Highlights In a bleeding postmenopausal woman with didelphys uterus, endometrial biopsy should be taken from both uterine cavities. Sentinel lymph node mapping has not been previously described in the setting of endometrial cancer and uterine didelphys. Routine sentinel lymph node mapping was successfully performed in a patient with endometrial cancer and uterine didelphys.

Background: Endometrioid endometrial cancer (EC), the most common histologic type of EC, is underpinned by somatic genetic alterations of genes in the PI3K and MAPK pathways, and ARID1A. Here we sought to evaluate the genomic landscape of stage IA, low-volume, non-myometrial invasive ECs to define the earliest drivers and tumor evolution of endometrioid EC. Methods: Early-stage, non-myoinvasive EC of patients who underwent surgery at our institution and were subjected to clinical tumor-normal targeted sequencing of up to 505 cancer-related genes. Tumor volume was determined based on the gross and histologic measurements, and the tumors were dichotomized into low- and high- volume disease using a cutoff of 1 cm3. Genomic data was extracted and cancer cell fractions (CCF) of the somatic mutations were determined using ABSOLUTE integrating the mutational variant allele frequency and copy number alterations. Results: A total of 160 non-invasive, stage I ECs were identified, of which the majority (n=128; 80%) were FIGO grade 1 (15.6% grade 2, 3.1% grade 3). The median calculated volume of disease was 1.3 cm3, and 18 cases did not have gross disease with only microscopic disease identified on pathologic evaluation. The tumors on average had 6.7 somatic mutations affecting known EC cancer-related genes, including PTEN (n=127; 79%), ARID1A (n=84; 52.5%), PIK3CA (n=81; 50.6%), CTNNB1 (n=61; 38.1%), PIK3R1 (n=59; 36.8%) and KRAS (n=48; 30%). None of the alterations showed a significant correlation with tumor volume. Analysis of the CCF revealed that in 43% of ECs (69/160)[WB1] , PTEN mutations were clonal and had the highest CCF, indicating that PTEN mutation is most likely the initiating carcinogenic event. In other subsets of the early-stage ECs studied, the highest CCFs were mutations affecting PIK3CA/PIK3R1 (n=26), ARID1A (n=17), and KRAS (n=16). We also found that 109 (68%) ECs harbored activating mutations involving at least two genes in the PI3K/AKT/mTOR pathway, with the most common combination being PTEN/PIK3CA or PTEN/PIK3R1. AKT1 mutations occurred in absence of other PI3K/AKT/mTOR pathway alterations. Unsupervised Louvain clustering based on somatic mutations and gene copy number alterations revealed 5 distinct clusters of ECs with unique mutational drivers. Conclusion: Stage IA non-myoinvasive ECs show considerable genomic heterogeneity suggestive of multiple evolutionary pathways. Further studies are warranted to define whether this heterogeneity is a sign of early genomic drift (i.e. neutral tumor evolution) or whether the various evolutionary pathways all confer similar selective advantages. Citation Format: Sara Moufarrij, Carol Aghajanian, Nadeem Abu Rustum, Lora Ellenson, Britta Weigelt, Amir Momeni. Distinct mutational drivers and evolutionary pathways in stage IA non-myoinvasive endometrioid endometrial cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 130.

No abstract