1. Introduction
Lynch syndrome (LS), also known as Hereditary Non-Polyposis Colorectal Cancer syndrome (HNPCC), is an autosomal dominant inherited cancer susceptibility syndrome with a medium to high degree of penetrance (30–70%). It is caused by a germline mutation in one of the genes in the DNA mismatch repair gene (MMR) family, which includes MSH2, MLH1, MSH6, and less commonly PMS1 and PMS2. The prevalence of this syndrome is 0.9 up to 2.7% [
1].
LS is associated with a very high risk of developing colorectal cancer, which is typically diagnosed at an early age and with a proximal colonic predilection; 70–85% of colorectal cancers in LS are next to the splenic flexure. According to the literature, the lifetime risk of colorectal cancer ranges from 25–83% in females [
2]. LS is associated with an elevated risk of multiple extracolonic cancers, including cancer of the endometrium, ovary, stomach, small bowel, hepatobiliary tract, transitional cell carcinoma of the ureter and renal pelvis, brain, and skin tumors of the Muir-Torre Syndrome [
3].
LS accounts for approximately 3% of all colon cancers and may account for a similar number of ECs. The LS phenotypes include a propensity for cancers of the proximal colon, poor tumor differentiation with mucinous or signet-ring cell histologic features or a medullary growth pattern, abundant infiltrating lymphocytes in the tumor, and synchronous and metachronous colorectal cancers [
4].
For patients with LS, the lifetime risk of developing endometrial cancer (EC) varies from 30 to 70%, and the lifetime risk of developing ovarian cancer (OC) is 12–15% [
5].
Extracolonic cancers are more often observed in MSH2 mutations compared to MLH1 mutation families [
6,
7]. The risk of developing EC and OC in LS varies depending on which gene has mutated, as described by the Prospective Lynch Syndrome Data base (PLSD) [
8], which reported an incidence of EC and OC at 75 years of age of 37% and 11% for MLH1 carriers, 49% and 17% for MHS2 carriers, and 41% and 11% for MSH6 carriers, respectively [
9]. Ovarian cancer, particularly of the endometrioid type, is related to MLH1 mutations [
10]. The synchronous diagnosis of an endometrioid EC and an endometrioid OC is a relatively common situation observed in LS [
11].
Furthermore, women with a mutation in the MSH6 gene probably have a milder clinical phenotype with a later onset of both colorectal cancer and EC [
12,
13]. The lifetime risk for women with a mutation in the PMS2 gene is unknown, but studies have suggested that these patients have a milder phenotype compared to women with mutations in the MLH1 and MSH2 genes [
14,
15,
16].
In more than half of the mutated patients with metachronous colorectal and gynecological cancers, EC or OC are the tumors that are diagnosed first, making these the “sentinel cancers” of the syndrome [
17]. Before DNA mismatch repair gene mutations were used to determine germline genetic defects in families with LS, clinical criteria (called Amsterdam criteria) based on an early age at cancer onset and the presence of more cancers among family members, defined individuals with LS [
18,
19]. Similar criteria were included in the Bethesda Guidelines that were developed in 1997 and revised in 2004 [
20]. If the Amsterdam Criteria or Bethesda Guidelines are met, molecular pathology testing of the cancer for alterations typical of LS is indicated. This includes testing for microsatellite instability (MSI) and MMR protein immunohistochemistry (IHC). Although the Amsterdam criteria is commonly used, it has poor sensitivity for the detection of LS, which is often underdiagnosed [
21].
The diagnosis of LS allows clinicians to tailor treatment and clinical management, and to optimize counselling and cancer surveillance for patients and their families. With regards to surveillance for colon cancer, the early detection of lesions by colonoscopy is associated with improved survival [
8]. For gynecological cancers, there is no clear data that points to the benefit of surveillance on survival [
22]. However, multicenter studies assessing the benefits of surveillance in asymptomatic women with LS, considering their age, menopausal state, and surveillance interval, needs to be performed. Most recommend the use of transvaginal ultrasound and endometrial biopsy for premenopausal women, and transvaginal ultrasound alone for asymptomatic postmenopausal women [
23].
In contrast, risk-reducing hysterectomy and bilateral salpingo-oophorectomy have been shown to prevent gynecological cancer in women with LS, and should be recommended following the completion of childbearing for MLH1, MSH2, and MSH6 carriers over 35–40 years of age [
8].
Regarding outcomes, a favorable prognosis was suggested, which was probably related to the active local immune response. The release of peptides by MMR-deficient tumors allows the patients’ immune systems to better recognize them [
24].
Colorectal cancers with deficient mismatch repair are associated with an earlier stage at diagnosis, a lower propensity for metastasis, and consequently a significantly better prognosis than patients with stage-matched cancers with proficient mismatch repair [
4]. With regards to EC, the new prognostic molecular classification introduced by The Cancer Genome Atlas Research Network (TCGA) suggests a better prognosis for MSI hypermutated EC compared to TP53 tumors [
25]. Due to its prognostic role, MMR status has been recently added to EC management guidelines and is fundamental for the post-surgical treatment of early-stage EC. Due to its prognostic role, MMR status should be performed for all ECs, irrespective of the histologic subtype and of the patient’s age [
26].
As in ECs, deficient MMR endometrioid OCs are associated with better prognoses compared to those with mutated TP53 genes, according to some reports [
11].
Concerning treatment, new promising drugs specifically acting on MMR-deficient tumors have been evaluated. In particular, in 2017 the anti-PD-1 immune checkpoint antibody pembrolizumab was introduced for advanced MMR-deficient cancers, being the the first anti-neoplastic agent to be given a site agnostic license since it can be prescribed irrespective of the tumor site and depends only on the presence of MMR deficiency [
27]. Furthermore, for this reason, the identification of MMR-deficient tumors is so important for patient care [
24].
Although there are several epidemiologic studies of ECs and OCs in women with LS, detailed pathological analyses are lacking. It is not known with certainty whether ECs and OCs have unique pathological features in women with LS compared with sporadic tumors. Previous studies have suggested that in LS, EC is associated with a diagnosis at an earlier age [
28] and stage, and ECs have different histology compared with the general population [
29,
30]. Moreover, ECs associated with LS are thought to be preferentially located in the lower segment of the uterus, but not for MLH1-mutated ECs. With regards to OC, a higher incidence of non-serous tumors diagnosed in LS women was observed [
31].
The knowledge generated from these data is critical for understanding the natural history of EC and OC in this unique population of patients.
Therefore, the primary purpose of this study was to examine the pathological features of EC and OC in women with LS compared to sporadic cancers.
2. Materials and Methods
In our study we included 96 patients with ascertained LS, of which 82 were diagnosed with EC and 14 with OC. Data were collected in the databases of the departments of Medical Genetics of Molinette Hospital of Turin, Medical Genetics of San Luigi Gonzaga Hospital of Orbassano (Turin), Medical Genetics of IRCCs Candiolo Hospital, and IRCCs INT of Milan. Data related to diagnosis and treatment were found in the archives of the above institutions. The LS patients included in our study had genetic counselling because their personal or family history was suggestive of LS. The diagnosis was confirmed by observations of microsatellite instability and by IHC analysis. Pathologists classified the tumors according to the WHO (World Health Organization). Because of the study period, the grade and stage were defined according to the 1988 FIGO classification for OC and the 2009 FIGO classification for EC [
32,
33].
In the control group, we included 209 patients with EC and 187 patients with OC who did not have a family history suggestive of cancer, and who underwent surgery between 2006 and 2016 in the department of Gynecology and Obstetrics of the Umberto I Hospital of Turin. The women selected in the control group met the following inclusion criteria: a clinical and histopathological diagnosis of EC, or a clinical and histopathological diagnosis of OC at any age. Exclusion criteria included a definitive histological diagnosis of benign endometrial or ovarian pathology, a diagnosis of simple, typical complex, or atypical complex hyperplasia, and a family history of EC, OC, and/or colorectal cancer (to rule out the inclusion of possible mutation carriers, although they were untested for MMR gene mutations).
Statistical Analysis
Statistical analysis was performed using Fisher’s exact test and Chi-square tests (X2). Fisher’s exact test and X2 were used to compare and assess the significance of differences between the two groups (LS and the CG) for categorical variables (histological features). Continuous variables, such as the mean age at diagnosis of EC and OC in different groups, were analyzed using t tests. p values were two sided and p < 0.05 was considered significant for X2 and Fisher’s exact tests, whereas p < 0.01 was significant for t tests. Data were analyzed using the Statistical Package for Social Science (SPSS) version 24.0 (IBM, Armonk, NY, USA).
5. Conclusions
The lifetime risk of EC and OC is increased in patients with LS. EC and OC associated with LS occur at an earlier age than sporadic cancers, are predominantly present at an early stage, with a well or moderate grade of differentiation, and are often of endometrioid or clear cell histology.
The evaluation of a patient’s MMR status is becoming essential for all ECs because of its prognostic and therapeutic role.
There is a general consensus that healthy women with LS should be offered a risk-reducing prophylactic hysterectomy and bilateral salpingo-oophorectomy at around 40 years of age. Despite EC screening, the utility of OC surveillance in a healthy population is not yet evidence-based, nor is it evidence-based in patients with LS, in BRCA-mutated patients, or in the general population. Surveillance can be tailored to individual women. Recognition of these features and appropriate genetic testing enables the identification of gynecological cancers associated with LS, thereby allowing for tailored surveillance, treatment, or surgical prevention.