- Email: [email protected]
Cancer incidence and mortality after radioiodine treatment for hyperthyroidism: a population-based cohort study
ARTICLES
Cancer incidence and mortality after radioiodine treatment for hyperthyroidism: a population-based cohort study Jayne A Franklyn, Patrick Maisonneuve, Michael Sheppard, Joan Betteridge, Peter Boyle
Summary Background Radioiodine is used increasingly as first-line treatment for hyperthyroidism, but concerns remain about subsequent risk of cancer, especially in those treated at a young age. We investigated cancer incidence and mortality in patients treated with radioiodine for hyperthyroidism. Methods We did a population-based study in 7417 patients treated in Birmingham, UK, between 1950 and 1991. We compared details of all cancer diagnoses and deaths in 1971–91 from the UK Office for National Statistics with data on cancer incidence and mortality for England and Wales specific for age, sex, and period. Findings During 72 073 person-years of follow-up, 634 cancer diagnoses were made, compared with an expected number of 761 (standardised incidence ratio [SIR] 0·83 [95% CI 0·77–0·90]). The relative risk of cancer mortality was also decreased (observed cancer deaths 448, expected 499; standardised mortality ratio [SMR] 0·90 [0·82–0·98]). Incidence of cancers of the pancreas, bronchus, trachea, bladder, and lymphatic and haemopoietic systems was lowered. Mortality from cancers at all these sites was also reduced but findings were significant only for bronchus and trachea. There were significant increases in incidence and mortality for cancers of the small bowel (SIR 4·81 [2·16–10·72], SMR 7·03 [3·16–15·66]) and thyroid (SIR 3·25 [1·69–6·25], SMR 2·78 [1·16–6·67]), although absolute risk of these cancers was small. Interpretation The decrease in overall cancer incidence and mortality in those treated for hyperthyroidism with radioiodine is reassuring. The absolute risk of cancers of the small bowel and thyroid remain low, but the increased relative risk shows the need for long-term vigilance in those receiving radioiodine. Lancet 1999; 353: 2111–15
Department of Medicine, University of Birmingham, Birmingham, UK (Prof J A Franklyn FRCP, M Sheppard FRCP, J Betteridge RGN); and Division of Epidemiology and Biostatistics, European Institute of Oncology, Milan, Italy (P Maisonneuve PhD, Prof P Boyle PhD) Correspondence to: Prof Jayne A Franklyn, Department of Medicine, Queen Elizabeth Hospital, Edgbaston, Birmingham B15 2TH, UK
THE LANCET • Vol 353 • June 19, 1999
Introduction Radioiodine is used increasingly as first-line therapy for hyperthyroidism.1 Even though it has been used for this purpose for more than 50 years, concerns among patients and doctors remain about the subsequent risk of malignant disorders, especially among those treated at a young age.2 These concerns have been heightened by reports of increased incidence of cancers after the Chernobyl nuclear accident, especially for thyroid cancer among children exposed at younger than 10 years to the main contaminating isotope iodine-131.3,4 Although there have been several studies of cancer risk in patients treated with radioiodine for hyperthyroidism, results have been conflicting. Hall and colleagues 5 studied the incidence of leukaemia in a large population of Swedish patients exposed to iodine-131 during diagnostic scanning or treatment of hyperthyroidism or thyroid cancer. No significant excess risk of leukaemia was seen, but the risk of stomach cancer increased with time and increasing dose of radioactivity.6 Other studies that compared cancer risk in those treated by surgery and with radioiodine for hyperthyroidism7–9 showed no difference in cancer incidence or mortality at these or other specific sites. The small size of most studies and the low incidence of thyroid cancer has meant that an increase in thyroidcancer incidence after radioiodine therapy has not been convincingly confirmed. By contrast, a small increase of thyroid-cancer deaths was reported in a Swedish cohort treated for hyperthyroidism,10 and an analysis of patients treated with radioiodine in centres throughout the USA in 1946–64 also reported an increase in thyroid-cancer mortality.11 We investigated cancer incidence and mortality in patients treated with radioiodine for hyperthyroidism.
Patients and methods Patients All patients who had been treated for hyperthyroidism with radioiodine in the West Midlands region of the UK between 1950 and 1991 were eligible for inclusion. We recorded details of the date and dose of radioiodine treatment, together with patients’ demographic details, on the computerised Birmingham Thyroid Follow-up Register, 12,13 which was established to ensure regular biochemical testing and detection of hypothyroidism. The initial group comprised 8468 patients. Demographic details for these patients were sent to the UK Office for National Statistics (ONS) for tracing on the National Health Service Central Register. 452 patients were excluded because they had died before 1971. Of the remaining 8016 patients, 599 (7·5%) were excluded for the following reasons: one had no known date of birth, 518 not found on register, 32 emigrated from the UK, seven lived in Northern Ireland, 32 were not registered with a primary-care physician, seven died but had no record of cause of death, and two had no record of date of radioiodine therapy. The final cohort included 7417 patients. The characteristics of the excluded patients were similar to those remaining in the study. Because cancer-incidence data for England and Wales before 1971 could not be obtained from the ONS, and because cancer
2111
ARTICLES
registrations after 1991 are incomplete, follow-up (for cancer incidence and mortality) was restricted to 1971–91. Details of all cancer diagnoses and underlying causes of death during this follow-up period were obtained from the ONS and coded according to the ninth revision of the International Classification of Diseases.14 Cancer diagnoses had been collated by the ONS from data collected by the UK Regional Cancer Registries, whereas causes of death were collated independently from the national death-certificate scheme. Annual cancer-incidence data for England and Wales, specific for age and sex, were extracted from International Agency for Research on Cancer publications15–19 and cancer-mortality rates from the WHO databank. We compared site-specific data on cancer incidence and mortality from follow-up with data from England and Wales. This comparison was made in preference to comparison with incidence and mortality data from the West Midlands region so that case and control data would be collated by the ONS and because regional control data were incomplete in terms of site-specific and period-specific cancer mortality. In addition, about 20% of patients had moved from the region before completion of the study. We calculated the expected number of cancer diagnoses or deaths by multiplying the number of person-years in each stratum of age, sex, and calendar year by the corresponding incidence or mortality rates specific for age, sex, and period for England and Wales. The standardised incidence ratio (SIR, ratio of observed to expected number of cancer diagnoses) and standardised mortality ratio (SMR, ratio of observed to expected deaths due to cancer) were used as estimates of relative risk. 95% CI were calculated on the assumption that the observed number of cancer cases and deaths followed a Poisson distribution. We used multivariate Poisson regression to assess trends in incidence and mortality according to time from treatment, age at treatment, and cumulative dose of radioiodine given. Data analysis was done with SAS version 6.12.
Results Characteristics of the cohort are shown in table 1. Most patients (84·9%) received one dose of radioiodine (mean age at first treatment 56·6 years [SD 12·7]). 634 diagnoses of cancer were made in 613 patients between 1971 and 1991 (table 2). The expected number of cancer diagnoses for the total number of person-years at risk Site of cancer
Number of diagnoses Observed
Number of patients
Mean (SD) cumulative dose of iodine-131 (MBq)
Mean (SD) age (years)
All patients
7417
308 (232)
56·6 (12·7)
72 073
Sex Male Female
1228 (16·6%) 6189 (83·4%)
305 (214) 308 (235)
54·1 (12·2) 57·1 (12·8)
11 514 60 559
Cumulative dose of iodine-131 (MBq) ⭐220 3657 (49·3%) 221–480 2521 (34·0%) ⭓481 1228 (16·6%) Unknown 11 (0·1%)
162 (43) 328 (66) 701 (300) ··
55·7 (15·1) 57·6 (13·0) 57·5 (12·0) 55·5 (13·0)
35 826 26 271 9888 89
Number of doses 1 2 3 Unknown
6297 (84·9%) 925 (12·5%) 184 (2·5%) 11 (0·1%)
260 (157) 525 (289) 849 (533) ··
57·0 (12·8) 54·5 (12·3) 53·6 (11·4) 55·5 (15·1)
60 424 9507 2052 89
Period of first treatment 1950–59 1960–69 1970–79 1980–89 1990–91
439 (5·9%) 1526 (20·6%) 2030 (27·4%) 2757 (37·2%) 665 (8·9%)
519 (412) 351 (257) 241 (196) 287 (174) 359 (200)
53·5 (7·9) 55·9 (9·1) 57·6 (11·0) 57·0 (14·7) 56·0 (17·0)
5616 21 415 27 229 17 160 653
Age at first treatment (years) ⭐49 2141 (28·9%) 50–59 2270 (30·6%) 60–69 1769 (23·9%) ⭓70 1237 (16·6%)
289 (209) 312 (249) 317 (240) 321 (221)
41·8 (7·0) 54·4 (2·8) 64·2 (2·9) 75·6 (4·8)
23 047 26 370 15 520 7136
Duration of follow-up (years) * 1 5452 (73·5%) 2–9 6344 (85·5%) 10–19 3760 (50·7%) ⭓20 1356 (18·3%)
279 (189) 288 (207) 307 (255) 354 (273)
57·1 (13·8) 56·7 (12·6) 55·0 (9·5) 51·9 (7·4)
5119 36 342 22 509 8103
*Between 1971 and 1991, when cancer-incidence data were available (eg, a patient who was first treated in 1968 will not contribute to 1-year follow-up data because he would have had 3 years of follow-up when observation began).
Table 1: Demographic and treatment characteristics of 7417 patients treated with radioiodine for hyperthyroidism
(72 073) was 761. The relative risk of mortality from cancer was reduced. Although a reduction in cancer incidence was seen in men (SIR 0·65 [95% CI 0·54–0·79], p<0·0001) and women (SIR 0·85 [0·78–0·92], p<0·001), a significant reduction in cancer
SIR (95% CI)
p
Expected 0·19
188 39 23 1·2
0·90 (0·77–1·04) 0·97 (0·71–1·34) 0·47 (0·26–0·85) 4·81 (2·16–10·72)
68 65
112 106
222 132
Genitourinary organs Total Bladder
97 18
Other sites* Total Brain Thyroid
15 4 9
Lymphatic and haemopoietic Total Lymphoma Leukaemia
Respiratory and intrathoracic organs Total Bronchus and trachea Bone, connective tissue, skin, and breast Total Breast
All sites
5
9·0
169 38 11 6
Number of deaths Observed
0·55 (0·23–1·33)
Lip, oral cavity, and pharynx Digestive organs and peritoneum Total Stomach Pancreas Small bowel
Person-years of follow-up
SMR (95% CI)
p
Expected
2
5·4
0·37 (0·09–1·48)
0·16
0·16 0·86 0·01 0·0001
151 39 17 6
157 37 24 0·8
0·96 (0·82–1·12) 1·05 (0·77–1·44) 0·72 (0·45–1·16) 7·03 (3·16–15·66)
0·64 0·76 0·18 0·0001
0·60 (0·48–0·77) 0·61 (0·48–0·78)
0·01 0·0001
79 78
104 100
0·76 (0·61–0·95) 0·78 (0·62–0·97)
0·01 0·03
226 124
0·98 (0·86–1·12) 1·07 (0·90–1·27)
0·77 0·45
74 69
81 74
0·92 (0·73–1·15) 0·93 (0·74–1·18)
0·46 0·58
128 28
0·76 (0·62–0·92) 0·63 (0·40–1·00)
0·006 0·05
62 7
78 14
0·80 (0·62–1·02) 0·49 (0·23–1·03)
0·07 0·06
12 7·4 2·8
1·28 (0·77–2·13) 0·54 (0·20–1·44) 3·25 (1·69–6·25)
0·33 0·22 0·0004
14 8 5
1·61 (0·89–2·55) 1·20 (0·60–2·40) 2·78 (1·16–6·67)
0·12 0·60 0·02
24 12 10
39 16 13
0·61 (0·41–0·91) 0·73 (0·41–1·28) 0·75 (0·40–1·39)
0·02 0·27 0·36
25 7 13
30 11 11
0·84 (0·57–1·24) 0·64 (0·31–1·34) 1·17 (0·68–2·01)
0·38 0·23 0·58
634
761
0·83 (0·77–0·90)
0·0001
448
499
0·90 (0·82–0·98)
0·02
8·7 6·7 1·8
*International Classification of Diseases (ninth revision) codes 190–194, excluding 195–199 (secondary cancers).
Table 2: Cancer diagnoses and deaths
2112
THE LANCET • Vol 353 • June 19, 1999
ARTICLES Patient
Sex
Thyroid cancer 1* Female
Iodine-131 therapy Date
Dose (MBq) 111 126 296 296 333 56 122 185 185 170 122 370
Age at Date of cancer first dose diagnosis (years)
Date of death
2*
Female
3*† 4*†
Female Female
5*†
Female
6*†
Female
7*
Female
May, 1975 May, 1976 September, 1967 August, 1969 September, 1969 May, 1972 May, 1973 May, 1981 May, 1983 May, 1967 April, 1968 October, 1979
8*†
Female
February, 1977
185
74
October, 1985
November, 1985
9*
Male
Cause of death
59
December, 1980
December, 1981
Papillary thyroid
Ischaemic heart disease
60
June, 1978
March, 1990
Follicular thyroid
Coronary atherosclerosis
64 46
February, 1977 August, 1973
June, 1977 January, 1975
Anaplastic thyroid Papillary thyroid
Carcinoma thyroid Carcinoma thyroid
64
January, 1989
September, 1989
Carcinoma thyroid
50
November, 1971
December, 1971
Carcinoma thyroid (not further specified) Anaplastic thyroid
77
September, 1983
September, 1985
July, 1971
259
45
February, 1986
May, 1987
Small-bowel cancer 10* Female 11* Female 12*† Female 13*† Female 14*† Female 15*† Female 16†‡ Male
March, 1982 May, 1984 May, 1962 December, 1953 January, 1968 February, 1966 January, 1969
185 130 296 296 185 444 304
59 33 51 58 74 57 64
January, 1988 January, 1987 August, 1985 March, 1976 January, 1980 January, 1978 1969
June, 1988 Alive August, 1985 June, 1987 February, 1980 April, 1978 February, 1974
17†‡
May, 1972
296
65
1963
December, 1972
Female
Cancer morphology
Carcinoma thyroid (not further specified) Carcinoma thyroid (not further specified) Follicular thyroid Adenocarcinoma duodenum Carcinoma small intestine Carcinoma ileum Mucocarcinoid tumour ileum Carcinoma duodenum Carcinoma duodenum Not recorded at ONS (before 1971) Not recorded at ONS (before 1971)
Carcinoma thyroid Hypertensive heart disease Carcinoma thyroid Carcinoma unknown primary site Carcinomatosis Alive Carcinoma ileum Carcinoma ileum Carcinoma duodenum Carcinoma duodenum Carcinoma small bowel Carcinoma small intestine
Cancer morphology was recorded at time of cancer registration and underlying cause of death reflects death certification. *Contributes to cancer-incidence data. †Contributes to cancer-mortality data. ‡Diagnosis of small-bowel cancer made before or within 1 year of iodine-131 treatment.
Table 3: Details of radioiodine treatment and cancer diagnoses and deaths in patients with thyroid and small-bowel cancers
mortality was seen only in men (SMR 0·76 [0·62–0·97], p<0·01; women 0·94 [0·84–1·04]). When data from the cohort were compared with cancer incidence and mortality data for the West Midlands region (rather than with England and Wales control data), findings were similar (data not shown). There were significant decreases in incidence of cancers of the pancreas, bronchus and trachea, bladder, and lymphatic and haemopoietic systems (the latter accounted for by non-significant reductions in incidence of lymphomas and leukaemias, table 2). Mortality due to cancers at each of these sites was also reduced, but cancer mortality was significant only for bronchus and trachea. By contrast, there were significant increases in diagnoses and deaths from cancers of the thyroid and small intestine. A similar pattern of results for cancer incidence and mortality was seen when findings in men and women were analysed separately (data not shown). A diagnosis of thyroid cancer was made a minimum of 15 months after radioiodine therapy, and in all but one patient after an interval of more than 4 years (table 3). Hospital records showed that each patient had typical clinical and biochemical features of hyperthyroidism at the time of radioiodine treatment in the absence of any clinical symptoms of thyroid cancer. A diagnosis of smallbowel cancer was made at least 3 years after administration of radioiodine for hyperthyroidism in the six patients contributing incidence data. However, this diagnosis preceded or was within 1 year of radioiodine treatment in two of six patients, reflected in small-bowel cancer mortality (table 3). Although only a few patients had thyroid or small-bowel cancer, there were no obvious differences from the other patients in terms of dose of radioiodine given or age at radioiodine treatment. The observed reduction in cancer diagnoses and deaths (at all sites) in the cohort was not significantly associated with the cumulative dose of radioiodine, age at first
THE LANCET • Vol 353 • June 19, 1999
treatment, years since first treatment, or period of treatment with radioiodine (data not shown). There was no association with characteristics of radioiodine therapy for cancer of the bronchus, pancreas, or lymphatic and haemopoietic systems. By contrast, a significant reduction in overall incidence of bladder cancer was accompanied by a positive association between bladder-cancer incidence and cumulative dose of radioiodine (⭐220 MBq, SIR 0·24 [0·08–0·73]; 221–480 MBq, SIR 0·54 [0·24–1·2]; ⭓481 MBq, SIR 1·96 [1·02–3·77]; p=0·0005 for trend in incidence with cumulative dose). SIR for uterine cancer was not different from expected, but a positive association with radioiodine dose was seen (⭐220 MBq, SIR 0·52 [0·28–0·96]; 221–480 MBq, SIR 0·73 [0·41–1·32]; ⭓481 MBq, SIR 2·11 [1·2–3·7]; p=0·002 for trend in incidence with cumulative dose).
Discussion Significant decreases in overall cancer incidence and mortality were seen, which in turn reflected significant decreases in incidence and mortality of lung cancer and lymphatic and haemopoietic malignant disorders, together with non-significant reductions in incidence and mortality from pancreatic and bladder cancer. By contrast, significant increases in incidence and mortality rates for thyroid cancer and cancer of the small bowel were seen. Data on cancer incidence and mortality (for patients and controls) were derived from different sources, which were unlikely to share similar biases. The general agreement between findings for incidence and mortality thus lends support to their validity. In our previous study of a similar group treated with radioiodine in 1950–89 (which included more person-years at risk because followup before 1971 was not excluded), we did not calculate cancer incidence or report deaths due to cancer at specific sites, but did report an increase in all-cause mortality,
2113
ARTICLES
which was secondary to an increase in risk of death from cardiovascular and cerebrovascular disease, thyroid disease, and fracture of the femur.20 By contrast with the increased risk of death secondary to vascular disease in those treated with radioiodine for hyperthyroidism,20 the absence of cancer risk in those treated with radioiodine in this study is reassuring. Our findings are in accord with two reports from the Mayo Clinic (Rochester, NY, USA), which compared 1005 patients treated with radioiodine with a group of patients who were treated by surgery and found no increase in cancer incidence or mortality. 7,8 Our findings also support those of the Massachusetts General Hospital (Boston, MA, USA), which showed a non-significant decrease in cancer mortality (SMR 0·9 [95% CI 0·7–1·1]) in 1762 patients treated with radioiodine.21 Our findings conflict, however, with those of a Swedish study,6,10 in which a small but significant increase in overall cancer incidence and mortality was seen (SIR 1·06 [1·01–1·11], SMR 1·09 [1·03–1·16]). Furthermore, subgroup analysis of patients surviving more than 10 years after radioiodine therapy in the Swedish cohort showed significant increases in incidence for cancers of the stomach, kidney, and brain, and significant increases in mortality from cancers of the digestive tract and respiratory organs. A striking finding in our study was the reduction in incidence and mortality from lung cancer, and, to a lesser extent, from other smoking-related cancers,22 notably bladder and pancreas. The explanation for these findings is unknown, but they may reflect a difference in smoking habits between the patients and the control population. One possibility is exclusion of smokers from the radioiodine-treated group. We previously reported an increase in deaths due to cardiovascular and cerebrovascular diseases in patients similar to those reported here.23 Since smoking is a major risk factor for mortality from vascular disease, as well as specific malignant disorders such as carcinoma of bronchus,22 it seems unlikely that those with a history of previous or current smoking were excluded from radioiodine treatment, but the absence of smoking history in this and previous studies means that this possibility cannot be explored. Our study provides no evidence for the increase in incidence or mortality from cancers of the stomach, kidney, or brain suggested by Holm and colleagues6 or cancers of the digestive tract or respiratory organs suggested by Hall and colleagues.10 Similarly, there was no increase in risk of breast cancer, in accord with Hoffman and colleagues.9 The reduction in lymphatic and haemopoietic malignant disorders was similar to that reported in the Swedish study6 and similar to a nonsignificant decrease in lymphatic malignant disorders in the Mayo Clinic radioiodine group when compared with the surgery group.7 The mechanism of this reduction remains unexplained. We noted a significant increase in incidence and mortality related to small-bowel cancer. Again, the validity of the results is supported by the agreement of the data for incidence and mortality, which relied on different sources of information. Two patients who died from small-bowel cancer were given radioiodine for hyperthyroidism within 1 year of cancer diagnosis, although given the rarity of this type of cancer, incidence and mortality remained significant after exclusion of these two cases. The absence of an association between
2114
radioiodine therapy and stomach cancer may argue against a cause-and-effect relation for small-bowel cancer, since a higher topical dose of radiation is more likely to be received by the gastric mucosa than by the small bowel. We found significant increases in incidence and mortality of thyroid cancer. Careful inspection of the records of patients with thyroid cancer confirmed the original diagnosis of hyperthyroidism and absence of symptoms of thyroid neoplasia. Our observations agreed with other studies16–20 of SIR and SMR, and incident and fatal cases of thyroid cancer reflected the age distribution of patients in terms of person-years at risk and increasing mortality relative to incidence with increasing age. A striking finding was the incidence of two cases of anaplastic thyroid cancer, with a further two patients for whom death from thyroid cancer occurred within 9 months of diagnosis, suggesting poorly differentiated tumours. Our findings for incidence of thyroid cancer differ from those from the Swedish study,6 in which a non-significant increase in incidence was seen (SIR 1·29 [0·76–2·03]). Although overall thyroid-cancer mortality was increased in that study, no significant increase in thyroid cancer SMR remained when thyroid cancer deaths occurring in the first year after radioiodine were excluded.10 Dobyns and colleagues23 reported data from 25 centres in the USA and one in the UK that together treated 21 714 patients with radioiodine for hyperthyroidism between 1946 and 1964. More than 1 year after radioiodine therapy, 19 thyroid cancers were found, which did not differ significantly from the number of diagnoses made in those treated by thyroidectomy or medical therapy alone. Nonetheless, five anaplastic carcinomas were found in the radioiodine group, as well as an increased incidence of benign thyroid tumours in the subgroup treated with radioiodine at younger than 20 years. A follow-up study of this group has been published.11 Cancer-incidence data were not available, but in these patients (who were treated with a combination of radioiodine, surgery, and antithyroid drugs) there was an increase in thyroid-cancer mortality (SMR 2·77 [1·85–3·98]), which was more striking in the subgroup receiving any radioiodine (SMR 3·94 [·52–5·86]). The numbers of thyroid-cancer deaths in the US study, and the numbers of thyroid-cancer diagnoses and deaths in our study suggest that it is not possible to establish a relation between cancer incidence or mortality and radioiodine dose, time from treatment, or age at treatment, and that findings for thyroid cancer may reflect an association with thyrotoxicosis rather than radioiodine exposure. Despite about a three-fold increase in thyroidcancer incidence and mortality described in our study, the absolute risk of thyroid cancer remained small. Interest in thyroid carcinogenesis and radiation exposure stems from the increased incidence of thyroid cancer among atom-bomb survivors, although only among those aged younger than 20 years at exposure.24 Significantly increased risks of thyroid cancer have also been described in children given radiotherapy for thymic enlargement25 and a non-significant increase was reported in children exposed to fallout from nuclear-weapons testing in the Nevada desert.26 These findings, together with data from the Chernobyl accident,3,4 suggest a difference in the response of the thyroid according to age at exposure, with the youngest being at greatest risk. Our cohort included small numbers of patients treated with
THE LANCET • Vol 353 • June 19, 1999
ARTICLES
iodine-131 who were younger than 30 years. With increasing trends for treating young patients with radioiodine,1 maintenance of follow-up in terms of cancer risk in all age-groups is important. Contributors J A Franklyn and M C Sheppard planned the project and supervised execution, analysis, and writing. P Maisonneuve and P Boyle advised on study design and did the statistical analyses. J Betteridge advised on study design and collated the data. P Maisonneuve, P Boyle, and J Betteridge contributed to the writing and preparation of the paper.
Acknowledgments This project was supported by the UK MRC, the BUPA Foundation, the Research Committee of the West Midlands R&D Directorate, the Endowment Fund of the Former United Birmingham Hospitals, and the Associazione Italiana per la Ricerca sul Cancro.
References 1 2
3 4
5 6 7
8
9 10 11
Franklyn JA. The management of hyperthyroidism. N Engl J Med 1994; 330: 1731–38. Baxter M, Stewart PM, Daykin J, Sheppard MC, Franklyn JA. Radioiodine therapy in young patients: perception of risk and use. QJM 1993; 86: 495–99. Baverstock K, Egloff B, Pinchera A, Ruchti C, Williams D. Thyroid cancer after Chernobyl. Nature 1992; 359: 21–22. Pacini F, Vorontsova T, Demidchik EP, et al. Post-Chernobyl thyroid carcinoma in Belarus children and adolescents: comparison with naturally occurring thyroid carcinoma in Italy and France. J Clin Endocrinol Metab 1997; 82: 3563–69. Hall P, Boice JD Jr, Berg G, et al. Leukaemia incidence after iodine-131 exposure. Lancet 1992; 340: 1–4. Holm L-E, Hall P, Wiklund K, et al. Cancer risk after iodine-131 therapy for hyperthyroidism. J Natl Cancer Inst 1991; 83: 1072–77. Hoffman DA, McConahey WM, Fraumeni JF, Kurland LT. Cancer incidence following treatment of hyperthyroidism. Int J Epidemiol 1982; 11: 218–24. Hoffman DA, McConahey WM, Diamond EL, Kurland LT. Mortality in women treated for hyperthyroidism. Am J Epidemiol 1982; 115: 243–54. Hoffman DA, McConahey WM. Breast cancer following iodine-131 therapy for hyperthyroidism. J Natl Cancer Inst 1983; 70: 63–67. Hall P, Berg G, Bjelkengren G, et al. Cancer mortality after iodine131 therapy for hyperthyroidism. Int J Cancer 1992; 50: 886–90. Ron E, Doody MM, Becker DV, et al. Cancer mortality following treatment for adult hyperthyroidism. JAMA 1998; 280: 347–55.
THE LANCET • Vol 353 • June 19, 1999
12 Barber SG, Carter DJ, Bishop JM. System for long-term review of patients at risk of becoming hypothyroid: further experience. Lancet 1977; ii: 967–70. 13 Franklyn JA, Daykin J, Drolc Z, Farmer M, Sheppard M. Long-term follow-up of treatment of thyrotoxicosis by three different methods. Clin Endocrinol (Oxf) 1991; 34: 71–77. 14 International classification of diseases: manual of the international classification of diseases, injuries, and causes of death, 9th rev, vol 1. Geneva: World Health Organization, 1977. 15 Waterhouse J, Muir CS, Shanmugaratnam K, Powell J, eds. Cancer incidence in five continents, vol III. IARC scientific publication no 15. Lyon: IARC, 1976. 16 Waterhouse J, Muir CS, Shanmugaratnam K, Powell J, eds. Cancer incidence in five continents, vol IV. IARC scientific publication no 42. Lyon: IARC, 1982. 17 Muir CS, Waterhouse J, Mack T, Powell J, Whelan SL, eds. Cancer incidence in five continents, vol V. IARC scientific publication no 88. Lyon: IARC, 1987. 18 Parkin DM, Muir CS, Whelan SL, Gao YT, Ferlay J, Powell J, eds. Cancer incidence in five continents, vol VI. IARC scientific publication no 120. Lyon: IARC, 1992. 19 Parkin DM, Whelan SL, Ferlay J, Raymond L, Young J, eds. Cancer incidence in five continents, vol VII. IARC scientific publication no 143. Lyon: IARC, 1997. 20 Franklyn JA, Maisonneuve P, Sheppard MC, Betteridge J, Boyle P. Mortality after the treatment of hyperthyroidism with radioactive iodine. N Engl J Med 1998; 338: 712–18. 21 Goldman MB, Maloof F, Monson RR, Aschengrau A, Cooper DS, Ridgway EC. Radioactive iodine therapy and breast cancer. Am J Epidemiol 1988; 127: 969–980. 22 Boyle P. Cancer, cigarette smoking and premature death in Europe: a review including the recommendations of European cancer experts consensus meeting, Helsinki October 1996. Lung Cancer 1997; 17: 1–60. 23 Dobyns BM, Sheline GE, Workman JB, Tompkins EA, McConahey WM, Becker DV. Malignant and benign neoplasms of the thyroid in patients treated for hyperthyroidism: a report of the cooperative thyrotoxicosis therapy follow-up study. J Clin Endocrinol Metab 1974; 38: 976–98. 24 Thompson DE, Mabuchi K, Ron E, et al. Cancer incidence in atomic bomb survivors, part II: solid tumors, 1958–1987. Radiat Res 1994; 137 (suppl 2): S17–67. 25 Shore RE, Woodard E, Hildreth N, Dvoretsky P, Hempelmann L, Pasternack B. Thyroid tumors following thyroid irradiation. J Natl Cancer Inst 1985; 74: 1177–84. 26 Kerber RA, Till JE, Simon SL, et al. A cohort study of thyroid disease in relation to fall-out from nuclear weapons testing. JAMA 1993; 270: 2076–82.
2115
Cancer incidence and mortality after radioiodine treatment for hyperthyroidism: a population-based cohort study Jayne A Franklyn, Patrick Maisonneuve, Michael Sheppard, Joan Betteridge, Peter Boyle
Summary Background Radioiodine is used increasingly as first-line treatment for hyperthyroidism, but concerns remain about subsequent risk of cancer, especially in those treated at a young age. We investigated cancer incidence and mortality in patients treated with radioiodine for hyperthyroidism. Methods We did a population-based study in 7417 patients treated in Birmingham, UK, between 1950 and 1991. We compared details of all cancer diagnoses and deaths in 1971–91 from the UK Office for National Statistics with data on cancer incidence and mortality for England and Wales specific for age, sex, and period. Findings During 72 073 person-years of follow-up, 634 cancer diagnoses were made, compared with an expected number of 761 (standardised incidence ratio [SIR] 0·83 [95% CI 0·77–0·90]). The relative risk of cancer mortality was also decreased (observed cancer deaths 448, expected 499; standardised mortality ratio [SMR] 0·90 [0·82–0·98]). Incidence of cancers of the pancreas, bronchus, trachea, bladder, and lymphatic and haemopoietic systems was lowered. Mortality from cancers at all these sites was also reduced but findings were significant only for bronchus and trachea. There were significant increases in incidence and mortality for cancers of the small bowel (SIR 4·81 [2·16–10·72], SMR 7·03 [3·16–15·66]) and thyroid (SIR 3·25 [1·69–6·25], SMR 2·78 [1·16–6·67]), although absolute risk of these cancers was small. Interpretation The decrease in overall cancer incidence and mortality in those treated for hyperthyroidism with radioiodine is reassuring. The absolute risk of cancers of the small bowel and thyroid remain low, but the increased relative risk shows the need for long-term vigilance in those receiving radioiodine. Lancet 1999; 353: 2111–15
Department of Medicine, University of Birmingham, Birmingham, UK (Prof J A Franklyn FRCP, M Sheppard FRCP, J Betteridge RGN); and Division of Epidemiology and Biostatistics, European Institute of Oncology, Milan, Italy (P Maisonneuve PhD, Prof P Boyle PhD) Correspondence to: Prof Jayne A Franklyn, Department of Medicine, Queen Elizabeth Hospital, Edgbaston, Birmingham B15 2TH, UK
THE LANCET • Vol 353 • June 19, 1999
Introduction Radioiodine is used increasingly as first-line therapy for hyperthyroidism.1 Even though it has been used for this purpose for more than 50 years, concerns among patients and doctors remain about the subsequent risk of malignant disorders, especially among those treated at a young age.2 These concerns have been heightened by reports of increased incidence of cancers after the Chernobyl nuclear accident, especially for thyroid cancer among children exposed at younger than 10 years to the main contaminating isotope iodine-131.3,4 Although there have been several studies of cancer risk in patients treated with radioiodine for hyperthyroidism, results have been conflicting. Hall and colleagues 5 studied the incidence of leukaemia in a large population of Swedish patients exposed to iodine-131 during diagnostic scanning or treatment of hyperthyroidism or thyroid cancer. No significant excess risk of leukaemia was seen, but the risk of stomach cancer increased with time and increasing dose of radioactivity.6 Other studies that compared cancer risk in those treated by surgery and with radioiodine for hyperthyroidism7–9 showed no difference in cancer incidence or mortality at these or other specific sites. The small size of most studies and the low incidence of thyroid cancer has meant that an increase in thyroidcancer incidence after radioiodine therapy has not been convincingly confirmed. By contrast, a small increase of thyroid-cancer deaths was reported in a Swedish cohort treated for hyperthyroidism,10 and an analysis of patients treated with radioiodine in centres throughout the USA in 1946–64 also reported an increase in thyroid-cancer mortality.11 We investigated cancer incidence and mortality in patients treated with radioiodine for hyperthyroidism.
Patients and methods Patients All patients who had been treated for hyperthyroidism with radioiodine in the West Midlands region of the UK between 1950 and 1991 were eligible for inclusion. We recorded details of the date and dose of radioiodine treatment, together with patients’ demographic details, on the computerised Birmingham Thyroid Follow-up Register, 12,13 which was established to ensure regular biochemical testing and detection of hypothyroidism. The initial group comprised 8468 patients. Demographic details for these patients were sent to the UK Office for National Statistics (ONS) for tracing on the National Health Service Central Register. 452 patients were excluded because they had died before 1971. Of the remaining 8016 patients, 599 (7·5%) were excluded for the following reasons: one had no known date of birth, 518 not found on register, 32 emigrated from the UK, seven lived in Northern Ireland, 32 were not registered with a primary-care physician, seven died but had no record of cause of death, and two had no record of date of radioiodine therapy. The final cohort included 7417 patients. The characteristics of the excluded patients were similar to those remaining in the study. Because cancer-incidence data for England and Wales before 1971 could not be obtained from the ONS, and because cancer
2111
ARTICLES
registrations after 1991 are incomplete, follow-up (for cancer incidence and mortality) was restricted to 1971–91. Details of all cancer diagnoses and underlying causes of death during this follow-up period were obtained from the ONS and coded according to the ninth revision of the International Classification of Diseases.14 Cancer diagnoses had been collated by the ONS from data collected by the UK Regional Cancer Registries, whereas causes of death were collated independently from the national death-certificate scheme. Annual cancer-incidence data for England and Wales, specific for age and sex, were extracted from International Agency for Research on Cancer publications15–19 and cancer-mortality rates from the WHO databank. We compared site-specific data on cancer incidence and mortality from follow-up with data from England and Wales. This comparison was made in preference to comparison with incidence and mortality data from the West Midlands region so that case and control data would be collated by the ONS and because regional control data were incomplete in terms of site-specific and period-specific cancer mortality. In addition, about 20% of patients had moved from the region before completion of the study. We calculated the expected number of cancer diagnoses or deaths by multiplying the number of person-years in each stratum of age, sex, and calendar year by the corresponding incidence or mortality rates specific for age, sex, and period for England and Wales. The standardised incidence ratio (SIR, ratio of observed to expected number of cancer diagnoses) and standardised mortality ratio (SMR, ratio of observed to expected deaths due to cancer) were used as estimates of relative risk. 95% CI were calculated on the assumption that the observed number of cancer cases and deaths followed a Poisson distribution. We used multivariate Poisson regression to assess trends in incidence and mortality according to time from treatment, age at treatment, and cumulative dose of radioiodine given. Data analysis was done with SAS version 6.12.
Results Characteristics of the cohort are shown in table 1. Most patients (84·9%) received one dose of radioiodine (mean age at first treatment 56·6 years [SD 12·7]). 634 diagnoses of cancer were made in 613 patients between 1971 and 1991 (table 2). The expected number of cancer diagnoses for the total number of person-years at risk Site of cancer
Number of diagnoses Observed
Number of patients
Mean (SD) cumulative dose of iodine-131 (MBq)
Mean (SD) age (years)
All patients
7417
308 (232)
56·6 (12·7)
72 073
Sex Male Female
1228 (16·6%) 6189 (83·4%)
305 (214) 308 (235)
54·1 (12·2) 57·1 (12·8)
11 514 60 559
Cumulative dose of iodine-131 (MBq) ⭐220 3657 (49·3%) 221–480 2521 (34·0%) ⭓481 1228 (16·6%) Unknown 11 (0·1%)
162 (43) 328 (66) 701 (300) ··
55·7 (15·1) 57·6 (13·0) 57·5 (12·0) 55·5 (13·0)
35 826 26 271 9888 89
Number of doses 1 2 3 Unknown
6297 (84·9%) 925 (12·5%) 184 (2·5%) 11 (0·1%)
260 (157) 525 (289) 849 (533) ··
57·0 (12·8) 54·5 (12·3) 53·6 (11·4) 55·5 (15·1)
60 424 9507 2052 89
Period of first treatment 1950–59 1960–69 1970–79 1980–89 1990–91
439 (5·9%) 1526 (20·6%) 2030 (27·4%) 2757 (37·2%) 665 (8·9%)
519 (412) 351 (257) 241 (196) 287 (174) 359 (200)
53·5 (7·9) 55·9 (9·1) 57·6 (11·0) 57·0 (14·7) 56·0 (17·0)
5616 21 415 27 229 17 160 653
Age at first treatment (years) ⭐49 2141 (28·9%) 50–59 2270 (30·6%) 60–69 1769 (23·9%) ⭓70 1237 (16·6%)
289 (209) 312 (249) 317 (240) 321 (221)
41·8 (7·0) 54·4 (2·8) 64·2 (2·9) 75·6 (4·8)
23 047 26 370 15 520 7136
Duration of follow-up (years) * 1 5452 (73·5%) 2–9 6344 (85·5%) 10–19 3760 (50·7%) ⭓20 1356 (18·3%)
279 (189) 288 (207) 307 (255) 354 (273)
57·1 (13·8) 56·7 (12·6) 55·0 (9·5) 51·9 (7·4)
5119 36 342 22 509 8103
*Between 1971 and 1991, when cancer-incidence data were available (eg, a patient who was first treated in 1968 will not contribute to 1-year follow-up data because he would have had 3 years of follow-up when observation began).
Table 1: Demographic and treatment characteristics of 7417 patients treated with radioiodine for hyperthyroidism
(72 073) was 761. The relative risk of mortality from cancer was reduced. Although a reduction in cancer incidence was seen in men (SIR 0·65 [95% CI 0·54–0·79], p<0·0001) and women (SIR 0·85 [0·78–0·92], p<0·001), a significant reduction in cancer
SIR (95% CI)
p
Expected 0·19
188 39 23 1·2
0·90 (0·77–1·04) 0·97 (0·71–1·34) 0·47 (0·26–0·85) 4·81 (2·16–10·72)
68 65
112 106
222 132
Genitourinary organs Total Bladder
97 18
Other sites* Total Brain Thyroid
15 4 9
Lymphatic and haemopoietic Total Lymphoma Leukaemia
Respiratory and intrathoracic organs Total Bronchus and trachea Bone, connective tissue, skin, and breast Total Breast
All sites
5
9·0
169 38 11 6
Number of deaths Observed
0·55 (0·23–1·33)
Lip, oral cavity, and pharynx Digestive organs and peritoneum Total Stomach Pancreas Small bowel
Person-years of follow-up
SMR (95% CI)
p
Expected
2
5·4
0·37 (0·09–1·48)
0·16
0·16 0·86 0·01 0·0001
151 39 17 6
157 37 24 0·8
0·96 (0·82–1·12) 1·05 (0·77–1·44) 0·72 (0·45–1·16) 7·03 (3·16–15·66)
0·64 0·76 0·18 0·0001
0·60 (0·48–0·77) 0·61 (0·48–0·78)
0·01 0·0001
79 78
104 100
0·76 (0·61–0·95) 0·78 (0·62–0·97)
0·01 0·03
226 124
0·98 (0·86–1·12) 1·07 (0·90–1·27)
0·77 0·45
74 69
81 74
0·92 (0·73–1·15) 0·93 (0·74–1·18)
0·46 0·58
128 28
0·76 (0·62–0·92) 0·63 (0·40–1·00)
0·006 0·05
62 7
78 14
0·80 (0·62–1·02) 0·49 (0·23–1·03)
0·07 0·06
12 7·4 2·8
1·28 (0·77–2·13) 0·54 (0·20–1·44) 3·25 (1·69–6·25)
0·33 0·22 0·0004
14 8 5
1·61 (0·89–2·55) 1·20 (0·60–2·40) 2·78 (1·16–6·67)
0·12 0·60 0·02
24 12 10
39 16 13
0·61 (0·41–0·91) 0·73 (0·41–1·28) 0·75 (0·40–1·39)
0·02 0·27 0·36
25 7 13
30 11 11
0·84 (0·57–1·24) 0·64 (0·31–1·34) 1·17 (0·68–2·01)
0·38 0·23 0·58
634
761
0·83 (0·77–0·90)
0·0001
448
499
0·90 (0·82–0·98)
0·02
8·7 6·7 1·8
*International Classification of Diseases (ninth revision) codes 190–194, excluding 195–199 (secondary cancers).
Table 2: Cancer diagnoses and deaths
2112
THE LANCET • Vol 353 • June 19, 1999
ARTICLES Patient
Sex
Thyroid cancer 1* Female
Iodine-131 therapy Date
Dose (MBq) 111 126 296 296 333 56 122 185 185 170 122 370
Age at Date of cancer first dose diagnosis (years)
Date of death
2*
Female
3*† 4*†
Female Female
5*†
Female
6*†
Female
7*
Female
May, 1975 May, 1976 September, 1967 August, 1969 September, 1969 May, 1972 May, 1973 May, 1981 May, 1983 May, 1967 April, 1968 October, 1979
8*†
Female
February, 1977
185
74
October, 1985
November, 1985
9*
Male
Cause of death
59
December, 1980
December, 1981
Papillary thyroid
Ischaemic heart disease
60
June, 1978
March, 1990
Follicular thyroid
Coronary atherosclerosis
64 46
February, 1977 August, 1973
June, 1977 January, 1975
Anaplastic thyroid Papillary thyroid
Carcinoma thyroid Carcinoma thyroid
64
January, 1989
September, 1989
Carcinoma thyroid
50
November, 1971
December, 1971
Carcinoma thyroid (not further specified) Anaplastic thyroid
77
September, 1983
September, 1985
July, 1971
259
45
February, 1986
May, 1987
Small-bowel cancer 10* Female 11* Female 12*† Female 13*† Female 14*† Female 15*† Female 16†‡ Male
March, 1982 May, 1984 May, 1962 December, 1953 January, 1968 February, 1966 January, 1969
185 130 296 296 185 444 304
59 33 51 58 74 57 64
January, 1988 January, 1987 August, 1985 March, 1976 January, 1980 January, 1978 1969
June, 1988 Alive August, 1985 June, 1987 February, 1980 April, 1978 February, 1974
17†‡
May, 1972
296
65
1963
December, 1972
Female
Cancer morphology
Carcinoma thyroid (not further specified) Carcinoma thyroid (not further specified) Follicular thyroid Adenocarcinoma duodenum Carcinoma small intestine Carcinoma ileum Mucocarcinoid tumour ileum Carcinoma duodenum Carcinoma duodenum Not recorded at ONS (before 1971) Not recorded at ONS (before 1971)
Carcinoma thyroid Hypertensive heart disease Carcinoma thyroid Carcinoma unknown primary site Carcinomatosis Alive Carcinoma ileum Carcinoma ileum Carcinoma duodenum Carcinoma duodenum Carcinoma small bowel Carcinoma small intestine
Cancer morphology was recorded at time of cancer registration and underlying cause of death reflects death certification. *Contributes to cancer-incidence data. †Contributes to cancer-mortality data. ‡Diagnosis of small-bowel cancer made before or within 1 year of iodine-131 treatment.
Table 3: Details of radioiodine treatment and cancer diagnoses and deaths in patients with thyroid and small-bowel cancers
mortality was seen only in men (SMR 0·76 [0·62–0·97], p<0·01; women 0·94 [0·84–1·04]). When data from the cohort were compared with cancer incidence and mortality data for the West Midlands region (rather than with England and Wales control data), findings were similar (data not shown). There were significant decreases in incidence of cancers of the pancreas, bronchus and trachea, bladder, and lymphatic and haemopoietic systems (the latter accounted for by non-significant reductions in incidence of lymphomas and leukaemias, table 2). Mortality due to cancers at each of these sites was also reduced, but cancer mortality was significant only for bronchus and trachea. By contrast, there were significant increases in diagnoses and deaths from cancers of the thyroid and small intestine. A similar pattern of results for cancer incidence and mortality was seen when findings in men and women were analysed separately (data not shown). A diagnosis of thyroid cancer was made a minimum of 15 months after radioiodine therapy, and in all but one patient after an interval of more than 4 years (table 3). Hospital records showed that each patient had typical clinical and biochemical features of hyperthyroidism at the time of radioiodine treatment in the absence of any clinical symptoms of thyroid cancer. A diagnosis of smallbowel cancer was made at least 3 years after administration of radioiodine for hyperthyroidism in the six patients contributing incidence data. However, this diagnosis preceded or was within 1 year of radioiodine treatment in two of six patients, reflected in small-bowel cancer mortality (table 3). Although only a few patients had thyroid or small-bowel cancer, there were no obvious differences from the other patients in terms of dose of radioiodine given or age at radioiodine treatment. The observed reduction in cancer diagnoses and deaths (at all sites) in the cohort was not significantly associated with the cumulative dose of radioiodine, age at first
THE LANCET • Vol 353 • June 19, 1999
treatment, years since first treatment, or period of treatment with radioiodine (data not shown). There was no association with characteristics of radioiodine therapy for cancer of the bronchus, pancreas, or lymphatic and haemopoietic systems. By contrast, a significant reduction in overall incidence of bladder cancer was accompanied by a positive association between bladder-cancer incidence and cumulative dose of radioiodine (⭐220 MBq, SIR 0·24 [0·08–0·73]; 221–480 MBq, SIR 0·54 [0·24–1·2]; ⭓481 MBq, SIR 1·96 [1·02–3·77]; p=0·0005 for trend in incidence with cumulative dose). SIR for uterine cancer was not different from expected, but a positive association with radioiodine dose was seen (⭐220 MBq, SIR 0·52 [0·28–0·96]; 221–480 MBq, SIR 0·73 [0·41–1·32]; ⭓481 MBq, SIR 2·11 [1·2–3·7]; p=0·002 for trend in incidence with cumulative dose).
Discussion Significant decreases in overall cancer incidence and mortality were seen, which in turn reflected significant decreases in incidence and mortality of lung cancer and lymphatic and haemopoietic malignant disorders, together with non-significant reductions in incidence and mortality from pancreatic and bladder cancer. By contrast, significant increases in incidence and mortality rates for thyroid cancer and cancer of the small bowel were seen. Data on cancer incidence and mortality (for patients and controls) were derived from different sources, which were unlikely to share similar biases. The general agreement between findings for incidence and mortality thus lends support to their validity. In our previous study of a similar group treated with radioiodine in 1950–89 (which included more person-years at risk because followup before 1971 was not excluded), we did not calculate cancer incidence or report deaths due to cancer at specific sites, but did report an increase in all-cause mortality,
2113
ARTICLES
which was secondary to an increase in risk of death from cardiovascular and cerebrovascular disease, thyroid disease, and fracture of the femur.20 By contrast with the increased risk of death secondary to vascular disease in those treated with radioiodine for hyperthyroidism,20 the absence of cancer risk in those treated with radioiodine in this study is reassuring. Our findings are in accord with two reports from the Mayo Clinic (Rochester, NY, USA), which compared 1005 patients treated with radioiodine with a group of patients who were treated by surgery and found no increase in cancer incidence or mortality. 7,8 Our findings also support those of the Massachusetts General Hospital (Boston, MA, USA), which showed a non-significant decrease in cancer mortality (SMR 0·9 [95% CI 0·7–1·1]) in 1762 patients treated with radioiodine.21 Our findings conflict, however, with those of a Swedish study,6,10 in which a small but significant increase in overall cancer incidence and mortality was seen (SIR 1·06 [1·01–1·11], SMR 1·09 [1·03–1·16]). Furthermore, subgroup analysis of patients surviving more than 10 years after radioiodine therapy in the Swedish cohort showed significant increases in incidence for cancers of the stomach, kidney, and brain, and significant increases in mortality from cancers of the digestive tract and respiratory organs. A striking finding in our study was the reduction in incidence and mortality from lung cancer, and, to a lesser extent, from other smoking-related cancers,22 notably bladder and pancreas. The explanation for these findings is unknown, but they may reflect a difference in smoking habits between the patients and the control population. One possibility is exclusion of smokers from the radioiodine-treated group. We previously reported an increase in deaths due to cardiovascular and cerebrovascular diseases in patients similar to those reported here.23 Since smoking is a major risk factor for mortality from vascular disease, as well as specific malignant disorders such as carcinoma of bronchus,22 it seems unlikely that those with a history of previous or current smoking were excluded from radioiodine treatment, but the absence of smoking history in this and previous studies means that this possibility cannot be explored. Our study provides no evidence for the increase in incidence or mortality from cancers of the stomach, kidney, or brain suggested by Holm and colleagues6 or cancers of the digestive tract or respiratory organs suggested by Hall and colleagues.10 Similarly, there was no increase in risk of breast cancer, in accord with Hoffman and colleagues.9 The reduction in lymphatic and haemopoietic malignant disorders was similar to that reported in the Swedish study6 and similar to a nonsignificant decrease in lymphatic malignant disorders in the Mayo Clinic radioiodine group when compared with the surgery group.7 The mechanism of this reduction remains unexplained. We noted a significant increase in incidence and mortality related to small-bowel cancer. Again, the validity of the results is supported by the agreement of the data for incidence and mortality, which relied on different sources of information. Two patients who died from small-bowel cancer were given radioiodine for hyperthyroidism within 1 year of cancer diagnosis, although given the rarity of this type of cancer, incidence and mortality remained significant after exclusion of these two cases. The absence of an association between
2114
radioiodine therapy and stomach cancer may argue against a cause-and-effect relation for small-bowel cancer, since a higher topical dose of radiation is more likely to be received by the gastric mucosa than by the small bowel. We found significant increases in incidence and mortality of thyroid cancer. Careful inspection of the records of patients with thyroid cancer confirmed the original diagnosis of hyperthyroidism and absence of symptoms of thyroid neoplasia. Our observations agreed with other studies16–20 of SIR and SMR, and incident and fatal cases of thyroid cancer reflected the age distribution of patients in terms of person-years at risk and increasing mortality relative to incidence with increasing age. A striking finding was the incidence of two cases of anaplastic thyroid cancer, with a further two patients for whom death from thyroid cancer occurred within 9 months of diagnosis, suggesting poorly differentiated tumours. Our findings for incidence of thyroid cancer differ from those from the Swedish study,6 in which a non-significant increase in incidence was seen (SIR 1·29 [0·76–2·03]). Although overall thyroid-cancer mortality was increased in that study, no significant increase in thyroid cancer SMR remained when thyroid cancer deaths occurring in the first year after radioiodine were excluded.10 Dobyns and colleagues23 reported data from 25 centres in the USA and one in the UK that together treated 21 714 patients with radioiodine for hyperthyroidism between 1946 and 1964. More than 1 year after radioiodine therapy, 19 thyroid cancers were found, which did not differ significantly from the number of diagnoses made in those treated by thyroidectomy or medical therapy alone. Nonetheless, five anaplastic carcinomas were found in the radioiodine group, as well as an increased incidence of benign thyroid tumours in the subgroup treated with radioiodine at younger than 20 years. A follow-up study of this group has been published.11 Cancer-incidence data were not available, but in these patients (who were treated with a combination of radioiodine, surgery, and antithyroid drugs) there was an increase in thyroid-cancer mortality (SMR 2·77 [1·85–3·98]), which was more striking in the subgroup receiving any radioiodine (SMR 3·94 [·52–5·86]). The numbers of thyroid-cancer deaths in the US study, and the numbers of thyroid-cancer diagnoses and deaths in our study suggest that it is not possible to establish a relation between cancer incidence or mortality and radioiodine dose, time from treatment, or age at treatment, and that findings for thyroid cancer may reflect an association with thyrotoxicosis rather than radioiodine exposure. Despite about a three-fold increase in thyroidcancer incidence and mortality described in our study, the absolute risk of thyroid cancer remained small. Interest in thyroid carcinogenesis and radiation exposure stems from the increased incidence of thyroid cancer among atom-bomb survivors, although only among those aged younger than 20 years at exposure.24 Significantly increased risks of thyroid cancer have also been described in children given radiotherapy for thymic enlargement25 and a non-significant increase was reported in children exposed to fallout from nuclear-weapons testing in the Nevada desert.26 These findings, together with data from the Chernobyl accident,3,4 suggest a difference in the response of the thyroid according to age at exposure, with the youngest being at greatest risk. Our cohort included small numbers of patients treated with
THE LANCET • Vol 353 • June 19, 1999
ARTICLES
iodine-131 who were younger than 30 years. With increasing trends for treating young patients with radioiodine,1 maintenance of follow-up in terms of cancer risk in all age-groups is important. Contributors J A Franklyn and M C Sheppard planned the project and supervised execution, analysis, and writing. P Maisonneuve and P Boyle advised on study design and did the statistical analyses. J Betteridge advised on study design and collated the data. P Maisonneuve, P Boyle, and J Betteridge contributed to the writing and preparation of the paper.
Acknowledgments This project was supported by the UK MRC, the BUPA Foundation, the Research Committee of the West Midlands R&D Directorate, the Endowment Fund of the Former United Birmingham Hospitals, and the Associazione Italiana per la Ricerca sul Cancro.
References 1 2
3 4
5 6 7
8
9 10 11
Franklyn JA. The management of hyperthyroidism. N Engl J Med 1994; 330: 1731–38. Baxter M, Stewart PM, Daykin J, Sheppard MC, Franklyn JA. Radioiodine therapy in young patients: perception of risk and use. QJM 1993; 86: 495–99. Baverstock K, Egloff B, Pinchera A, Ruchti C, Williams D. Thyroid cancer after Chernobyl. Nature 1992; 359: 21–22. Pacini F, Vorontsova T, Demidchik EP, et al. Post-Chernobyl thyroid carcinoma in Belarus children and adolescents: comparison with naturally occurring thyroid carcinoma in Italy and France. J Clin Endocrinol Metab 1997; 82: 3563–69. Hall P, Boice JD Jr, Berg G, et al. Leukaemia incidence after iodine-131 exposure. Lancet 1992; 340: 1–4. Holm L-E, Hall P, Wiklund K, et al. Cancer risk after iodine-131 therapy for hyperthyroidism. J Natl Cancer Inst 1991; 83: 1072–77. Hoffman DA, McConahey WM, Fraumeni JF, Kurland LT. Cancer incidence following treatment of hyperthyroidism. Int J Epidemiol 1982; 11: 218–24. Hoffman DA, McConahey WM, Diamond EL, Kurland LT. Mortality in women treated for hyperthyroidism. Am J Epidemiol 1982; 115: 243–54. Hoffman DA, McConahey WM. Breast cancer following iodine-131 therapy for hyperthyroidism. J Natl Cancer Inst 1983; 70: 63–67. Hall P, Berg G, Bjelkengren G, et al. Cancer mortality after iodine131 therapy for hyperthyroidism. Int J Cancer 1992; 50: 886–90. Ron E, Doody MM, Becker DV, et al. Cancer mortality following treatment for adult hyperthyroidism. JAMA 1998; 280: 347–55.
THE LANCET • Vol 353 • June 19, 1999
12 Barber SG, Carter DJ, Bishop JM. System for long-term review of patients at risk of becoming hypothyroid: further experience. Lancet 1977; ii: 967–70. 13 Franklyn JA, Daykin J, Drolc Z, Farmer M, Sheppard M. Long-term follow-up of treatment of thyrotoxicosis by three different methods. Clin Endocrinol (Oxf) 1991; 34: 71–77. 14 International classification of diseases: manual of the international classification of diseases, injuries, and causes of death, 9th rev, vol 1. Geneva: World Health Organization, 1977. 15 Waterhouse J, Muir CS, Shanmugaratnam K, Powell J, eds. Cancer incidence in five continents, vol III. IARC scientific publication no 15. Lyon: IARC, 1976. 16 Waterhouse J, Muir CS, Shanmugaratnam K, Powell J, eds. Cancer incidence in five continents, vol IV. IARC scientific publication no 42. Lyon: IARC, 1982. 17 Muir CS, Waterhouse J, Mack T, Powell J, Whelan SL, eds. Cancer incidence in five continents, vol V. IARC scientific publication no 88. Lyon: IARC, 1987. 18 Parkin DM, Muir CS, Whelan SL, Gao YT, Ferlay J, Powell J, eds. Cancer incidence in five continents, vol VI. IARC scientific publication no 120. Lyon: IARC, 1992. 19 Parkin DM, Whelan SL, Ferlay J, Raymond L, Young J, eds. Cancer incidence in five continents, vol VII. IARC scientific publication no 143. Lyon: IARC, 1997. 20 Franklyn JA, Maisonneuve P, Sheppard MC, Betteridge J, Boyle P. Mortality after the treatment of hyperthyroidism with radioactive iodine. N Engl J Med 1998; 338: 712–18. 21 Goldman MB, Maloof F, Monson RR, Aschengrau A, Cooper DS, Ridgway EC. Radioactive iodine therapy and breast cancer. Am J Epidemiol 1988; 127: 969–980. 22 Boyle P. Cancer, cigarette smoking and premature death in Europe: a review including the recommendations of European cancer experts consensus meeting, Helsinki October 1996. Lung Cancer 1997; 17: 1–60. 23 Dobyns BM, Sheline GE, Workman JB, Tompkins EA, McConahey WM, Becker DV. Malignant and benign neoplasms of the thyroid in patients treated for hyperthyroidism: a report of the cooperative thyrotoxicosis therapy follow-up study. J Clin Endocrinol Metab 1974; 38: 976–98. 24 Thompson DE, Mabuchi K, Ron E, et al. Cancer incidence in atomic bomb survivors, part II: solid tumors, 1958–1987. Radiat Res 1994; 137 (suppl 2): S17–67. 25 Shore RE, Woodard E, Hildreth N, Dvoretsky P, Hempelmann L, Pasternack B. Thyroid tumors following thyroid irradiation. J Natl Cancer Inst 1985; 74: 1177–84. 26 Kerber RA, Till JE, Simon SL, et al. A cohort study of thyroid disease in relation to fall-out from nuclear weapons testing. JAMA 1993; 270: 2076–82.
2115