Radiation and Public Perception - ACS Publications - American

---

Downloaded via TUFTS UNIV on July 11, 2018 at 23:03:56 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

9 Cancer Incidence and Mortality after Iodine-131 Therapy for Hyperthyroidism Per Hall and Lars-Erik Holm Department of General Oncology, Radiumhemmet, Karolinska Hospital, Stockholm, Sweden

Cancer risk was studied in 10,552 Swedish hyperthyroid pa­ tients treated with I between 1950 and 1975. Patients were followed for an average of 15 years (range 1-35 years) and were matched with the Swedish Cancer Register (SCR) and the Swedish Cause of Death Register (SCDR). The overall stan­ dardizedincidence ratio (SIR) was 1.06 [95% confidence inter­ val(CI) = 1.01-1.11], and the overall standardized mortality ratio(SMR)was 1.09 (95% CI = 1.03-1.16). The stomach was the only site for which cancer risk increased over time (p < 0.05) and with increasing activity of I administered (p = not significant). No increased incidence of leukemia was found, which adds further support to the view that a radiation dose deliv­ ered gradually over time is less carcinogenic than the same to­ tal dose received over a short time. A possible excess owing to radiation was suggested only for stomach cancer. 131

131

IoDINE-131

THERAPY FOR HYPERTHYROD IS IM WAS FIRST N ITRODUCED in

the 1940s (1, 2) and, in many clinics, is considered to be the treatment of choice for this disease, largely because serious side effects are uncommon. However, concern still exists as to the possible carcinogenic effect of I. Reports on increased risks of breast (3, 4) and thyroid cancer (3) among hyperthyroid patients treated with I are in contrast to others 131

131

0065-2393/95/0243-0103$08.00/0 © 1995 American Chemical Society

Young and Yalow; Radiation and Public Perception Advances in Chemistry; American Chemical Society: Washington, DC, 1995.

104

RADIATION A N D PUBLIC PERCEPTION

that failed to detect increased cancer risks among such patients (5-8). In a recent study of 1762 hyperthyroid women, 80% of whom received I, overall mortality was significantly elevated, but the standardized mortality ratio (SMR) for cancer did not differ from unity (4). Although several studies of patients treated with I were conducted, no clear pattern of risk was observed. Leukemia was never found to be in excess following I therapy for hyperthyroidism. However, the risk of leukemia was elevated in three studies of thyroid cancer patients treated with larger doses of I (9-11), but the number of leukemias was small in these studies with 2, 3, and 4 cases, respectively. The purpose of the present study was to analyze the incidence and mortality of cancer and leukemia in a large Swedish population treated with I for hyperthyroidism. 131

131

131

131

131

Subjects and Methods The patients were admitted to seven university hospitals between 1950 and 1975 and were all under the age of 75 years at the time of first I treatment. Ninety-four cases were excluded due to insufficient information on names and dates of birth. Mean age at the time of first I treatment was 57 years (range 13-74 years). Case records from the hospitals were used to obtain information on thyroid disorder and treatment. Some patients had previously received external radiotherapy toward the head-neck region (3%), thyroid hormone supplement for nontoxic goiter (2%), surgery for nontoxic goiter (3%), antithyroid drugs (24%), or surgery for hyperthyroidism (14%). Fifty-nine percent of the patients received only one I treatment, and 41% received two or more treatments. The mean total activity administered was 506 M B q (range 37-19,980 MBq). A total I activity of 220 M B q or less (mean 150 MBq) was given to 30% of the patients, 221-480 M B q (mean 315 MBq) to 38%, and >480 MBq (mean 1063 MBq) to 32%. The mean number of treatments in each group was 1.1, 1.5, and 2.3, respectively. The dose to the thyroid gland aimed at 60-100 Gy. In calculating mean organ doses, the International Commission on Radiological Protection (ICRP) tables (12), mean administered activity of I, and mean 24-hour uptake of I were used. The stomach wall received a mean dose of 0.25 Gy, and the urinary bladder wall and small intestine each received 0.14 Gy. No other organ received more than 0.10 Gy. The mean total body dose was estimated to be 0.08 Gy and the mean bone marrow dose to be 0.05 Gy. 131

131

131

131

131

131

Young and Yalow; Radiation and Public Perception Advances in Chemistry; American Chemical Society: Washington, DC, 1995.

9.

HALL & HOLM

Cancer Incidence and Mortality

105

The total cohort was matched with the Swedish Cancer Register (SCR) from 1958 to 1985 and the Swedish Cause of Death Register (SCDR) from 1952 to 1986 to identify cancer incidence and mortality in the cohort. The SCR receives notifications on newly diagnosed can­ cers, not only from clinicians but also from pathologists and cytologists. More than 96% of all cancers in Sweden are reported to the SCR (13). All deaths are certified by a physician. The matching con­ cept for both record linkages was the unique identification number that is given to all individuals in Sweden. Patients were considered to be at risk 1 year after the initial I treatment until death or end of follow-up period. Attained age, sex, and calendar year were taken into consideration when the expected incidence and mortality were calculated using data from the SCR and the SCDR. The expected incidence and mortality were thus based on findings from the whole Swedish population (i.e., indirect standard­ ization). Standardized incidence ratio (SIR) and SMR were defined as the ratio between observed and expected numbers and were calcu­ lated using the methods suggested by Breslow and Day (14). The 95% CI was determined by assuming the observed number of cases to be distributed as a Poisson variable. 131

Results Within the first year of follow-up, 345 patients died, and the analyses were thus based on 10,207 patients. The patients were followed for an average of 15 years (range: 1-35 years). A total of 1543 cancers were observed more than 1 year after ex­ posure (SIR = 1.06; 95% CI = 1.01-1.11). More than 10 years after exposure, 830 cancers were seen (SIR = 1.10; 95% CI = 1.02-1.17; Table I), and significantly elevated risks were seen for cancer of the stomach (SIR = 1.33; 95% CI = 1.01-1.71; η = 58), kidney (SIR = 1.51; 95% CI = 1.06-2.08; η = 37), and nervous system (SIR = 1.63; 95% CI = 1.10-2.32; η = 30). Deaths due to cancer or leukemia were observed in 977 cases. Fifty-three percent of the diagnoses were confirmed at autopsy, and an additional 46% were reported from hospitals but not based on au­ topsy findings. The overall SMR for malignant tumors and leukemia was 1.09 (95% CI = 1.03-1.16). Sites significantly elevated after 10 years of follow-up were cancer of the stomach (SMR = 1.41; 95% CI = 1.06-1.85; η = 54; Table I) and lung (SMR = 1.80; 95% CI = 1.39-2.31; η = 63). A total of 37 leukemias occurred more than 1 year after exposure, and the SIR was 0.81 (95% CI = 0.57-1.12; Table II). Chronic lym­ phatic leukemia (CLL), a condition not known to be increased after

Young and Yalow; Radiation and Public Perception Advances in Chemistry; American Chemical Society: Washington, DC, 1995.

106

RADIATION A N D PUBLIC PERCEPTION

Table I. Observed Number of Cancers and Deaths from Cancer, 10 Years after Exposure, in 10,207 Hyperthyroid Patients Receiving I, SIR, SMR, and 95% CIforSelected Organs Mortality Incidence Cancer Site 95% CI Obsd SMR SIR Obsd 95% Cl 1.06- -1.85 Stomach 1.41 58 54 1.33 1.01- 1.71 1.39- -2.31 Lung 50 1.80 1.17 0.87- 1.54 63 Breast 134 1.04 39 0.77 0.54--1.05 0.87- 2.33 Kidney 0.90 0.51- -1.49 37 1.51 15 1.06- •2.08 Bladder 28 1.13 8 0.71 0.31^1.40 0.75- 1.63 Nervous system 30 0.97 0.42--1.91 8 1.63 1.10- -2.32 Thyroid 9 0.08--2.37 2 0.66 1.32 0.61- 2.50 131

Total*

830

1.10

1.02- 1.17

510

1.14

1.04--1.24

Also includes all sites not listed in the table.

e

Table Π. Observed Number of Leukemias 1 Year after Exposure in 10,207 Hyperthyroid Patients Receiving I, SIR, and 95% CI in Relation to Type 1 of Leukemia Type of Leukemia Observed SIR 95% C7 Non-CLL 0.55-1.25 0.85 25 0.39-1.30 CLL 12 0.75 0.57-1.12 All leukemias 0.81 37 13I

irradiation, had approximately the same risk (SIR = 0.75) as non-CLL (SIR = 0.85). Risk of leukemia did not vary by sex, age, time, or dose of I . Table III shows the cancer incidence for some selected organs in relation to follow-up. Except for stomach cancer there were no sig­ nificant time trends for any of the cancer sites or for all cancers com­ bined. The mortality for stomach cancer and for all cancers combined in­ creased with the increasing administration of I activity (Table IV). These trends, however, were not statistically significant. No trend was seen for lung cancer, although the highest risk was seen in patients receiving >481 MBq. Similar patterns were seen when incidences were analyzed. The elevated risk for thyroid cancer (n = 8) among patients given >481 M B q included six thyroid cancers diagnosed during the first 5 years of follow-up. 13I

131

Discussion Patients receiving I therapy for hyperthyroidism had an overall can­ cer incidence (6%) and cancer mortality (9%) slightly greater than ex131

Young and Yalow; Radiation and Public Perception Advances in Chemistry; American Chemical Society: Washington, DC, 1995.

Young and Yalow; Radiation and Public Perception Advances in Chemistry; American Chemical Society: Washington, DC, 1995.

0

713 1.02 Total "Includes cancer sites not listed in the table.

131

0.95-1.10 394 1.10

0.99-1.21

436

1.10

1.00-1.20

Table ΠΙ. Observed Number of Deaths from Cancer 1 Year after Exposure in 10,207 Hyperthyroid Patients Receiving I, SIR, and 95% CI in Relation to Duration of Follow-up Yearsof Follow-up 1-9 10-14 >15 Observed SIR 95% CI Cancer Site Observed SIR 95% CI 95% CI SIR Observed 34 0.77 0.53-1.08 1.07-2.14 23 Stomach 1.10 1.54 0.70-1.65 35 1.49 1.12-1.93 55 20 0.90-1.90 Lung 0.99 30 0.61-1.53 1.33 1.01 0.85-1.20 135 0.82-1.33 64 1.01 Breast 0.78-1.29 70 1.06 29 1.26 0.85-1.81 23 0.60-1.83 Kidney 1.95 14 1.10 1.24-2.93 1.11 0.70-1.66 23 14 0.58-1.76 Bladder 1.22 14 1.05 0.67-2.05 0.89 0.53-1.41 18 17 Nervous system 1.74 0.69-2.21 1.30 1.02-2.78 13 0.57-2.38 9 1.25 5 0.32-2.39 Thyroid gland 1.49 4 1.15 0.49-3.48

3

"1

ο

a-

1

I

Ο

ο r 2

r

> r

Young and Yalow; Radiation and Public Perception Advances in Chemistry; American Chemical Society: Washington, DC, 1995.

a

0

0.98 274 Total Also includes all sites not listed in the table.

131

0.86-1.10

388

1.09

0.98-1.19

315

1.13

1.01-1.26

Table TV. Observed Number of Deaths from Cancer 1 Year after Exposure in 10,207 Hyperthyroid Patients Receiving I, SMR, and 95% CI in Relation to Administered Activity 481 MBq Observed SMR 95% CI Cancer Site Observed SMR 95% CI 95% CI Observed SMR 1.00 25 0.65-1.48 0.84-1.72 Stomach 35 1.05 1.22 0.73-1.46 33 Lung 33 1.20 0.83-1.69 33 1.12-2.11 0.95 1.56 0.66-1.34 41 Breast 20 0.58 0.35-0.89 46 0.57-1.25 1.08 0.87 0.79-1.44 28 Female genital organs 36 0.83-1.65 41 1.19 0.92-1.81 1.11 0.79-1.50 37 1.31 Male genital organs 5 0.62 0.20-1.44 18 0.37-1.69 1.68 0.86 0.99-2.65 8 Kidney 12 1.17 0.61-2.05 20 0.69-2.22 1.53 1.30 0.94-2.37 13 Bladder 5 0.82 0.26-1.90 9 1.12 0.63 0.17-1.62 4 0.51-2.13 Nervous system 6 1.01 0.37-2.20 4 0.21-1.96 0.56 0.76 0.15-1.44 4 2 1.08 0.13-3.91 1.82-8.30 Thyroid 3 1.25 4.21 0.26-3.64 8 Lymphomas 5 0.66 0.21-1.54 10 1.04 0.54 0.15-1.40 0.50-1.91 4

I

δ

3

w

r ο *ό m

α

ζ σ *ΰ

>

ο

α

I—*

9.

HALL & HOLM

Cancer Incidence and Mortality

109

pected in the general population. Only stomach cancer revealed both increased incidence and mortality figures and was also the only site where the risk increased with increasing dose. The stomach wall received the highest dose, 0.25 Gy, and it is possible that the I explains the excess. The increased medical surveillance of the patients could have led to detection of cancers not causing symptoms, thus indicating a higher risk for cancer. Many studies showing elevated cancer risks after exposure to low doses of ionizing radiation have been based on studies of children. Only 5% of the patients in the present study were younger than 40 years at the time of I exposure. The risk associated with protracted whole-body exposure of 0.08 Gy from I may not be comparable to the high dose rate situations in other studies in which increased cancer mortality was observed. The latency period for solid tumors is usually at least 10 years (15, 16). The mean follow-up period was 18 years among the 7818 patients surviving 10 years, and it is possible that this follow-up period was not sufficiently long to detect an increased cancer mortality due to radiation exposure. The study population represented a select group, as young patients and women of fertile age generally were not likely to receive I because of the potential hazards of such therapy. The study group also consisted of patients unfit for surgery because of cardiac or respiratory diseases. Some of the risk factors for these diseases, such as diet and smoking, might have influenced the cancer risk. A reference population of nonirradiated patients with hyperthyroidism would have been preferable, instead of the country as a whole, even if this method also would have included a selection bias because there was always a reason why some patients were given radiotherapy and others not. The strengths of the study were the few patients lost to follow-up, the accuracy of the SCR and SCDR, and data on individually administered I activity. The puzzling finding that more patients died of lung cancer (n = 63) than received the diagnosis (n = 50) is explained by the fact that the autopsy often defined a more specified diagnosis than was indicated in the clinical records. It was shown that when an autopsy, disproving the original diagnosis, was delayed, the death certificate was not amended (17). The SMR for cancer of the stomach and lung was highest among those receiving the highest I activity. In a study by Darby et al. (18) patients receiving radiotherapy for ankylosing spondylitis, a benign condition of the spine, showed increased mortality from esophagus, stomach, and lung cancer. These organs received >10 Gy, and no decreasing risk with increasing age at exposure was noticed. A selection bias could partly explain our findings because the prevalence 131

131

131

131

131

131

Young and Yalow; Radiation and Public Perception Advances in Chemistry; American Chemical Society: Washington, DC, 1995.

110

RADIATION A N D PUBLIC PERCEPTION

of smokers is expected to be higher among patients with hyperthyroidism than in the normal population (19). In a study of thyroid cancer patients (11), an elevated risk for stomach cancer was also found in those treated with I in contrast to thyroid cancer patients receiving other types of treatment. The risk increased over time, and the mean dose to the stomach was 2.1 Gy compared with 0.25 Gy in the present study. A mean dose of approximately 0.5 Gy to the thyroid gland was received by 35,000 patients examined with I (20). No increased risk related to the exposure was found. The dose to the thyroid in the present study aimed at 60-100 Gy, and it is likely that this dose had a cell-killing instead of a carcinogenic effect because no increased mortality was seen in patients followed for more than 10 years after exposure. Because the cure rate of thyroid cancer is high, incidence data instead of mortality data should be used. However, no significantly increased incidence of thyroid cancer was noticed. Patients receiving the highest I activity had the highest cancer mortality and probably also the most severe hyperthyroidism, as reflected by the higher number of treatments. In a recent study of the present cohort, elevated risks for most causes of death were found among those receiving the highest amount of I (21). It was concluded that the underlying disease rather than the I therapy was the reason for this. If these patients also had a cancer, although not the cause of death, they would have been reported as dying from a malignant disease. This probably contributes to our findings of a slighdy increased overall risk in the group receiving the highest I activity. The induction of leukemia by ionizing radiation has been well documented, and excess mortality seems to reach a peak within 10 years after exposure (15, 16). Among the atomic bomb survivors, elevated risks were found among those with an absorbed dose to the bone marrow of >0.5 Gy but not among those with absorbed doses lower than 0.5 (22). In patients treated with X-ray for ankylosing spondylitis, elevated risks of leukemia were found (18), and excess mortality became detectable within 2 years of exposure and peaked within the first 5 years. In our study 37 leukemias were found more than 1 year after exposure, and SIR was 0.81. Using data from the atomic bomb survivors (22) of an excess relative mortality per Gy organ-absorbed dose of 5.21 and an estimated bone marrow dose of 0.05 Gy, the SMR in our study would be 1.26. The lack of correspondence is probably explained by the large difference in dose rate, because the biological half-life of I is at least 8 days. Our observations suggest that (a) low doses of ionizing radiation are less effective in inducing cancer than higher doses, (b) the protracted nature of the I exposure makes the isotope a less effective 131

131

131

131

131

131

131

131

Young and Yalow; Radiation and Public Perception Advances in Chemistry; American Chemical Society: Washington, DC, 1995.

9.

HALL & HOLM

Cancer Incidence and Mortality

111

carcinogen probably because of the opportunity of cellular repair of the radiation damage, (c) the dose was so low that a detectable increase in cancer or leukemia was unlikely, and (d) extrapolating from high doses to low doses and dose rates does not seem to underestimate the risk.

Acknowledgments This study was performed in cooperation with the following: G . Lundell, Department of General Oncology, and K. Wiklund, Department of Cancer Epidemiology, Radiumhemmet, Karolinska Hospital, Stockholm, Sweden; M . Lidberg, Department of Hospital Physics, South Hospital, Stockholm, Sweden; E . Cederquist and J . Tennvall, Department of General Oncology, University Hospital, Lund, Sweden; G. Bjelkengren, Department of General Oncology, and U.-B. Ericsson, Department of Internal Medicine, Malmo General Hospital, Malmô, Sweden; G . Berg, Department of General Oncology, and S. Lindberg, Nuclear Medicine Division, Sahlgren's Hospital, Gothenburg, Sweden; H . Wicklund, Department of General Oncology, University Hospital, Uppsala, Sweden; A. Hallquist and L . - G . Larsson, Department of General Oncology, University Hospital, Umeâ, Sweden; and J. D . Boice, Jr., Epidemiology and Biostatistics Program, Division of Cancer Etiology, National Cancer Institute, Bethesda, M D . The study was supported by Public Health Service Contract No. N01-CP-51034 from the National Cancer Institute, Bethesda, M D . The authors wish to thank Elisabeth Bjurstedt for valuable assistance in various aspects of our study.

References 1. Hamilton, J. G.; Lawrence, J. H. J. Clin. Invest. 1942, 21, 624. 2. Hertz, S. Roberts, A. J. Clin. Invest. 1942, 21, 624. 3. Ron, E.; Curtis, R.; Hoffman, D. Α.; Flannary, J. T. Br. J. Cancer 1984, 49, 87-92. 4. Goldman, M. B.; Maloof, F.; Monson, R. R.; Aschengrau, Α.; Cooper, S.; Ridgway, E. C. Am. J. Epidemiol. 1988, 127, 969-980. 5. Hoffman, D. Α.; McConahey, W. M. J. Natl. Cancer Inst. 1983, 70, 6367. 6. Hoffman, D. A. In Radiation Carcinogenesis: Epidemiology and Biolog­ ical Significance; Boice, J. D., Jr.; Fraumeni, J. F., Jr., Eds.; Raven Press: New York, 1984; pp 273-280. 7. Holm, L.-E. In Radiation Carcinogenesis: Epidemiology and Biological Significance; Boice, J. D., Jr.; Fraumeni, J. F., Jr., Eds.; Raven Press: New York, 1984; pp 263-271. 8. Holm, L.-E.; Hall, P.; Wiklund, K. Lundell, G. Berg, G.; Bjelkengren, G.; Cederquist, E.; Ericsson, U.-B.; Hallquist, Α.; Larsson, L.-G.; Lid­;

D.

;

;

Young and Yalow; Radiation and Public Perception Advances in Chemistry; American Chemical Society: Washington, DC, 1995.

112

RADIATION A N D PUBLIC PERCEPTION

berg, M.; Lindberg S.; Tennvall, J.; Wicklund, H.; Boice, J. D., Jr. J. Natl. Cancer Inst. 1991, 83, 1072-1077. 9. Brincker, H.; Hansen, H. S.; Andersen, A. P. Br. J. Cancer 1973, 28, 232-237. 10. Edmonds, C. J.; Smith, T. Br. J. Radiol. 1986, 59, 45-51. 11. Hall, P.; Holm, L.-E.; Lundell, G.; Bjelkengren, G.; Larsson, L.-G.; Lindberg, S.; Tenvall, J.; Wicklund, H.; Boice, J. D., Jr. Br. J. Cancer 1991, 64, 159-163. 12. International Commission on Radiological Protection; Radiation Dose to Patients from Radiopharmaceuticals; Annals of the ICRP: Publication 53; Pergamon Press: Oxford, United Kingdom, 1988; Vol. 18. 13. Mattsson, B.; Wallgren, A. Acta Radiol. Oncol. 1984, 23, 305-313. 14. Breslow, Ν. E.; Day, Ν. E. In The Design and Analysis of Cohort Stud­ ies; International Agency for Research on Cancer: Lyon, France, 1987; Vol. II. 15. United Nations Scientific Committee on the Effects of Atomic Radiation; Sources, Effects and Risks of Ionizing Radiation; 1988 Report to the Gen­ eral Assembly, with annexes; United Nations: New York, 1988. 16. Committee on the Biological Effects of Ionizing Radiations; Health Effects of Exposure to Low Levels of Ionizing Radiation;BEIRV;National Acad­ emy Press: Washington, DC, 1990. 17. Mattsson, B.; Rutqvist, L.-E. Radiother. Oncol. 1985, 4, 63-70. 18. Darby, S. C.; Doll, R.; Gill, S. K. Smith, P. G. Br. J. Cancer 1987, 55, 179-190. 19. Bartalena, L.; Martino, E.; Marcocci, C., et al. J. Endocrinol. Invest. 1989, 12, 733-737. 20. Holm, L.-E.; Wiklund, Κ. E.; Lundell, G. E., et al. J. Natl. Cancer Inst. 1989, 81, 302-306. 21. Hall, P.; Lundell, G.; Holm, L.-E. Acta Endocrinol. 1993, 128, 230. 22. Shimizu, Y.; Kato, H.; Schull, W. J. Radiat. Res. 1990, 121, 120-141. ;

for review October 2, 1992.

RECEIVED

ACCEPTED

revised manuscript March

25, 1993.

Young and Yalow; Radiation and Public Perception Advances in Chemistry; American Chemical Society: Washington, DC, 1995.