Only By Products From Many Adjuvant Breast Cancer Studies

Publisher Website
Google Scholar
Abstract
To the Editor: The paper by Lippman et al prompts me to send you the following letter that might induce chairpersons of cooperative studies to reflect on the possibility that they may lose years of study time due to inadequate study designs. The report of Lippman et al shows a typical adjuvant breast cancer study where, on the grounds of the study design, only by-products can be expected as results. Surgery-only control subjects were omitted from the study design based on the hypothetical but until today never-proven efficacy of doxorubicincyclophosphamide combination chemotherapy in the adjuvant situation. The main aim of any adjuvant study, if treatment is better than nothing, can not be answered by such a study design. Thus, secondary problems such as the tolerance of cytotoxic drugs after irradiation are studied. In another recently published study, by 1974 surgery-only control subjects were already omitted in an adjuvant cyclophosphamide, methotrexate, 5-fluorouracil (CMF) trial. The three treatment arms (CMF, CMF plus tamoxifen, and CMF plus tamoxifen plus BCG) showed no difference in time to relapse. To extract at least some useful information the prognostic value of progesterone receptors were analyzed. However, the hundreds of CMF cycles were unnecessary for this information. The similarly designed 1976 Swiss study (chlorambucil, methotrexate, and fluorouracil [LMF] for six months v LMF for 24 months) will only demonstrate the known fact of the oncogenicity of chlorambucil. The lessons from such insufficient study designs unfortunately have not been learned in the eighties: The Ludwig Institute V trial on perioperative and adjuvant treatment lacks adequate surgery-only groups for an important prognostic subset (node-positive postmenopausal patients). For this subset the recent publication of the side effects will probably be the only useful information; however, this type of result never justifies the chemotherapy-induced suffering of these women. The new Swiss trial implemented in 1982 that compares CMF v LMF without surgery-only control subjects in one to three node-positive patients is an example of unnecessary chemotherapy (CMF in node-positive postmenopausal women.) From all the studies without adequate control subjects, the best results we can expect are the type of results presented by Lippman et al. With respect to advances in therapeutic decision making, the years spent on these studies are lost.

This publication has 0 references indexed in Scilit: