometimes leading to the gross enlargement of the organ. In
the lymph nodes and thymus, malignant lymphoma led to the loss of distinguishable cortical and medullary regions,
with the entire node appearing to contain only a single type of cell. Involved lymph nodes were typically grossly
enlarged. In the liver and kidney, aggregates of neoplastic lymphocytes disrupted the normal arrangement of the
parenchyma, and in the lungs, neoplastic lymphocytes were often found expanding the bronchial-associated
lymphoid tissue. Malignant lymphoma in the bone marrow resulted in a hypercellular marrow cavity with a
monotonous population of malignant lymphocytes rather than the typical mix of erythrocytes and leukocytes in
various stages of maturity.
TABLE 9
Incidences of Malignant Lymphoma in Female Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
Malignant Lymphomaa
Overall rateb 2/90 (2%) 13/90 (14%) 9/90 (10%) 6/90 (7%)
Adjusted ratec 2.5% 15.6% 10.7% 7.1%
Terminal rated 1/67 (1%) 12/72 (17%) 5/69 (7%) 3/72 (4%)
First incidence (days) 604 731 516 590
Poly-3 teste P=0.474 P=0.004 P=0.035 P=0.153
a Historical control incidence for 2-year studies (all routes) (mean ± standard deviation): 89/590 (16.0% ± 8.3%), range 2%-36% b Number of animals with neoplasm per number of animals necropsied c Poly-3 estimated neoplasm incidence after adjustment for intercurrent mortality d Observed incidence at terminal euthanasia e Beneath the sham control incidence is the P value associated with the trend test. Beneath the exposed group incidence are the P values
corresponding to pairwise comparisons between the sham controls and that exposed group. The Poly-3 test accounts for differential mortality
in animals that do not reach terminal euthanasia.
82 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
Other Tissues: Several tissues had significantly increased incidences of lesions in one, or even two exposed groups
of males or females. Many of them, such as lymphocytic infiltration or inflammation in various tissues, are common
findings in aged mice and the incidences and severities recorded in this study were not considered exposure related.
The incidence of other lesions lacked an exposure concentration response and were considered sporadic occurrences
or of unknown importance.
Two hibernomas of the mediastinum occurred in 5 W/kg males (sham control, 0/90; 2.5 W/kg, 0/89; 5 W/kg, 2/90;
10 W/kg, 0/90; Table A1). These are unusual neoplasms of brown adipose tissue. Hibernomas were composed of
round cells with moderate amounts of cytoplasm filled with tiny vacuoles, and small, round nuclei. Two benign
ovarian teratomas occurred in 5 W/kg females, and one in 10 W/kg females (0/75, 0/86, 2/82, 1/80; Table B1).
Neither of these neoplasms occurred in the sham controls, nor have they occurred in the overall historical control
populations [males: mediastinum, hibernoma (0/589); females: ovary, benign teratoma (0/590)]. However, benign
teratomas have been reported in the literature to occur in B6C3F1 mice (Alison et al., 1987). Both the hibernomas
and the teratomas were considered sporadic occurrences of rare neoplasms, and while unusual, were not considered
exposure related.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 83
Peer Review Draft NOT FOR ATTRIBUTION
CDMA
28-DAY STUDY
All mice survived to the end of the study (Table 10). Weekly mean body weights of exposed groups of males and
females were similar to those of the sham controls at all time points (Table 10 and Figure 7). There were no clinical
signs related to exposure to CDMA cell phone RFR.
TABLE 10
Mean Body Weights and Survival of Mice Exposed to CDMA-Modulated Cell Phone RFR for 28 Days
Sham Control 5 W/kg 10 W/kg 15 W/kg
Day
Av. Wt.
(g)
No. of
Survivors
Av. Wt.
(g)
Wt. (% of
Controls)
No. of
Survivors
Av. Wt.
(g)
Wt. (% of
Controls)
No. of
Survivors
Av. Wt.
(g)
Wt. (% of
Controls)
No. of
Survivors
Male
1 20.2 10 20.4 100.7 10 20.4 101.0 10 20.9 103.2 10
8 21.8 10 21.8 100.0 10 22.2 101.6 10 22.4 102.4 10
15 22.8 10 22.4 98.3 10 23.0 100.9 10 23.3 102.3 10
22 24.0 10 23.5 98.0 10 23.9 99.6 10 24.2 101.0 10
29 24.9 10 24.3 97.6 10 25.2 101.2 10 25.1 101.1 10
Female
1 18.1 10 18.2 100.5 10 17.9 99.2 10 17.6 97.5 10
8 18.9 10 19.0 100.8 10 18.7 99.3 10 18.7 99.0 10
15 20.1 10 20.1 99.6 10 20.0 99.4 10 19.8 98.2 10
22 21.0 10 21.0 99.9 10 20.8 99.1 10 20.5 97.4 10
30 21.7 10 21.7 99.7 10 21.6 99.4 10 21.2 97.5 10
84 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
FIGURE 7
Growth Curves for Mice Exposed to CDMA-Modulated Cell Phone RFR for 28 Days
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 85
Peer Review Draft NOT FOR ATTRIBUTION
Similar to what was seen in mice exposed to GSM cell phone RFR, body temperatures were significantly higher in
males and significantly lower in females at several time points (Table 11).
TABLE 11
Mean Body Temperatures of Mice Exposed to CDMA-Modulated Cell Phone RFR for 28 Daysa
Sham Control 5 W/kg 10 W/kg 15 W/kg
Day Temperature
(° C)
No.
Measured
Temperature
(° C)
No.
Measured
Temperature
(° C)
No.
Measured
Temperature
(° C)
No.
Measured
Male
0 37.0 ± 0.2 10 37.0 ± 0.1 10 38.0 ± 0.2** 10 37.8 ± 0.2** 10
2 35.7 ± 0.1 10 36.1 ± 0.1 10 37.0 ± 0.3** 10 36.5 ± 0.2** 10
4 36.2 ± 0.2 10 36.7 ± 0.2 10 37.0 ± 0.2** 10 37.1 ± 0.2** 10
7b 36.6 ± 0.2 9 36.4 ± 0.2 10 37.3 ± 0.3 10 37.3 ± 0.2 10
14b 35.5 ± 0.3 10 35.8 ± 0.1 10 36.1 ± 0.2 10 36.0 ± 0.1 10
17 36.0 ± 0.3 10 36.2 ± 0.3 10 36.8 ± 0.4 10 37.2 ± 0.3* 10
20 36.5 ± 0.3 10 36.4 ± 0.2 10 37.3 ± 0.3* 10 37.6 ± 0.2** 10
27 35.8 ± 0.4 9 36.5 ± 0.3 10 37.4 ± 0.3** 10 36.8 ± 0.3 10
2-27c 36.0 ± 0.2 10 36.3 ± 0.1 10 37.1 ± 0.1** 10 36.9 ± 0.1** 10
Female
0 38.1 ± 0.1 10 37.5 ± 0.1* 9 38.3 ± 0.1 10 38.0 ± 0.2 10
2 37.5 ± 0.2 10 37.0 ± 0.2 9 38.1 ± 0.2 10 37.5 ± 0.2 10
4 37.0 ± 0.2 10 37.2 ± 0.2 10 37.7 ± 0.2 10 37.5 ± 0.2 10
7b 38.6 ± 0.1 10 37.9 ± 0.2** 9 38.0 ± 0.1* 10 38.3 ± 0.1 10
14b 36.9 ± 0.1 10 36.5 ± 0.2 9 37.0 ± 0.1 10 37.0 ± 0.2 10
17 37.9 ± 0.1 10 37.1 ± 0.2** 10 37.6 ± 0.1 10 37.4 ± 0.2 10
20 37.7 ± 0.2 10 37.2 ± 0.1 10 37.5 ± 0.2 10 37.9 ± 0.1 10
27 37.8 ± 0.1 10 37.4 ± 0.3 10 37.9 ± 0.2 10 38.0 ± 0.3 10
2-27 37.6 ± 0.1 10 37.2 ± 0.1** 10 37.7 ± 0.1 10 37.7 ± 0.1 10
* Significantly different (P≤0.05) from the sham control group by Williams’ or Dunnett’s test
** P≤0.01 a Temperatures are given as mean ± standard error. b All temperatures were recorded within 5 minutes of the exposure cessation, except for the measurements on days 7 and 14, which were
recorded at least 1 hour after exposure. c Average of days 2 to 27, excluding days 7 and 14
86 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
There were no exposure-related effects on organ weights of males exposed to CDMA cell phone RFR (Table G3).
The absolute kidney weight of 15 W/kg females was significantly less (12%) than that of the sham controls
(Table G3); however, because there was no similar effect on relative kidney weight and no associated
histopathologic findings, the biological significance of this finding was unknown.
There were no histopathologic lesions related to the effects of exposure to CDMA cell phone RFR.
Exposure Level Selection Rationale: In male and female mice exposed for 5 days to cell phone RFR up to 12 W/kg,
only sporadic increases were observed in body temperature, regardless of the sex or age of the animals (Wyde et al.,
2018). Because no significant effects of cell phone RFR were observed in body temperature at 12 W/kg, a higher
upper exposure level was selected for the 28-day studies. Due to limits on the maximum capacity of the exposure
system to generate high RF fields, the maximum achievable exposure level capacity was 15 W/kg, which was
selected as the highest exposure level for the 28-day studies. Selection of the highest exposure level for the 2-year
studies was also limited by the power capacity of the exposure system to generate maximum RF fields. Based on
the technical limitations and increased body temperatures at various time points that were similarly observed at 10
and 15 W/kg in the 28-day studies, the exposure levels selected for the 2-year studies were 2.5, 5, and 10 W/kg.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 87
Peer Review Draft NOT FOR ATTRIBUTION
2-YEAR STUDY
Survival
Estimates of 2-year survival probabilities for male and female mice are shown in Table 12 and in the Kaplan-Meier
survival curves (Figure 8). Survival was significantly higher in 2.5 W/kg males compared to that in the sham
controls. Survival of males and females in all other exposed groups was generally similar to that of the sham
controls.
TABLE 12
Survival of Mice Exposed to CDMA-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
Male
Animals initially in study 105 106 105 105
14-week interim evaluationa 15 15 15 15
Accidental deathb 0 0 1 0
Moribund 8 2 5 3
Natural deaths 16 6 13 16
Animals surviving to study termination 66 83 71 71
Percent probability of survival at end of studyc 73 91 80 79
Mean survival (days)d 687 715 706 704
Survival analysise P=1.000 P=0.003N P=0.343N P=0.482N
Female
Animals initially in study 105 104 105 105
14-week interim evaluation 15 15 15 15
Moribund 9 5 4 4
Natural deaths 14 9 16 14
Animals surviving to study termination 67f 75g 70h 72h
Percent probability of survival at end of study 74 84 77 79
Mean survival (days) 704 715 715 712
Survival analysis P=0.758N P=0.168N P=0.702N P=0.517N
a Excluded from survival analysis b Censored in the survival analysis c Kaplan-Meier determinations d Mean of all deaths (uncensored, censored, and terminal euthanasia) e The result of the life table trend test (Tarone, 1975) is in the sham control column, and the results of the life table pairwise comparisons (Cox,
1972) with the sham controls are in the exposed group columns. A negative trend or lower mortality in an exposure group is indicated by N. f Includes one animal that died during the last week of the study g Includes three animals that died during the last week of the study; one of these was censored in the survival analysis h Includes one animal that died during the last week of the study; this animal was censored in the survival analysis
88 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
FIGURE 8
Kaplan-Meier Survival Curves for Mice Exposed to CDMA-Modulated Cell Phone RFR for 2 Years
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 89
Peer Review Draft NOT FOR ATTRIBUTION
Body Weights and Clinical Observations
Mean body weights of exposed groups of males and females were similar to those of the sham controls throughout
the study (Tables 13 and 14; Figure 9). In males, there were higher occurrences of the clinical signs
mass-torso/lateral and mass-torso/ventral in the 10 W/kg group. In females, more occurrences of ruffled fur were
recorded in the 5 and 10 W/kg groups and more occurrences of thin were recorded in all exposed groups. These
findings were not correlated with differences in body weights or incidences of neoplasms in exposed animals.
90 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
TABLE 13
Mean Body Weights and Survival of Male Mice Exposed to CDMA-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
Day
Av. Wt.
(g)
No. of
Survivors
Av. Wt.
(g)
Wt. (% of
Controls)
No. of
Survivors
Av. Wt.
(g)
Wt. (% of
Controls)
No. of
Survivors
Av. Wt.
(g)
Wt. (% of
Controls)
No. of
Survivors
0 20.6 105 20.4 99.1 106 20.3 99.0 105 20.4 99.3 105
8 21.9 104 21.7 99.0 106 21.9 100.0 105 21.9 100.2 105
15 22.9 104 22.9 99.9 106 23.2 101.3 105 22.7 99.2 105
22 24.1 104 24.1 100.0 106 24.4 101.2 105 24.0 99.6 105
29 25.1 104 25.1 99.9 106 25.2 100.3 105 25.1 99.8 105
36 26.3 104 26.2 99.9 106 26.2 100.0 104 26.3 100.1 105
43 27.3 104 27.3 99.8 106 27.2 99.7 104 27.4 100.3 105
50 28.1 104 28.2 100.5 106 28.0 99.8 104 28.3 100.8 105
57 29.3 104 29.4 100.2 106 28.9 98.5 104 29.5 100.5 105
64 30.5 104 30.2 99.1 106 29.8 97.8 104 30.7 100.5 105
71 31.7 104 31.3 98.9 106 31.2 98.3 104 32.2 101.7 105
79 32.9 104 32.6 99.0 106 32.3 98.1 104 33.4 101.6 105
86 33.6 104 33.4 99.2 106 33.0 98.0 104 34.6 102.8 105
93 34.3 94 33.8 98.5 96 33.6 98.0 94 35.3 102.8 95
121 38.6 89 38.3 99.0 91 38.3 99.1 89 39.5 102.4 90
149 42.0 89 41.4 98.6 91 42.2 100.6 89 43.2 103.0 90
177 44.3 89 44.0 99.4 91 44.6 100.7 89 45.2 102.0 90
205 46.0 89 45.8 99.6 91 46.5 101.0 89 46.4 101.0 90
233 47.3 89 46.8 99.1 91 47.1 99.6 89 47.0 99.4 90
261 47.6 89 47.5 99.8 90 47.8 100.5 89 47.7 100.3 90
289 48.2 88 48.3 100.2 90 48.5 100.5 89 48.1 99.8 90
317 48.8 88 48.7 99.7 90 49.0 100.4 89 48.7 99.8 90
345 49.4 88 49.0 99.1 90 49.4 100.1 89 49.3 99.7 90
373 50.0 87 49.9 99.9 90 50.1 100.1 89 49.7 99.4 90
401 50.4 86 50.4 99.8 90 50.7 100.5 89 50.5 100.2 90
429 50.7 85 50.8 100.1 90 51.1 100.8 89 50.8 100.1 89
457 51.1 84 51.5 100.7 90 51.4 100.5 89 51.1 100.1 89
485 51.5 84 51.9 100.8 90 51.6 100.3 89 51.6 100.3 87
513 50.5 83 51.7 102.4 90 51.5 102.0 89 51.2 101.3 86
541 49.7 83 51.2 103.0 89 50.8 102.3 89 50.5 101.7 85
569 50.4 82 51.7 102.7 88 51.5 102.3 87 50.8 101.0 85
597 50.9 81 52.4 103.0 87 52.1 102.4 84 51.4 101.1 84
625 50.7 78 52.5 103.4 86 52.1 102.7 84 50.9 100.3 83
639 49.8 78 52.5 105.5 85 51.3 103.1 83 50.3 101.0 80
653 49.0 78 52.1 106.4 85 50.9 103.8 82 49.9 101.8 79
667 49.0 76 52.3 106.6 84 51.3 104.6 80 49.9 101.7 76
681 49.2 74 51.9 105.6 84 51.1 104.0 78 49.8 101.3 75
695 48.7 71 51.4 105.5 84 50.5 103.8 76 49.3 101.2 73
709 48.5 69 51.0 105.2 84 49.5 102.0 75 48.8 100.5 72
723 48.4 67 51.0 105.5 83 48.6 100.5 73 48.3 99.9 71
Mean for Weeks
1-13 27.3 27.1 99.6 27.0 99.4 27.4 100.5
14-52 44.7 44.4 99.3 44.7 100.1 45.0 101.0
53-105 49.9 51.5 103.3 50.9 102.1 50.3 100.8
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 91
Peer Review Draft NOT FOR ATTRIBUTION
TABLE 14
Mean Body Weights and Survival of Female Mice Exposed to CDMA-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
Day
Av. Wt.
(g)
No. of
Survivors
Av. Wt.
(g)
Wt. (% of
Controls)
No. of
Survivors
Av. Wt.
(g)
Wt. (% of
Controls)
No. of
Survivors
Av. Wt.
(g)
Wt. (% of
Controls)
No. of
Survivors
0 17.4 105 17.4 99.7 104 17.4 100.2 105 17.5 100.4 105
8 18.4 105 18.4 100.0 104 18.5 100.5 105 18.4 100.2 105
15 19.4 105 19.6 100.7 104 19.5 100.3 105 19.3 99.0 105
22 20.2 105 20.3 100.5 104 20.2 99.7 105 20.1 99.6 105
29 20.8 105 21.0 100.8 104 20.8 99.8 105 20.7 99.6 105
36 21.5 105 21.6 100.4 103 21.5 99.9 105 21.5 99.8 105
43 22.0 105 22.0 100.0 103 21.9 99.7 105 21.8 99.4 105
50 22.5 105 22.4 99.4 103 22.3 99.0 105 22.0 97.8 105
57 22.8 105 22.7 99.8 103 22.9 100.3 105 22.6 99.1 105
64 23.3 105 23.4 100.2 103 23.4 100.4 105 23.2 99.4 105
71 23.4 105 23.7 101.2 103 23.9 102.2 105 23.8 101.6 105
79 23.9 105 24.4 101.8 103 24.7 103.1 105 24.5 102.4 105
86 24.0 105 24.1 100.5 103 24.6 102.6 105 24.5 102.2 105
93 24.3 95 24.1 99.2 93 24.9 102.4 95 24.8 102.2 95
121 26.3 90 26.4 100.2 88 26.9 102.2 90 27.0 102.5 90
149 28.8 90 29.5 102.3 88 29.9 103.9 90 30.2 104.8 90
177 30.8 90 32.2 104.6 88 32.3 105.0 90 33.0 107.1 90
205 33.4 90 35.6 106.5 88 35.1 105.2 90 35.8 107.1 90
233 36.8 90 38.2 103.7 88 38.2 103.9 89 39.1 106.3 89
261 38.4 90 40.4 105.1 88 40.8 106.1 89 42.1 109.5 89
289 40.3 90 43.6 108.1 87 43.3 107.4 89 44.1 109.4 89
317 42.3 90 45.6 108.0 87 46.0 108.9 89 46.8 110.8 89
345 45.0 90 48.0 106.7 87 48.7 108.3 89 49.0 109.0 89
373 47.6 90 50.4 105.8 87 50.7 106.4 89 51.0 107.0 89
401 49.9 90 52.2 104.5 87 52.6 105.3 89 52.7 105.6 89
429 51.4 90 53.5 104.2 87 54.3 105.8 89 53.6 104.3 89
457 53.3 89 55.2 103.7 87 55.6 104.5 89 54.8 102.9 89
485 55.0 89 56.2 102.3 87 56.8 103.3 89 56.0 102.0 88
513 54.5 87 55.9 102.6 86 56.8 104.3 87 56.7 104.1 86
541 51.9 87 54.0 104.1 86 54.7 105.4 86 54.1 104.4 85
569 52.2 83 54.2 103.8 86 55.2 105.7 85 54.5 104.5 85
597 55.3 80 56.8 102.7 85 57.5 104.0 84 56.7 102.5 84
625 56.3 76 56.9 101.1 85 57.7 102.6 84 57.1 101.5 83
639 54.8 75 55.8 101.9 83 56.2 102.5 84 55.7 101.6 82
653 54.5 71 55.3 101.4 83 55.5 101.8 82 55.7 102.1 81
667 55.1 70 55.3 100.5 82 55.1 100.1 81 55.0 99.9 81
681 54.6 70 54.6 99.9 82 54.7 100.1 80 53.9 98.6 77
695 54.0 69 54.0 99.9 80 53.6 99.1 77 53.7 99.3 73
709 53.7 68 53.2 99.1 77 53.3 99.3 76 53.2 99.0 72
723 53.0 68 52.3 98.7 75 53.4 100.7 72 52.7 99.4 72
737 52.2 67 51.3 98.3 75 53.2 101.9 69 52.2 99.9 71
Mean for Weeks
1-13 22.4 22.6 100.5 22.6 100.7 22.5 100.1
14-52 34.6 36.4 104.4 36.6 105.3 37.2 106.9
53-107 53.3 54.3 101.9 54.8 102.9 54.4 102.1
92 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
FIGURE 9
Growth Curves for Mice Exposed to CDMA-Modulated Cell Phone RFR for 2 Years
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 93
Peer Review Draft NOT FOR ATTRIBUTION
14-Week Interim Evaluation
There were no changes to the hematology variables attributable to CDMA cell phone RFR exposure (Table F2).
At the 14-week interim evaluation, mean body weights of exposed groups of males and females were similar to
those of the sham controls (Table G4). The absolute right and left kidney weights were significantly lower (7% and
8%, respectively) in 5 W/kg males, and the absolute left kidney weight was significantly lower (8%) in 10 W/kg
males (Table G4). The relative right and left kidney weights were significantly lower in 10 W/kg males. The
absolute liver weight was significantly lower (10%) in 5 W/kg males, and the relative liver weight was significantly
lower in 10 W/kg males. The changes in the liver weights were considered small and sporadic and therefore not
toxicologically relevant; there were no histopathologic lesions that would account for changes in liver weights.
Although the absolute thymus weight of 10 W/kg males was 22% higher than that of the sham controls, the relative
thymus weight was not higher in the 10 W/kg males, nor were there any histopathologic lesions in the thymus.
There were no significant changes in organ weights in females.
In males, there were no exposure-related effects on reproductive organ weights, testis spermatid concentrations,
caudal epididymal sperm concentrations, or sperm motility (Table H4). In females, there were no exposure-related
effects on estrous cyclicity (Tables H5 and H6; Figure H2). Compared to the sham controls, there were statistically
significant differences for extended estrous in the 2.5 W/kg group and extended diestrus in the 5 W/kg group;
however, these changes were considered sporadic due to the lack of an exposure-related response.
In the kidney of 10 W/kg females, there was a significantly higher incidence of interstitial lymphocytic cellular
infiltration (sham control, 0/10; 2.5 W/kg, 1/10, 5 W/kg, 1/10; 10 W/kg, 5/10; Table D4). The lesions were minimal
to mild in severity, and consisted of clusters of lymphocytes within the interstitium.
94 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
Pathology and Statistical Analyses
This section describes the statistically significant or biologically noteworthy changes in the incidences of malignant
lymphoma and neoplasms and/or nonneoplastic lesions of the liver, pituitary gland, and uterus in the 2-year study.
Summaries of the incidences of neoplasms and nonneoplastic lesions, statistical analyses of primary neoplasms that
occurred with an incidence of at least 5% in at least one animal group, and historical incidences for the neoplasms
mentioned in this section are presented in Appendix C for male mice and Appendix D for female mice.
Liver: There was a significantly higher incidence of hepatoblastoma in 5 W/kg males (Tables 15, C1, and C2). In
2.5 W/kg males, there was a significantly higher incidence of hepatocellular adenoma and a significantly lower
incidence of hepatocellular carcinoma. When these neoplasms were combined (hepatocellular adenoma,
hepatocellular carcinoma, or hepatoblastoma), there were no significant differences in the incidences between
exposed and sham control groups of males. Hepatocellular adenomas were well-circumscribed lesions that
compressed the surrounding liver parenchyma. Most were considerably larger than a hepatic lobule, and when
located at the edge of the liver would usually cause an outward protrusion of the liver surface. They were made up
of hepatocytes that lacked the normal architectural arrangement; while portal areas might be found near the edge of
a hepatocellular adenoma, they were typically lacking within the center of the neoplasm. Most adenomas lacked
cellular pleomorphism and contained few, if any, mitotic figures. Hepatocellular carcinomas were usually large
lesions, typically larger than hepatocellular adenomas, and frequently contained areas of necrosis. They were often
multinodular and compressive, and were composed of trabeculae of neoplastic hepatocytes that were arranged at
least three cells wide (in contrast to normal hepatic trabeculae, which are a single hepatocyte wide). Cells within
hepatocellular carcinomas had higher mitotic rates and more pleomorphism when compared to hepatocellular
adenomas. Hepatoblastomas were composed of small cells with scant cytoplasm and hyperchromatic, oval nuclei,
often arranged in nests and whorls. Hepatoblastomas frequently arose from within a hepatocellular adenoma or
carcinoma; when this occurred, only the hepatoblastoma was recorded.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 95
Peer Review Draft NOT FOR ATTRIBUTION
TABLE 15
Incidences of Neoplasms of the Liver in Male Mice
Exposed to CDMA-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
Number Examined Microscopically 90 89 90 90
Hepatocellular Adenomaa
Overall rateb 52/90 (58%) 66/89 (74%) 55/90 (61%) 62/90 (69%)
Adjusted ratec 62.3% 75.4% 64.9% 72.7%
Terminal rated 45/66 (68%) 64/83 (77%) 51/71 (72%) 54/71 (76%)
First incidence (days) 393 625 656 478
Poly-3 teste P=0.199 P=0.043 P=0.428 P=0.096
Hepatocellular Carcinomaf
Overall rate 28/90 (31%) 18/89 (20%) 25/90 (28%) 31/90 (34%)
Adjusted rate 34.2% 20.6% 29.0% 36.2%
Terminal rate 18/66 (27%) 16/83 (19%) 18/71 (25%) 22/71 (31%)
First incidence (days) 608 629 559 461
Poly-3 test P=0.177 P=0.033N P=0.287N P=0.459
Hepatoblastoma, Multipleg 0 0 1 0
Hepatoblastoma (includes multiple)h
Overall rate 6/90 (7%) 6/89 (7%) 16/90 (18%) 7/90 (8%)
Adjusted rate 7.5% 6.9% 18.9% 8.5%
Terminal rate 5/66 (8%) 6/83 (7%) 14/71 (20%) 7/71 (10%)
First incidence (days) 711 729 (T) 679 729 (T)
Poly-3 test P=0.328 P=0.562N P=0.026 P=0.523
Hepatocellular Adenoma, Hepatocellular Carcinoma, or Hepatoblastomai
Overall rate 68/90 (76%) 70/89 (79%) 69/90 (77%) 75/90 (83%)
Adjusted rate 80.3% 79.6% 79.8% 85.6%
Terminal rate 52/66 (79%) 67/83 (81%) 59/71 (83%) 61/71 (86%)
First incidence (days) 393 625 559 461
Poly-3 test P=0.175 P=0.532N P=0.548N P=0.230
(T) Terminal euthanasia a Historical control incidence for 2-year studies (all routes) (mean ± standard deviation): 308/589 (51.9% ± 10.3%), range 34%-70% b Number of animals with neoplasm per number of animals with liver examined microscopically c Poly-3 estimated neoplasm incidence after adjustment for intercurrent mortality d Observed incidence at terminal euthanasia e Beneath the sham control incidence is the P value associated with the trend test. Beneath the exposed group incidence are the P values
corresponding to pairwise comparisons between the sham controls and that exposed group. The Poly-3 test accounts for differential mortality
in animals that do not reach terminal euthanasia. A lower incidence in an exposure group is indicated by N. f Historical control incidence: 164/589 (27.6% ± 8.3%), range 16%-42% g Number of animals with neoplasm h Historical control incidence: 19/589 (3.0% ± 2.2%), range 0%-7% i Historical control incidence: 408/589 (68.8% ± 8.6%), range 53%-80%
Malignant Lymphoma: Compared to the sham controls, the incidences of malignant lymphoma were higher in all
exposed groups of females compared to the controls, and the increase in the 2.5 W/kg group was statistically
significant (Tables 16, D1, and D2). This was similar to the pattern seen in females exposed to GSM cell phone
RFR in that the incidences of malignant lymphoma in groups exposed to cell phone RFR (either CDMA or GSM)
96 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
TABLE 16
Incidences of Malignant Lymphoma in Female Mice
Exposed to CDMA-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
Malignant Lymphomaa
Overall rateb 2/90 (2%) 9/89 (10%) 6/90 (7%) 7/90 (8%)
Adjusted ratec 2.5% 10.7% 7.2% 8.4%
Terminal rated 1/67 (2%) 8/74 (11%) 4/69 (6%) 4/71 (6%)
First incidence (days) 604 689 716 635
Poly-3 teste P=0.220 P=0.035 P=0.152 P=0.094
a Historical control incidence for 2-year studies (all routes) (mean ± standard deviation): 89/590 (16.0% ± 8.3%), range 2%-36% b Number of animals with neoplasm per number of animals necropsied c Poly-3 estimated neoplasm incidence after adjustment for intercurrent mortality d Observed incidence at terminal euthanasia e Beneath the sham control incidence is the P value associated with the trend test. Beneath the exposed group incidence are the P values
corresponding to pairwise comparisons between the sham controls and that exposed group. The Poly-3 test accounts for differential mortality
in animals that do not reach terminal euthanasia.
were similar and increasingly higher exposures did not have increasingly higher incidences. The incidence in the
sham control group, shared by the GSM- and CDMA-modulated cell phone RFR studies, was at the low end of the
range for malignant lymphoma in historical controls (Tables 16 and D3). All of the incidences in the exposed
groups fell within the overall historical control range. Malignant lymphoma in the CDMA cell phone RFR-exposed
groups was similar in appearance, and in the organs that were involved, to that observed in the sham controls and the
GSM cell phone RFR-exposed groups.
Other Tissues: Several tissues had significantly increased incidences of lesions in one, or even two, exposed groups
of males or females. Some of these lesions are common background lesions and were not considered toxicologically
important; the incidences of others lacked a dose response and were considered sporadic occurrences and not related
to treatment.
In 5 W/kg males, two adenomas (0/86, 0/84, 2/89, 0/83) and one carcinoma (0/86, 0/84, 1/89, 0/83) occurred in the
pars distalis of the pituitary gland (Table C1); no neoplasms of the pituitary gland pars distalis occurred in the sham
control group or in the other exposed groups of males, including those in the GSM study (Table A1). Only two
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 97
Peer Review Draft NOT FOR ATTRIBUTION
adenomas of the pituitary gland (pars distalis) have been recorded in the current (August 2017) historical control
database of 576 male mice (all routes), and no carcinomas of the pars distalis have been recorded.
In the uterus of female mice, there were one or two occurrences of adenocarcinoma (sham control, 0/89; 2.5 W/kg,
2/89; 5 W/kg, 0/88; 10 W/kg, 1/90) or leiomyosarcoma (0/89, 1/89, 1/88, 2/90) in most of the exposed groups; these
neoplasms did not occur in the sham control group (Table D1). Neither uterine adenocarcinomas nor
leiomyosarcomas have been recorded in the current historical control database (0/590). These neoplasms were
considered sporadic occurrences, and not related to exposure.
GENETIC TOXICOLOGY
Twenty tissue samples obtained from animals in the 14-week interim evaluation study were evaluated for DNA
damage using the comet assay (two sexes, two cell phone RFR modulations, five tissues). Results are based on the
standard 100-cell scoring approach in use at the time these data were collected; data obtained using a 150-cell
scoring approach, recommended in a recently adopted international guideline for the in vivo comet assay, are noted
here for the few instances where results differed between the two methods. The complete 100-cell and 150-cell data
are presented in Appendix E data tables. Significant increases in DNA damage (percent tail DNA) were observed in
cells of the frontal cortex of male mice exposed to both modulations, CDMA and GSM (Tables E1 and E2).
Positive results were also obtained for male mouse frontal cortex (CDMA and GSM) (Table E3) using the 150-cell
approach. Of note is the low percent comet tail DNA value in the frontal cortex of sham control mice. There is no
appropriate historical control database to provide context for this response, but bonafide changes in DNA damage
levels in a treatment group should remain constant relative the control value. No technical aspects of the study that
may have influenced this control value independently of the treated group values (e.g., % agarose gel, duration of
electrophoresis, electromagnetic field strength, slide position in the electrophoresis tank) were identified. Technical
factors that influence control levels have not been shown to alter sensitivity to detect effects in treated groups (Recio
et al., 2012). No other tissues showed evidence of a treatment-related effect in male mice. In female mice exposed
to the CDMA modulation, significant increases in DNA damage were seen in blood leukocytes using both scoring
approaches (Tables E4 and E6). In female mouse liver samples exposed to either modulation, the mean percent
comet tail DNA was elevated above the sham control for all exposures when evaluated using either scoring
98 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
approach. Results of the 100-cell scoring approach were judged to be negative (Tables E4 and E6); scoring
150 cells resulted in a negative call for GSM-exposed female mice (Table E5) but in CDMA-exposed female mouse
liver, a significant increase (P=0.010) in percent comet tail DNA was seen in the 5 W/kg group, resulting in an
equivocal call for this dataset (Table E6).
In the micronucleus assay for male mice exposed to CDMA (Table E7), although a significant trend was observed
for micronucleated polychromatic erythrocytes (PCEs) (P=0.013), the absolute increase was quite small and fell
within the laboratory’s historical control range. In addition, no corresponding increase in micronucleated
normochromatic erythrocytes was observed; the mature erythrocyte population ought to be in steady state
equilibrium after continuous 14 weeks of exposure, such as occurred in this study. Thus, the overall result in the
micronucleus assay for male mice exposed to CDMA was judged to be negative. No other significant effects on
either micronucleus frequency or % PCEs were seen in male or female mice exposed to either modulation of cell
phone RFR.
99
Peer Review Draft NOT FOR ATTRIBUTION
DISCUSSION AND CONCLUSIONS
The Food and Drug Administration (FDA) nominated the radio frequency radiation (RFR) emissions of wireless
communication devices for toxicology and carcinogenicity testing based on several factors. Current exposure
guidelines are based on protection from acute injury from thermal effects, and little is known about the potential for
health effects of long-term exposure. Epidemiology and toxicology studies have not definitively demonstrated an
association between cell phone RFR exposure and any specific health problems in humans; however, the results of
these studies are mixed and further complicated by confounding factors (including potential recall biases of the
study participants that could impact the assessment of exposure). For epidemiology studies, exposures in the
general population may not have occurred for a long enough period of time to accommodate the long latency period
for some types of cancers in humans. Studies in laboratory animals have been complicated by limitations that
researchers have faced in conducting robust studies designed to characterize the toxicity and carcinogenicity of cell
phone RFR.
To improve on the existing methods of exposing laboratory animals to RFR, the NTP worked in collaboration with
experts from the Radio-Frequency Fields Group at the National Institute of Standards and Technology (NIST,
Boulder, CO) and IT’IS Foundation (Zurich, Switzerland) to design, construct, and validate a novel system of
delivering RFR exposure that improved on the designs of previous exposure systems. Together with NIST and the
IT’IS Foundation, the NTP identified and constructed an exposure system designed to uniformly expose
unrestrained, individually housed animals to a uniform field of cell phone RFR at frequencies and modulations that
reflect those currently in use in wireless communication devices (GSM and CDMA). The exposure facility was
installed at IIT Research Institute (Chicago, IL) where all animal studies were conducted following system testing
and RFR exposure validation.
Studies were designed to evaluate the toxicology and carcinogenicity of whole-body exposure to cell phone RFR in
individually housed, unrestrained animals. Studies for both GSM- and CDMA-modulated cell phone RFR were
100 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
conducted simultaneously with a common control group in a sham chamber. Exposures were conducted in
10 minute periods, followed by 10 minutes of rest with no cell phone RFR exposure. The exposure system ran
continuously, alternating each 10 minute block of active exposure between the GSM- and CDMA-exposed mice
over the course of approximately 18 hours a day, 7 days per week. Based on the on/off cycling scheme, the actual
daily exposure time to cell phone RFR was approximately 9 hours per day.
Studies were conducted in multiple phases. The first phase comprised a series of short-term toxicity studies
conducted in young and aged B6C3F1/N mice and Hsd:Harlan Sprague Dawley SD rats to characterize the effects of
cell phone RFR exposure on body temperature and the potential impact of animal size. The impact of cell phone
RFR exposure during pregnancy was also evaluated in rats. These studies demonstrated that rats were more
sensitive to the heating effects of cell phone RFR than were the mice (Wyde et al., 2018). In both young and aged
male and female mice, body temperatures were only sporadically increased at exposures to cell phone RFR up to
12 W/kg (GSM and CDMA). These data suggest that exposures of up to 12 W/kg did not markedly alter the
thermoregulatory capacity in mice. It must be noted, however, that core body temperature is a general surrogate for
the heating effects of RFR and that these results do not address the issue of potential changes in temperature that
may occur in localized areas within some tissues.
The findings from these short-term studies were used to guide the selection of cell phone RFR exposure levels for
the 28-day and 2-year studies. Because no significant effect of cell phone RFR exposure up to 12 W/kg was
observed in the body temperature of mice in these thermal pilot studies, a higher level of cell phone RFR exposure
(15 W/kg) was selected for the highest exposure group in the 28-day studies. The selection of 15 W/kg was
determined by the technical limitations of the exposure system to deliver higher cell phone RFR fields in the 28-day
studies. Results from the 28-day studies demonstrated some increases in core body temperature at various time
points at 10 and 15 W/kg. Based on the observed increases in body temperature and the power limitations of the
system to generate maximum RFR fields for the large numbers of mice that were required for the 2-year studies, the
highest exposure level for the 2-year studies was 10 W/kg.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 101
Peer Review Draft NOT FOR ATTRIBUTION
The effects of whole-body exposure to GSM- or CDMA-modulated cell phone RFR at 1,900 MHz for 14 weeks or
2 years were studied in B6C3F1/N mice at specific absorption rates (SARs) of 2.5, 5, and 10 W/kg, with a common
sham control group for both GSM- and CDMA-modulated signals. At SAR exposures up to 10 W/kg, there were no
exposure-related effects on survival or mean body weights in either modulation (GSM or CDMA).
In both the GSM and CDMA studies, the incidences of malignant lymphoma in all exposed female groups were
higher than that in the sham controls. These incidences were significantly increased only in the GSM groups at 2.5
and 5 W/kg, and in the CDMA group at 2.5 W/kg compared to sham controls. The 2% incidence of lymphoma in
the concurrent sham controls was the lowest incidence observed thus far in female B6C3F1/N mice. The incidence
is well below the overall historical control mean of 16%, and appreciably lower than the lower end of the range of
overall historical control values in other studies (10% to 36%). Additionally, the incidences of malignant lymphoma
in all exposed groups were within the range observed in overall historical controls. These considerations reduce the
confidence that these increases in incidences were attributable to the RFR exposure, so these were considered
equivocal findings. In NTP conclusions, such uncertain responses in the absence of other clearer effects on
carcinogenicity would be referred to as equivocal evidence of carcinogenicity (i.e., may have been related to
exposure).
In males, there were no common lesions observed between the two modulations. Potential cell phone
RFR-mediated effects observed in the lung and the skin of males were specific to the GSM modulation. In the lung,
there was a positive trend in the combined incidence of alveolar/bronchiolar adenoma or carcinoma in male mice,
but there was no significant effect in any of the individual groups compared to controls. The combined incidences at
the upper two exposure levels exceeded the historical control range (16% to 34%). Despite a significant trend in the
combined incidence of alveolar/bronchiolar adenoma or carcinoma, the observation that the incidences were only
marginally outside the historical range reduce the confidence that the increased incidences were attributable to the
RFR exposure. Therefore, these were considered equivocal findings.
The combined incidences of fibrosarcoma, sarcoma, or malignant fibrous histiocytoma in the skin were higher in the
5 and 10 W/kg GSM males but were not statistically different than that of the sham controls. Malignant fibrous
102 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
histiocytoma was the predominant neoplasm in this combination. There was also a lack of an increased exposure
level response. However, the incidences in both groups were above the historical control range for malignant
fibrous histiocytoma. Additionally, there was one occurrence of a sarcoma in the 2.5 W/kg GSM males and one
occurrence of a fibrosarcoma in the 10 W/kg GSM males. While the incidences in the 5 and 10 W/kg GSM males
were not significant versus the current sham controls, the increases were seen in the top two exposure groups and
were outside the historical range. This suggests that the increases in incidences observed may have been attributable
to the RFR exposure, so these were considered equivocal findings.
At 2 years in the CDMA study only, there was a significantly increased incidence of hepatoblastoma in males
exposed to 5 W/kg. The incidence at 5 W/kg exceeded the historical control; however, no increases were observed
in males at 10 W/kg. Additionally, when all liver neoplasms (hepatocellular adenoma, hepatocellular carcinoma, or
hepatoblastoma) were combined, there were no significant differences between any of the exposed groups compared
to the sham controls. The isolated increase in only the 5 W/kg group and overall lack of exposure response reduces
the confidence that the increase in incidence of hepatoblastoma observed was attributable to the RFR exposure,
therefore, this was considered an equivocal finding.
Subsets of male and female mice from the 2-year studies were examined at 14 weeks to evaluate biomarkers of
genotoxicity. Chromosomal damage was evaluated using the peripheral blood erythrocyte micronucleus (MN)
assay, and DNA damage was evaluated in the frontal cortex, hippocampus, cerebellum, liver, and peripheral blood
using the comet assay. Results of the MN assays were negative, but significant increases in DNA damage were
observed in cells of the frontal cortex of male mice exposed to both modulations (GSM and CDMA) and in blood
leukocytes of female mice (CDMA only).
Unlike ionizing radiation or ultraviolet light, cell phone RFR is not sufficiently energetic, by several orders of
magnitude, to directly damage macromolecules (IARC, 2013), and little is known about the mechanisms by which
RFR could induce DNA damage in the absence of thermal effects. Proposed mechanisms include, for example,
induction of oxygen radicals and interference with DNA repair mechanisms (Ruediger, 2009; Yakymenko et al.,
2016).
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 103
Peer Review Draft NOT FOR ATTRIBUTION
No histopathologic assessments of cytotoxicity (apoptosis and necrosis) were conducted in the male mouse brain
tissues that were examined for DNA damage, which leaves open the possibility that apoptosis or necrosis may have
confounded the comet assay results. However, this seems unlikely as brain sections from other groups of mice in
this interim 14-week study and in the 2-year study did undergo histopathologic assessment and no significant
evidence of cytotoxicity was observed.
Although increases in DNA damage were observed in the frontal cortex of male mice, there were no increases
observed in the incidences of any type of neoplasm in the brain of males in the 2 year study. Similarly, while
increased DNA damage was observed in blood leukocytes of female mice exposed to CDMA-modulated cell phone
RFR, there were no increased incidences of related neoplasms. Therefore, no association was established between
DNA damage appearing early in the studies and neoplasm development in these tissues.
CONCLUSIONS
Under the conditions of these 2-year studies, there was equivocal evidence of carcinogenic activity of
GSM-modulated cell phone RFR at 1,900 MHz in male B6C3F1/N mice based on the combined incidences of
fibrosarcoma, sarcoma, or malignant fibrous histiocytoma in the skin and the incidences of alveolar/bronchiolar
adenoma or carcinoma (combined) in the lung. There was equivocal evidence of carcinogenic activity of
GSM-modulated cell phone RFR at 1,900 MHz in female B6C3F1/N mice based on the incidences of malignant
lymphoma (all organs). There was equivocal evidence of carcinogenic activity of CDMA-modulated cell phone
RFR at 1,900 MHz in male B6C3F1/N mice based on the incidences of hepatoblastoma of the liver. There was
equivocal evidence of carcinogenic activity of CDMA-modulated cell phone RFR at 1,900 MHz in female
B6C3F1/N mice based on the incidences of malignant lymphoma (all organs).
Exposure to GSM- or CDMA-modulated cell phone RFR at 1,900 MHz did not increase the incidence of any
nonneoplastic lesions in male or female B6C3F1/N mice.
* Explanation of Levels of Evidence of Carcinogenic Activity is on page 14.
104 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
105
Peer Review Draft NOT FOR ATTRIBUTION
REFERENCES
Adair, R.K. (2003). Biophysical limits on athermal effects of RF and microwave radiation. Bioelectromagnetics 24,
39-48.
Alison, R.H., Morgan, K.T., Haseman, J.K., and Boorman, G.A. (1987). Morphology and classification of ovarian
neoplasms in F344 rats and (C57BL/6 × C3H)F1 mice. J. Natl. Cancer Inst. 78, 1229-1243.
Ammari, M., Lecomte, A., Sakly, M., Abdelmelek, H., and de Seze, R. (2008). Exposure to GSM 900 MHz
electromagnetic fields affects cerebral cytochrome c oxidase activity. Toxicology 250, 70-74.
Ammari, M., Gamez, C., Lecomte, A., Sakly, M., Abdelmelek, H., and de Seze, R. (2010). GFAP expression in the
rat brain following sub-chronic exposure to a 900 MHz electromagnetic field signal. Int. J. Radiat. Biol. 86,
367-375.
Asanami, S., and Shimono, K. (1997). High body temperature induces micronuclei in mouse bone marrow. Mutat.
Res. 390, 79-83.
Bailer, A.J., and Portier, C.J. (1988). Effects of treatment-induced mortality and tumor-induced mortality on tests
for carcinogenicity in small samples. Biometrics 44, 417-431.
Belyaev, I.Y., Koch, C.B., Terenius, O., Roxström-Lindquist, K., Malmgren, L.O.G., Sommer, W.H, Salford, L.G.,
and Persson, B.R.R. (2006). Exposure of rat brain to 915 MHz GSM microwaves induces changes in gene
expression but not double stranded DNA breaks or effects on chromatin conformation. Bioelectromagnetics 27,
295-306.
Bieler, G.S., and Williams, R.L. (1993). Ratio estimates, the delta method, and quantal response tests for increased
carcinogenicity. Biometrics 49, 793-801.
Boorman, G.A., Montgomery, C.A., Jr., Eustis, S.L., Wolfe, M.J., McConnell, E.E., and Hardisty, J.F. (1985).
Quality assurance in pathology for rodent carcinogenicity studies. In Handbook of Carcinogen Testing (H.A.
Milman and E.K. Weisburger, Eds.), pp. 345-357. Noyes Publications, Park Ridge, NJ.
Borbély, A.A., Huber, R., Graf, T., Fuchs, B., Gallmann, E., and Achermann, P. (1999). Pulsed high-frequency
electromagnetic field affects human sleep and sleep electroencephalogram. Neurosci. Lett. 275, 207-210.
Brendler-Schwaab, S., Hartmann, A., Pfuhler, S., and Speit, G. (2005). The in vivo comet assay: Use and status in
genotoxicity testing. Mutagenesis 20, 245-254.
Brix, A.E., Hardisty, J.F., and McConnell, E. E. (2010). Combining Neoplasms for Evaluation of Rodent
Carcinogenesis Studies. In Cancer Risk Assessment (C.-H. Hsu and T. Stedeford, Eds.), Chapter 28. John Wiley
& Sons, Inc., Hoboken, NJ.
Brusick, D., Albertini, R., McRee, D., Peterson, D., Williams, G., Hanawalt, P., and Preston. J. (1998).
Genotoxicity of radiofrequency radiation. DNA/Genetox Expert Panel. Environ. Mol. Mutagen. 32, 1-16.
Buddhikot, M.M., Kennedy, I., Mullany, F., and Viswanathan, H. (2009). Ultra-broadband femtocells via
opportunistic reuse of multi-operator and multi-service spectrum. Bell Labs Tech. J. 13, 129-143.
106 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
Burlinson, B., Tice, R.R., Speit, G., Agurell, E., Brendler-Schwaab, S.Y., Collins, A.R., Escobar, P., Honma, M.,
Kumaravel, T.S., Nakajima, M., Sasaki, Y.F., Thybaud, V., Uno, Y., Vasquez, M., and Hartmann, A. (2007).
Fourth International Work Group on Genotoxicity Testing: Results of the in vivo comet assay workgroup. Mutat.
Res. 627, 31-35.
Capstick, M., Kuster, N., Kuehn, S., Gong, Y., Wilson, P., Ladbury, J., Koepke, G., McCormick, D.L., Gauger, J.,
and Melnick, R.L. (2017). A radio frequency radiation exposure system for rodents based on reverberation
chambers. IEEE Trans. Electromagn. Compatibil. 59, 1041-1052.
Cardis, E., Richardson, L., Deltour, I., Armstrong, B., Feychting, M., Johansen, C., Kilkenny, M., McKinney, P.,
Modan, B., Sadetzki, S., Schüz, J., Swerdlow, A., Vrijheid, M., Auvinen, A., Berg, G., Blettner, M., Bowman, J.,
Brown, J., Chetrit, A., Christensen, H.C., Cook, A., Hepworth, S., Giles, G., Hours, M., Iavarone, I.,
Jarus-Hakak, A., Klaeboe, L., Krewski, D., Lagorio, S., Lönn, S., Mann, S., McBride, M., Muir, K., Nadon, L.,
Parent, M.E., Pearce, N., Salminen, T., Schoemaker, M., Schlehofer, B., Siemiatycki, J., Taki, M.,
Takebayashi, T., Tynes, T., van Tongeren, M., Vecchia, P., Wiart, J., Woodward, A., and Yamaguchi, N. (2007).
The INTERPHONE study: Design, epidemiological methods, and description of the study population. Eur.
J. Epidemiol. 22, 647-664.
Cellular Telecommunications Industry Association (CTIA) (2017). Annual wireless industry survey.
Chagnaud, J.L., and Veyret, B. (1999). In vivo exposure of rats to GSM-modulated microwaves: Flow cytometry
analysis of lymphocyte subpopulations and of mitogen stimulation. Int. J. Radiat. Biol. 75, 111-113.
Chia, S.E., Chia, H.P., and Tan, J.S. (2000). Prevalence of headache among handheld cellular telephone users in
Singapore: A community study. Environ. Health Perspect. 108, 1059-1062.
Code of Federal Regulations (CFR) 21, Part 58.
Code of Federal Regulations (CFR) 47 §1.1310.
Collins, A.R. (2004). The comet assay for DNA damage and repair. Principles, applications and limitations. Mol.
Biotechnol. 26, 246-261.
Cook, A., Woodward, A., Pearce, N., and Marshall, C. (2003). Cellular telephone use and time trends for brain,
head and neck tumours. N. Z. Med. J. 116, U457.
Cox, D.R. (1972). Regression models and life-tables. J. R. Stat. Soc. B34, 187-220.
Crawford, B.D. (1985). Perspectives on the somatic mutation model of carcinogenesis. In Advances in Modern
Environmental Toxicology. Mechanisms and Toxicity of Chemical Carcinogens and Mutagens (M.A. Mehlman,
W.G. Flamm, and R.J. Lorentzen, Eds.), pp. 13-59. Princeton Scientific Publishing Co., Inc., Princeton, NJ.
Czerninski, R., Zini, A., and Sgan-Cohen, H.D. (2011). Risk of parotid malignant tumors in Israel (1970–2006).
Epidemiology 22, 130-131.
Dasdag, S., Akdag, M.Z., Ulukaya, E., Uzunlar, A.K., and Ocak, A.R. (2009). Effect of mobile phone exposure on
apoptotic glial cells and status of oxidative stress in rat brain. Electromagn. Biol. Med. 28, 342-354.
de Gannes, F.P., Billaudel, B., Taxile, M., Haro, E., Ruffié, G., Lévêque, P., Veyret, B., and Lagroye, I. (2009).
Effects of head-only exposure of rats to GSM-900 on blood-brain barrier permeability and neuronal degeneration.
Radiat. Res. 172, 359-367.
Deltour, I., Johansen, C., Auvinen, A., Feychting, M., Klaeboe, L., and Schüz, J. (2009). Time trends in brain tumor
incidence rates in Denmark, Finland, Norway, and Sweden, 1974–2003. J. Natl. Cancer Inst. 101, 1721-1724.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 107
Peer Review Draft NOT FOR ATTRIBUTION
de Vocht, F., Burstyn, I., and Cherrie, J.W. (2011). Time trends (1998–2007) in brain cancer incidence rates in
relation to mobile phone use in England. Bioelectromagnetics 32, 334-339.
Dixon, W.J., and Massey, F.J., Jr. (1957). Introduction to Statistical Analysis, 2nd ed., pp. 276-278, 412.
McGraw-Hill Book Company, Inc., New York.
Dunn, O.J. (1964). Multiple comparisons using rank sums. Technometrics 6, 241-252.
Dunnett, C.W. (1955). A multiple comparison procedure for comparing several treatments with a control. J. Am.
Stat. Assoc. 50, 1096-1121.
Eberhardt, J.L., Persson, B.R.R., Brun, A.E., Salford, L.G., and Malmgren, L.O.G. (2008). Blood-brain barrier
permeability and nerve cell damage in rat brain 14 and 28 days after exposure to microwaves from GSM mobile
phones. Electromagn. Biol. Med. 27, 215-229.
Elekes, E., Thuróczy, G., and Szabó, L.D. (1996). Effect on the immune system of mice exposed chronically to
50 Hz amplitude-modulated 2.45 GHz microwaves. Bioelectromagnetics 17, 246-248.
Eşmekaya, M.A., Seyhan, N., and Ömeroğlu, S. (2010). Pulse modulated 900 MHz radiation induces
hypothyroidism and apoptosis in thyroid cells: A light, electron microscopy and immunohistochemical study. Int.
J. Radiat. Biol. 86, 1106-1116.
Frey, A.H. (1998). Headaches from cellular telephones: Are they real and what are the implications? Environ.
Health Perspect. 106, 101-103.
Fritze, K., Wiessner, C., Kuster, N., Sommer, C., Gass, P., Hermann, D.M., Kiessling, M., and Hossmann, K.A.
(1997). Effect of global system for mobile communication microwave exposure on the genomic response of the rat
brain. Neuroscience 81, 627-639.
Gart, J.J., Chu, K.C., and Tarone, R.E. (1979). Statistical issues in interpretation of chronic bioassay tests for
carcinogenicity. JNCI 62, 957-974.
Gatta, L., Pinto, R., Ubaldi, V., Pace, L., Galloni, P., Lovisolo, G.A., Marino, C., and Pioli, C. (2003). Effects of
in vivo exposure to GSM-modulated 900 MHz radiation on mouse peripheral lymphocytes. Radiat. Res. 160,
600-605.
Girard, D.M., and Sager, D.B. (1987). The use of Markov chains to detect subtle variation in reproductive cycling.
Biometrics 43, 225-234.
Gong, Y., Capstick, M., Kuehn, S., Wilson, P., Ladbury, J., Koepke, G., McCormick, D.L., Melnick, R.L., and
Kuster, N. (2017). Life-time dosimetric assessment for mice and rats exposed in reverberation chambers of the
2-year NTP cancer bioassay study on cell phone radiation. IEEE Trans. Electromagn. Compatibil. 59, 1798-1808.
Grafström, G., Nittby, H., Brun, A., Malmgren, L., Persson, B.R.R., Salford, L.G., and Eberhardt, J. (2008).
Histopathological examinations of rat brains after long-term exposure to GSM-900 mobile phone radiation. Brain
Res. Bull. 77, 257-263.
Guney, M., Ozguner, F., Oral, B., Karahan, N., and Mungan, T. (2007). 900 MHz radiofrequency-induced
histopathologic changes and oxidative stress in rat endometrium: Protection by vitamins E and C. Toxicol. Ind.
Health 23, 411-420.
Guy, A.W., Chou, C.-K., Kunz, L.L., Crowley, J., and Krupp, J. (1985). Effects of long-term low-level
radiofrequency radiation exposure on rats. Volume 9. Summary. Bioelectromagnetics Research Laboratory,
Department of Rehabilitation Medicine, School of Medicine, University of Washington, Seattle, WA.
108 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
Hardell, L., Hallquist, A., Hansson Mild, K., Carlberg, M., Gertzén, H., Schildt, E.B., and Dahlqvist, A. (2004). No
association between the use of cellular or cordless telephones and salivary gland tumours. Occup. Environ. Med. 61,
675-679.
Hardell, L., Eriksson, M., Carlberg, M., Sundström, C., and Mild, K.H. (2005). Use of cellular or cordless
telephones and the risk for non-Hodgkin’s lymphoma. Int. Arch. Occup. Environ. Health 78, 625-632.
Hardell, L., Carlberg, M., Ohlson, C.G., Westberg, H., Eriksson, M., and Hansson Mild, K. (2007). Use of cellular
and cordless telephones and risk of testicular cancer. Int. J. Androl. 30, 115-122.
Hardell, L., Carlberg, M., Hansson Mild, K., and Eriksson, M. (2011). Case-control study on the use of mobile and
cordless phones and the risk for malignant melanoma in the head and neck region. Pathophysiology 18, 325-333.
Hartmann, A., Agurell, E., Beevers, C., Brendler-Schwaab, S., Burlinson, B., Clay, P., Collins, A., Smith, A.,
Speit, G., Thybaud, V., and Tice, R.R. (2003). Recommendations for conducting the in vivo alkaline Comet assay.
Mutagenesis 18, 45-51.
Haseman, J.K. (1992). Value of historical controls in the interpretation of rodent tumor data. Drug Inf. J. 26,
191-200.
Haseman, J.K. (1995). Data analysis: Statistical analysis and use of historical control data. Regul. Toxicol.
Pharmacol. 21, 52-59.
Haseman, J.K., and Rao, G.N. (1992). Effects of corn oil, time-related changes, and inter-laboratory variability on
tumor occurrence in control Fischer 344 (F344/N) rats. Toxicol. Pathol. 20, 52-60.
Heddle, J.A., Hite, M., Kirkhart, B., Mavournin, K., MacGregor, J.T., Newell, G.W., and Salamone, M.F. (1983).
The induction of micronuclei as a measure of genotoxicity. A report of the U.S. Environmental Protection Agency
Gene-Tox Program. Mutat. Res. 123, 61-118.
Hocking, B., and Westerman, R. (2000). Neurological abnormalities associated with mobile phone use. Occup.
Med. (London) 50, 366-368.
Huber, R., Graf, T., Cote, K.A., Wittmann, L., Gallmann, E., Matter, D., Schuderer, J., Kuster, N., Borbély, A.A.,
and Achermann, P. (2000). Exposure to pulsed high-frequency electromagnetic field during waking affects human
sleep EEG. Neuroreport 11, 3321-3325.
Huber, R., Treyer, V., Borbély, A.A., Schuderer, J., Gottselig, J.M., Landolt, H.P., Werth, E., Berthold, T.,
Kuster, N., Buck, A., and Achermann, P. (2002). Electromagnetic fields, such as those from mobile phones, alter
regional cerebral blood flow and sleep and waking EEG. J. Sleep Res. 11, 289-295.
Huber, R., Schuderer, J., Graf, T., Jütz, K., Borbély, A.A., Kuster, N., and Achermann, P. (2003). Radio frequency
electromagnetic field exposure in humans: Estimation of SAR distribution in the brain, effects on sleep and heart
rate. Bioelectromagnetics 24, 262-276.
Hung, C.S., Anderson, C., Horne, J.A., and McEvoy, P. (2007). Mobile phone ‘talk-mode’ signal delays
EEG-determined sleep onset. Neurosci. Lett. 421, 82-86.
Hunt, C.R., Pandita, R.K., Laszlo, A., Higashikubo, R., Agarwal, M., Kitamura, T., Gupta, A., Rief, N.,
Horikoshi, N., Baskaran, R., Lee, J.-H., Löbrich, M., Paull, T.T., Roti Roti, J.L., and Pandita, T.K. (2007).
Hyperthermia activates a subset of ataxia-telangiectasia mutated effectors independent of DNA strand breaks and
heat shock protein 70 status. Cancer Res. 67, 3010-3017.
Ilhan, A., Gurel, A., Armutcu, F., Kamisli, S., Iraz, M., Akyol, O., and Ozen, S. (2004). Ginkgo biloba prevents
mobile phone-induced oxidative stress in rat brain. Clin. Chim. Acta. 340, 153-162.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 109
Peer Review Draft NOT FOR ATTRIBUTION
Imge, E.B., Kiliçoğlu, B., Devrim, E., Cetin, R., and Durak, I. (2010). Effects of mobile phone use on brain tissue
from the rat and a possible protective role of vitamin C – a preliminary study. Int. J. Radiat. Biol. 86, 1044-1049.
Inskip, P.D., Hoover, R.N., and Devesa, S.S. (2010). Brain cancer incidence trends in relation to cellular telephone
use in the United States. Neuro-oncology 12, 1147-1151.
International Agency for Research on Cancer (IARC) (2011). IARC Report to the Union for International Cancer
Control (UICC) on the Interphone Study. IARC, Lyon, France.
International Agency for Research on Cancer (IARC) (2013). IARC Monographs on the Evaluation of Carcinogenic
Risks to Humans. Non-ionizing Radiation, Part 2: Radiofrequency Electromagnetic Fields, Vol. 102. IARC, Lyon,
France.
INTERPHONE Study Group (2010). Brain tumour risk in relation to mobile telephone use: Results of the
INTERPHONE international case-control study. Int. J. Epidemiol. 39, 675-694.
INTERPHONE Study Group (2011). Acoustic neuroma risk in relation to mobile telephone use: Results of the
INTERPHONE international case-control study. Cancer Epidemiol. 35, 453-464.
Johansen, C., Boice, J.D., Jr., McLaughlin, J.K., and Olsen, J.H. (2001). Cellular telephones and cancer – a
nationwide cohort study in Denmark. J. Natl. Cancer Inst. 93, 203-207.
Jonckheere, A.R. (1954). A distribution-free k-sample test against ordered alternatives. Biometrika 41, 133-145.
Kampinga, H.H., and Dikomey, E. (2001). Hyperthermic radiosensitization: Mode of action and clinical relevance.
Int. J. Radiat. Biol. 77, 399-408.
Kaplan, E.L., and Meier, P. (1958). Nonparametric estimation from incomplete observations. J. Am. Stat. Assoc.
53, 457-481.
Kaufman, D.W., Anderson, T.E., and Issaragrisil, S. (2009). Risk factors for leukemia in Thailand. Ann. Hematol.
88, 1079-1088.
Kissling, G.E., Dertinger, S.D., Hayashi, M., and MacGregor, J.T. (2007). Sensitivity of the erythrocyte
micronucleus assay: Dependence on number of cells scored and inter-animal variability. Mutat. Res. 634, 235-240.
Kolomytseva, M.P., Gapeev, A.B., Sadovnikov, V.B., and Chemeris, N.K. (2002). Suppression of nonspecific
resistance of the body under the effect of extremely high frequency electromagnetic radiation of low intensity [in
Russian]. Biofizika 47, 71-77.
Komae, N., Hibino, Y., and Sugano, N. (1999). Analysis of micronuclei induced under hyperthermic conditions in
human lymphocyte culture by fluorescence in situ hybridization (FISH) and spectral karyotyping (SKY) methods [in
Japanese]. Yakugaku Zasshi 119, 763-772.
Kwon, M.S., and Hämäläinen, H. (2011). Effects of mobile phone electromagnetic fields: Critical evaluation of
behavioral and neurophysiological studies. Bioelectromagnetics 32, 253-272.
La Regina, M., Moros, E.G., Pickard, W.F., Straube, W.L., Baty, J., and Roti Roti, J.L. (2003). The effect of
chronic exposure to 835.62 MHz FDMA or 847.74 MHz CDMA radiofrequency radiation on the incidence of
spontaneous tumors in rats. Radiat. Res. 160, 143-151.
Lee, H.J., Pack, J.K., Kim, T.H., Kim, N., Choi, S.Y., Lee, J.S., Kim, S.H., and Lee, Y.S. (2010). The lack of
histological changes of CDMA cellular phone-based radio frequency on rat testis. Bioelectromagnetics 31, 528-534.
110 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
Lee, H.J., Jin, Y.B., Lee, J.S., Choi, S.Y., Kim, T.H., Pack, J.K., Choi, H.D., Kim, N., and Lee, Y.S. (2011).
Lymphoma development of simultaneously combined exposure to two radiofrequency signals in AKR/J mice.
Bioelectromagnetics 32, 485-492.
Linet, M.S., Taggart, T., Severson, R.K., Cerhan, J.R., Cozen, W., Hartge, P., and Colt, J. (2006). Cellular
telephones and non-Hodgkin lymphoma. Int. J. Cancer 119, 2382-2388.
Lönn, S., Klaeboe, L., Hall, P., Mathiesen, T., Auvinen, A., Christensen, H.C., Johansen, C., Salminen, T.,
Tynes, T., and Feychting, M. (2004). Incidence trends of adult primary intracerebral tumors in four Nordic
countries. Int. J. Cancer 108, 450-455.
Lönn, S., Ahlbom, A., Christensen, H.C., Johansen, C., Schüz, J., Edström, S., Henriksson, G., Lundgren, J.,
Wennerberg, J., and Feychting, M. (2006). Mobile phone use and risk of parotid gland tumor. Am. J. Epidemiol.
164, 637-643.
Lorenzo, Y., Costa, S., Collins, A.R., and Azqueta, A. (2013). The comet assay, DNA damage, DNA repair and
cytotoxicity: Hedgehogs are not always dead. Mutagenesis 28, 427-432.
Loughran, S.P., Wood, A.W., Barton, J.M., Croft, R.J., Thompson, B., and Stough, C. (2005). The effect of
electromagnetic fields emitted by mobile phones on human sleep. Neuroreport 16, 1973-1976.
Lowden, A., Ã…kerstedt, T., Ingre, M., Wiholm, C., Hillert, L., Kuster, N., Nilsson, J.P., and Arnetz, B. (2011). Sleep
after mobile phone exposure in subjects with mobile phone-related symptoms. Bioelectromagnetics 32, 4-14.
Lushnikov, K.V., Gapeev, A.B., Sadovnikov, V.B., and Cheremis, N.K. (2001). Effect of extremely high frequency
electromagnetic radiation of low intensity on parameters of humoral immunity in healthy mice [in Russian].
Biofizika 46, 753-760.
McQuade, J.M.S., Merritt, J.H., Miller, S.A., Scholin, T., Cook, M.C., Salazar, A., Rahimi, O.B., Murphy, M.R.,
and Mason, P.A. (2009). Radiofrequency-radiation exposure does not induce detectable leakage of albumin across
the blood-brain barrier. Radiat. Res. 171, 615-621.
Mailankot, M., Kunnath, A.P., Jayalekshmi, H., Koduru, B., and Valsalan, R. (2009). Radio frequency
electromagnetic radiation (RF-EMR) from GSM (0.9/1.8GHz) mobile phones induces oxidative stress and reduces
sperm motility in rats. Clinics (Sao Paulo) 64, 561-565.
Mann, K., and Röschke, J. (1996). Effects of pulsed high-frequency electromagnetic fields on human sleep.
Neuropsychobiology 33, 41-47.
Maronpot, R.R., and Boorman, G.A. (1982). Interpretation of rodent hepatocellular proliferative alterations and
hepatocellular tumors in chemical safety assessment. Toxicol. Pathol. 10, 71-80.
Masuda, H., Sanchez, S., Dulou, P.E., Haro, E., Anane, R., Billaudel, B., Lévêque, P., and Veyret, B. (2006). Effect
of GSM-900 and -1800 signals on the skin of hairless rats. I: 2-Hour acute exposures. Int. J. Radiat. Biol. 82,
669-674.
Masuda, H., Ushiyama, A., Takahashi, M., Wang, J., Fujiwara, O., Hikage, T., Nojima, T., Fujita, K., Kudo, M., and
Ohkubo, C. (2009). Effects of 915 MHz electromagnetic-field radiation in TEM cell on the blood-brain barrier and
neurons in the rat brain. Radiat. Res. 172, 66-73.
Meral, I., Mert, H., Mert, N., Deger, Y., Yoruk, I., Yetkin, A., and Keskin, S. (2007). Effects of 900-MHz
electromagnetic field emitted from cellular phone on brain oxidative stress and some vitamin levels of guinea pigs.
Brain Res. 1169, 120-124.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 111
Peer Review Draft NOT FOR ATTRIBUTION
Miller, J.A., and Miller, E.C. (1977). Ultimate chemical carcinogens as reactive mutagenic electrophiles. In Origins
of Human Cancer (H.H. Hiatt, J.D. Watson, and J.A. Winsten, Eds.), pp. 605-627. Cold Spring Harbor Laboratory,
Cold Spring Harbor, NY.
Muscat, J.E., Hinsvark, M., and Malkin, M. (2006). Mobile telephones and rates of brain cancer.
Neuroepidemiology 27, 55-56.
Nasta, F., Prisco, M.G., Pinto, R., Lovisolo, G.A., Marino, C., and Pioli, C. (2006). Effects of GSM-modulated
radiofrequency electromagnetic fields on B-cell peripheral differentiation and antibody production. Radiat. Res.
165, 664-670.
National Toxicology Program (2018). Toxicology and Carcinogenesis Studies of GSM- and CDMA-Modulated
Cell Phone Radio Frequency Radiation at 900 MHz in Hsd:Sprague Dawley SD Rats (Whole-Body Exposure).
Technical Report Series No. 595. National Institutes of Health, Public Health Service, U.S. Department of Health
and Human Services, Research Triangle Park, NC (in preparation).
Nelson, P.D., Toledano, M.B., McConville, J., Quinn, M.J., Cooper, N., and Elliott, P. (2006). Trends in acoustic
neuroma and cellular phones: Is there a link? Neurology 66, 284-285.
Nittby, H., Widegren, B., Krogh, M., Grafström, G., Berlin, H., Rehn, G., Eberhardt, J.L., Malmgren, L.,
Persson, B.R.R., and Salford, L.G. (2008). Exposure to radiation from global system for mobile communications at
1,800 MHz significantly changes gene expression in rat hippocampus and cortex. Environmentalist 28, 458-465.
Nittby, H., Brun, A., Eberhardt, J., Malmgren, L., Persson, B.R.R., and Salford, L.G. (2009). Increased blood-brain
barrier permeability in mammalian brain 7 days after exposure to the radiation from a GSM-900 mobile phone.
Pathophysiology 16, 103-112.
Nittby, H., Brun, A., Strömblad, S., Moghadam, M.K., Sun, W., Malmgren, L., Eberhardt, J., Persson, B.R., and
Salford, L.G. (2011). Nonthermal GSM RF and ELF EMF effects upon rat BBB permeability. Environmentalist 31,
140-148.
Nomura, E., Ioka, A., and Tsukuma, H. (2011). Trends in the incidence of primary intracranial tumors in Osaka,
Japan. Jpn. J. Clin. Oncol. 41, 291-294.
Novoselova, E.G., Fesenko, E.E., Makar, V.R., and Sadovnikov, V.B. (1999). Microwaves and cellular immunity.
II. Immunostimulating effects of microwaves and naturally occurring antioxidant nutrients. Bioelectrochem.
Bioenerg. 49, 37-41.
Oberto, G., Rolfo, K., Yu, P., Carbonatto, M., Peano, S., Kuster, N., Ebert, S., and Tofani, S. (2007).
Carcinogenicity study of 217 Hz pulsed 900 MHz electromagnetic fields in Pim1 transgenic mice. Radiat. Res. 168,
316-326.
Office of Engineering and Technology (OET) (1999). Questions and answers about biological effects and potential
hazards of radiofrequency electromagnetic fields. OET Bulletin 56, 4th ed. Federal Communication Commission,
Washington, DC.
Oktem, F., Ozguner, F., Mollaoglu, H., Koyu, A., and Uz, E. (2005). Oxidative damage in the kidney induced by
900-MHz-emitted mobile phone: Protection by melatonin. Arch. Med. Res. 36, 350-355.
Oral, B., Guney, M., Ozguner, F., Karahan, N., Mungan, T., Comlekci, S., and Cesur, G. (2006). Endometrial
apoptosis induced by a 900-MHz mobile phone: Preventive effects of vitamins E and C. Adv. Ther. 23, 957-973.
Organisation for Economic Cooperation and Development (OECD) (2014). OECD Guideline for the Testing of
Chemicals. In vivo Mammalian Alkaline Comet Assay. Testing Guideline 489. OECD Publishing, Paris.
112 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
Ozguner, F., Altinbas, A., Ozaydin, M., Dogan, A., Vural, H., Kisioglu, A.N., Cesur, G., and Yildirim, N.G.
(2005a). Mobile phone-induced myocardial oxidative stress: Protection by a novel antioxidant agent caffeic acid
phenethyl ester. Toxicol. Ind. Health 21, 223-230.
Ozguner, F., Oktem, F., Armagan, A., Yilmaz, R., Koyu, A., Demirel, R., Vural, H., and Uz, E. (2005b).
Comparative analysis of the protective effects of melatonin and caffeic acid phenethyl ester (CAPE) on mobile
phone-induced renal impairment in rat. Mol. Cell. Biochem. 276, 31-37.
Ozguner, F., Bardak, Y., and Comlekci, S. (2006). Protective effects of melatonin and caffeic acid phenethyl ester
against retinal oxidative stress in long-term use of mobile phone: A comparative study. Mol. Cell. Biochem. 282,
83-88.
Ozgur, E., Güler, G., and Seyhan, N. (2010). Mobile phone radiation-induced free radical damage in the liver is
inhibited by the antioxidants N-acetyl cysteine and epigallocatechin-gallate. Int. J. Radiat. Biol. 86, 935-945.
Paparini, A., Rossi, P., Gianfranceschi, G., Brugaletta, V., Falsaperla, R., De Luca, P., and Romano Spica, V.
(2008). No evidence of major transcriptional changes in the brain of mice exposed to 1800 MHz GSM signal.
Bioelectromagnetics 29, 312-323.
Pew Research Center Survey (2017). Mobile fact sheet. Internet and Technology, Pew Research Center.
Washington, D.C.
Pfuhler, S., and Wolf, H.U. (1996). Detection of DNA-crosslinking agents with the alkaline comet assay. Environ.
Mol. Mutagen. 27, 196-201.
Piegorsch, W.W., and Bailer, A.J. (1997). Statistics for Environmental Biology and Toxicology, Section 6.3.2.
Chapman and Hall, London.
Portier, C.J., and Bailer, A.J. (1989). Testing for increased carcinogenicity using a survival-adjusted quantal
response test. Fundam. Appl. Toxicol. 12, 731-737.
Portier, C.J., Hedges, J.C., and Hoel, D.G. (1986). Age-specific models of mortality and tumor onset for historical
control animals in the National Toxicology Program’s carcinogenicity experiments. Cancer Res. 46, 4372-4378.
Prohofsky, E.W. (2004). RF absorption involving biological macromolecules. Bioelectromagnetics 25, 441-451.
Propp, J.M., McCarthy, B.J., Davis, F.G., and Preston-Martin, S. (2006). Descriptive epidemiology of vestibular
schwannomas. Neurooncology 8, 1-11.
Pyrpasopoulou, A., Kotoula, V., Cheva, A., Hytiroglou, P., Nikolakaki, E., Magras, I.N., Xenos, T.D.,
Tsiboukis, T.D., and Karkavelas, G. (2004). Bone morphogenetic protein expression in newborn rat kidneys after
prenatal exposure to radiofrequency radiation. Bioelectromagnetics 25, 216-227.
Recio, L., Hobbs, C., Caspary, W., and Witt, K.L. (2010). Dose-response assessment of four genotoxic chemicals in
a combined mouse and rat micronucleus (MN) and Comet assay protocol. J. Toxicol. Sci. 35, 149-162.
Recio, L., Kissling, G.E., Hobbs, C.A., and Witt, K.L. (2012). Comparison of Comet assay dose-response for ethyl
methanesulfonate using freshly prepared versus cryopreserved tissues. Environ. Mol. Mutagen. 53, 101-113.
Regel, S.J., Tinguely, G., Schuderer, J., Adam, M., Kuster, N., Landolt, H.P., and Achermann, P. (2007). Pulsed
radio-frequency electromagnetic fields: Dose-dependent effects on sleep, the sleep EEG and cognitive performance.
J. Sleep Res. 16, 253-258.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 113
Peer Review Draft NOT FOR ATTRIBUTION
Repacholi, M.H., Lerchl, A., Röösli, M., Sienkiewicz, Z., Auvinen, A., Breckenkamp, J., d’Inzeo, G., Elliott, P.,
Frei, P., Heinrich, S., Lagroye, I., Lahkola, A., McCormick, D.L., Thomas, S., and Vecchia, P. (2012). Systematic
review of wireless phone use and brain cancer and other head tumors. Bioelectromagnetics 33, 187-206.
Röösli, M., Michel, G., Kuehni, C.E., and Spoerri, A. (2007). Cellular telephone use and time trends in brain
tumour mortality in Switzerland from 1969 to 2002. Eur. J. Cancer Prev. 16, 77-82.
Ruediger, H.W. (2009). Genotoxic effects of radiofrequency electromagnetic fields. Pathophysiology 16, 89-102.
Rundell, M.S., Wagner, E.D., and Plewa, M.J. (2003). The comet assay: Genotoxic damage or nuclear
fragmentation? Environ. Mol. Mutagen. 42, 61-67.
Saika, K., and Katanoda, K. (2011). Comparison of time trends in brain and central nervous system cancer mortality
(1990–2006) between countries based on the WHO mortality database. Jpn. J. Clin. Oncol. 41, 304-305.
Sanchez, S., Masuda, H., Billaudel, B., Haro, E., Anane, R., Lévêque, P., Ruffie, G., Lagroye, I., and Veyret, B.
(2006). Effect of GSM-900 and -1800 signals on the skin of hairless rats. II: 12-Week chronic exposures. Int.
J. Radiat. Biol. 82, 675-680.
Sanchez, S., Masuda, H., Ruffie, G., De Gannes, F.P., Billaudel, B., Haro, E., Lévêque, P., Lagroye, I., and
Veyret, B. (2008). Effect of GSM-900 and -1800 signals on the skin of hairless rats. III: Expression of heat shock
proteins. Int. J. Radiat. Biol. 84, 61-68.
Sandström, M., Wilen, J., Oftedal, G., and Hansson Mild, K. (2001). Mobile phone use and subjective symptoms.
Comparison of symptoms experienced by users of analogue and digital mobile phones. Occup. Med. (London) 51,
25-35.
Santini, R., Santini, P., Danze, J.M., Le Ruz, P., and Seigne, M. (2002a). Investigation on the health of people
living near mobile telephone relay stations: I/Incidence according to distance and sex [in French]. Pathol. Biol.
(Paris) 50, 369-373.
Santini, R., Seigne, M., Bonhomme-Faivre, L., Bouffet, S., Defrasne, E., and Sage, M. (2002b). Symptoms
experienced by users of digital cellular phones: A study of a French engineering school. Electromagn. Biol. Med.
21, 81-88.
Sasaki, Y.F., Sekihashi, K., Izumiyama, F., Nishidate, E., Saga, A., Ishida, K., and Tsuda, S. (2000). The comet
assay with multiple mouse organs: Comparison of comet assay results and carcinogenicity with 208 chemicals
selected from the IARC monographs and U.S. NTP Carcinogenicity Database. Crit. Rev. Toxicol. 30, 629-799.
Schmid, W. (1975). The micronucleus test. Mutat. Res. 31, 9-15.
Schoemaker, M.J., and Swerdlow, A.J. (2009). Risk of pituitary tumors in cellular phone users: A case-control
study. Epidemiology 20, 348-354.
Schüz, J., Jacobsen, R., Olsen, J.H., Boice, J.D., Jr., McLaughlin, J.K., and Johansen, C. (2006). Cellular telephone
use and cancer risk: Update of a nationwide Danish cohort. J. Natl. Cancer. Inst. 98, 1707-1713.
Schüz, J., Elliott, P., Auvinen, A., Kromhout, H., Poulsen, A.H., Johansen, C., Olsen, J.H., Hillert, L.,
Feychting, M., Fremling, K., Toledano, M., Heinävaara, S., Slottje, P., Vermeulen, R., and Ahlbom, A. (2011). An
international prospective cohort study of mobile phone users and health (Cosmos): Design considerations and
enrolment. Cancer Epidemiol. 35, 37-43.
Sheppard, A.R., Swicord, M.L., and Balzano, Q. (2008). Quantitative evaluations of mechanisms of radiofrequency
interactions with biological molecules and processes. Health Phys. 95, 365-396.
114 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
Shirley, E. (1977). A non-parametric equivalent of Williams’ test for contrasting increasing dose levels of a
treatment. Biometrics 33, 386-389.
Smialowicz, R.J., Rogers, R.R., Garner, R.J., Riddle, M.M., Luebke, R.W., and Rowe, D.G. (1983). Microwaves
(2,450 MHz) suppress murine natural killer cell activity. Bioelectromagnetics 4, 371-381.
Smith, P., Kuster, N., Ebert, S., and Chevalier, H.J. (2007). GSM and DCS wireless communication signals:
Combined chronic toxicity/carcinogenicity study in the Wistar rat. Radiat. Res. 168, 480-492.
Sokolovic, D., Djindjic, B., Nikolic, J., Bjelakovic, G., Pavlovic, D., Kocic, G., Krstic, D., Cvetkovic, T., and
Pavlovic, V. (2008). Melatonin reduces oxidative stress induced by chronic exposure of microwave radiation from
mobile phones in rat brain. J. Radiat. Res. 49, 579-586.
Sommer, A.M., Streckert, J., Bitz, A.K., Hansen, V.W., and Lerchl, A. (2004). No effects of GSM-modulated
900 MHz electromagnetic fields on survival rate and spontaneous development of lymphoma in female AKR/J mice.
BMC Cancer 4, 77.
Sommer, A.M., Bitz, A.K., Streckert, J., Hansen, V.W., and Lerchl, A. (2007). Lymphoma development in mice
chronically exposed to UMTS-modulated radiofrequency electromagnetic fields. Radiat. Res. 168, 72-80.
Sonmez, O.F., Odaci, E., Bas, O., and Kaplan, S. (2010). Purkinje cell number decreases in the adult female rat
cerebellum following exposure to 900 MHz electromagnetic field. Brain. Res. 1356, 95-101.
Speit, G., and Schütz, P. (2013). Hyperthermia-induced genotoxic effects in human A549 cells. Mutat. Res.
747-748, 1-5.
Speit, G., Schütz, P., and Hoffmann, H. (2007). Genotoxic effects of exposure to radiofrequency electromagnetic
fields (RF-EMF) in cultured mammalian cells are not independently reproducible. Mutat. Res. 626, 42-47.
Speit, G., Gminski, R., and Tauber, R. (2013). Genotoxic effects of exposure to radiofrequency electromagnetic
fields (RF-EMF) in HL-60 cells are not reproducible. Mutat. Res. 755, 163-166.
Stagg, R.B., Hawel, L.H., III, Pastorian, K., Cain, C., Adey, W.R., and Byus, C.V. (2001). Effect of immobilization
and concurrent exposure to a pulse-modulated microwave field on core body temperature, plasma ACTH and
corticosteroid, and brain ornithine decarboxylase, Fos and Jun mRNA. Radiat. Res. 155, 584-592.
Stang, A., Schmidt-Pokrzywniak, A., Lash, T.L., Lommatzsch, P.K., Taubert, G., Bornfeld, N., and Jöckel, K.H.
(2009). Mobile phone use and risk of uveal melanoma: Results of the risk factors for uveal melanoma case-control
study. J. Natl. Cancer Inst. 101, 120-123.
Straus, D.S. (1981). Somatic mutation, cellular differentiation, and cancer causation. JNCI 67, 233-241.
Takebayashi, T., Varsier, N., Kikuchi, Y., Wake, K., Taki, M., Watanabe, S., Akiba, S., and Yamaguchi, N. (2008).
Mobile phone use, exposure to radiofrequency electromagnetic field, and brain tumour: A case-control study. Br.
J. Cancer 12, 652-659.
Tarone, R.E. (1975). Tests for trend in life table analysis. Biometrika 62, 679-682.
Tice, R.R., Agurell, E., Anderson, D., Burlinson, B., Hartmann, A., Kobayashi, H., Miyamae, Y., Rojas, E.,
Ryu, J.-C., and Sasaki, Y.F. (2000). Single cell gel/Comet assay: Guidelines for in vitro and in vivo genetic
toxicology testing. Environ. Mol. Mutagen. 35, 206-221.
Tillmann, T., Ernst, H., Ebert, S., Kuster, N., Behnke, W., Rittinghausen, S., and Dasenbrock, C. (2007).
Carcinogenicity study of GSM and DCS wireless communication signals in B6C3F1 mice. Bioelectromagnetics 28,
173-187.
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 115
Peer Review Draft NOT FOR ATTRIBUTION
Tomruk, A., Güler, G., and Dincel, A.S. (2010). The influence of 1800 MHz GSM-like signals on hepatic oxidative
DNA and lipid damage in nonpregnant, pregnant, and newly born rabbits. Cell Biochem. Biophys. 56, 39-47.
United States Code (USC) 42 §4321 et seq.
Utteridge, T.D., Gebski, V., Finnie, J.W., Vernon-Roberts, B., and Kuchel, T.R. (2002). Long-term exposure of
Eµ-Pim1 transgenic mice to 898.4 MHz microwaves does not increase lymphoma incidence. Radiat. Res. 158,
357-364.
Verschaeve, L., Juutilainen, J., Lagroye, I., Miyakoshi, J., Saunders, R., de Seze, R., Tenforde, T., van Rongen, E.,
Veyret, B., and Xu, Z. (2010). In vitro and in vivo genotoxicity of radiofrequency fields. Mutat. Res. 705, 252-268.
Veyret, B., Bouthet, C., Deschaux, P., de Seze, R., Geffard, M., Joussot-Dubien, J., le Diraison, M., Moreau, J.M.,
and Caristan, A. (1991). Antibody responses of mice exposed to low-power microwaves under combined,
pulse-and-amplitude modulation. Bioelectromagnetics 12, 47-56.
Vijayalaxmi, and Prihoda, T.J. (2012). Genetic damage in human cells exposed to non-ionizing radiofrequency
fields: A meta-analysis of the data from 88 publications (1990-2011). Mutat. Res. 749, 1-16.
Volkow, N.D., Tomasi, D., Wang, G.J., Vaska, P., Fowler, J.S., Telang, F., Alexoff, D., Logan, J., and Wong, C.
(2011). Effects of cell phone radiofrequency signal exposure on brain glucose metabolism. JAMA 305, 808-813.
Wagner, P., Röschke, J., Mann, K., Hiller, W., and Frank, C. (1998). Human sleep under the influence of pulsed
radiofrequency electromagnetic fields: A polysomnographic study using standardized conditions.
Bioelectromagnetics 19, 199-202.
Wagner, P., Röschke, J., Mann, K., Fell, J., Hiller, W., Frank, C., and Grözinger, M. (2000). Human sleep EEG
under the influence of pulsed radio frequency electromagnetic fields. Results from polysomnographies using
submaximal high power flux densities. Neuropsychobiology 42, 207-212.
Williams, D.A. (1971). A test for differences between treatment means when several dose levels are compared with
a zero dose control. Biometrics 27, 103-117.
Williams, D.A. (1972). The comparison of several dose levels with a zero dose control. Biometrics 28, 519-531.
Williams, D.A. (1986). A note on Shirley’s nonparametric test for comparing several dose levels with a zero-dose
control. Biometrics 42, 183-186.
Witt, K.L., Knapton, A., Wehr, C.M., Hook, G.J., Mirsalis, J., Shelby, M.D., and MacGregor, J.T. (2000).
Micronucleated erythrocyte frequency in peripheral blood of B6C3F1 mice from short-term, prechronic, and chronic
studies of the NTP Carcinogenesis Bioassay Program. Environ. Mol. Mutagen. 36, 163-194.
Witt, K.L., Livanos, E., Kissling, G.E., Torous, D.K., Caspary, W., Tice, R.R., and Recio, L. (2008). Comparison of
flow cytometry- and microscopy-based methods for measuring micronucleated reticulocyte frequencies in rodents
treated with nongenotoxic and genotoxic chemicals. Mutat. Res. 649, 101-113.
Wyde, M.E., Horn, T.L., Capstick, M.H., Ladbury, J.M., Koepke, G., Wilson, P.F., Kissling, G.E., Stout, M.D.,
Kuster, N., Melnick, R.L., Gauger, J., Bucher, J.R., and McCormick, D.L. (2018). Pilot Studies of the National
Toxicology Program’s Cell Phone Radiofrequency Radiation Reverberation Chamber Exposure System (publication
pending).
Yakymenko, I., Tsybulin, O., Sidorik, E., Henshel, D., Kyrylenko, O., and Kyrylenko, S. (2016). Oxidative
mechanisms of biological activity of low-intensity radiofrequency radiation. Electromagn. Biol. Med. 35, 186-202.
116 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
A-1
Peer Review Draft NOT FOR ATTRIBUTION
APPENDIX A
SUMMARY OF LESIONS IN MALE MICE
EXPOSED TO GSM-MODULATED
CELL PHONE RFR FOR 2 YEARS
TABLE A1 Summary of the Incidence of Neoplasms in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years.............................................................A-2
TABLE A2 Statistical Analysis of Primary Neoplasms in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years.............................................................A-8
TABLE A3a Historical Incidence of Skin Neoplasms in Control Male B6C3F1/N Mice................................. A-11
TABLE A3b Historical Incidence of Alveolar/bronchiolar Neoplasms
in Control Male B6C3F1/N Mice ....................................................................................................A-11
TABLE A4 Summary of the Incidence of Nonneoplastic Lesions in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years...........................................................A-12
A-2 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
TABLE A1
Summary of the Incidence of Neoplasms in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Yearsa
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
Disposition Summary
Animals initially in study 105 105 105 105
14-Week interim evaluation 15 15 15 15
Early deaths
Accidental death 1
Moribund 8 6 2 6
Natural deaths 16 19 8 12
Survivors
Died last week of study 1 4
Terminal euthanasia 66 63 79 68
Missing 1
Animals examined microscopically 100 100 100 100
14-Week Interim Evaluation
Nervous System
Brain (10) (10) (10) (10)
Hamartoma, lipomatous 1 (10%)
Systems Examined with No Neoplasms Observed
Alimentary System
Cardiovascular System
Endocrine System
General Body System
Genital System
Hematopoietic System
Integumentary System
Musculoskeletal System
Respiratory System
Special Senses System
Urinary System
2-Year Study
Alimentary System
Esophagus (88) (87) (88) (90)
Gallbladder (73) (66) (74) (79)
Intestine large, cecum (81) (77) (84) (78)
Leiomyoma 1 (1%)
Intestine large, colon (84) (83) (85) (84)
Intestine large, rectum (84) (85) (86) (84)
Intestine small, duodenum (77) (77) (83) (79)
Adenocarcinoma 1 (1%) 1 (1%)
Adenoma 1 (1%)
Intestine small, ileum (81) (79) (85) (80)
Intestine small, jejunum (79) (79) (82) (79)
Adenocarcinoma 2 (3%) 1 (1%)
Hepatocellular carcinoma, metastatic,
liver 1 (1%)
Hepatocholangiocarcinoma, metastatic,
liver 1 (1%)
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 A-3
Peer Review Draft NOT FOR ATTRIBUTION
TABLE A1
Summary of the Incidence of Neoplasms in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
2-Year Study (continued)
Alimentary System (continued)
Liver (90) (89) (90) (90)
Adenocarcinoma, metastatic,
Harderian gland 1 (1%)
Carcinoma, metastatic, islets, pancreatic 1 (1%)
Hemangioma 1 (1%)
Hemangiosarcoma 1 (1%) 3 (3%) 2 (2%) 2 (2%)
Hepatoblastoma 6 (7%) 3 (3%) 8 (9%) 1 (1%)
Hepatoblastoma, multiple 1 (1%)
Hepatocellular adenoma 25 (28%) 28 (31%) 20 (22%) 26 (29%)
Hepatocellular adenoma, multiple 27 (30%) 33 (37%) 46 (51%) 29 (32%)
Hepatocellular carcinoma 26 (29%) 23 (26%) 28 (31%) 19 (21%)
Hepatocellular carcinoma, multiple 2 (2%) 2 (2%) 2 (2%) 3 (3%)
Hepatocholangiocarcinoma 1 (1%) 4 (4%)
Malignant fibrous histiocytoma,
metastatic, skin 1 (1%)
Mesentery (12) (14) (13) (17)
Hemangiosarcoma 1 (8%) 1 (8%)
Hepatocholangiocarcinoma, metastatic,
liver 1 (7%)
Malignant fibrous histiocytoma,
metastatic, skin 1 (8%)
Fat, hepatocholangiocarcinoma,
metastatic, liver 1 (8%) 1 (7%)
Fat, lipoma 1 (8%)
Pancreas (87) (88) (88) (86)
Hepatocholangiocarcinoma, metastatic,
liver 1 (1%) 2 (2%)
Salivary glands (90) (89) (89) (89)
Stomach, forestomach (88) (87) (89) (87)
Squamous cell papilloma 1 (1%) 2 (2%)
Stomach, glandular (87) (86) (88) (85)
Malignant fibrous histiocytoma,
metastatic, skin 1 (1%)
Tooth (27) (26) (16) (20)
Cardiovascular System
Aorta (89) (89) (89) (87)
Alveolar/bronchiolar carcinoma,
metastatic, lung 1 (1%)
Hepatocholangiocarcinoma, metastatic,
liver 1 (1%)
Blood vessel (1) (0) (0) (0)
Heart (90) (89) (90) (90)
Alveolar/bronchiolar carcinoma,
metastatic, lung 1 (1%) 1 (1%) 2 (2%)
Hemangiosarcoma 1 (1%) 1 (1%)
Hepatocholangiocarcinoma, metastatic,
liver 1 (1%) 2 (2%)
A-4 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
TABLE A1
Summary of the Incidence of Neoplasms in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
2-Year Study (continued)
Endocrine System
Adrenal cortex (90) (89) (89) (88)
Bilateral, malignant fibrous histiocytoma,
metastatic, skin 1 (1%)
Subcapsular, adenoma 3 (3%) 3 (3%)
Adrenal medulla (90) (88) (88) (86)
Islets, pancreatic (88) (88) (90) (89)
Adenoma 2 (2%)
Adenoma, multiple 1 (1%)
Carcinoma 1 (1%) 1 (1%)
Parathyroid gland (68) (68) (67) (66)
Pituitary gland (86) (85) (87) (85)
Thyroid gland (89) (88) (88) (88)
General Body System
Peritoneum (1) (0) (0) (0)
Hepatocholangiocarcinoma, metastatic,
liver 1 (100%)
Tissue NOS (0) (0) (0) (1)
Genital System
Coagulating gland (2) (2) (0) (4)
Epididymis (90) (89) (90) (90)
Hemangioma 1 (1%)
Hepatocholangiocarcinoma, metastatic,
liver 1 (1%)
Preputial gland (89) (88) (90) (89)
Prostate (90) (87) (90) (87)
Seminal vesicle (90) (88) (90) (90)
Fibroma 1 (1%)
Malignant fibrous histiocytoma,
metastatic, skin 1 (1%)
Testis (90) (88) (90) (90)
Hemangioma 1 (1%)
Interstitial cell, adenoma 2 (2%)
Hematopoietic System
Bone marrow (90) (88) (90) (90)
Hemangiosarcoma 1 (1%) 1 (1%)
Lymph node (6) (8) (7) (9)
Sarcoma, metastatic, skin 1 (13%)
Axillary, hepatocholangiocarcinoma,
metastatic, liver 1 (17%)
Lymph node, mandibular (72) (61) (63) (60)
Lymph node, mesenteric (85) (82) (88) (83)
Hemangioma 1 (1%)
Hepatocholangiocarcinoma, metastatic,
liver 1 (1%)
Malignant fibrous histiocytoma,
metastatic, skin 1 (1%)
Spleen (87) (88) (89) (88)
Hemangiosarcoma 4 (5%) 1 (1%) 1 (1%)
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 A-5
Peer Review Draft NOT FOR ATTRIBUTION
TABLE A1
Summary of the Incidence of Neoplasms in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
2-Year Study (continued)
Hematopoietic System (continued)
Thymus (75) (83) (81) (72)
Hepatocellular carcinoma, metastatic,
liver 1 (1%)
Hepatocholangiocarcinoma, metastatic,
liver 3 (4%)
Thymoma benign 1 (1%)
Integumentary System
Mammary gland (2) (5) (2) (8)
Skin (90) (89) (90) (90)
Keratoacanthoma 1 (1%)
Pilomatrixoma 1 (1%)
Sebaceous gland, adenoma 1 (1%)
Subcutaneous tissue, fibrosarcoma 1 (1%)
Subcutaneous tissue, hemangioma 1 (1%)
Subcutaneous tissue, hemangiosarcoma 1 (1%) 2 (2%)
Subcutaneous tissue, lipoma 1 (1%)
Subcutaneous tissue, liposarcoma 1 (1%)
Subcutaneous tissue,
malignant fibrous histiocytoma 1 (1%) 4 (4%) 3 (3%)
Subcutaneous tissue,
malignant fibrous histiocytoma,
multiple 1 (1%)
Subcutaneous tissue, sarcoma 1 (1%)
Musculoskeletal System
Bone (90) (88) (90) (90)
Hepatocholangiocarcinoma, metastatic,
liver 1 (1%)
Skeletal muscle (90) (89) (90) (90)
Hepatocellular carcinoma, metastatic,
liver 1 (1%) 1 (1%)
Hepatocholangiocarcinoma, metastatic,
liver 1 (1%) 2 (2%)
Malignant fibrous histiocytoma,
metastatic, skin 1 (1%)
Sarcoma 1 (1%)
Nervous System
Brain (90) (89) (90) (90)
Hepatocholangiocarcinoma, metastatic,
liver 1 (1%)
Brain trigeminal ganglion (69) (79) (72) (79)
Nerve trigeminal (67) (53) (66) (63)
Peripheral nerve, sciatic (89) (89) (90) (89)
Spinal cord (90) (89) (90) (90)
A-6 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
TABLE A1
Summary of the Incidence of Neoplasms in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
2-Year Study (continued)
Respiratory System
Lung (90) (89) (90) (90)
Adenocarcinoma, metastatic,
Harderian gland 1 (1%)
Alveolar/bronchiolar adenoma 11 (12%) 13 (15%) 16 (18%) 15 (17%)
Alveolar/bronchiolar adenoma, multiple 2 (2%) 2 (2%) 1 (1%)
Alveolar/bronchiolar carcinoma 11 (12%) 12 (13%) 15 (17%) 17 (19%)
Alveolar/bronchiolar carcinoma, multiple 2 (2%) 1 (1%) 1 (1%)
Carcinoma, metastatic, islets, pancreatic 1 (1%)
Hepatoblastoma, metastatic, liver 1 (1%) 1 (1%) 1 (1%)
Hepatocellular carcinoma, metastatic,
liver 11 (12%) 8 (9%) 6 (7%) 5 (6%)
Hepatocholangiocarcinoma, metastatic,
liver 1 (1%) 3 (3%)
Sarcoma, metastatic, skin 1 (1%)
Mediastinum (0) (0) (2) (1)
Alveolar/bronchiolar carcinoma,
metastatic, lung 1 (100%)
Hibernoma 2 (100%)
Nose (90) (89) (90) (89)
Trachea (90) (89) (89) (90)
Special Senses System
Eye (90) (89) (90) (90)
Adenocarcinoma, metastatic,
Harderian gland 1 (1%)
Harderian gland (88) (89) (90) (90)
Adenocarcinoma 3 (3%) 2 (2%) 1 (1%)
Adenoma 6 (7%) 7 (8%) 11 (12%) 5 (6%)
Urinary System
Kidney (90) (89) (90) (89)
Alveolar/bronchiolar carcinoma,
metastatic, lung 1 (1%)
Hepatocellular carcinoma, metastatic,
liver 1 (1%)
Hepatocholangiocarcinoma, metastatic,
liver 1 (1%) 2 (2%)
Malignant fibrous histiocytoma,
metastatic, skin 1 (1%)
Renal tubule, adenoma 1 (1%) 1 (1%)
Urinary bladder (87) (88) (90) (89)
Hemangioma 2 (2%)
Urothelium, papilloma 2 (2%)
Systemic Lesions
Multiple organsb (90) (89) (90) (90)
Histiocytic sarcoma 1 (1%) 2 (2%)
Leukemia granulocytic 1 (1%)
Lymphoma malignant 6 (7%) 4 (4%) 3 (3%) 4 (4%)
Mast cell tumor 1 (1%)
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 A-7
Peer Review Draft NOT FOR ATTRIBUTION
TABLE A1
Summary of the Incidence of Neoplasms in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
2-Year Study (continued)
Neoplasm Summary
Total animals with primary neoplasmsc
14-Week interim evaluation 1
2-Year study 79 82 82 77
Total primary neoplasms
14-Week interim evaluation 1
2-Year study 144 152 182 140
Total animals with benign neoplasms
14-Week interim evaluation 1
2-Year study 61 67 77 61
Total benign neoplasms
14-Week interim evaluation 1
2-Year study 77 91 109 81
Total animals with malignant neoplasms
2-Year study 49 47 53 45
Total malignant neoplasms
2-Year study 66 61 73 59
Total animals with metastatic neoplasms
2-Year study 14 15 6 10
Total metastatic neoplasms
2-Year study 34 37 7 12
Total animals with uncertain neoplasmsbenign
or malignant
2-Year study 1
Total uncertain neoplasms
2-Year study 1
a Number of animals examined microscopically at the site and the number of animals with neoplasm b Number of animals with any tissue examined microscopically c Primary neoplasms: all neoplasms except metastatic neoplasms
A-8 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
TABLE A2
Statistical Analysis of Primary Neoplasms in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
Harderian Gland: Adenoma
Overall ratea 6/90 (7%) 7/89 (8%) 11/90 (12%) 5/90 (6%)
Adjusted rateb 7.5% 8.7% 12.7% 6.0%
Terminal ratec 6/66 (9%) 5/63 (8%) 11/80 (14%) 4/72 (6%)
First incidence (days) 729 (T) 672 729 (T) 689
Poly-3 testd P=0.415N P=0.506 P=0.194 P=0.470N
Harderian Gland: Adenoma or Carcinoma
Overall rate 9/90 (10%) 9/89 (10%) 12/90 (13%) 5/90 (6%)
Adjusted rate 11.2% 11.1% 13.9% 6.0%
Terminal rate 8/66 (12%) 5/63 (8%) 12/80 (15%) 4/72 (6%)
First incidence (days) 690 651 729 (T) 689
Poly-3 test P=0.160N P=0.588N P=0.386 P=0.179N
Liver: Hepatocellular Adenoma
Overall rate 52/90 (58%) 61/89 (69%) 66/90 (73%) 55/90 (61%)
Adjusted rate 62.3% 73.8% 75.3% 64.7%
Terminal rate 45/66 (68%) 52/63 (83%) 61/80 (76%) 49/72 (68%)
First incidence (days) 393 533 605 614
Poly-3 test P=0.526N P=0.072 P=0.044 P=0.437
Liver: Hepatocellular Carcinoma
Overall rate 28/90 (31%) 25/89 (28%) 30/90 (33%) 22/90 (24%)
Adjusted rate 34.2% 30.0% 34.1% 25.9%
Terminal rate 18/66 (27%) 15/63 (24%) 25/80 (31%) 17/72 (24%)
First incidence (days) 608 547 604 538
Poly-3 test P=0.169N P=0.340N P=0.556N P=0.157N
Liver: Hepatocellular Adenoma or Carcinoma
Overall rate 67/90 (74%) 68/89 (76%) 74/90 (82%) 64/90 (71%)
Adjusted rate 79.1% 79.9% 83.4% 74.3%
Terminal rate 51/66 (77%) 52/63 (83%) 66/80 (83%) 54/72 (75%)
First incidence (days) 393 533 604 538
Poly-3 test P=0.232N P=0.526 P=0.296 P=0.281N
Liver: Hepatoblastoma
Overall rate 6/90 (7%) 3/89 (3%) 9/90 (10%) 1/90 (1%)
Adjusted rate 7.5% 3.7% 10.4% 1.2%
Terminal rate 5/66 (8%) 3/63 (5%) 8/80 (10%) 1/72 (1%)
First incidence (days) 711 729 (T) 667 729 (T)
Poly-3 test P=0.105N P=0.244N P=0.350 P=0.054N
Liver: Hepatocellular Carcinoma or Hepatoblastoma
Overall rate 32/90 (36%) 27/89 (30%) 35/90 (39%) 23/90 (26%)
Adjusted rate 39.1% 32.4% 39.7% 27.1%
Terminal rate 22/66 (33%) 17/63 (27%) 29/80 (36%) 18/72 (25%)
First incidence (days) 608 547 604 538
Poly-3 test P=0.089N P=0.230N P=0.534 P=0.067N
Liver: Hepatocellular Adenoma, Hepatocellular Carcinoma, or Hepatoblastoma
Overall rate 68/90 (76%) 68/89 (76%) 74/90 (82%) 65/90 (72%)
Adjusted rate 80.3% 79.9% 83.4% 75.4%
Terminal rate 52/66 (79%) 52/63 (83%) 66/80 (83%) 55/72 (76%)
First incidence (days) 393 533 604 538
Poly-3 test P=0.243N P=0.553N P=0.367 P=0.276N
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 A-9
Peer Review Draft NOT FOR ATTRIBUTION
TABLE A2
Statistical Analysis of Primary Neoplasms in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
Lung: Alveolar/bronchiolar Adenoma
Overall rate 13/90 (14%) 13/89 (15%) 18/90 (20%) 16/90 (18%)
Adjusted rate 16.0% 16.0% 20.7% 19.0%
Terminal rate 9/66 (14%) 10/63 (16%) 16/80 (20%) 14/72 (19%)
First incidence (days) 488 663 604 658
Poly-3 test P=0.297 P=0.583 P=0.279 P=0.380
Lung: Alveolar/bronchiolar Carcinoma
Overall rate 13/90 (14%) 12/89 (13%) 16/90 (18%) 18/90 (20%)
Adjusted rate 16.1% 14.7% 18.5% 21.2%
Terminal rate 12/66 (18%) 8/63 (13%) 16/80 (20%) 14/72 (19%)
First incidence (days) 568 594 729 (T) 614
Poly-3 test P=0.165 P=0.488N P=0.418 P=0.259
Lung: Alveolar/bronchiolar Adenoma or Carcinoma
Overall rate 23/90 (26%) 24/89 (27%) 32/90 (36%) 34/90 (38%)
Adjusted rate 28.1% 29.2% 36.8% 39.9%
Terminal rate 18/66 (27%) 17/63 (27%) 30/80 (38%) 28/72 (39%)
First incidence (days) 488 594 604 614
Poly-3 test P=0.040 P=0.506 P=0.149 P=0.074
Skin (Subcutaneous Tissue): Malignant Fibrous Histiocytoma
Overall rate 1/90 (1%) 0/89 (0%) 5/90 (6%) 3/90 (3%)
Adjusted rate 1.2% 0.0% 5.8% 3.6%
Terminal rate 0/66 (0%) 0/63 (0%) 4/80 (5%) 3/72 (4%)
First incidence (days) 674 —e 654 729 (T)
Poly-3 test P=0.127 P=0.499N P=0.124 P=0.321
Skin (Subcutaneous Tissue): Fibrosarcoma, Sarcoma, or Malignant Fibrous Histiocytoma
Overall rate 1/90 (1%) 1/89 (1%) 5/90 (6%) 4/90 (4%)
Adjusted rate 1.2% 1.2% 5.8% 4.7%
Terminal rate 0/66 (0%) 0/63 (0%) 4/80 (5%) 3/72 (4%)
First incidence (days) 674 523 654 488
Poly-3 test P=0.093 P=0.758N P=0.124 P=0.197
Spleen: Hemangiosarcoma
Overall rate 0/87 (0%) 4/88 (5%) 1/89 (1%) 1/88 (1%)
Adjusted rate 0.0% 5.0% 1.2% 1.2%
Terminal rate 0/66 (0%) 3/63 (5%) 1/80 (1%) 0/72 (0%)
First incidence (days) — 672 729 (T) 681
Poly-3 test P=0.538N P=0.065 P=0.515 P=0.507
All Organs: Hemangiosarcoma
Overall rate 2/90 (2%) 6/89 (7%) 6/90 (7%) 2/90 (2%)
Adjusted rate 2.5% 7.4% 6.9% 2.4%
Terminal rate 0/66 (0%) 4/63 (6%) 6/80 (8%) 1/72 (1%)
First incidence (days) 702 667 729 (T) 681
Poly-3 test P=0.394N P=0.141 P=0.163 P=0.677N
All Organs: Hemangioma or Hemangiosarcoma
Overall rate 3/90 (3%) 10/89 (11%) 7/90 (8%) 3/90 (3%)
Adjusted rate 3.7% 12.3% 8.1% 3.6%
Terminal rate 1/66 (2%) 8/63 (13%) 7/80 (9%) 2/72 (3%)
First incidence (days) 702 667 729 (T) 681
Poly-3 test P=0.277N P=0.042 P=0.195 P=0.641N
A-10 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
TABLE A2
Statistical Analysis of Primary Neoplasms in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
All Organs: Malignant Lymphoma
Overall rate 6/90 (7%) 4/89 (4%) 3/90 (3%) 4/90 (4%)
Adjusted rate 7.3% 4.9% 3.5% 4.8%
Terminal rate 4/66 (6%) 1/63 (2%) 3/80 (4%) 3/72 (4%)
First incidence (days) 263 609 729 (T) 690
Poly-3 test P=0.307N P=0.375N P=0.222N P=0.359N
(T) Terminal euthanasia a Number of neoplasm-bearing animals/number of animals examined. Denominator is number of animals examined microscopically for liver,
lung, and spleen; for other tissues, denominator is number of animals necropsied. b Poly-3 estimated neoplasm incidence after adjustment for intercurrent mortality c Observed incidence at terminal euthanasia d Beneath the sham control incidence is the P value associated with the trend test. Beneath the exposed group incidence are the P values
corresponding to pairwise comparisons between the sham controls and that exposed group. The Poly-3 test accounts for differential mortality
in animals that do not reach terminal euthanasia. A negative trend or a lower incidence in an exposure group is indicated by N. e Not applicable; no neoplasms in animal group
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 A-11
Peer Review Draft NOT FOR ATTRIBUTION
TABLE A3a
Historical Incidence of Skin Neoplasms in Control Male B6C3F1/N Micea
Fibrous Histiocytoma
Malignant Fibrous
Histiocytoma
Fibrosarcoma,
Sarcoma, or All
Fibrous Histiocytoma
Overall Historical Incidence: All Routes
Total (%) 1/589 (0.2%) 1/589 (0.2%) 5/589 (0.9%)
Mean ± standard deviation 0.2% ± 0.6% 0.1% ± 0.3% 0.8% ± 1.0%
Range 0%-2% 0%-1% 0%-2%
a Data as of August 2017
TABLE A3b
Historical Incidence of Alveolar/bronchiolar Neoplasms in Control Male B6C3F1/N Micea
Adenoma Carcinoma
Adenoma
or Carcinoma
Overall Historical Incidence: All Routes
Total (%) 84/589 (14.3%) 66/589 (11.2%) 142/589 (24.1%)
Mean ± standard deviation 14.3% ± 5.4% 11.0% ± 4.4% 24.0% ± 5.3%
Range 8%-24% 4%-20% 16%-34%
a Data as of August 2017
A-12 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
TABLE A4
Summary of the Incidence of Nonneoplastic Lesions in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Yearsa
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
Disposition Summary
Animals initially in study 105 105 105 105
14-Week interim evaluation 15 15 15 15
Early deaths
Accidental death 1
Moribund 8 6 2 6
Natural deaths 16 19 8 12
Survivors
Died last week of study 1 4
Terminal euthanasia 66 63 79 68
Missing 1
Animals examined microscopically 100 100 100 100
14-Week Interim Evaluation
Alimentary System
Liver (10) (10) (10) (10)
Inflammation, focal 2 (20%) 4 (40%)
Pancreas (10) (10) (10) (10)
Infiltration cellular, lymphocyte 1 (10%)
Inflammation, chronic 1 (10%)
Genital System
Prostate (10) (10) (10) (10)
Infiltration cellular, lymphocyte 2 (20%) 1 (10%)
Hematopoietic System
Lymph node, mandibular (5) (7) (10) (8)
Hemorrhage 2 (20%)
Nervous System
Brain (10) (10) (10) (10)
Hemorrhage 1 (10%) 1 (10%)
Respiratory System
Lung (10) (10) (10) (10)
Congestion 1 (10%) 1 (10%)
Hemorrhage 2 (20%) 3 (30%) 2 (20%) 2 (20%)
Nose (10) (10) (10) (10)
Respiratory epithelium, hyperplasia 2 (20%) 1 (10%)
Urinary System
Kidney (10) (10) (10) (10)
Nephropathy, chronic progressive 1 (10%) 2 (20%) 1 (10%)
Interstitium, infiltration cellular,
lymphocyte 2 (20%) 1 (10%) 1 (10%) 1 (10%)
a Number of animals examined microscopically at the site and the number of animals with lesion
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 A-13
Peer Review Draft NOT FOR ATTRIBUTION
TABLE A4
Summary of the Incidence of Nonneoplastic Lesions in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
14-Week Interim Evaluation (continued)
Systems Examined with No Lesions Observed
Cardiovascular System
Endocrine System
General Body System
Integumentary System
Musculoskeletal System
Special Senses System
2-Year Study
Alimentary System
Esophagus (88) (87) (88) (90)
Gallbladder (73) (66) (74) (79)
Inflammation, acute 1 (2%)
Intestine large, cecum (81) (77) (84) (78)
Intestine large, colon (84) (83) (85) (84)
Intestine large, rectum (84) (85) (86) (84)
Intestine small, duodenum (77) (77) (83) (79)
Intestine small, ileum (81) (79) (85) (80)
Peyer’s patch, hyperplasia, lymphoid 1 (1%) 1 (1%) 1 (1%) 1 (1%)
Peyer’s patch, infiltration cellular,
plasma cell 1 (1%)
Intestine small, jejunum (79) (79) (82) (79)
Inflammation, granulomatous 1 (1%)
Epithelium, cyst 1 (1%)
Peyer’s patch, hyperplasia, lymphoid 1 (1%) 1 (1%)
Liver (90) (89) (90) (90)
Angiectasis 2 (2%)
Basophilic focus 1 (1%) 2 (2%) 4 (4%) 3 (3%)
Clear cell focus 28 (31%) 34 (38%) 41 (46%) 31 (34%)
Eosinophilic focus 4 (4%) 4 (4%) 8 (9%) 1 (1%)
Extramedullary hematopoiesis 2 (2%) 2 (2%) 2 (2%)
Fatty change 37 (41%) 31 (35%) 35 (39%) 35 (39%)
Fibrosis 1 (1%)
Hemorrhage 1 (1%)
Infiltration cellular, lymphocyte 2 (2%) 2 (2%)
Infiltration cellular, mixed cell 1 (1%)
Inflammation, focal 1 (1%) 1 (1%) 3 (3%)
Inflammation, chronic 2 (2%)
Inflammation, chronic active 2 (2%) 1 (1%) 1 (1%)
Mixed cell focus 2 (2%) 3 (3%) 7 (8%) 4 (4%)
Necrosis 6 (7%) 6 (7%) 4 (4%) 3 (3%)
Bile duct, cyst 2 (2%) 1 (1%)
Hepatocyte, fatty change, focal 1 (1%) 2 (2%)
Mesentery (12) (14) (13) (17)
Artery, inflammation, chronic active 1 (7%) 2 (12%)
Fat, hemorrhage 1 (7%)
Fat, inflammation, granulomatous 1 (6%)
Fat, mineral 1 (8%)
Fat, necrosis 8 (67%) 11 (79%) 12 (92%) 13 (76%)
A-14 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
TABLE A4
Summary of the Incidence of Nonneoplastic Lesions in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
2-Year Study (continued)
Alimentary System (continued)
Pancreas (87) (88) (88) (86)
Hemorrhage 1 (1%)
Infiltration cellular, lymphocyte 3 (3%) 5 (6%) 3 (3%) 1 (1%)
Infiltration cellular, mixed cell 1 (1%)
Inflammation, granulomatous 1 (1%)
Inflammation, acute 1 (1%)
Inflammation, chronic active 1 (1%)
Acinus, atrophy 1 (1%)
Duct, cyst 1 (1%) 2 (2%)
Duct, fibrosis 1 (1%)
Salivary glands (90) (89) (89) (89)
Infiltration cellular, lymphocyte 58 (64%) 59 (66%) 65 (73%) 65 (73%)
Stomach, forestomach (88) (87) (89) (87)
Cyst, squamous 1 (1%) 3 (3%)
Hyperkeratosis 1 (1%) 2 (2%)
Infiltration cellular, lymphocyte 1 (1%)
Inflammation, chronic 1 (1%)
Epithelium, hyperplasia, focal 3 (3%) 2 (2%) 2 (2%)
Epithelium, hyperplasia, diffuse 1 (1%) 1 (1%)
Stomach, glandular (87) (86) (88) (85)
Accumulation, hyaline droplet 2 (2%)
Cyst 1 (1%)
Hemorrhage 1 (1%)
Ulcer 1 (1%)
Epithelium, hyperplasia, focal 1 (1%)
Tooth (27) (26) (16) (20)
Dysplasia 26 (96%) 26 (100%) 14 (88%) 20 (100%)
Inflammation, suppurative 2 (7%) 2 (13%)
Inflammation, chronic active 1 (5%)
Cardiovascular System
Aorta (89) (89) (89) (87)
Blood vessel (1) (0) (0) (0)
Inflammation, chronic 1 (100%)
Heart (90) (89) (90) (90)
Bacteria 1 (1%) 2 (2%)
Cardiomyopathy 10 (11%) 2 (2%) 1 (1%) 2 (2%)
Inflammation, acute 1 (1%)
Inflammation, chronic active 2 (2%) 2 (2%) 1 (1%)
Thrombus 1 (1%) 2 (2%) 1 (1%)
Artery, inflammation, chronic active 1 (1%) 2 (2%) 3 (3%)
Endocardium, mineral 1 (1%)
Endothelium, hyperplasia 1 (1%) 1 (1%)
Epicardium, inflammation, chronic 1 (1%)
Epicardium, mineral 1 (1%)
Myocardium, hemorrhage 1 (1%)
Myocardium, mineral 2 (2%) 2 (2%) 1 (1%) 1 (1%)
Myocardium, necrosis 1 (1%) 2 (2%)
GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596 A-15
Peer Review Draft NOT FOR ATTRIBUTION
TABLE A4
Summary of the Incidence of Nonneoplastic Lesions in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
2-Year Study (continued)
Endocrine System
Adrenal cortex (90) (89) (89) (88)
Accessory adrenal cortical nodule 1 (1%)
Angiectasis 1 (1%)
Hyperplasia, focal 3 (3%) 1 (1%) 6 (7%) 6 (7%)
Hypertrophy, focal 2 (2%) 8 (9%) 9 (10%) 1 (1%)
Infiltration cellular, mononuclear cell 1 (1%)
Bilateral, hyperplasia, focal 1 (1%)
Bilateral, hypertrophy, focal 1 (1%) 5 (6%) 4 (4%) 1 (1%)
Subcapsular, hyperplasia 69 (77%) 72 (81%) 80 (90%) 72 (82%)
Adrenal medulla (90) (88) (88) (86)
Islets, pancreatic (88) (88) (90) (89)
Atrophy 1 (1%)
Hyperplasia 18 (20%) 20 (23%) 16 (18%) 10 (11%)
Infiltration cellular, lymphocyte 2 (2%) 1 (1%) 2 (2%)
Parathyroid gland (68) (68) (67) (66)
Cyst 2 (3%) 4 (6%) 1 (2%)
Pituitary gland (86) (85) (87) (85)
Pars distalis, angiectasis 1 (1%)
Pars distalis, cyst 3 (3%) 4 (5%) 3 (3%) 4 (5%)
Pars distalis, hyperplasia, focal 1 (1%) 2 (2%) 1 (1%)
Thyroid gland (89) (88) (88) (88)
Infiltration cellular, lymphocyte 1 (1%) 1 (1%)
General Body System
Peritoneum (1) (0) (0) (0)
Tissues NOS (0) (0) (0) (1)
Genital System
Coagulating gland (2) (2) (0) (4)
Cyst 2 (100%) 1 (50%) 3 (75%)
Bilateral, inflammation, chronic active 1 (50%)
Epididymis (90) (89) (90) (90)
Granuloma sperm 1 (1%) 1 (1%) 1 (1%) 1 (1%)
Infiltration cellular, lymphocyte 29 (32%) 17 (19%) 22 (24%) 28 (31%)
Spermatocele 1 (1%)
Bilateral, duct, atrophy 1 (1%)
Preputial gland (89) (88) (90) (89)
Infiltration cellular, lymphocyte 43 (48%) 32 (36%) 38 (42%) 33 (37%)
Inflammation, suppurative 1 (1%)
Inflammation, chronic active 1 (1%) 1 (1%)
Bilateral, hyperplasia 1 (1%)
Bilateral, duct, dilation 6 (7%) 2 (2%) 9 (10%) 2 (2%)
Duct, dilation 10 (11%) 6 (7%) 11 (12%) 4 (4%)
Duct, inflammation, chronic active 1 (1%)
Duct, necrosis 1 (1%)
Prostate (90) (87) (90) (87)
Hyperplasia, focal 1 (1%)
Infiltration cellular, lymphocyte 4 (4%) 3 (3%) 6 (7%) 9 (10%)
Inflammation, acute 1 (1%) 5 (6%)
Inflammation, chronic active 1 (1%) 1 (1%)
A-16 GSM- and CDMA-Modulated Cell Phone RFR, NTP TR 596
NOT FOR ATTRIBUTION Peer Review Draft
TABLE A4
Summary of the Incidence of Nonneoplastic Lesions in Male Mice
Exposed to GSM-Modulated Cell Phone RFR for 2 Years
Sham Control 2.5 W/kg 5 W/kg 10 W/kg
2-Year Study (continued)
Genital System (continued)
Seminal vesicle (90) (88) (90) (90)
Dilation 4 (4%) 4 (5%) 5 (6%) 4 (4%)
Hyperplasia 1 (1%)
Inflammation, chronic active 1 (1%) 1 (1%)
Bilateral, atrophy 1 (1%)
Bilateral, dilation 27 (30%) 26 (30%) 23 (26%) 29 (32%)
Bilateral, fibrosis 1 (1%)
Bilateral, inflammation, acute 1 (1%)
Bilateral, inflammation, chronic 1 (1%) 1 (1%)
Bilateral, inflammation, chronic active 1 (1%)
Testis (90) (88) (90) (90)
Bilateral, germ cell, degeneration 1 (1%)
Germ cell, degeneration 2 (2%) 1 (1%) 1 (1%)
Hematopoietic System
Bone marrow (90) (88) (90) (90)
Hypercellularity 3 (3%) 2 (2%) 3 (3%)
Lymph node (6) (8) (7) (9)
Bronchial, infiltration cellular, mixed cell 1 (11%)
Iliac, erythrophagocytosis 1 (13%)
Iliac, hemorrhage 1 (13%)
Iliac, hyperplasia, lymphoid 1 (13%) 2 (22%)
Iliac, infiltration cellular, histiocyte 2 (25%) 2 (29%)
Iliac, infiltration cellular, plasma cell 1 (11%)
Iliac, pigment 2 (29%)
Lumbar, hemorrhage 1 (11%)
Mediastinal, hyperplasia, lymphoid 1 (11%)
Mediastinal, infiltration cellular,
plasma cell 1 (13%)
Pancreatic, hyperplasia, lymphoid 2 (29%)
Renal, hemorrhage 1 (17%) 1 (14%)
Renal, hyperplasia, lymphoid 1 (14%)
Renal, infiltration cellular, mixed cell 1 (11%)
Lymph node, mandibular (72) (61) (63) (60)
Hemorrhage 1 (2%)
Hyperplasia, lymphoid 2 (3%)
Infiltration cellular, histiocyte 1 (1%) 1 (2%)
Lymph node, mesenteric (85) (82) (88) (83)
Erythrophagocytosis 1 (1%) 5 (6%) 4 (5%) 1 (1%)
Hemorrhage 10 (12%) 11 (13%) 7 (8%) 13 (16%)
Hyperplasia, lymphoid 4 (5%) 2 (2%) 2 (2%) 5 (6%)
Infiltration cellular, histiocyte 8 (9%) 7 (9%) 5 (6%) 4 (5%)
Infiltration cellular, mixed cell 2 (2%)
Infiltration cellular, plasma cell 1 (1%) 1 (1%) 1 (1%) 1 (1%)
Spleen (87) (88) (89) (88)
Extramedullary hematopoiesis 15 (17%) 15 (17%) 13 (15%) 12 (14%)
Hyperplasia, lymphoid 5 (6%) 2 (2%) 5 (6%) 3 (3%)
White pulp, atrophy 1 (1%)
Thymus (75) (83) (81) (72)
Atrophy 11 (15%) 16 (19%) 4 (5%) 14 (19%)
Cyst 11 (15%) 16 (19%) 26 (32%) 15 (21%)
Hemorrhage 1 (1%) 1 (1%) 1 (1%)
Infiltration cellular, histiocyte 1 (1%)
