SusByecT No. 889. This 55-year-old woman had ductal carcinomaof the breast with positive axillary nodes diagnosed in 1980. She was treated at Straub Clinic and Hospital and returned to Ebeye to be placed on chemotherapy, but expired late in 1980. Suspsect No. 1554. Diabetes and senility were the clinical problems of this 62-year-old woman whodied in 1981 on Ebeye. SUBJECT No. 1571. Neurologic abnormalities detected on the 1982 examination led to the diagnosis of a spinal cord tumor(astrocytoma) in this 28-year-old woman. She died at Tripler ArmyHospital following surgery for the tumor in 1982. SusBsecT No. 945. This 57-year-old woman had severe pulmonary disease (FEV: = 0.5) and a history of cough and dyspnea for many years. She died in 1982 after being admitted to Ebeye Hospital with increasing cough andchestpain. Tuberculosis had not been confirmedin earlier evaluations, and the cause of the lung disease was not ascertained. HEMATOLOGY No hematologic malignancies were diagnosed in 1980-1982. Mean neutrophil counts (Figure 2a) in the Rongelap and Ailingnae groups remain, as in most years, slightly lower than control values. Lymphocyte counts (Figure 2b) are low only in the small Ailingnae group, although mean Rongelap values were below “9 a) — wl 3 d control levels during the early years of surveillance. Platelet counts (Figure 2c and d) are currently near control levels, although in retrospect one can argue that it may have taken about 20 years for this to occur in the Rongelap group. Hematocrit values have always been within a few percent of control levels and are not shown. There have been few statistically significant differences in blood counts between exposed and unexposed groups on a year-to-year basis. The relative constancy of the differences over many years, however, raises the possibility of long-term constraints on hematopoiesis in the _Rongelap and Ailingnae groups. The following table is an analysis of group differences in the blood cell counts of Figure 2. The entries are p values for tests of trend of blood cell counts over time,’ the counts of the exposed groups being less than the comparison group in all instances. Rongelap vs Ailingnae vs 0.04 NS 0.04 0.04 0.04 0.004 NS NS Comparison Neutrophils Lymphocytes Platelets (females) Platelets (males) Comparison NS = notsignificant The nonparametric test used in this analysis is one of low sensitivity, and a more detailed analysis is in preparation. In particular, the effects of mortality on trend will be investigated. Although there could have been inherently different counts amongthe groupsirrespective of radiation exposure, the significance of the latter is suggested by the observation that three cell lines reflect the same trend. Thepossibility that there may be such a long-term depression of hematopoietic elements stands in contrast to data from other sources. Occasionaldifferences in blood counts have been noted between radiation-exposed and control populations in Japan, but the differences ‘“‘were small and too irregular with respect to age, sex, and time of exposure to be attributed conclusively to radiation exposure.” Quantitative recovery of hematopoietic tissue from acute radiation injury is often complete within 2-3 months.’ There is no reason to infer clinical significance from the present findings; the variations, on both a group and an individual basis, are minor, and there is no evidencesofarof increased suscepti. bility to infection in exposed persons. MARKERS OF POSSIBLE SUBCLINICAL NEOPLASIA A variety of tests have been performed as a part of surveillance efforts to detect neoplastic or paraneoplastic processes which might remain subclinical for extended periods. In 1981, 400cell leukocyte differentials were doneto look for changes in low frequency cells, particularly monocytes and basophils (Table 1). Macrocytosis and polycythemia are routinely evaluated whenclinically indicated, and grouped values