How the US National Academy of Sciences misled the world community on cancer risk assessment: new findings challenge historical foundations of the linear dose response

Abstract

This paper extends several recent publications indicating that Hermann J. Muller: (1) Made deceptive statements during his Noble Prize Lecture on December 12, 1946, that were intended to promote the acceptance of the linear dose-response model for risk assessment for ionizing radiation and (2) that such actions of Muller were masked by a series of decisions by Muller’s long-time colleague and esteemed radiation geneticist Curt Stern, affecting key publications in the mutation literature. Such actions further enhanced acceptance of the linearity dose-response model while preventing Muller’s deceptions from being discovered. This paper provides documentation that Muller reinforced such practices within the scientific literature in the early 1950s, by supporting scientifically questionable actions of Stern. Detailed documentation is provided that demonstrates how these actions affected national and international risk assessment policy for ionizing radiation and chemical carcinogens via the recommendations of the National Academy of Sciences Biological Effects of Atomic Radiation committee in 1956, to adopt the linear dose-response model.

Appendix

Stern–Muller temporal letter exchange concerning the aged-stored sperm control mutation rate (Source: Lilly Library, Stern–Muller correspondence)

Curt Stern wrote a letter to Hermann J. Muller on January 22, 1947 (American Philosophical Society 1947a), informing him that “At the present time it looks as if our new control data (probably the results of the first 3 months of the first Uphoff experiment; note that her first month’s reading was an especially low mutation rate of 0.005 %) for aged sperm are considerably below those of Caspari’s.” He then asked Muller to “send me your figures on rate of sex-linked lethal in sperm aged several weeks, (most desirably, if you have them, data on 3 weeks), in comparison to control data from non-aged sperm?”

On February 3, 1947 (Lilly Library 1947b, February 3), Muller answered by stating that “…. sperm of males which are about a week old and have been copulating freely (as in Caspari’s experiment) during that period have only about .07 or .08 % of lethal. Thus, the latter sperm, after 3 weeks, should contain something like .28 % of lethal.”

On July 23, 1947 (American Philosophical Society 1947b), Stern writes Muller again stating that “I have mislaid your letter of some months ago (February 3, 1947, letter) in which you gave me some details of your own on the mutation rate under various physiological conditions. May I therefore ask you two questions and will you permit me to use your answers in a report which I am just preparing for the Manhattan Project? Obviously, full credit for it would be given. The questions are: (1) What is the spontaneous mutation rate in sperm derived from Canton-special males of from 3- to 6 days old? (2) What is the weekly increase in mutation rate of sperm from such males stored in females?”

On August 4, 1947 (Lilly Library 1947c), Muller responds “When sperm were stored in females, there was a weekly increase in the mutation frequency of about 0.07 %, on the average.” On August 7, 1947 (American Philosophical Society 1947c), Stern cabled Muller asking him the temperature used and on August 8, 1947 (American Philosophical Society 1947d), Muller answered via cable indicating “25 °C.” A subsequent undated letter, but most likely prior to September 9, 1947 (American Philosophical Society 1947e), Muller noted “A recalculation of my data gives the figure of 0.08 % instead of 0.07 % as the frequency of lethal accumulating in mature sperm per week.” Since Uphoff and Stern (1947) did not include this correction in their report to the AEC it suggests that this undated letter was received after submittal of their report to the AEC.

The control value therefore used by Uphoff and Stern (1947) of 0.07 % for the estimated mutation rate of the sperm stored in the spermatheca was based on the earlier letter correspondence-supplied estimates of Muller (Lilly Library 1947b, c, February 3 and August 4) which Muller later clarified as being slightly in error.

The Caspari and Uphoff studies used Drosophila melanogaster fruit flies, breeding Canton-wild-type (S) males with Muller-5 females. Muller claimed (Lilly Library 1947c, August 4) that he never conducted mutation experiments with aged males of the Canton-wild-type stock. Muller stated that he had tested the aged sperm mutation frequency in “a number of different stocks (of Drosophila males) without finding any difference.” The rate of increase on a weekly basis was said to be 0.07 % on average. This value of 0.07 % is believed to be prior to the correction to 0.08 %. This suggests that Muller did not observe significant inter-stock variation in mutation rates of the stored sperm.

Stern seems to have completed his Uphoff and Stern (1947) paper for the Manhattan Project during August, 1947. Stern knew that Uphoff’s mean mutation frequency was 0.1682 % (0.1365–0.2097 %). This suggests a weekly mean increase in mutation rate of 0.0227 % (0.0122–0.0366 %), far lower than the 0.07 or 0.08 % mean weekly increase in Muller. When Stern wrote to Muller on September 9, 1947, he stated that for the Canton-special stock “…the weekly increase is considerably less than that found by you and others. It seems to be much more of the order of 0.03–0.05.” This September 9, 1947, letter was written probably just after the submission of the Uphoff and Stern (1947) paper to the AEC, and definitely before the submission of the Caspari and Stern (1948) paper for publication by Genetics (i.e., November 25, 1947). Thus, the judgments of Uphoff and Stern that found that Uphoff’s data were “uninterpretable” and that supported the reliability of the Caspari control data were made with the information provided by Muller during the summer of 1947. The apparent argument that Stern seems to be suggesting in his September 9, 1947, letter to Muller is that the Canton-wild-type stored sperm in the female may yield uniquely lower control mutation values. The argument is tenuous as the far higher weekly rate was consistently shown by multiple investigators, and with multiple Drosophila stocks, only being low in two Uphoff experiments. In fact, significant inter-strain differences on the frequency of dominant lethal mutations as induced by radiation were not reported in various Drosophila strains, including the Canton-special wild-type strain (Demerec and Fano 1944; Strömnaes 1951). This suggestion by Stern was not included in the Uphoff and Stern (1947) report.

This letter exchange between Stern and Muller fails to provide support for the later statements of Muller that Caspari’s control group was unusually high. The Muller data and statements also do not provide support for the conclusion that the low Uphoff control data were in a normal range. None of this information was provided by Stern in his Science publication to permit the scientific community to better evaluate the Uphoff and Caspari control group data.