To Test or Re-Test, that is the Question: Comparison of the Mismatch Repair Deficiency between Primary and Recurrent Sites of Uterine Cancers (088)

Objectives: Mismatch repair deficiency (MMR-D) is a common genotype in uterine cancer with prognostic and therapeutic implications. Prior studies have demonstrated tumoral evolution with respect to MMR deficiency between primary and recurrent gastrointestinal cancers. Whether uterine cancers undergo similar clonal evolution between primary and recurrent sites is less well understood. The purpose of this study was to determine if MMR deficiency concordance exists between primary and recurrent disease sites in patients with uterine cancer who later develop recurrent disease. Methods: We identified patients with a history of primary and recurrent uterine cancer from a single high-volume cancer center. Histologic slides were reviewed to confirm the diagnosis of primary and recurrent cancer, and formalin-fixed paraffin-embedded tissue blocks were requested from archives. Immunohistochemistry (IHC) was performed using antibodies against mismatch repair proteins (MMR), including MLH1, PMS2, MSH2, and MSH6. Results: A total of 86 primary and recurrent matched specimens were identified from 43 patients. Eleven patients (26%) did not have sufficient tissue for analysis from the primary or recurrent site and were therefore excluded from the analysis. The majority (28/32, 88%) were at stage I at the time of diagnosis; 11/32 (34%) were high-grade epithelial, and 1/32 (3%) had carcinosarcoma. Based on IHC expression of MMR proteins, 6/32 (19%) of primary tumors had MMR deficiency (MMR-D), while 9/32 (28%) of recurrent tumors had MMR-D. All patients (6/6, 100%) with MMR-D tumors in the primary setting retained MMR-D in the recurrent setting, while 3/26 (12%) patients who were MMR proficient in the primary setting were MMR-D in the recurrent setting. Of patients with MMR-D, 5/6 (83%) were deficient due to no MLH1 and PMS2 in the primary tumor, while one tumor had no MSH2 and MSH6 expression. All six patients with MMR-D in the primary tumor had an identical IHC phenotype in the recurrent setting. Of patients whose tumor became MMR-D, 3/3 (100%) had a loss of MLH1, and 2/3 (67%) had a loss of PMS2. Two of the three patients (67%) who developed MMR-D were at grade 1 epithelial at the time of initial diagnosis. In total, 23/32 (72%) patients showed proficient expression in both the primary and recurrent tumors, and 29/32 (91%) had concordance between primary and recurrent tumors. Conclusions: Rates of MMR-D in the primary setting were 19%, consistent with rates previously reported. Patients with MMR-D uterine tumors in the primary setting are likely to remain MMR-D in the recurrent setting. We identified 12% of patients with recurrent uterine cancer who developed MMR-D in the recurrent setting. This clonal evolution represents a biomarker that has therapeutic implications for patients with recurrent uterine cancer. Objectives: Mismatch repair deficiency (MMR-D) is a common genotype in uterine cancer with prognostic and therapeutic implications. Prior studies have demonstrated tumoral evolution with respect to MMR deficiency between primary and recurrent gastrointestinal cancers. Whether uterine cancers undergo similar clonal evolution between primary and recurrent sites is less well understood. The purpose of this study was to determine if MMR deficiency concordance exists between primary and recurrent disease sites in patients with uterine cancer who later develop recurrent disease. Methods: We identified patients with a history of primary and recurrent uterine cancer from a single high-volume cancer center. Histologic slides were reviewed to confirm the diagnosis of primary and recurrent cancer, and formalin-fixed paraffin-embedded tissue blocks were requested from archives. Immunohistochemistry (IHC) was performed using antibodies against mismatch repair proteins (MMR), including MLH1, PMS2, MSH2, and MSH6. Results: A total of 86 primary and recurrent matched specimens were identified from 43 patients. Eleven patients (26%) did not have sufficient tissue for analysis from the primary or recurrent site and were therefore excluded from the analysis. The majority (28/32, 88%) were at stage I at the time of diagnosis; 11/32 (34%) were high-grade epithelial, and 1/32 (3%) had carcinosarcoma. Based on IHC expression of MMR proteins, 6/32 (19%) of primary tumors had MMR deficiency (MMR-D), while 9/32 (28%) of recurrent tumors had MMR-D. All patients (6/6, 100%) with MMR-D tumors in the primary setting retained MMR-D in the recurrent setting, while 3/26 (12%) patients who were MMR proficient in the primary setting were MMR-D in the recurrent setting. Of patients with MMR-D, 5/6 (83%) were deficient due to no MLH1 and PMS2 in the primary tumor, while one tumor had no MSH2 and MSH6 expression. All six patients with MMR-D in the primary tumor had an identical IHC phenotype in the recurrent setting. Of patients whose tumor became MMR-D, 3/3 (100%) had a loss of MLH1, and 2/3 (67%) had a loss of PMS2. Two of the three patients (67%) who developed MMR-D were at grade 1 epithelial at the time of initial diagnosis. In total, 23/32 (72%) patients showed proficient expression in both the primary and recurrent tumors, and 29/32 (91%) had concordance between primary and recurrent tumors. Conclusions: Rates of MMR-D in the primary setting were 19%, consistent with rates previously reported. Patients with MMR-D uterine tumors in the primary setting are likely to remain MMR-D in the recurrent setting. We identified 12% of patients with recurrent uterine cancer who developed MMR-D in the recurrent setting. This clonal evolution represents a biomarker that has therapeutic implications for patients with recurrent uterine cancer.