- [9 detectable levels, the degree of damage may <hen preclude neoplasmidevelopment. The need for a very sensitive TSH assay was obvious in an experimental|design to test the subtle effects that TSH might have in irradiated animals that should be developing neoplasms some of which were given supplemental T,: Some of our observations on the Marshallese patients have prappted us to questior the part played by TSH in the origin of neoplasms, because when first Marshallese began to develop nodules, they were given prophylactic thyroid horfone (T,) to inhibi the development of more neoplasms. It has subsequently been sh people became at least mildly hypothyroid after many years if supplement (Larsen,et al, including this principle investigation, that many of these were not given Ty press). Asnear ‘as we can tell, most of the Marshallese have taken the Ty tablets taken, the T, levels (at first PBI's) have been at a respectable 1 duals. In spite of the administration of Ty for more than 15 yearp, we have continue to find new lesions at almost every armmual examination of these observations have prompted same skepticism as to whether TSH i T, adminstration had been as successfully followed as was inten sed people. The neoplasms or whe Using the rat mod in which neoplasms were expected to develop, we have used a very itive rat TSH as on serum collected at the time animals were sacrificed. (Experim al data still imec” Radioimmmoassay for TSH and its Application To The Study of Neoplasm Induction We have modified and tested a radioimmmoassay (RIA) for rat |TSH in our laborat: under this contract. This has been done according to the method df Kieffer, et al, 1974, using NIAMD materials, and with advise of Dr. P. Reed Larsem who has been a collaborator on the Marshallese work. laboratory and is reliable. Our assay has become well The NIAMD Rat TSH: iodination and RAT TSH RP I-2 for standard. I-2 (potency 35 Our range of normal qstablished in our ].U./mg) was used f Yalues on a large number of male control animals was 40 to 800 ng/ml with a mean of [412 SD ~ 263 ng/ml 1”)