Prophylactic mastectomy for BRCA1/2 carriers: progress and more questions.

Abstract

In the mid 1990s, genetic testing to identify carriers of BRCA alterations came into the clinical arena. But once a high-risk carrier was identified, what did we have to offer to reduce her risk? Regarding prophylactic mastectomy, a high-risk woman was often confronted with strongly held views— both for and against the procedure—with no supporting data. The literature contained several case reports of chest wall cancer occurring after prophylactic mastectomy [1-3]. These, along with a widely cited negative study of prophylactic surgery in a rat mammary model of chemical carcinogenesis [4], led to understandable concerns that the procedure might be ineffective in high-risk women. The 1997 Consensus Statement for BRCA1/2 carriers by the Cancer Genetics Studies Consortium [5] stated that no recommendation for prophylactic mastectomy could be made because of lack of evidence of benefit. Since that time, there have been retrospective and prospective studies of prophylactic mastectomy showing a high degree of risk reduction with the procedure [6-8]. The current report adds additional information regarding the efficacy of prophylactic mastectomy in BRCA1/2 carriers [9]. Rebbeck et al [9] have assembled a group of 105 BRCA1/2 mutation carriers who had bilateral prophylactic mastectomy. They followed this surgical group for 5.5 years and two of the women subsequently developed breast cancer at 2.3 and 9.2 years after prophylactic mastectomy. Both had subcutaneous mastectomy, and the authors say that one of these cancers developed in “substantial residual” breast tissue. Some points to take away from this study: 1) If your patient decides to have prophylactic mastectomy, do not leave substantial breast tissue behind. The total mastectomy is the procedure of choice; 2) their findings support the procedure’s efficacy. Only 1.9% of these highest-risk women developed breast cancer after prophylactic mastectomy over 5.5 years of follow-up. What might have been expected in these women had surgery not been performed? Enter the control group and the weakest link in this study. The investigators identified 378 matched controls (matched by study center, BRCA1 versus BRCA2 mutation, and age). Over a mean follow-up of 6.4 years, 184 of them (49%) developed breast cancer. That is a remarkably high likelihood of breast cancer, even in mutation carriers. What, besides their mutation status, might explain this? In their overall analysis, all events that had occurred in cases and controls before their visit to the study center could be included (their group 1). Women who had previously had bilateral prophylactic mastectomy (BPM) were included. Controls had to be cancer-free when their matched cases had prophylactic surgery, but they could have developed breast cancer before being seen at the center. It is certainly conceivable that the development of a breast cancer, or a recurrence of breast cancer, would motivate women to be seen at a major medical center. This selection bias could artificially increase the number of breast cancers in the control group and, thus, overestimate the benefit of the procedure. A hypothetical example may be helpful: hypothetical case 1 (Fig 1) had her BPM in 1988. She was first seen at her respective high-risk center in 1991 and she was last followed in 1996, cancer-free. Control number 1 (Fig 2) had to be cancer-free in 1988, when her matched case had her surgery. But this control developed breast cancer in 1994 and was seen subsequently at a major center. Recognizing this potential bias, the authors went a step further to their analysis 3, in which they excluded events that had occurred before a participant’s first center visit. This drops the number of cases from 102 to 57, and the number of controls from 378 to 107. Now the outcomes clock is reset, and only events after the first clinic visit count. But does this eliminate bias in the controls? What, besides JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 22 NUMBER 6 MARCH 15 2004