Commentary: With Regards to the WISDOM trial.

In a recent publication in JAMA, Laura Esserman and her colleagues report on the results of the WISDOM trial [1]. This extensive study of more than 28,000 women aged 40 to 74 compared traditional annual mammographic screening with risk-based screening, i.e., screening tailored to a woman’s risk of developing breast cancer. In the risk-based screening arm, women were offered one of four different screening protocols, depending upon whether they were judged to be at low risk, average risk, elevated risk or high risk for developing breast cancer. Risk was determined using a combination of family history, mammographic density, polygenic risk score (PRS) and genetic testing for pathogenic variants in three high-risk genes (BRCA1, BRCA2, and PALB2) and in six moderate-penetrance genes (CHEK2, ATM, CDH1, PTEN, STK11 and TP53.) There were 291 women assigned to the high-risk group (2%) who were offered yearly MRIs. These included 159 women with a BRCA1, BRCA2 or PALB2 mutation. The other women were offered mammograms, starting at age 40 or 50.
In total, there were 260 cancers diagnosed in the annual screening arm and 263 in the risk-based arm. There were fewer mammograms in the risk-based arm, but about the same number of biopsies. In the risk-based arm, 164 cancers (63%) were diagnosed in the average risk group (who underwent annual mammography), and 25 cancers were diagnosed in the high-risk group. The differences in the stage at diagnosis of the 263 cancers and the 260 cancers were small. In the annual mammography arm, there were 31 cancers diagnosed at stage 2b or higher, compared to 21 cancers in the risk-based arm. Much of the benefit seemed to derive from MRI screening, leading to down-staging. All 25 cancers in the high-risk group were Stage IIA or lower, suggesting that MRI was helpful. It is too early to compare the number of deaths.
It is of interest to consider those women who were found to carry a pathogenic variant and who were assigned to the high-risk group. In an accompanying article, Fergus et al supply the details [2]. Among the 23,098 women who were tested, 714 (3.1%) carried a pathogenic variant, including 33 in BRCA1, 82 in BRCA2 and 44 in PALB2. 70% of the carriers of these three high-penetrance genes would have qualified for genetic testing under normal circumstances (i.e., they had a first- or second-degree relative with breast or ovarian cancer or male breast cancer or were Jewish) and 30% of them had no prior indication for testing. Among the 10,144 women with no family history, 44 were found to have a mutation in BRCA1, BRCA2 or PALB2 (0.4%). In contrast, 115 mutations were found in the 3393 women who had an indication for testing (3.4%). This implies that 72% of the high-risk mutations were found in the subgroup of 25% of the patients who would qualify for testing under normal circumstances and 28% of the mutations were found in women who would not qualify.
This raises the important question of whether or not testing for high-penetrance mutations should be extended to all women. We have good reasons to consider doing so. MRI screening has been shown to be superior to mammography in women with a BRCA1 or BRCA2 mutation [3]. For women with a mutation, options go well beyond screening. In North America, the majority of unaffected women with a BRCA mutation will opt for risk-reducing bilateral salpingo-oophorectomy (BSO) and about one-half will opt for risk-reducing mastectomy [4]. BSO is recommended for BRCA1 carriers at age 35–40 and for BRCA2 carriers at age 45. BSO has been shown to be highly effective in preventing ovarian and fallopian tube cancer and in reducing all-cause mortality [5]. It appears to be safe for these women to alleviate the symptoms of surgical menopause with menopausal hormone therapy [6]. After risk-reducing mastectomy, the chances of dying of breast cancer fall to 1% at 15 years [7]. Further, for those who do develop breast cancer, olaparib (a PARP inhibitor) is now indicated for high-risk BRCA-positive breast cancer patients and has been shown to reduce mortality [8]. Survival is better for BRCA1 and BRCA2 mutation carriers who know their mutation status at the time of diagnosis versus those who were tested after [9]. Chemoprevention is also an option, but despite encouragement by WISDOM study personnel, few women in the elevated- and high-risk groups chose to take it (4% in the elevated-risk group and 10% in the high-risk group). Perhaps this is due to concern regarding the side effects of tamoxifen and the lack of evidence that chemoprevention is effective in BRCA1 and BRCA2 carriers. In our study of 4,578 unaffected BRCA carriers, only 5% chose to take tamoxifen or raloxifene and the benefit of the intervention did not achieve statistical significance (HR = 0.64; 95% CI 0.40-1.03; P = 0.07) [10].
One important topic the trial doesn’t address is the paradigm of cascade testing. How many of the 159 women who were found to have a high penetrance mutation had a healthy female relative who subsequently went for testing? Once a mutation is identified in a woman, unaffected relatives are eligible. Cascade testing should increase the mutation detection count and increase the return on the investment, although up until now, cascade testing has been floundering. If one could identify one additional unaffected carrier per mutation-positive family, this would double the return per dollar invested.
If we extend genetic testing from those with an indication to all women, we will increase the testing rate by four-fold but only increase the mutation yield by 40%. Is it worth it? I think so - the key statistic here is the cost of a test. If the cost of testing can be reduced to the cost of a mammogram, say 150 dollars, I would argue that we should it offer it to all women. A genetic test needs to be done once, whereas a mammogram might be repeated 34 times! To sum up, my opinion is that we should offer genetic testing to everybody. We should encourage those women who have the resources to pay for it to do so and we should strive to see that there are means of accessing testing for those who do not. We should put much more effort into cascade testing.