TABLE 2: Thyroid Nodules Diagnosed at Surgery through 1991 Adenomatous nodules Adenomas Papillary cancers Follicular cancers Occult cancers Rongelap (67)° 18 1 5 . - Ailingnae (19) 4 1 : . 1 Utirik (167) 10 5 4 1° 6 Comparison (277)° 4 1 2 - 24 NOT INCLUDEDarethe following unoperated (and therefore unconfirmed) nodules: Rongelap--1: Ailingnae - - 1: Utirik -- 1: Comparison -- 5. INCLUDEDare all consensus diagnoses of a panel of consultant pathologists: two different lesions were detected in one person from Rongelap, one from Ailingnae, and two from Utirik. a b c d. Numberof persons (including those in utero) who were originally exposed. This numberincludesall persons who have been in the Comparison group since 1957. Some have not been seen for many years; others were added as recently as 1976. No thyroid surgeries have been performed on this group since 1985. Equally divided opinion in onecase;follicular carcinomavs. atypical adenoma. Majority opinion in one case; occult papillary carcinomavs. follicular carcinoma. The samepatient had lymphocytic thyroiditis. characteristic of the type of thyroid injury sustained by the Marshallese. Two alternative explanations are 1) the “epidemic” of thyroid nodule formation is virtually over, and 2) the recent decrease in nodule incidence is due to random fluctuation and therefore temporary. Although time may tell which of the above explanations is correct, the respite in new cases provides an opportunity to bring together information on thyroid nodules collected by the Marshall Islands Medical Program over almost three decades and to draw tentative conclusions on several issues that may be relevant to inadvertent radiation exposures elsewhere. Radiation risks to the thyroid: Oneaspect of radiation-induced thyroid injury that has been repeatedly assessed is the dose of radiation required to induce it. Data available from the Marshall Islands Medical Program have been recently summarized (Robbins and Adams, 1988), with the following conclusions: 1) Therisk coefficient for thyroid nodules, adjusted for their occurrence in the Comparison population, was 8.3 per 10° persons, per cGy, per year. 2) The risk coefficient for thyroid cancer was 1.5 per 10° persons, per cGy, per year. 3) The contribution of '34I to the thyroid absorbed dose was relatively small, in the range of 1015%, the remainder being due to short-lived radioiodines. Perhaps as a consequence, the radiation-induced risk for developing nodular disease in the exposed Marshallese appears similar to that predicted if the total thyroid dose had been from external irradiation alone. Since the above analysis includedall the nodules up to the present, and since the Marshallese thyroid dose data have provided no insight into radiationinduced risk of thyroid carcinoma that was not already available from other sources, no further comment on dose-response and risk of thyroid disease will be made in this summary. 15