Specialist oncological surgery for removal of the ovaries and fallopian tubes in BRCA1 and BRCA2 pathogenic variant carriers may reduce primary peritoneal cancer risk to very low levels

Abstract Risk‐reducing bilateral salpingo‐oophorectomy (RRBSO) is highly effective for the prevention of high‐grade serous ovarian cancer (HGSOC) in BRCA1/2 pathogenic variant carriers (PVCs), but does not completely eliminate future risk of primary peritoneal cancer (PPC). The requirement to completely remove fallopian tubes at RRBSO and carefully exclude occult cancer/serous tubal intraepithelial carcinoma (STIC) lesions may not have been appreciated historically. We calculated rates of HGSOC and PPC in confirmed BRCA1/2 PVCs registered on the regional database in those who did (cases) and did not (controls) undergo RRBSO after genetic testing. Expected annual rates of ovarian/peritoneal cancer were 1% for BRCA1 ≥ 35 years and 0.5% for BRCA2 ≥ 45 years. Follow‐up before 35/45 years was “risk free” and lead time excluded RRBSO <35 years and <45 years for BRCA1 and BRCA2, respectively. Women were followed from personal mutation report (controls) or RRBSO (cases) to death, ovarian/peritoneal cancer or last follow‐up, whichever was sooner. In total, 891 cases (BRCA1 = 468, BRCA2 = 423) and 1302 controls had follow‐up ≥35 years (BRCA1 = 736) and ≥45 years (BRCA2 = 566), respectively, over a total of 7261.1 risk eligible years (mean = 8.15 years). Twenty‐one occult ovarian cancers were found at RRBSO (2.4%), 16 at stage 1. Post RRBSO, 56.97 ovarian/peritoneal cancers were expected but only 3 were observed (HR = 0.053; 95% CI = 0.013‐0.14), with combined Kaplan‐Meier analysis HR = 0.029 (95% CI = 0.009‐0.100, P < .001). Risk reduction was greater in specialist (HR = 0.03; 95% CI = 0.001‐0.13) compared to non‐specialist centres (HR = 0.11; 95% CI = 0.02‐0.37) (P = .07). In controls, 23.35 ovarian/peritoneal cancers were expected with 32 observed (HR = 1.37; 95% CI = 0.95‐1.91). RRBSO <35/<45 years reduces the risk of ovarian/peritoneal cancer by 95% in BRCA1/2 PVCs and may be greater in specialist centres.


| INTRODUCTION
High-grade serous ovarian cancer (HGSOC) is a strongly heritable cancer, with a 3-fold increase in risk of developing the disease in women with first-degree relatives with ovarian cancer. 1 Carriers of germline pathogenic variants (PVs) in BRCA1 or BRCA2 have a high lifetime risk of ovarian cancer, in particular high-grade serous pathology. 2 Women with PVs in BRCA1 or BRCA2 have a cumulative lifetime risk of ovarian cancer of 44% to 61% and 17% to 24%, respectively. [3][4][5] As early detection of ovarian cancer using serum CA125 and transvaginal ultrasound scans has not been effective at reducing sufficient cancers to early stage (1-2), 6,7 and cancers detected on such screening are still associated with high mortality, 8 carriers of PVs in these genes are strongly advised to undergo risk-reducing bilateral salpingo-oophorectomy (RRBSO). 9 This surgery is usually encouraged at or just before the main ovarian cancer risk period starts at 35 to 40 years for BRCA1 and 40 to 45 years for BRCA2. 10 Female carriers of PVs in BRCA1/2 undertaking RRBSO have increased life expectancy mainly due to reduction in ovarian cancer risk, although there may be some reduction in breast cancer risk and reduced mortality from previous breast cancer. [11][12][13] Women undergoing RRBSO are warned of a residual risk of primary peritoneal cancer (PPC), which was first described in 1982. 14 A meta-analysis of studies assessing this risk suggested only a 79% reduction in risk of an "ovarian" type cancer after RRBSO in BRCA1/2 PV carriers 15 (hazard ratio = 0.21; 95% CI = 0.12-0.39). Subsequent review suggested that this residual risk may be mitigated by earlier surgery and was predominantly seen in BRCA1 PV carriers. 16 The origin of PPC has been hypothesised to be due to one of three sources.
Firstly, "ovarian rest cells" displaced during the embryological journey in the abdominal cavity may be a primary origin. Secondly, cells from the fimbria/ovaries could be displaced during adulthood into the peritoneum or thirdly through dissemination at the time of RRBSO.
The latter two possibilities have become more prominent since the description of serous tubal in-situ carcinoma (STIC) lesions in the fimbrial end of the fallopian tube as the probable precursor lesion for most high-grade serous ovarian/tubal cancers. 17 Indeed, the presence of STIC lesions was predictive of future PPC in one single institution study with 2 of 7 (28.6%) developing subsequent PPC compared to only 1 of 287 (0.3%) without STIC. 18 The theory regarding potential prevention of ovarian/tubal cancers by removal of the fallopian tubes has resulted in a number of pilot studies to assess the potential benefits of early tubal surgery and delayed oophorectomy to mitigate the effects of early surgical menopause on subsequent health and quality of life. 19,20 We have previously published an early series of RRBSO where we noted that there were no cases of PPC following 300 surgeries that included 160 BRCA1/2 PV carriers. 21 The recent development of a second PPC in a BRCA2 PV carrier operated on outside a gynaecological oncology centre prompted us to revisit the PPC risk in 891 BRCA1/2 PV carriers who have tested positive in our region and have undergone RRBSO.

What's new?
Women who carry genetic variants in BRCA1/2 that are linked to high-grade serous ovarian cancer are advised to undergo risk-reducing prophylactic bilateral salpingooophorectomy (RRBSO). RRBSO, however, is associated with a significant residual risk of primary peritoneal cancer (PPC). Here, the rate of PPC following prophylactic surgery for high-grade serous ovarian cancer was investigated among 2,193 BRCA1/2 mutation carriers who had undergone RRBSO. Estimates indicate that prophylactic surgery reduced the overall risk of PPC by 95 percent, a far greater reduction than previously described. Residual risk was lowest when RRBSO was carried out at specialist centers compared to non-specialist centers.

| PATIENTS AND METHODS
Female carriers of PVs in BRCA1 or BRCA2 were identified from our regional register covering a population of 5 million in Northwest England as previously described. 10,11 A total of 3653 female PV carriers were identified from 1753 families. Women were eligible if they had undergone RRBSO without any evidence on CA125 and ultrasound of the prior presence of ovarian cancer. The controls were women who had not undergone RRBSO including any time post genetic testing before RRBSO (to avoid bias of not including this follow-up). Cases were followed from date of RRBSO to date of death, PPC or date of last follow-up, whichever was earlier. Controls were followed from date of personal mutation report to date of death, ovarian/peritoneal cancer or date of last follow-up, whichever was earlier. Cases were censored at date of surgery if ovarian cancer was identified as an occult lesion. As we were not aware of any PPC cases postsurgery in any of our BRCA1/2 families and all RRBSO were recorded in the specialty centres for high-risk women since 1980, 21,22 we included follow-up from surgery in those identified as PV carriers after RRBSO.

| Ovarian/tubal/peritoneal primary risk
We chose a conservative estimate of ovarian cancer risk based on the recent prospective series showing a 44% and 17% risk for BRCA1 and BRCA2, respectively, to age 80 years. 4 This is lower than our previous in-house estimates. 10,21 Risk for BRCA1 was considered at 1% annual risk from age 35 years (45% risk to age 80 years) and 0.5% risk for BRCA2 from age 45 years (17.5% risk to age 80 years). Follow-up before 35 years was considered to be "risk free" and calculation of "lead time" excluded RRBSO before the risk period for both genes and <45 years for BRCA2. We also performed an analysis using 5-year cumulative risks from a typical pedigree in BOADICEA v.3 (https:// pluto.srl.cam.ac.uk/cgi-bin/bd3/v3/bd.cgi), but this substantially  Table 1  protocol. The surgical procedures in nonspecialty hospitals did not follow standardised protocols (eg, "bagging" of tubes and ovaries) or submit surgical specimens for specialist gynaecological pathology review.

| Expected ovarian cancers
A total of 56.97 ovarian cancers were expected using life tables (Table 1)

| PPC in whole series
We assessed the proportion of all ovarian cancers in the regional register who had a proven PV. The low level of PPC in the specialty series fits well with the concept that most PPC derives from fimbrial STIC cells that get displaced from the fimbrial ends of the fallopian tubes into the peritoneal cavity. 26  Although the results from the meta-analysis show only a 79% reduction in overall ovarian type cancer risk with RRBSO, some reviews have perhaps misled those counselling about risk after surgery that these are small and of the order of just 1% to 2%. 29 The last review quoted a 1.53% incidence of PPC (28/1830); however, this represents short-term follow-up rather than lifetime risk. 27 The two papers that assessed risk at 20 years came up with risks of 4.3% based on 7 PPCs post RRBSO 30 and 3.9% in BRCA1 carriers based on 5 PPCs. 23 If incidence continued at the same rates, these would be equivalent to at least a risk of 8% by age 80 for a BRCA1 carrier having surgery before 40 years of age. The present study shows convincing evidence for the first time that risks can be reduced below these levels when RRBSO is carried out in a careful, oncologically driven fashion, as the upper 95% CI excludes 8% as described earlier.
Another publication that may have confused the issue was an early report on fallopian tube and peritoneal cancers that suggested risks of only 0.6% and 1.3%, respectively, based on identification of BRCA1/2 in 5/29 (17.2%) and 9/22 (40%) of primary-site tumours analysed. 31 Clearly these risk estimates are incorrect for fallopian tube malignancy 27 and also do not reflect the fact that so many BRCA1/2 PV carriers are identified and opt for RRBSO and that PPC may be due to implantation of cells at the time of surgery.
The current study also identified a high rate of occult Stage There are some limitations to the present study. We do not have full details of procedures carried out beyond the two specialty centres. However, we have good long-term follow-up of a very large series of BRCA1 and BRCA2 PV carriers showing a very low rate of PPC when surgery is carried out in an oncology-driven fashion in a specialty centre. We have not carried out a formal statistical adjustment for factors that might increase ovarian cancer risk.
However, as these factors including an ovarian cancer family history and not having had chemotherapy were significantly more likely in the RRBSO group, this should mean our results are even more robust.
In conclusion, we have shown a very low rate PPC in women undergoing careful RRBSO in a specialist oncology centre. Although even longer follow-up is required to confirm this low PPC rate especially with PPC occurring at older ages, we can already exclude the point estimate of only a 79% reduction in risk from the meta-analysis. 15 Therefore, the rate of PPC may be lower than previously thought if surgery is carried out early in the risk period, performed in a specialty centre and no STIC lesions identified following careful pathological assessment of entire fallopian tubes. Women undergoing such surgery should be told that this reduces their risk by 90% to 95%, and this may be more in those undergoing surgery before the risk period.