The purpose of this study was to describe the unique characteristics and long-term outcome data of a cohort of hMLH1+ tumors and compare them with tumors with intact MMR expression and deficient MMR expression (presumed Lynch syndrome) in a largely suburban/rural patient population. The results confirm prior reports that hMLH1+ tumors have poor prognostic indicators, including higher grade and LVSI, and are associated with older age.[9,10] In addition, RFS among hMLH1+ patients is reduced in the current study, similar to other reports. The current study demonstrates that MLH1 promoter hypermethylation is a risk factor for recurrence in patients who would otherwise be considered to have a low probability for poor outcomes.
This work expands on previous studies evaluating the impact of MSI-H status in endometrial cancer as it relates to survival. Initial research demonstrated an improved disease-free survival and disease-specific survival with MSI-H tumors, consistent with data from the colorectal cancer literature.[11,12] Others contradicted these findings, and still others found no difference in overall survival benefit or disease-free survival.[13,14] These early studies culminated in a meta-analysis in which pooled analysis did not show any association between MSI-H and worse outcomes.
In contrast, two newer studies evaluating molecular risk stratification for endometrial cancers have also analyzed outcomes of MSI-H tumors. The Cancer Genome Atlas Research Network analyzed MSI-H tumors as one of the four main clusters of their study, and overall, this group had worse progression-free survival. They report that most MSI-H tumors are attributable to MLH1 hypermethylation but do not further distinguish these outcomes. A larger NRG/GOG study classified over 900 cases and found MMR-deficient tumors were associated with higher grade, presence of LVSI, and decreased progression-free survival. MLH1 hypermethylation specifically is not addressed in this study.
However, few studies have specifically looked at MLH1 promoter hypermethylation patients as a distinct subgroup of MSI-H tumors. We chose to look at only endometrioid type to help control for subtypes already known to be at increased risk for recurrence, such as serous tumors. In addition, none of the serous tumors from this time period at our institution had MLH1 promoter hypermethylation present. McMeekin et al demonstrated that progression-free survival was worse in patients with hMLH1+ endometrial tumors compared with those who had intact MMR expression in a univariate analysis, although this effect did not remain significant in multivariate analysis. Cosgrove et al demonstrated that patients with hMLH1+ endometrial tumors had a decreased RFS and an increased rate of recurrence among advanced-stage endometrioid adenocarcinomas. Despite having similar numbers in each category, our study did not show significance in recurrence rate among advanced-stage tumors. This may be due to differences in adjuvant treatment, as 43.7% of our MLH1+ patients received adjuvant treatment overall, compared with 30.5% in Cosgrove et al. However, our study does not describe adjuvant treatment specifically for the advanced-stage group.
Contrary to prior studies, in the subset of patients with endometrioid histology, early stage hMLH1+ patients recurred significantly more often. This early stage group was further divided to analyze the cases contributing to the increased recurrence rate. Initially, the high-intermediate-risk (HIR) and low-risk (LR) groups represented the two extremes of patients more and less likely to recur among stage I and II disease. The low-intermediate-risk (LIR) group developed as a significant proportion of patients who did not meet criteria for either group. Recurrence rates among hMLH1+ patients were increased for all three of these groups—11.1% in the LR group, 12.9% in the LIR group, and 21.1% in the HIR group. Although the recurrence rate was only significantly different in LIR and HIR groups, sample size is likely a factor in the inability to detect a difference in the LR group. These early stage patients are thought to be at low baseline risk for recurrence, but hMLH1+ status may be an independent risk factor for recurrence, and evaluating these patients for additional treatment in the adjuvant setting may be warranted, as suggested in previous studies. These findings confirm results reported by Backes et al, in which HIR patients had increased recurrence rates and decreased RFS. These studies highlight the use of MLH1+ status as a poor prognostic indicator.
The strengths of the current study include a large population of patients with MMR distributions similar to other studies from institutions that initiated early universal screening for MMR defects. Limitations include those inherent to a retrospective, single-institution study. Also, 110 patients were excluded from analysis during this study timeframe due to early poor compliance following implementation of universal screening for MMR defects at our institution. To help address this, the patients included in this study were compared with those excluded due to incomplete MMR testing, shown in Supplement A (all supplemental materials can be found at American Journal of Clinical Pathology online). The groups were largely similar, although the excluded group did have increased depth of invasion and increased positive lymph nodes. Ideally, this would be addressed in a future study by completing MMR testing on the tumors to both ensure any patients with Lynch syndrome are identified and complete universal testing.
Another question that has been raised previously is the treatment efficacy of various therapies in MSI-H endometrial cancers. Two examples are highlighted. Bilbao et al concluded that MSI-H status may predict success of radiotherapy in early stage endometrioid adenocarcinomas. Alternatively, Resnick et al showed an improvement in progression-free survival with advanced-stage intact MMR expression tumors receiving chemotherapy, suggesting that MMR-deficient tumors may have an increased risk of platinum resistance. If these patients do recur more frequently, whether in the early stage setting as shown in our study or the advanced-stage setting as shown by Cosgrove et al, they may benefit from additional or more tailored adjuvant therapy. It has been suggested that molecular classification of endometrial cancers may help with the planning of adjuvant therapy.
This study adds further evidence that molecular profiling of tumors can predict tumor behavior and possibly guide treatment decisions. Even patients with low-grade, early stage disease recurred more frequently when found to have MLH1 promoter hypermethylation. These patients also recurred faster than their advanced-stage comparisons, suggesting that differences in adjuvant treatment could have an impact. Given the high prevalence of this epigenetic event and maturing data from centers that began universal screening for these defects, confirmatory studies are feasible to obtain and should be performed. Prospective clinical trials are ongoing to better characterize response to additional adjuvant treatment, notably NRG-GY020 to study pembrolizumab in addition to radiation in newly diagnosed HIR endometrial cancers and NRG-GY018 to study pembrolizumab in addition to standard chemotherapy for advanced or recurrent endometrial cancer.
Am J Clin Pathol. 2022;157(1):90-97. © 2022 American Society for Clinical Pathology