Increased prevalence of BRCA1/2 mutations in women with macrotextured breast implants and anaplastic large cell lymphoma of the breast

TO THE EDITOR:

Women with a germline mutation in the BRCA1 or BRCA2 genes have high cumulative risks of developing breast cancer before the age of 80 years (ie, ∼72% and 69%, respectively).¹  To reduce risk, an increasing proportion of BRCA1/2 mutation carriers opt for preventive mastectomy and reconstruction with breast implants. However, breast implants are associated with a strongly increased relative risk (odds ratio = 400) of anaplastic large-cell lymphoma (BIA-ALCL), with a low absolute risk of 1/7000 at age 75 years.^2-4  Host susceptibility factors for BIA-ALCL are largely unknown. As we observed several women with BRCA1/2 mutations, implants and BIA-ALCL, we examined whether BRCA1/2 mutation carriership increases the risk of BIA-ALCL in women with implants.

In December 2018, we identified 49 confirmed cases of BIA-ALCL (median age, 55 years; range, 29-75) via the Dutch nationwide Pathology Database; methods were detailed previously.²  Reasons for breast implants were cosmetic (n = 32), reconstruction after breast cancer surgery (n = 15), or prophylactic mastectomy (n = 2). All BIA-ALCL cases with reconstruction after breast cancer received macrotextured implants, whereas cosmetic cases received other implant types (Table 1). Median interval between insertion of implants to development of BIA-ALCL was 11 years (range, 3-39). Based on medical records of all BIA-ALCL cases, 6 women had BRCA1/2 mutations (BRCA1, n = 4; BRCA2, n = 2). Of the 15 BIA-ALCL cases following breast cancer reconstruction, 4 (26.7%; 95% confidence interval [95% CI], 7.8-55.1) carried BRCA1/2 mutations (median age at breast cancer diagnosis, 51; range, 26-60) (Table 1). To further examine the prevalence of BRCA1/2 mutation carriers in our cohort, we analyzed germline DNA from 18/49 women with BIA-ALCL (supplemental Methods, available on the Blood Web site). Biopsy material of 1 of 6 known BRCA1/2 mutation carriers was included and the mutation confirmed. No germline mutations were observed in the remaining women. Therefore, the prevalence of BRCA1/2 mutations in our entire BIA-ALCL series is at least 12.2% (6/49; 95% CI, 4.6-24.8).

We compared the 26.7% prevalence of BRCA1/2 mutations in BIA-ALCL cases after reconstruction for breast cancer (∼30% of our cohort) with the expected prevalence, based on recently published age-specific prevalence rates of BRCA1/2 mutations in an unselected Dutch breast cancer cohort diagnosed before 50 years.⁵  However, 8/15 women in our cohort were diagnosed with breast cancer after age 49 (median age, 54; range, 50-60). Because no literature is available on BRCA1/2 prevalence for this age group, we chose to apply the estimate for women aged 45 to 49 years as the best available approximation (Table 2).⁵  Based on these data, 5.1% (95% CI 4.6-5.7) of BIA-ALCL cases with breast implants after breast cancer surgery would be expected to carry a BRCA1/2 mutation.⁵  This is significantly lower than our observed estimate of 26.7% (P = .006). Because the prevalence of BRCA1/2 mutations decreases with older age at breast cancer diagnosis,^5,6  the calculated expected 5.1% prevalence overestimates the true expected BRCA1/2 prevalence in breast cancer patients in our cohort of women with BIA-ALCL, rendering the true difference with our observed prevalence an underestimation.

Subsequently, to determine the risk of BIA-ALCL in BRCA1/2 mutation carriers and noncarriers, we calculated the expected proportion of BRCA1/2 mutations in women with breast implants in the general population (supplemental Methods). For women with implants for cosmetic reasons (∼70% of the cohort), we assumed the prevalence to be similar to the general population, for which we used a recently reported estimate of 0.5% (95% CI, 0.5-0.6) based on 50,726 women of predominantly European ancestry⁶  with BRCA1/2 mutations, as classified in ClinVar.⁷  This estimate is in line with other similar studies.^8-10  By combining the expected BRCA1/2 prevalence rates for cosmetic and reconstructive cases with our previously reported overall cumulative risk of BIA-ALCL of 1/7000 at the age of 75 years,²  we estimated the number of women with breast implants with and without BRCA1/2 mutations. Based on (at least) 4 BRCA1/2 mutation carriers with BIA-ALCL and 43 noncarrier BIA-ALCL cases, we then determined the absolute risk of developing BIA-ALCL in BRCA1/2 mutation carriers to be ∼1/1551 (95% CI, 1/5692−1/606) before the age of 75 years, compared with 1/7507 (95% CI, 1/10,373 - 1/5573) in noncarriers with a breast implant (odds ratio = 4.8; 95% CI. 1.7-13.5; P = .012). The BIA-ALCL risk of 1/1551 for women with a BRCA1/2 mutation may be underestimated because (1) the expected age-specific BRCA1/2 mutation prevalence in women with breast cancer aged 50 to 60 was overestimated and (2) we could only determine BRCA1/2 mutation status in 18/49 BIA-ALCL cases.

We excluded the 2 BRCA1/2 cases with bilateral prophylactic mastectomy (BPM) from the risk calculation given previously because BRCA1/2 mutation carriership was the a priori indication for BPM and subsequent breast reconstruction. Nationwide data from the Hereditary Breast and Ovarian Cancer Research Group Netherlands indicate that 1950 Dutch BRCA1/2 mutation carriers underwent BPM, with ∼75% having a reconstruction with implants.¹¹  Therefore, the observation of 2 women with BIA-ALCL in this population (∼1/730) further supports our findings of increased risk of BIA-ALCL in BRCA1/2 mutation carriers.

The currently estimated risk for BIA-ALCL in women with BRCA1/2 mutations applies to the Dutch population; these findings need to be validated in other BIA-ALCL series. Recently, a prospective single institution study from Memorial Sloan-Kettering Cancer Center, NY, NY, presented an exceptionally high risk for BIA-ALCL in women with implants after breast cancer surgery (1/355).¹²  At least 5 of 10 BIA-ALCL cases had a previous contralateral prophylactic mastectomy.¹³  Possibly, this high risk is at least partly related to specific features, including genetic characteristics, of the patient population in the adherence area of this single institution.

Our study has several limitations. First, if BRCA1/2 mutation carriers with breast cancer would more often undergo mastectomy (with reconstruction) than lumpectomy, we may have overestimated BIA-ALCL risk in carriers compared with noncarriers. However, a recent Dutch study shows that breast cancer recurrence rates in BRCA1/2 mutation carriers (and noncarriers) do not differ between mastectomy and lumpectomy, suggesting that this bias may be small.¹⁴  Second, we did not account for the number of implants per woman, although BRCA1/2 mutation carriers with breast cancer likely have a higher rate of bilateral implants than non-BRCA1/2 breast cancer patients because of increased rates of contralateral breast cancer and prophylactic contralateral mastectomy.¹⁵  Higher bilateral implant prevalence may have led to some overestimation of our calculated BIA-ALCL risk in BRCA1/2 mutation carriers. The extent of this bias is unclear, however, because we actually do not know whether bilateral implants increase risk of BIA-ALCL compared with unilateral implants. Third, BRCA1/2 mutation testing could only be performed in 18/49 women; as a consequence, our risk estimates are conservative. Strengths of our study include the complete nationwide ascertainment of BIA-ALCL cases, histopathological confirmation of all cases, and the availability of complete clinical data, including implant type. Because all breast cancer patients in this study, both BRCA1/2 carriers and noncarriers, had macrotextured breast implants, confounding by “high-risk” implant types can be excluded.^16-19 

This study has been performed in the context of a breast cancer population with macrotextured breast implants. If validated in larger international cohorts, the results of this study may have important implications for breast reconstruction options after breast cancer surgery and prophylactic mastectomy in women with established BRCA1/2 mutations. Such implications would include personalized patient information for BRCA1/2 mutation carriers opting for implants and promotion of alternative autologous breast reconstruction procedures.