Year: 2000
Type: article
Abstract: The last sentence of the letter of Rockhill et al. states, “more attention to the risk/benefit calculations of Gail et al. is needed before prophylactic use of this drug [tamoxifen] becomes widespread.” We agree. The first sentence in their letter quotes our commentary (1) by including the important word “potentially” after the word “prevention.” Despite this clearly qualified statement, they wished to offer a different perspective. We wish they also had put this important issue (net public health impact of tamoxifen) into the proper perspective. Their letter missed the major theme of our commentary (i.e., the proof of the principle of cancer chemoprevention), which also included many other major issues, such as the development of promising new agents (e.g., those involving new selective estrogen receptor modulators). The letter of Rockhill et al. sharply criticizes Dr. Fisher's published stance on tamoxifen's net public health impact (2). We believe that it is appropriate to leave it to Dr. Fisher to defend his position. Rockhill et al. did not acknowledge that the net public health issue appeared in the section entitled “Unresolved [italics added] Tamoxifen Prevention Issues.” We never stated categorically that tamoxifen will provide a net public health benefit. Several other misinterpretations and overinterpretations in the letter of Rockhill et al. include the statement that tamoxifen use for African-Americans “is likely to cause a net loss of public health in nearly all age-risk groups,” ignoring the paper by Gail et al. (these estimates are “subject to greater uncertainty”) (3), an accompanying editorial (these risk/benefit rates are “misleading”) (4), and the American Society of Clinical Oncology technology assessment (“It is reasonable to apply the same criteria for tamoxifen use across all ethnic groups. . . .”) (5); the unfounded implication that we think tamoxifen is risk-free prevention for every woman, ignoring our painstaking discussions of the risks (1); and a total disregard of certain favorable indications, such as for women who have had hysterectomies (e.g., favorable risk–benefit ratios as stated by Gail et al. for every age-risk category of such women 35–59 years old) (3). According to Rockhill et al., the Nurses' Health Study (NHS) indicates that U.S. women within the positive tamoxifen risk–benefit categories of Gail et al. account for few breast cancers and are very few in number. The letter bases this view on percentages involving women 50–69 years old and on risk–benefit cutoff figures from Table 10 of Gail et al. (3), although Table 12 in the same article would have been more relevant for comparisons with NHS “healthy volunteers.” On the other hand, let us consider only the Breast Cancer Prevention Trial (BCPT) women who were less than 50 years old—approximately 5300 women (approximately 40% of the 13388 total). Tables 10 and 11 in the article by Gail et al. (3) indicate that all of these women were in positive risk–benefit categories. A rough extrapolation suggests that there are from 2 to 4 million such tamoxifen-eligible women in the United States [relevant data: 23% of eligible screened women were accrued to BCPT (6); 35–49 years old = 56% of BCPT women <60 years old (6); and 9 million 35- to 59-year-old U.S. women fit BCPT eligibility (McCaskill-Stevens W, Ford LG: personal communication)]. Any counting based on this line of reasoning would result in at least hundreds of thousands of U.S. women with positive tamoxifen risk–benefit ratios. Tamoxifen use in the BCPT was associated with 28 fewer breast cancers (versus placebo) per 1000 women (<50 years old) in 5 years (2). Therefore, the numbers of breast cancers in U.S. women potentially prevented by tamoxifen in 5 years run from 2800 (if there are only 100000 eligible women) to 56000 (if 2 million women are eligible). Obviously, the two lines of reasoning outlined above show a disconnect between the indications of the BCPT and the NHS. This disconnection is one of many fascinating issues raised by our commentary, and we look forward to hearing more from both sides.
Authors: S. M. Lippman, P. H. Brown
Language: English
Subfield: Genetics
Domain: Life Sciences
Open Access status: bronze
OpenAlex ID: W2034976643