Abstract 551P
Background
Tumor marker, CT, MRI, and cervical smear were conventional methods for early diagnosis and predict prognosis in endometrial cancer, but they still had limitations. Cervical smear samples, which are easy to obtain, and reflect the tumor environment, are sequenced in this study.
Methods
Patients who underwent endometrial surgery were prospectively enrolled since January 2021. Cervical smear samples were obtained preoperatively via vaginal sampling. Matched blood samples were obtained simultaneously. Tumor genomic DNA (gDNA) and cell-free DNA were extracted and analyzed with a custom panel that covers 100 endometrial cancer-related genes. Prepared libraries were sequenced using NovaSeq 6000 System (Illumina) and analyzed using PiSeq analysis (Dxome).
Results
Cervical smear samples were obtained from 101 patients, including 76 cancer patients with predominantly early-stage cancer (i.e., 40 stage IA patients and 17 stage III or IV patients) and 25 non-cancer patients. Overall, cervical swab-based gDNA detected cancer with a sensitivity of 66% and specificity of 96%. Investigation of matched blood and matched PAP smear among endometrial cancer patients showed that the detection rate of cervical swab-based gDNA was higher than either of the two modalities. Also, we identified patients effectively with loss of MSH2 or MSH6 and aberrant p53 expression based on immunohistochemistry. Genomic landscape analysis identified PTEN, PIK3CA, TP53, ARID1A, KRAS, and CTNNB1 as the most frequently altered genes. There were co-occurring genes, such as PTEN with ARID1A and PIK3CA, and TP53 was captured exclusively with CTNNB1, ARD1A, and PTEN. 50 patients were classified according to the Proactive Molecular Risk Classifier for Endometrial Cancer: MMRd (n=6, 12%), POLE (n=3, 6%), p53 abnormal (n=10, 20%) and p53 wildtype (n=31, 62%). CTNNB1 mutation occurred in 8 patients in the p53 wildtype group.
Conclusions
Cervical swab-based gDNA showed moderate sensitivity and high specificity, with an improved detection rate compared to either whole blood ctDNA or conventional PAP smear. Furthermore, it allowed patients to be classified according to ProMisE classification, having both predictive and prognostic implications.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.