LAUSR

Ovarian highgrade serous carcinoma (HGSC) is thought to mainly arise from the Fallopian tubes from a precursor lesion known as serous tubal intraepithelial carcinoma (STIC). 2 Germline pathogenic variants (GPVs) in BRCA1/2 are the most important causes of HGSC and riskreducing salpingooophorectomy (RRSO) is offered to BRCA1/2 heterozygotes. We report a BRCA1 heterozygote who developed a pelvic HGSC 15 years postRRSO, following an occult Fallopian tube HGSC that was missed on initial pathological examination. Exome sequencing of DNA derived from normal and tumour tissues showed that the pelvic tumour is a recurrence from the original Fallopian tube tumour. BRCA1 and BRCA2 GPVs are the strongest known risk factors for HGSC, and for this reason, RRSO is offered to women who are heterozygous for BRCA1/2 GPVs as standard of care, usually between the ages of 35 and 50 years. For women who have undergone RRSO, there persists a small but significant risk for peritoneal cancer. A systematic review found that 1.2% of BRCA1/2 patients undergoing RRSO had occult tubal carcinoma at the time of surgery. Moreover, 0.54% of these patients went on to develop peritoneal cancer postRRSO, of which only 17% (4/24) had STICs at initial diagnosis. Incomplete removal of adnexa, lack of standard pathological sectioning protocol and early precursors unidentifiable by routine histological examination potentially explain the lack of detection of early carcinomas at the time of RRSO. Nonetheless, the failure to find a precursor for all recurrences, as well as the oftensignificantly delayed presentation of the peritoneal cancer after RRSO, highlights some unresolved aspects to the predominant model that most, if not all, pelvic HGSC arise from the Fallopian tube. 4 Several studies have elucidated identical TP53 mutations in a primary HGSC and a second HGSC arising in the same woman, suggesting a common origin for both tumours. 6 More extensive genetic and clonal analyses, however, have not yet been published and are presented in our case report. Here, we report the case of a woman in her 60s, heterozygous for the common Ashkenazi Jewish founder BRCA1 GPV (NM_007294.4:c.68_69delAG; p.Glu23fs), who underwent RRSO with total hysterectomy 15 years previously. At the time of RRSO, both the adnexa and uterus were reported to be benign on pathology review and no further treatment was offered. The ovarian cancer tumour marker CA125 was first measured 3 years after surgery. It remained below 35 U/ mL until 15 years after the prophylactic surgery, when the patient was found to have a CA125 of 513 U/mL. Subsequent CT of the abdomen and pelvis revealed a 5 cm lower right quadrant mass. Biopsy confirmed a HGSC of Müllerian origin, and debulking of the pelvic mass with pelvic lymph node dissection was performed. On pathological examination, the pelvic tumour showed typical morphology of HGSC and abnormal (diffuse aberrant) staining for P53 on immunohistochemistry (IHC; figure 1A,B). The patient’s previous pathology material from RRSO with hysterectomy was reviewed and a tiny focus (0.8 mm) of HGSC with adjacent STIC was identified, showing highgrade nuclear atypia with abnormal (diffuse aberrant) staining for P53 (figure 1C,D). Given the long interval, it was debated whether the second tumour was a recurrence or an independent primary lesion, hence molecular studies were undertaken to address this question. DNA extraction, wholeexome sequencing (WES) and data analysis are described in online supplemental methods. Analysis of somatic copy number alteration (SCNA) profiles (figure 2A) derived from WES shows loss of heterozygosity (LOH) of chromosome 17 in both the Fallopian tube carcinoma (FTC) and the pelvic carcinoma (PC). This indicates that the alternate (wildtype) allele to the BRCA1 GPV that is present in both the normal and the tumour tissue (figure 2B) has a second hit event (LOH) in both the FTC and the PC. Whole genome duplication is seen in the PC (ploidy 4.3) compared with the FTC (ploidy 2.1). Consistent with this, a higher frequency of amplifications was observed in the PC compared with the FTC. We found greater genomic instability in the PC than the FTC, shown by greater number of SCNA segments in the PC versus the FTC (figure 2A). The overall pattern of these SCNAs suggests that the tumours are related. The PC appears to have evolved from the FTC with whole genome duplication being a critical step in its appearance 15 years after the FTC. It is possible that both the FTC and the PC are Cancer genetics