Mutational DNA profile in non-atypical endometriomas using Next-generation-sequencing

Endometriosis is a common gynecologic disorder, affecting at least 6 – 10% of reproductive-aged women. Endometriotic implants are observed all over the pelvis but the endometriosis-associated cancers typically arise in endometriomas. This observations rise up the question whether the special ovarian cyst environment, especially oxidative stress caused by free „catalytic“ iron, induces cancer driver gene mutations in endometriotic implants that may lead to carcinogenic transformation in these cysts.

The study was approved by the cantonal ethics committee. Formalin-fixed and paraffin-embedded (FFPE) tissues of 16 ovarian endometriotic cysts (endometrioma) were included in the study. Three tissue cylinders were punched from each tissue block. Genomic DNA was extracted and next generation sequencing (NGS) was performed using targeted amplification with the AmpliSeq Comprehensive Cancer panel (Ion Torrent, Life Technologies) which includes all exons of 409 cancer genes. Alignment, variant calling, and annotation were done using Ion Reporter 4.2. To focus on variants with potential cancer driver properties, we searched specifically for variants with entries in the databases Cosmic, ClinVar, and dbSNP. Furthermore, SNV and MNV variants were filtered in for deleterious amino acid changes (SIFT = 0.85). The dbSNP variants were selected to gain information about potential predisposing germ-line mutations (including only dbSNPs with an occurrence in the population of less than 1% frequency). Validation of NGS variants was done by Sanger sequencing. Germ-line mutations were detected by analysis of adjacent normal tissue of the patient.

One of the 16 cases showed low quality metrics and was therefore not suitable for the use in the subsequent NGS workflow. The mean depth of the target regions in the remaining 15 cases ranged from 181x – 556x total coverage and the uniformity from 79.4% – 87.51%. A SNV in KMT2C and an INDEL mutation in KAT6B were verified as true-positive variants in one patient each, but analysis of adjacent normal tissue identified them as germ-line mutations. Analyzing the mutational landscape in the endometrioma of one patient more closely revealed three TP53 mutations: i) p.R248Q; variant allele ratio: 0.08; COSM10662; dbSNP: rs11540652; ClinVar: with pathogenic allele, ii) p.P190L; variant allele ratio: 0.17; COSM43657, and iii) p.P191fs*56; variant allele ratio: 0.14; COSM45341. Due to limited material, only the mutation giving rise to p.R248Q was investigated by Sanger sequencing. Interestingly, this mutation was verified by Sanger sequencing whereas no mutation was detected in DNA derived from adjacent normal tissue classifying it as a potential somatic tumor driver mutation.

Our results demonstrate that despite persistent oxidative stress caused by free iron, endometriomas do generally not harbor somatic gene mutations. Nevertheless, somatic cancer driver mutations, such as mutations in TP53 in one of our cases, may be found in benign endometriomas. TP53 mutations are frequently found in high-grade serous carcinomas, which are not considered as typically endometriosis-associated ovarian cancer, but have also been described in endometriosis-associated high-grade endometrioid and clear cell carcinomas in several studies. However, at the present time, NGS seems not to represent a diagnostic tool for the detection of patients with increased risk of endometriosis-associated ovarian cancer.