Molecular, Pathologic, and Clinical Features of Early-Onset Endometrial Cancer: Identifying Presumptive Lynch Syndrome Patients

Lynch syndrome (also known as hereditary nonpolyposis colon cancer; ref. 1) has been recognized for almost a century as a familial clustering of cancers of the colon, endometrium, stomach, ovary, and ureter (2). The development of a reliable set of clinical criteria for the recognition of Lynch syndrome continues to the present day with the serial introduction of the Amsterdam criteria, Amsterdam criteria II, and the Bethesda Guidelines showing increasing sensitivity but with loss of specificity (3–7). The underlying defect in Lynch syndrome is a dominantly inherited germ-line mutation in one of four DNA mismatch repair (MMR) genes, MLH1, MSH2, MSH6, and PMS2, the products of which cooperate in a protein complex that binds to and repairs small replicative DNA sequence errors (8, 9). The majority of mutations detected in Lynch syndrome families (85-90%) are seen in MLH1 and MSH2 in approximately equal measure(10), with the remaining 10% to 15% occurring in MSH6 (11) and rare mutations in PMS2 (12). A varying propensity to develop endometrial cancer (EC) has been reported, with several studies showing that EC is more likely to occur in the setting of MSH2 and MSH6 germ-line mutation carriers (13–15).

EC carries a lifetime risk of 0.9% in the United Kingdom and 2.6% in the United States with an average age of onset of 68 years (16). However, for patients with Lynch syndrome, the lifetime risk of EC increases to 40% to 60% (17, 18), and the average age of onset falls to 48 years. Some authors have calculated that the risk of developing EC within Lynch syndrome is equal to or in excess of that associated with colorectal cancer in female mutation carriers, highlighting young-onset EC as a “sentinel cancer” for this disorder (19). In support of this premise, recent studies have shown that although the lifetime risk of colorectal cancer in male mutation carriers approaches 90% (18), in females it is considerably less at 52% and with EC equally common at 54% (20). However, not all reports have returned this finding (21). Individuals with Lynch syndrome are also more likely to develop multiple primary cancers during their lifetime; 54% to 61% of patients develop a second primary tumor and 15% to 23% have three or more primary tumors (16). Because of these high risks for cancer, screening and prophylactic surgery may be offered to individuals with Lynch syndrome (22–24). Detection of deleterious mutations, in particular MMR genes, can unequivocally establish a diagnosis of Lynch syndrome. However, in 40% to 50% of putative Lynch syndrome families, no mutations can be found, and although such families are empirically at less risk than proven mutation carriers, their risk remains substantial (16).

A major advance in the diagnosis of Lynch syndrome has been the establishment in recent years of immunohistochemistry techniques to detect deficiency of a MMR protein in tumors in routine laboratory settings. This can be achieved with readily available paraffin-embedded tissue sections, is relatively inexpensive and rapid because it involves no DNA extraction steps, and works equally well on colorectal or endometrial tumors. Loss of MMR protein MSH2, MSH6, or PMS2 alone gives a high likelihood that a family has Lynch syndrome. In contrast, loss of MLH1 may arise either as a consequence of a germ-line mutation or as a result of silencing of MLH1 expression by methylation of CpG islands within its promoter (25). Although a minority of Lynch syndrome ECs have been reported with methylation of MLH1, the overwhelming majority do not show this feature (26). Combining the techniques of MMR immunohistochemistry and differential methylation of MLH1 with family history analysis allows the triage of presumptive Lynch syndrome EC patients. However, some level of diagnostic uncertainty remains in the absence of a confirmed germ-line mutation.

In this study, we examine the proportion of young-onset EC likely to have Lynch syndrome using tumor analysis and identify specific clinical and pathology features that may assist in the recognition of Lynch syndrome at presentation.

The study was conducted with approval from the Ethics Committee of the Royal Brisbane and Women's Hospital and the Queensland Institute of Medical Research. Patients were identified from the Queensland Centre for Gynaecological Cancer database and were included if they matched the following criteria: (a) diagnosed with EC between June 1988 and June 2006; (b) primary treatment at Queensland Centre for Gynaecological Cancer; and (c) ages ≤50 y at time of diagnosis of EC. One consultant histopathologist (M.C.C.) reviewed original H&E sections from all patients. Immunohistochemistry for MMR proteins, microsatellite instability (MSI) testing, and MLH1 promoter methylation was done on archival EC tumor samples as described in the following sections.

Immunohistochemistry. Paraffin sections (4 μm) were affixed to Superfrost Plus adhesive slides (Menzel-Gläser) and air-dried overnight at 37°C. After dewaxing in xylol and rehydration through descending graded alcohols to distilled water, sections were transferred to Target Retrieval Solution pH 9.0 (DAKO), and antigen retrieval was done for 20 min at 121°C using a commercial pressure cooker (Decloaker, Biocare Medical). The sections were cooled and transferred to TBS with Tween 20 (0.05 mol/L Tris, 0.15 mol/L NaCl, 0.1% Tween 20; TBS-T), pH 7.2-7.4. The subsequent immunohistochemical protocol was done on an automated immunostainer (AutoStainer, DAKO). Endogenous peroxidase activity was quenched by incubating the sections with 1.0% H₂O₂ in TBS for 10 min. Following washing in two changes of TBS-T, nonspecific antibody binding was inhibited by incubating the sections with serum-free protein block for 10 min. Excess blocking reagent was blown from the sections and the primary antibody applied as detailed in Table 1. Following this and subsequent incubations, the sections were washed in two changes of TBS-T. Sections were then incubated with MACH3 Mouse Probe⁺ (Biocare Medical) for 20 min, and then with MACH3 M-Polymer-HRP⁺ (BioCare Medical) for 10 min. Color was developed using Liquid DAB⁺ substrate chromogen system (DAKO) for 6 to 8 min. Then sections were washed in running tap water, lightly counterstained in Mayer's hematoxylin, dehydrated through ascending graded alcohols, cleared in xylene, and mounted using DePeX (BDH Gurr).

MSI testing. MSI was analyzed from formalin-fixed paraffin-embedded normal and tumor paired tissue DNA using a 10-marker panel as has previously been described (27). Briefly, the 10 MSI markers included the mononucleotide markers BAT25, BAT26, BAT34C4, and BAT40; dinucleotide markers D5S346, D17S250, ACTC, D18S55, and D10S197; and one penta-mono-tetra compound marker, MYCL. Individual markers were amplified by PCR using a touchdown cycling protocol. PCR products were then diluted accordingly for each marker and then equal amounts of diluted PCR product were pooled into one of two pools. One microliter of pooled product was added to 8.5 μL of formamide and 0.2 μL of ROX 500 size marker (Applied Biosystems, Inc.) and denatured at 95°C for 2 min with rapid cool. MSI analysis was done with the Genescan analysis software after separation on the ABI 3100 capillary based sequencer (Genescan v 3.7, Applied Biosystems). Tumors were classified as MSI high (MSI-H) if ≥30% of the markers showed instability, MSI low (MSI-L) if <30% showed instability, and microsatellite stable (MSS) if no marker exhibited instability. Only cases with five or more evaluable markers were considered.

MLH1 methylation assay. Sodium bisulfite conversion of DNA derived from formalin-fixed paraffin-embedded tumor tissue was done with the Methyleasy High Throughput DNA Bisulfite modification kit (Human Genetic Signatures) as per manufacturer's instructions. After sodium bisulfite conversion, the converted tumor DNA was amplified by fluorescence-based, real-time quantitative PCR using locus-specific PCR primers that flank an oligonucleotide probe with a 5′ fluorescent reporter dye (6FAM) and a 3′ minor groove binder nonfluorescent quencher (MGBNFQ, Applied Biosystems), also known as a TaqMan probe (Applied Biosystems). The 5′ nuclease activity of the Taq DNA polymerase cleaves the probe releasing the reporter dye and fluorescing. The fluorescence from the TaqMan reaction was detected on a Corbett Rotor-Gene 6000 (Corbett Research). The reaction conditions for MLH1 and Alu MethyLight reaction were as follows: a 30-μL reaction volume with 1× buffer A, 200 μmol/L deoxynucleotide triphosphates, 500 nmol/L forward and reverse PCR primers, 200 nmol/L probe, 3.5 mmol/L MgCl₂, 3 μL of a stabilizer consisting of 0.01% Tween 20 and 0.05% gelatin, 0.1 unit of AmpliTaq Gold polymerase (Applied Biosystems), and 2 μL of converted DNA (estimated to be at ∼20 ng/μL). The PCR cycling conditions were a standard quantitative PCR program of 95°C for 10 min, then 50 cycles of 95°C for 15 s, followed by 60°C for 1 min. The MLH1 reaction was run in duplicate whereas the Alu reaction was run singularly for individual samples. MLH1 methylation was recorded as positive based on the amplification of the MLH1 and Alu reactions, and recorded as negative if no MLH1 amplification was seen but Alu amplification was present (28). The PCR primers and probes are listed in Table 2.

Patients were classified conservatively as having presumptive Lynch syndrome if they met the following criteria: loss of at least one MMR-gene protein by immunohistochemistry in their tumors, and, if there was immunohistochemical absence of MLH1, negative MLH1 hypermethylation status (26). Patients with normal immunohistochemistry and those with loss of MLH1 in combination with a positive MLH1 hypermethylation status were designated non-Lynch.

Since 1982, it has been a legislative requirement in Queensland that all newly diagnosed cancers are reported to the Queensland Cancer Registry. This database was used as the primary source of personal cancer history confirmed by both patient charts and the database of the Queensland Centre for Gynaecological Cancer, and supplemented by additional data sources where cancers were diagnosed outside the state of Queensland or before 1982. Information about family history of cancer was obtained from the patient charts and the Clinic Administration System database of the Queensland Centre for Gynaecological Cancer. Tumors occurring in first-degree relatives were categorized according to whether they were part of the Lynch syndrome spectrum of malignancies, specifically cancers of the colon, endometrium, stomach, ovary, ureter/renal pelvis, brain, small bowel, and hepatobiliary tract, and sebaceous skin cancer (19). Family histories were assessed to establish whether they met the Amsterdam criteria II for family member affectation status, namely that there are at least three relatives (one <50 y old) with a Lynch syndrome-associated cancer (colorectal, endometrial, small bowel, ureter, or renal pelvis), connected by first-degree relationships, and distributed across more than one generation. Nonmelanoma cutaneous tumors were excluded from the study due to the high incidence of these cancers in Australia, as were metastases and recurrences.

Statistical analysis was carried out using Statistical Package for Social Sciences (SPSS, version 14.0) and Statistica (Statsoft Corporation). Contingency tables were assessed using χ² or Fisher's exact test as appropriate. Differences between means were assessed using a t test. Regression analysis was carried out to test for independence between significant univariate parameters using the Statistica software package (Statsoft Corporation). P < 0.05 was considered significant. Sensitivity and specificity including 95% confidence intervals (95% CI) were calculated using Vassarstats Clinical Calculator One.

Patients. One hundred forty-six patients met the inclusion criteria for our study, and tumor analysis was completed on all cases. For 93 of the 146 (64%) patients, information on family history was available. Information about personal cancer history was obtained for all 146 patients. The median duration of follow-up was 4.2 years (range, 0-17 years; mean, 5 years). The mean age of all the patients at the time of diagnosis of EC was 45.1 years (range, 28-50 years). The average age at presentation for presumptive Lynch syndrome patients was 46.5 years (range, 39-50 years) and the average age at presentation for the balance of patients was 44.6 years (range, 28-50 years). Interestingly, patients with presumptive Lynch syndrome were not seen under 39 years of age.

Tumor analysis. The results of tumor analyses for MSI and immunohistochemistry are shown in Table 3. Representative cases stained for MMR proteins are illustrated in Fig. 1. In total, 38 of 146 (26%) EC tested showed loss of at least one MMR protein by immunohistochemistry. Of these 38, 18 (47%) tumors showed loss of both MLH1 and PMS2 proteins, 13 (34%) showed loss of both MSH2 and MSH6 proteins, whereas the remaining 7 (19%) tumors showed absence of MSH6 only. Of 18 tumors with immunohistochemical absence of MLH1, 6 (33%) were negative for MLH1 hypermethylation. The combination of immunohistochemistry and methylation analyses resulted in 26 (17.8%) patients classified as having presumptive Lynch syndrome. This group included all cases with immunohistochemical absence of MSH2 and/or MSH6, as well as cases with immunohistochemical absence of MLH1 where no methylation was detected. The remaining 120 (82.2%) patients were classified as non-Lynch. When patients ages between 39 and 50 years only were considered, presumptive Lynch syndrome comprised 26 of 129 (20.2%). Importantly, no patients ages <39 years showed MMR-deficient tumors. MSI testing was done for 120 of 146 (82.2%) tumors and showed 97.5% concordance with immunohistochemistry. Three discordant cases included two MSI-H tumors with normal immunohistochemistry and one MSS tumor lacking both MSH2 and MSH6.

Pathology review. A comparison of clinicopathologic features between presumptive Lynch syndrome EC and non-Lynch cases is shown in Table 4. Endometrioid histotype predominated in both groups. Presumptive Lynch tumors were more likely to be intermediate/high International Federation of Gynecology and Obstetrics (FIGO) stage (P = 0.029) and to have a high FIGO grade (P = 0.045), a higher average mitotic rate (P = 0.043), and deeper average myometrial invasion (P = 0.016), and to have tumor-infiltrating lymphocytes (P = 0.004). Regression analysis showed that tumor-infiltrating lymphocytes were an independent predictor of presumptive Lynch syndrome (P < 0.005), as was an increased mitotic rate (P = 0.04). There was no apparent association between presumptive Lynch syndrome status and overall histologic grade, nuclear grade, peritumoral lymphocytes, Crohn's-like aggregates, cervical invasion, lymphovascular invasion, or adenomyosis (P > 0.05).

Personal history of cancer. A comparison between the two groups of patients under study is shown in Table 5 with respect to previous, synchronous, and subsequent cancers. Of the total 146 patients, 9 (6.2%) had a previous history of cancer comprising colon (n = 1), breast (n = 3), and brain (n = 1) cancers and malignant melanoma (n = 4). The median time elapsed between the previous cancer and the EC was 10.5 years (range, 2-30.8 years; mean, 13.0 years). Synchronous cancers were recorded in 15 (10.2%) patients. Eleven of 15 (73%) were ovarian cancers. Five (3.4%) patients developed a cancer subsequent to the diagnosis of EC. These tumors developed after a median interval of 5.7 years (range, 2.9-11.6 years; mean, 6.0 years). Subsequent tumor types were colon cancer (n = 3), breast cancer (n = 1), and melanoma (n = 1). No significant difference was found with respect to the presence of previous, synchronous, or subsequent cancers between the two groups analyzed in this study.

Family history of cancer. Seven (7.6%) patients had a family history that met the Amsterdam criteria II, and 6 (85.7%) of these had been classified as presumptive Lynch syndrome (P < 0.001; Table 5). As expected, there was a significant difference in family history of a first-degree relative with a Lynch syndrome–related cancer (P = 0.002) and any type of family history (P = 0.043). In seven patients, it was not possible to definitively classify the cancer in a first-degree relative as within the Lynch syndrome spectrum. Interestingly, there was no significant difference between the two groups with respect to family history of a first-degree relative with cancer.

There were no significant differences between the average age of onset when patients who had a family history of cancer were compared with those who did not (45.8 versus 44.2 years; P > 0.05). Patients with any family history were more often of ages ≥40 years; however, this difference was not significant. Table 6 compares the tumor immunohistochemistry results of patients with and without a family history of any cancer. Patients with a family history of cancer were more likely to have a MMR-deficient cancer (29.3% versus 14.3%); however, this failed to attain statistical significance (P = 0.13). Interestingly, patients with a family history of cancer were significantly more likely to have deficiency of MSH2 and/or MSH6 than MLH1 and PMS2 (76% versus 20%; P = 0.039). Tumor features that significantly differentiated Lynch syndrome from the remainder of young-onset EC, such as tumor grade, mitotic rate, and myometrial invasion, were not significantly different between these two groups (P > 0.05). Sensitivity and specificity of family history measures to predict presumptive Lynch syndrome status were Amsterdam criteria II sensitivity 0.38 (95% CI, 0.16-0.64) and specificity 0.99 (95% CI, 0.91-1.00); first-degree relative with a Lynch syndrome–related cancer sensitivity 0.46 (95% CI 0.22-0.73) and specificity 0.90 (95% CI 0.80-0.96); first-degree relative with any type of cancer sensitivity 0.63 (95% CI 0.36-0.84) and specificity 0.60 (95% CI 0.47-0.71); and family history of any type of cancer sensitivity 0.88 (95% CI 0.60-0.98) and specificity 0.42 (0.31-0.53).

In patients where no family history information could be obtained, the characteristics of the tumors showed no significant differences from the cohort of patients where family history information was available for analysis (data not shown).

One in five women diagnosed with EC at or before the age of 50 years is likely to have underlying Lynch syndrome with clear implications for herself and her first-degree relatives. The recognition of Lynch syndrome among EC patients is important for preventing subsequent tumor development in the index patient and for presymptomatic testing and surveillance of family members (14). Several options have become available during the last decade (22, 29) to screen for, or reduce the risk of, gynecologic cancer in women who carry germ-line mutations in MMR genes. These options include annual transvaginal ultrasound, endometrial biopsy (23), and prophylactic hysterectomy and salpingooophorectomy (24, 30). In addition, clinical trials of oral contraceptive use are under way as a preventive measure for EC (31). Finally, a proportion of Lynch syndrome cases with EC will develop colonic neoplasia, and screening is recommended (29).

The incidence of Lynch syndrome in women with a newly diagnosed EC has been difficult to assess from published reports (14). Factors that have delayed the elucidation of an exact frequency have primarily arisen from the diversity of study designs, including positive family history-only case designs (32–35), MSI-type EC studies (36, 37), and groups without definitive MMR mutation results, a limitation also associated with our current study. The largest study thus far of unselected EC patients (n = 543) concluded, after comprehensive mutation testing and consistent with two previous reports (11, 35), that at least 1.8% of all EC could be attributed to Lynch syndrome (14). In that report, immunohistochemistry for MMR was used post hoc essentially to confirm MSI results, and investigations were not confined to the young-onset patient as is the case here. Our study suggests a one-in-five risk of Lynch syndrome in patients with young-onset EC, particularly those ages between 39 and 50 years. In assessing whether 18% represents a reasonable estimate of the proportion of young-onset EC likely to have Lynch syndrome, comparison might be made between our findings and those of Berends et al. (13). In a study of 58 EC patients ages <50 years, ∼9% were found to carry deleterious germ-line mutations in a MMR gene. In our study, we found that six of seven patients whose family history fulfilled the Amsterdam criteria II were classified as presumptive Lynch syndrome. In 30% to 50% of Amsterdam criteria II families, a germ-line mutation cannot be found (16). Therefore, if the mutation detection level of Berends et al. (13) represents ∼50% of Lynch syndrome EC, then our estimate of 18% using immunohistochemistry and methylation analysis is consistent with previous findings.

As a consequence of DNA MMR deficiency, Lynch syndrome tumors are characterized by the hallmark change of accumulated DNA sequence errors detected as MSI. MSI testing is neither inexpensive nor straightforward; it requires microdissection of tissue samples to enrich for tumor cells, subjective interpretation, and access to either a sequencing system or radioactive imaging, and is therefore an unsuitable assay for widespread use in clinical laboratories. Therefore, in assessing the prevalence of Lynch syndrome in this patient group, we used robust immunohistochemical techniques coupled with pathology and family history review that could be undertaken in a routine clinical setting. The balance of MMR-deficient patients showing absence of MLH1 with concomitant methylation of the gene promoter was excluded from the classification of presumptive Lynch syndrome, and this may also have introduced some bias; however, a previous report has suggested that in only a minority of cases would methylation of MLH1 be expected to inactivate the wild-type allele in Lynch syndrome (26). Therefore, it is likely that the approach taken in this study to use immunohistochemistry as a guide to selection of young-onset EC patients who might benefit from MMR mutation testing is a reasonable one. To seek mutations without initial tumor screening is likely to be expensive and inefficient due to the majority of young-onset patients having no indications of Lynch syndrome, the number of candidate genes, and the relatively poor success rate in the detection of mutations (16).

The current study also sought to identify histology features associated with MMR-deficient tumors. Although we were not able to identify any particular feature that would definitively identify Lynch syndrome EC (38), we found presumptive Lynch EC to be of a significantly higher FIGO stage and FIGO grade, to have a higher mitotic rate, to have deeper myometrial invasion, and, as has been found in Lynch syndrome colorectal cancers, to be more likely to have tumor-infiltrating lymphocytes. Broaddus et al. (38) observed that nonendometrioid EC occurred more often among the presumptive Lynch syndrome cases; however, this was not a significant difference. In contrast to this previous report, we did not find that nonendometrioid tumors were confined to the MSH2-deficient subgroup (38). Predicting Lynch syndrome status from histologic appearance in EC is considerably more difficult than in colorectal cancer (39). In addition, a definitive molecular assay analogous to somatic BRAF mutation in colorectal cancer is not yet available for EC (40). In colorectal cancer, the presence of a somatic BRAF mutation excludes Lynch syndrome in MSI-H tumors with a high degree of certainty (41). Currently, immunohistochemistry for the MMR genes is the most efficacious method for the presumptive identification of Lynch syndrome EC in a routine clinical setting.

We found no significant difference between presumptive Lynch and non-Lynch patients with regard to subsequent cancers. This finding may have resulted from a relatively short median follow-up time of only 4.2 years, whereas the subsequent cancers observed developed after a median of 5.7 years, raising the possibility that the follow-up interval in this series is not adequate to assess the risk of subsequent cancers. In a published report, Lu et al. (19) investigated 117 women who met the Amsterdam criteria II and had two primary cancers, and found that in 44 patients who had EC initially, the median time to developing a second cancer was 11 years. Alternatively, it is possible that there is no difference in the risk of developing subsequent primaries due to the age of the patients at presentation because a young age of onset for cancer itself carries an increased likelihood for genetic risk (42). The synchronous cancers in this study consisted mainly of ovarian cancers. In our study, 7.5% of all patients had this cancer at the same time as their EC. In a published report, Walsh et al. (43) investigated the frequency of coexisting ovarian malignancies in young EC patients ages ≤45 years. Among the 102 women included in their study, they found that 26 (25%) had a synchronous ovarian malignancy. Tran et al. (44) retrospectively investigated the profile and outcome of young women with EC ages ≤45 years. Among other things, they found that 7% of 41 patients had a synchronous ovarian malignancy, not including metastases. In similar studies, Evans-Metcalf et al. (45) reported synchronous ovarian malignancies in 11% of young women with EC ages ≤45 years. Thus, our percentage of 7.5% synchronous ovarian cancers is consistent with that reported by other investigators.

The Amsterdam criteria II in this study were tightly associated with prediction of presumptive Lynch syndrome cases. Six of the seven women whose family history met the Amsterdam criteria II showed loss of at least one MMR protein on immunohistochemistry. Five patients showed a loss of MSH2 and MSH6, and one showed a loss of MSH6. When considering family history of first-degree relatives with a Lynch syndrome–related cancer, the difference between the groups was also significant (P = 0.002). Because the Amsterdam criteria II were originally designed to identify patients with Lynch syndrome, and Lynch syndrome is clinically recognized by an aggregation of particular types of cancer in a family, the two findings about family history described above are in line with expectations and support the use of Amsterdam criteria II as a tool in the identification of Lynch syndrome among young-onset EC cases. As would be expected, the respective sensitivities of stringent criteria, such as Amsterdam criteria II, and the presence of a first-degree relative with a Lynch syndrome-spectrum cancer were high, in contrast to their relatively low sensitivity. With less stringent criteria such as a first-degree relative with any type of cancer or any family history of cancer, the sensitivity increased at the expense of the specificity.

We also examined the family history of a first-degree relative with any type of cancer and found no significant difference between our two groups. In a previous study, Suomi et al. (35) investigated the family history of 291 EC patients ages ≤60 years. They found that 45% (130 of 291) of patients in their study had at least one first-degree family member with cancer. This percentage is comparable to the 42% found in our study. Ollikainen et al. (35) have postulated that Lynch syndrome is unlikely to account in full for the observed excess of family cancer history seen in young-onset cases of EC, raising the possibility that novel syndromes of EC predisposition may account for these observations. The difference in family history in general (any family member with any type of cancer) between suspected Lynch syndrome and non-Lynch patients was also assessed. As expected, due to the extensive family history seen in Lynch syndrome kindreds, this showed a significant difference between the groups (P = 0.043). No significant differences in the prevalence of a MMR-deficient EC were seen between patients with and without a family history of cancer; however, the prevalence of MSH2/MSH6 absent tumors among those with a MMR-deficient tumor was significantly greater than that of MLH1/PMS2 absent tumors in patients with a family history of cancer, reflecting the cases of sporadic MLH1/PMS2 absent EC in those without a family history.

In conclusion, this work suggests that young-onset EC may herald significant implications for both the patient and her first-degree relatives. In contrast to colorectal cancer, histopathology alone is not a definitive predictor for Lynch syndrome in EC patients. Family history in this study was indicative of Lynch syndrome and confirmed the findings from the molecular studies. Interestingly, all presumptive Lynch syndrome patients presented at age ≥39 years. Approximately one in five EC patients ages ≤50 years showed features of having Lynch syndrome using laboratory and clinical features, which may be assessed at presentation. The magnitude of this risk warrants immunostaining for MMR proteins in all EC patients diagnosed at or before the age of 50 years.

We thank Lesley Jaskowski, Dan Jackson, and Aedan Roberts for their assistance in the preparation of data for this article, and Sullivan Nicolaides Pathology, QML Pathology, IQ Pathology, and Queensland Health Pathology and Scientific Services for supply of archived tissue for the project.