Introduction

Lynch et al. [1–8] at Creighton University may best be known for their characterization over 40 years ago of autosomal dominant transmission of a susceptibility trait for non-polyposis colorectal cancer with predilection to the right colon and high risks for endometrial and ovarian cancers in female family members and also increased risks for carcinomas of the stomach, small bowel and hepatobiliary tract, pancreas, renal pelvis, ureter and brain tumors, particularly glioblastomas. This hereditary disease, marked by no other grossly recognizable phenotype, is now termed “Lynch syndrome” [9]. Less well recognized are the concurrent studies that Lynch and his group published during that time showing multigenerational family pedigrees with inordinate numbers of breast cancers leading to their identification of autosomal dominant hereditary susceptibility to both breast and ovarian cancers at unusually young ages in some of these kindreds [10–16].

Soon after Hall et al. [17, 18] reported linkage of early-onset breast cancer to a site in chromosome 17q, Narod and associates demonstrated that hereditary breast ovarian cancer (HBOC) susceptibility in families from the Creighton Hereditary Cancer Registry was linked to a locus at 17q12–q23. This gene and a separate gene locus on chromosome 13q12.3, also linked to the HBOC syndrome, have been cloned and designated BRCA1 and BRCA2, respectively [19–22]. The Breast Cancer Consortium reported that over 90 % of their families with female breast cancer and two or more ovarian cancers were linked to mutations in BRCA1 or BRCA2 [23]. Meta-analysis of penetrance studies from Asia, North America and Europe reported a range of cumulative cancers by age 70 years of 47–66 % (mean 57 %) for breast cancer and 35–46 % (mean 40 %) for ovarian cancer in BRCA1 mutation carriers and of 40–57 % (mean 49 %) for breast cancer and 13–23 % (mean 18 %) for ovarian cancer in BRCA2 mutation carriers [24].

Because of the high risk for ovarian cancer and 50 % likelihood that the deleterious autosomal dominant susceptibility trait would be inherited in HBOC syndrome families, Lynch and colleagues recommended that prophylactic oophorectomy should be considered by the female members [12]. In spite of the intent to prevent the morbid effects of ovarian cancer in families prone to this disease by prophylactic oophorectomy, it became evident in early studies that, although extirpated ovaries appeared grossly and histologically normal, a fraction of the women with hereditary predilection to gynecologic cancer still developed intra-abdominal carcinomatosis indistinguishable from metastatic ovarian carcinoma [25–27].

Our study of HBOC syndrome families linked to BRCA1 or BRCA2 mutations in the Creighton Hereditary Cancer Registry found only five mutation carriers who were diagnosed with intra-abdominal carcinomatosis in 238 individuals who had undergone prophylactic oophorectomy, giving a calculated cumulative risk of 3.5 % over 20 years post surgery [28]. Histological examination of the submitted ovarian specimens from these five cases found no preinvasive or invasive cancers in three, but superficial papillary serous borderline tumors, one with possible early invasion, were found on ovarian sections in two cases [28]. Recent clinical-pathological and molecular genetic studies favor different carcinogenic pathways for most borderline ovarian tumors contrasted with high grade carcinomas, such as serous intra-abdominal carcinomatosis in BRCA1 and BRCA2 linked HBOC families [29–31]. This model would not preclude malignant deterioration of superficial implants from extraovarian preinvasive or invasive serous carcinomas or the advance of ovarian intraepithelial serous lesions and tumors into full blown cancers by natural progression or by spontaneous or epigenetic loss of heterozygosity (LOH) in women who carry inherited deleterious mutations that predispose to “ovarian” cancer [29–33]. Microscopic malignancies, such as these, overlooked during routine surgical pathological examination, whether primarily arising from ovarian surface epithelium or tiny surface implants from other malignant sites, could account for some of those few cases of intra-abdominal carcinomatosis that occur many months or years after intended prophylactic surgery.

Following the speculation of Foyle et al. [34] that widespread peritoneal papillary serous carcinoma in ten of their female patients with no evidence of ovarian carcinoma arose primarily from “extraovarian mesothelium with the same müllerian potential as ovarian surface (germinal) epithelium”, Tobacman et al. and Lynch et al. [25, 27] hypothesized that intra-abdominal carcinomatosis after oophorectomy in women at high genetic risk for ovarian cancers may result from malignant transformation of embryologically related epithelium of the müllerian system and coelomic epithelium of the peritoneum and ovarian surface. According with current concepts of normal human female embryology, primordial germ cells migrate from the yolk sac into mesenchyme posterior to the coelomic cavity during the fifth week of gestation to form bilateral thickened genital ridges just medial to the mesonephros and mesonephric ducts [35]. During the sixth week of gestation, cells from coelomic epithelium covering the genital ridges grow inward to invest the germ cells, and as embryo to fetal development progresses eventually surround germ cells to form primary follicles around individual oocytes. Meanwhile, a pair of paramesonephric ducts, more commonly called müllerian ducts in the female, form in a thickened band of proliferating mesoderm posterior to the coelomic cavity and lateral to the mesonephros and genital ridges [35]. As fetal growth and development continue, the primitive ovaries and bilateral müllerian systems descend caudally, both moving through coelomic lining epithelium, which becomes parietal peritoneum and the visceral peritoneum covering these pelvic organs and affiliated structures. Reaching the primitive pelvis, the distal müllerian ducts fuse in the midline and meet the invaginating urogenital sinus to form the upper vagina. The midline septum that resulted from fusion of the müllerian system then begins involution which continues cephalad until functionally resulting in a single uterine cervix and corpus and bilateral fallopian tubes [36]. Thereby, the epithelia of the endocervix, endometrium and endosalpinges derive from the müllerian system; while epithelium covering the pelvic viscera derives from coelomic epithelium.

Fallopian tube neoplasms

Piek et al. [37] reported that 94 % of ovarian (42/45), fallopian tube (3/3) and peritoneal cancers (2/2) in BRCA1 or BRCA2 mutation carriers from the Amsterdam Family Cancer Clinic were invasive serous carcinomas compared to just 62 % of cases with this pathological histotype in the population based Netherlands Cancer Registry. Two of Piek’s [37] three presumed cases of primary peritoneal serous carcinoma were in women who had undergone oophorectomy but not salpingectomy. Our 52 years review of gynecologic cancers in BRCA1 and BRCA2 mutation carriers from HBOC syndrome families in the Creighton Hereditary Cancer Registry found that 70 % of ovarian cancers (45/64), 86 % of fallopian tube cancers (12/14) and all of the cases classified as primary peritoneal cancers (8) were dedifferentiated serous or anaplastic carcinomas [38].

Among the early pedigrees of familial association of breast and ovarian carcinoma that Lynch and his associates reported in 1974 was one family member with papillary serous carcinoma involving both fallopian tubes and ovaries, and in 1975, Fraumeni et al. reported a primary fallopian tube carcinoma in a family with three breast and two ovarian cancers over two generations of first degree relatives [12, 39]. After linkage of most HBOC families to cancer-associated mutations in BRCA1 or BRCA2, several cases of microscopic invasive and in situ fallopian tube carcinomas were reported in BRCA1 and BRCA2 mutation carriers, some with malignant peritoneal cytology [40–45]. A sub-study reported in 2002 by the Breast Cancer Linkage Consortium estimated a highly significant (P < 0.001) increased relative risk of 49.94 (95 % CI 22.48–110.94) for fallopian tube cancer in mutation carriers from breast and/or ovarian cancer families with at least one individual known to carry a pathogenic mutation in BRCA1 compared with population incidence rates from the United Kingdom East Anglian Cancer Registry [46]. In 2003, Piek et al. [47] recounted that 37 % of the fallopian tubes that were prophylactically removed from women at high risk for breast and ovarian cancer showed dysplastic lesions, including one with in situ carcinoma, and they hypothesized that most hereditary serous carcinomas involving ovaries actually arise in fallopian tube epithelium Since then, the reports of occult invasive fallopian tube serous carcinomas and distinct noninvasive serous tubal intraepithelial carcinomas (STIC) in prophylactic surgical specimens from HBOC syndrome mutation carriers have continued to accumulate [48]. Many of these cases were associated with malignant cells in the peritoneal fluid, intra-abdominal carcinomatosis or later recurrent cancers [49–54]. Not only do STIC morphologically resemble high grade serous carcinomas (HGSC) involving pelvic and/or abdominal peritoneum and previously identified as “ovarian” or “primary peritoneal” cancers, but in more than 90 % of tested cases identical T53 mutations were shared by STIC and the co-existing HGSC [55–58]. Immunohistochemical (IHC) staining for p53 is quite common in STIC, though not consistent [55, 59]. However, when staining for p53 is found in HGSC, as a surrogate for T53 mutations, usually it is expressed in accompanying STIC, as well [55, 57]. On the other hand, complete absence of p53 expression in STIC correlates with frameshift mutations in T53, due to deletions, insertions, splicing junction and nonsense mutations [57]. Foci of p53 IHC straining found in otherwise morphologically normal tubal epithelium have been termed “p53 signatures”, and some investigators have interpreted these as possible precursor lesions from which STIC may develop in genetically susceptible individuals; though this remains controversial [48, 56, 59]. Finally, when the entire fallopian tube could be examined, STIC have been identified in more than half of the HGSC cases classified as primary peritoneal cancers [58, 60]. Such evidence as the foregoing has led to a growing consensus that many, if not most, cases of serous intra-abdominal carcinomatosis in HBOC syndrome mutation carriers, originate from fallopian tube neoplasms, often to involve adjacent ovaries and disseminate to the peritoneum and intraperitoneal organs (Fig. 1) [50, 58, 60].

Fig. 1
figure 1

Distal fallopian tube prophylactically removed from BRCA1 mutation carrier with high grade serous carcinoma arising in serous tubal intraepithelial carcinoma (STIC) (arrow) (H&E ×100). A single <1 mm implant was on the ipsilateral ovary. Four years after risk-reducing surgery the patient presented with disseminated serous intra-abdominal carcinomatosis

Full size image

Uterine neoplasms

Early attention to the possibility of increased susceptibility to uterine cancers in HBOC syndrome mutation carriers comes from a large study conducted by the multinational Breast Cancer Linkage Consortium, which included 11,847 individuals from 699 families with breast and/or ovarian and at least one member known to carry a pathogenic mutation in BRCA1 [46]. This group calculated estimated increased relative risks of 2.86 (95 % CI 1.69–4.16) for uterine corpus cancer and 3.72 (95 % CI 2.26–6.10) for uterine cervix cancers which were highly significant (P < 0.001) compared with general population cancer risks specific to country, calendar period and 5-year age groups [46]. In that study, the relative risks for uterine corpus and cervix cancers exceeded even those for colorectal, biliary and pancreatic cancers but not fallopian tube cancers in BRCA1 mutation linked families [46]. Although other hereditary factors could account for these increased cancer risks and the uterine cancer histotypes were not specified, this publication demands attention to uterine epithelium as a possible site for malignant transformation that may lead to intra-abdominal carcinomatosis [46].

Endometrium

Predilection of BRCA1 and BRCA2 mutation carriers to preinvasive and invasive serous fallopian tube carcinoma and a higher than expected overall 31 % (33/108) prevalence of BRCA1 or BRCA2 mutations in patients with fallopian tube cancers, especially a 59 % (10/17) prevalence in selected Jewish women who may carry one or another of the well defined Ashkenazi mutations in BRCA1 (185delAG and 5382insC) or BRCA2 (617delT) leads to question whether HBOC syndrome mutation carriers may be susceptible to serous carcinoma in other müllerian-derived epithelium which then seeds the peritoneum [31, 61]. Some 15 years ago, Hornreich et al. [62] reported cases of endometrial serous papillary carcinoma in a 53 year old woman of Ashkenazi descent and an ovarian papillary carcinoma in her 48 year old sister; both women carried the same BRCA1 5382insC mutation. A subsequent Israeli study of 20 consecutive Ashkenazi women with uterine serous papillary carcinomas found that four subjects (20 %), diagnosed at 65 years mean age, carried Ashkenazi BRCA1 mutations (three 185delAG and one 5382insC) [63]. Members of this study group with others later reported uterine serous papillary carcinoma in a 65 year old Ashkenazi women who carried a BRCA2 617delT mutation [64]. Albeit, two early studies of uterine serous cancers in Canada (1994–2000) and endometrial serous carcinomas in New York Jewish women (1986–1998) found no BRCA1 or BRCA2 mutations in 56 and 17 patients, respectively [65, 66]. And a collaborative Israeli study over 26 years (1982–2008) found no Ashkenazi BRCA1 or BRCA2 mutations in 34 patients with endometrial serous carcinoma [67]. However, two other series of endometrial serous carcinomas in Israel (updated 1997–2007 and 1999–2008) found that 25.8 % (8/31) and 15.7 % (8/51) of those women, respectively, carried an Ashkenazi mutation, which was significantly higher than the 2.5 % incidence of these mutations in their Ashkenazi population [68, 69]. A more recent American study of 151 women with uterine serous carcinoma for mutations in 30 tumor suppressor genes, including BRCA1 and BRCA2, found three subjects with mutations in BRCA1, a prevalence of 2 % in unselected women with this disease, which exceeds the expected frequency of 0.06 % BRCA1 mutations in the general population [70]. An interesting report from the surgical arm of Gynecologic Oncology Group study GOG-0199 found six primary endometrial cancers among 966 “high risk” patients, including 326 BRCA1 mutation carriers and 231 BRCA2 mutation carriers who underwent risk-reducing salpingo-oophorectomy (RRSO) [53]. Only 515 of these patients also had hysterectomy [53]. Two endometrial cancers, both serous carcinomas, were diagnosed in BRCA1 mutation carriers, and four endometrial cancers, two endometrioid and two mucinous carcinomas, were found in non-carriers. There were no endometrial cancers in BRCA2 mutation carriers [53]. We found that eight cases (7.9 %) of endometrial cancer could be confirmed in 101 BRCA1 and BRCA2 mutation carriers with gynecologic and peritoneal cancers from HBOC syndrome families accrued to the Creighton Hereditary Cancer Registry over the past 55 years [71]. Three of these eight cases of endometrial cancer in HBOC syndrome mutation carriers were high grade pure serous carcinomas or contained greater that 10 % high grade elements of serous carcinoma mixed with poorly differentiated endometrioid carcinoma (Fig. 2). The other five patients had well or moderately well differentiated endometrioid carcinomas (Fig. 3) [71].

Fig. 2
figure 2

Endometrial high grade serous carcinoma in tamoxifen treated BRCA1 mutation carrier presenting with disseminated serous intra-abdominal carcinomatosis (H&E ×200)

Full size image
Fig. 3
figure 3

Endometrial endometrioid adenocarcinoma, FIGO grade 1, in a BRCA1 mutation carrier previously treated with hormone replacement for 7 years (H&E ×100)

Full size image

High grade serous carcinomas arising in endometrium are rare and aggressive tumors, making up fewer than 5 % to no more than 12 % of uterine cancers in population based international cancer registries and a large American group trial [72–76]. Even when small, superficial or a minor component admixed with endometrioid carcinoma, endometrial high grade serous carcinomas are prone to lymphatic involvement and/or transtubal dissemination [74–84]. Most endometrial serous carcinomas, whether invasive or intraepithelial, are characterized by TP53 mutations, which often are shared with co-existing serous intra-abdominal carcinomatosis involving ovaries, fallopian tube serosa and peritoneum [81, 84]. Invasive endometrial serous carcinomas frequently coexist with endometrial intraepithelial carcinomas (EIC) [82]. Even when unaccompanied by invasive endometrial serous carcinoma, cases of EIC have been described variously with free-floating malignant cells in the fallopian tube, cancerous involvement of the ovary and metastases to pelvic and upper abdominal serosa and omentum [79, 81, 84] Simultaneous EIC and STIC have been reported with extrauterine extension or metastases [79, 82, 84], and shared TP53 mutations in EIN and STIC have been demonstrated in a couple of the cases [79]. Extensive sectioning of the entire fallopian tube and endometrium in a recent Dutch study discovered EIC in 9/60 (15 %) and STIC in 23/54 (43 %) of 60 cases operated for ovarian serous carcinoma [85]. In six cases these intraepithelial carcinomas coexisted, and in 19 cases there were invasive serous carcinomas of the fallopian tube and/or endometrium [85]. Four (17 %) of the 23 cases of STIC were non-fimbrial [85]. These observations and studies would not preclude transit of transformed neoplastic cells from primary fallopian tube sites, the ovary or peritoneum into the endometrium from which EIC and invasive serous carcinoma are manifest. However, tumorous involvement of endosalpinx and ovarian surface with upper abdominal dissemination have been described in cases of EIC without the presence of STIC [79, 81]; while EIC have been reported with and without synchronous STIC but accompanied by extrauterine serous carcinomas that carried TP53 mutations in common with EIC alone, STIC alone or both EIC and STIC [84]. Given the likelihood that EIC and STIC are preinvasive cancers capable of seeding peritoneum and intraperitoneal organs favors a concept that either lesion and their invasive counterparts could be sources of intra-abdominal carcinomatosis in susceptible individuals.

Endocervix

Uterine endocervical epithelium, like endometrium and fallopian tube epithelium, is a müllerian derivative [34] We reported three cases of uterine cervix carcinoma in 95 HBOC syndrome mutation carriers with invasive gynecologic cancers in the Creighton Hereditary Cancer Registry [38]. Two of these cases demonstrated endocervical papillary serous carcinoma, one of which extended to endometrium [38]. The other primary cervical cancer in an HBOC syndrome mutation carrier was a minimally invasive (<3 mm) squamous carcinoma [38]. Serous carcinoma, the rarest cervical cancer histotype, is aggressively capable of lymphatic spread and/or dissemination to the peritoneum and to pelvic and abdominal organs [86, 87] Endocervical invasive serous carcinoma accompanied by intraepithelial serous carcinoma and then the diagnosis of peritoneal carcinomatosis 24 months later has been reported in a young woman in her fourth decade whose identical twin and mother also died at young ages with peritoneal carcinomatosis and whose maternal grand mother and aunt had bilateral breast cancers; but no genetic testing for mutations in BRCA1 or BRCA2 were done on any of these individuals [87]. Whether or not germline mutations in BRCA1 or BRCA2 enhance susceptibility to either endocervical or squamous cell carcinoma of the uterine cervix, cancer-associated defects in these genes certainly should not be expected to convey immunity against those diseases.

Implications for risk-reducing surgery

The ultimate objectives of patient care are to prevent and relieve suffering from disease and to promote healthfulness. The present communication is directed toward preventing and relieving the morbidities of intra-abdominal carcinomatosis and mortality that may result therefrom in women with inherited susceptibility. This review has implicated several epithelial tissues as possible sites for primary malignant transformation from which intra-abdominal serous carcinomatosis may originate in mutation carriers from HBOC syndrome families: to wit, (1) ovarian surface and possibly inclusion cyst epithelium and the peritoneum which arise from the embryonic coelomic lining and (2) endosalpingeal, endometrial and endocervical epithelium that arise from the embryonic müllerian system.

Though several studies have shown statistically significant risk reductions for ovarian cancer with combined oral contraceptives (COC) used for more than a year in both BRCA1 and BRCA2 mutation carriers, use of these preparations for ovarian cancer chemoprophylaxis is guarded because they might be associated with increased risks for breast cancer in carriers of mutations in either gene [88, 89]. Also, the efficacy of presently available screening techniques intended to reduce morbidity and mortality attributed to serous ovarian cancers in high risk women is unproven [89, 90]. Therefore, prophylactic surgery intended to reduce the risk of gynecologic serous carcinomas remains the standard for cancer prevention in HBOC syndrome mutation carriers [89, 91].

For more than a dozen years prophylactic oophorectomy and/or salpingo-oophorectomy have been known to significantly reduce the risk for both breast cancer and serous peritoneal, fallopian tube or ovarian cancers in women from HBOC syndrome kindreds who carry BRCA1 or BRCA2 mutations [92, 93]. As reviewed above, accumulating molecular, pathological and translational studies of organ and tissue specimens from women who carry cancer-associated mutations in BRCA1 and BRCA2 have concluded that many or most pelvic and peritoneal serous carcinomatosis of gynecologic origin arise from fallopian tube epithelium. This has led some authors to speculate whether risk-reducing surgery in premenopausal HBOC mutation carriers during their fifth decades of life could be confined to salpingectomy while conserving the ovaries [94, 95]. Sparing the ovaries should avoid symptoms associated with an abrupt decline of endogenous ovarian hormones, such as vasomotor hot flushes, night sweats and sleep disturbance, diminished libido and eventual vaginal dryness and dyspareunia, that are experienced by some women following premature surgical menopause and may be protective against osteoporosis, cardiovascular disease and cognitive impairment with aging [94, 95]. On the other hand, whatever protection against breast cancer attributable to risk-reducing salpingo-oophorectomy (RRSO) in mutation carriers may be lost if the ovaries are left in situ [96]. Prophylactic mastectomies alone reduce the risk for first primary breast cancer in HBOC syndrome mutation carriers by at least 90 % [97–99]. While the addition of RRSO to prophylactic mastectomies may enhance this protection against breast cancer and reduce overall mortality [98, 100–102]. Published results from a Markov Monte Carlo simulation model developed to compare (1) bilateral RRSO at age 40 years with (2) bilateral salpingectomy alone at age 40 years or (3) with bilateral salpingectomy at age 40 years and delayed oophorectomy till age 50 years in BRCA1 and BRCA2 mutation carriers showed that RRSO at age 40 years was associated with higher life expectancy and lower cost than the other alternatives [96]. According to this model, when compared to the second alternative, bilateral salpingectomy at age 40 years; the strategy of bilateral RRSO at age 40 years reduced the risk for ovarian cancer by 22.8 % in BRCA1 and 20.5 % in BRCA2 mutation carriers, and the risk for breast cancer was reduced 22.6 % in BRCA1 and 39.7 % in BRCA2 mutation carriers [96]. Whereas, with bilateral salpingectomy at age 40 years and delayed oophorectomy at age 50 years, the risk for ovarian cancer was reduced only 14.6 % in BRCA1 and 13.1 % in BRCA2 mutation carriers, and the risk for breast cancer was reduced only 0.4 % in BRCA1 and 1.1 % in BRCA2 mutation carriers [96]. When quality-of-life measures were included in the model, salpingectomy at age 40 years with delayed oophorectomy at age 50 years yielded the highest quality-adjusted life-expectancy, but the simulations did not include hormone replacement therapy (HRT), which if added may have improved quality-adjusted life expectancy with RRSO at age 40 years to exceed the predicted quality-adjusted life-expectancy with prophylactic salpingectomy at age 40 years and delayed oophorectomy to age 50 years [96].

Earlier simulation models also demonstrated improved life expectancy in BRCA1 and BRCA2 mutation carriers who undergo prophylactic oophorectomy [103–108]. These gains begin to diminish when prophylactic surgery is performed after age 30 years, especially if oophorectomy is delayed beyond age 40 years [103–108]. It is beyond the scope of this communication to discuss the merits and weigh the outcomes of various decision models. Suffice it to note that the reviewed reports concur in demonstrating improvements in life expectancy for regimens that include oophorectomy by age 40 years in women at high risk for breast and ovarian cancers.

Pending definitive investigations into early salpingectomy and delayed oophorectomy, some have advocated that with careful discussion and patients’ understanding of risk, at least salpingectomy should be offered to mutation carriers who are set against premenopausal removal of their ovaries [95]. Surveys indicate interest for being engaged in such studies by women at increased hereditary risk for breast and ovarian cancers [109, 110], and at least one prospective trial is being set up to investigate cancer risk-reduction and quality of life with salpingectomy upon completion of childbearing and delayed oophorectomy and compare this with RRSO [111]. Already there is a published report of this planned approach in the care of a 41 year old carrier of the Ashkenazi 187delAG mutation in BRCA1, whose mother and maternal aunt carried the same mutation and were victims of breast cancer [112]. In spite of knowing this, the patient still was reluctant to have her ovaries removed; so after being advised of recommendations for RRSO, she instead chose minimally invasive salpingectomy [112]. Surgical pathology discovered a unilateral fallopian tube 2 mm serous carcinoma for which she was treated with hysterectomy-bilateral oophorectomy, pelvic and para-aortic lymphadenectomies and omentectomy, all with no other evident cancer [112]. After adjuvant platinum-based chemotherapy, the patient was reported to be disease-free 7 months following the diagnosis of occult cancer which likely would have gone unrecognized had not at least salpingectomy been carried out [112].

Review of cases that were reported during the past 5 years since adoption of rigorous surgical pathology protocols to thoroughly examine the entire fallopian tube and ovarian specimens of RRSO from BRCA1 and BRCA2 mutation carriers, finds that 71 % of the occult preinvasive and/or invasive serous carcinomas involved only fallopian tubes, 14 % involved both fallopian tubes and ovaries, and 15 % involved only ovaries (Table 1) [49–54]. When peritoneal fluid was taken for cytology at the time of RRSO in these cases, 24 % of the occult preinvasive and 40 % of the occult invasive carcinomas were positive for malignancy (Table 2) [49–54]. But only 7 % of the preinvasive carcinomas compared with 27 % of the occult invasive carcinomas discovered at RRSO were accompanied by metastatic disease or diagnosed with recurrent pelvic-abdominal cancer during follow-up [49–54]. Because the objective of prophylactic gynecologic surgery in HBOC syndrome mutation carriers is to reduce the risk of disseminated cancer, it is judicious to eliminate the organ tissues from which the malignancies may arise and to which they first extend. A meta-analysis published in 2009 concluded that salpingo-oophorectomy reduced the risk for “ovarian” cancer by at least 80 % and the risk for breast cancer by some 50 % in mutation carriers from HBOC syndrome families [91]. And subsequent publications have reported even more impressive risk reductions of ovarian and breast cancers in BRCA1 and BRCA2 mutation carriers after RRSO [90, 100–102].

Table 1 Proportion of occult intra-epithelial and invasive serous carcinomasa found involving fallopian tubes and/or ovaries from risk-reducing salpingo-oophorectomy on BRCA1 and BRCA2 mutation carriers
Full size table
Table 2 Positive peritoneal cytology and metastatic or recurrent cancer after documented occult preinvasive and invasive ovarian and fallopian tube carcinomas in risk-reducing salpingo-oophorectomies
Full size table

Observations and data support early removal of both fallopian tubes and ovaries by the end of the fifth decade of life to provide the most significant protective effect against ovarian cancer in HBOC syndrome mutation carriers. This is especially relevant for those who carry cancer-linked mutations in BRCA1 for whom the risk of ovarian cancer is higher and the age of diagnosis is younger than it is in BRCA2 mutation carriers [113, 114]. In a large international study of 1390 HBOC syndrome mutation carriers who had oophorectomy or salpingo-oophorectomy, 43 occult ovarian or fallopian tube cancers and one peritoneal cancer were diagnosed in BRCA1 mutation carriers at prophylactic surgery [102]. The youngest of these patients with occult cancer was age 34 years; three occult cancers were diagnosed before age 40 years, and 19 of the occult cancers were diagnosed between ages 40 and 49 years [102]. Thus, half of the occult cancers diagnosed at surgery in BRCA1 mutation carriers were discovered before age 50 years [102]. From a 1.5 % prevalence of occult cancers discovered in BRCA1 mutation carriers before age 40 years and the cumulative prevalence over the next 10 years, the authors estimated that the chance of being diagnosed with occult cancer rises to 14 % if RRSO is delayed until age 50 years [102]. Only two occult cancers were diagnosed in BRCA2 mutation carriers, both after age 60 years [102].

At least in the short term, HRT after oophorectomy probably does not increase the risk for breast cancer in HBOC syndrome mutation carriers. A collaborative cohort study of 462 BRCA1 and BRCA2 mutation carriers, adjusted for year of birth and mutated gene and prospectively followed for 3.6 years, demonstrated a significantly reduced hazards ratio 0.40 (95 % CI 0.18–0.92) for breast cancer in women who had undergone prophylactic oophorectomy at mean age 42.7 years (range 21.5–73.9 years) with 90 % being younger than 50 years [115]. Hormone replacement therapy did not alter the favorable results of reduced breast cancer risk in mutation carriers who had oophorectomies in that study [115]. In fact, women who underwent oophorectomy and took HRT had about one-third the risk for breast cancer as those who did not have oophorectomy and did not take HRT [115]. An earlier Markov decision model lumping together BRCA1 and BRCA2 mutation carriers also predicted that premenopausal oophorectomy lengthened life expectancy [106]. The simulated favorable effects associated with prophylactic oophorectomy were attributable mostly to reduced risks for both ovarian and breast cancers [106]. These good effects were somewhat obtunded by HRT, which was assumed to increase breast cancer risk over time, and the results improved when prophylactic mastectomy was concurrent with prophylactic oophorectomy [106]. In this model, life expectancy was not significantly affected when HRT was started after oophorectomy at age 35 years and continued until age 50 years; and life expectancy was slightly extended when oophorectomy was at age 40 years and HRT continued for 10 years, but this improvement did not approach the overall gain in life expectancy with prophylactic oophorectomy at age 40 years and no HRT [106]. Supporting expectations from the simulations, a case match control study of only BRCA1 mutation carriers actually found significantly lower odds ratio OR 0.58 (95 % CI 0.35–0.96, P = 0.02) for breast cancer in the group of subjects who ever used hormone therapy compared with those who had not [116]. Though the numbers of cases were few in which receptor status of the breast cancers were known, the findings in that study were not attributed to tumor type, as there was no significant difference in the use of hormone therapy reported by patients with ER-positive tumors 4/33 compared with those with ER-negative tumors 16/70 (P = 0.29) [116]. A study of 60 BRCA1 and BRCA2 mutation carriers disclosed the diagnosis of three breast cancers in 31 subjects (9.7 %) who took HRT after RRSO compared with nine breast cancers in 29 subjects (31 %) who did not use HRT after RRSO [117]. All three of the breast cancers in those who used HRT were in women who took estrogen only; two of these cancers were estrogen and progesterone receptor negative, and the receptor status of the other was unknown [117]. It is important to note that there are several known characteristics of breast cancers diagnosed in BRCA1 mutation carriers which differ from breast cancers in general populations, and there are differences between the characteristics of breast cancers in BRCA1 mutation carriers compared with BRCA2 mutation carriers, including the proportions of histologic phenotypes, mitotic activity, differentiation, expression of ER and PR receptors, p53, HER2 and many other properties, as well as ages of onset and variables depending on age [118, 119].

A recent review of breast cancer risk modifiers reported significant reductions of this disease in both BRCA1 and BRCA2 mutation carriers who had used tamoxifen [88]. While a current multivariate analysis of endometrial cancer risk factors in a case control study of BRCA1 and BRCA2 mutation carriers by the Hereditary Breast Cancer Study Group found highly significant increased OR 3.50 (95 % CI 1.51–8.10, P = 0.003) for the diagnosis of endometrial cancer in BRCA1 mutation carriers who had been treated for breast cancer or had any use of tamoxifen [120]. Though the OR for endometrial cancer diagnosis with tamoxifen treatment and use was increased in BRCA2 mutation carriers, this was not significant [120]. These findings confirmed a previous prospective cohort study, reported in 2007, which had found significantly increased risks for endometrial cancer in both BRCA1 and BRCA2 mutation carriers who used tamoxifen but non-significant increased risks for those who did not, and the authors concluded that the increased risks for endometrial cancer in mutation carriers could be largely explained by tamoxifen treatment for previous breast cancers [121, 122]. All six endometrial cancers in the continuing study were endometrioid carcinomas though one was confirmed to have a serous component in the later 2013 report [121, 122]. Three of seven endometrial cancers in BRCA1 mutation carriers from the Creighton Hereditary Cancer Registry were pure serous carcinomas (1) or mixed high grade serous and endometrioid carcinomas (2); the other four were grade 1 or grade 2 endometrioid carcinomas [71]. Two of these three patients with elements of serous carcinoma in their endometrial cancers had been prescribed tamoxifen for 5 and 7 years, respectively [71]. One BRCA2 mutation carrier with endometrial cancer in our registry had a grade 2 endometrioid carcinoma [71]. Without regard to hereditary risk, tamoxifen use has been associated with development of more aggressive types of endometrial cancer, especially serous and clear cell carcinomas after five or more years of treatment [123].

Endometrial cancer is the most common gynecologic malignancy reported from industrialized countries; though cervical cancer is still prevalent in the less developed world [72, 124]. Whether or not HBOC syndrome mutation carriers are at increased risks for endometrial cancer and in particularly its highly aggressive serous carcinoma histotypes, currently available observations and data indicate that women from families who inherit cancer-associated mutations in either BRCA1 or BRCA2 are expected to bear at least population risks for endometrial and cervical intraepithelial and invasive carcinomas. Moreover, women who have had extended chemoprophylaxis or treatment with tamoxifen and those who opt to forego or delay mastectomies remain at high risk for future breast cancers and likely treatment with tamoxifen, which significantly increases the risk of endometrial cancer [88, 90, 120, 125]. Because of the residing risks for even the most common, readily treated histotypes of endometrial and cervical cancers together with the formidable consequences of advanced pelvic serous carcinoma and disseminated intra-abdominal serous carcinomatosis, unless there are prevailing contraindications, it seems prudent to remove the uterus when risk-reducing surgery is elected by HBOC syndrome mutation carriers. Embryologically, the endosalpinx, endometrium and endocervix are müllerian derivatives and naturally remain in continuity through adult life. After childbearing, the uterus is an unnecessary and potentially dangerous organ in high risk patients. Including minimally invasive laparoscopic assisted vaginal hysterectomy (LAVH) along with RRSO can be accomplished by skilled laparoscopic surgeons with few if any complications, thereby prophylactically removing all müllerian derived epithelium at risk for malignant transformation [126, 127].

Expert opinion strongly favors that risk-reducing surgery for HBOC syndrome mutation carriers should consist of complete extirpation of the fallopian tubes and ovaries upon completion of childbearing and by age 40 years [89, 128, 129]. Several recent authoritative reviews have recommended short-term exogenous estrogen replacement for HBOC mutation carriers following risk-reducing removal of the ovaries during the fifth decade of life to alleviate acute menopausal symptoms and help protect against detrimental effects that can be associated with premature loss of endogenous hormones [89, 128, 129]. Whether effective or not, because of relatively high rates of osteopenia, osteoporosis and atraumatic fractures as well as their high risks for breast cancers and antineoplastic chemotherapy, many BRCA1 and BRCA2 mutation carriers who have undergone RRSO will be treated with tamoxifen [130–132]. Albeit, due to the adverse affects of tamoxifen on endometrium, this selective estrogen-receptor modulator (SERM) is largely being replaced in general populations by raloxifene, which may prove to be a better SERM for treatment of mutation carriers, as well [130–132]. Not only will RRSO–LAVH afford removal of the gynecological organs from which cancers may arise, be harbored and disseminate intra-abdominal carcinomatosis, but also removal of the uterus should simplify management decisions regarding HRT and treatment with SERMs.

Conclusions

Women of HBOC syndrome kindreds who carry cancer-associated mutations in BRCA1 or BRCA2 are highly susceptible to both breast cancers and intra-abdominal serous carcinomatosis, which was previously ascribed to primary ovarian cancers. Rarely, intra-abdominal carcinomatosis has been diagnosed in women who had undergone prophylactic oophorectomy because they were understood to be at increased risk for ovarian cancer. Embryologically, fallopian tube epithelium, endometrium and endocervical epithelium are müllerian derivatives and naturally remain in continuity through adult life. Ovarian surface and follicle epithelium arise from epithelium of the coelomic cavity. Preinvasive and early invasive serous carcinomas of fallopian tubes and endometrium, individually and concurrently, have been associated with malignant peritoneal cytology and serous carcinoma implants on ovaries and the peritoneum. Presently available evidence indicates that fallopian tube epithelium, especially the distal tube and fimbria, is the primary site of malignant transformation for the majority of these cases. Prophylactic mastectomy and salpingo-ophorectomy during the fifth decade of life have proven to be the most effective approach for reducing breast and ovarian cancer risk and mortality in BRCA1 and BRCA2 mutation carriers. Without mastectomy, RRSO alone by age 40 years reduces the risk for “ovarian” cancer by at least 80 % and breast cancer by 50 % in mutation carriers. Occult preinvasive and invasive cancers have been demonstrated in endometrial and ovarian sections from prophylactic surgery and cases of disseminated serous cancer and intra-abdominal carcinomatosis. Tamoxifen use and HRT with exogenous estrogens increase the risk for endometrial cancer. It is as yet unclear whether BRCA1 and BRCA2 mutation carriers are at increased risk for primary endometrial serous carcinoma, and there is insufficient data to preclude ovarian surface and/or cyst epithelium as primary sites for malignant transformation. At least in the short run, estrogen-replacement therapy does not affect the reduction of breast cancer risk and may alleviate acute symptoms, protect against accelerated deterioration and improve longevity after early premenopausal oophorectomy. RRSO–LAVH and other minimally invasive surgery can be performed with few if any complications. After childbearing, the uterus is unnecessary and the endometrium remains a potential primary site for cancer. Including hysterectomy with RRSO will simplify decisions concerning the prescription of SERMs and HRT and may further reduce the risk for gynecologic cancer and intra-abdominal carcinomatosis in HBOC syndrome mutation carriers.