Making a diagnosis of a rarely seen condition can be difficult for a pathologist, especially when key clinical and molecular observations are not available to them. We present an illustrative example. A 33-year-old woman presented to medical professionals with sudden onset of pelvic pain in October 2022. Investigations led to surgical removal of a 35.7-cm tumour, involving the right ovary. A right salpingo-oophorectomy, with associated lymph node dissection, was performed. The initial pathology reported the case as an ovarian endometrioid carcinoma, FIGO grade 1. There was endometriosis in the posterior cul-de-sac and the left pelvic side wall. Six removed pelvic and para-aortic lymph nodes were uninvolved by cancer. The final stage was pT1a N0; FIGO stage 1A. No further treatment was given. DNA testing of blood DNA for 40 different cancer susceptibility genes in 2023 (MyRIsk©, Myriad Genetics Laboratories, Salt Lake City) did not identify any germline pathogenic variants (GPVs). During follow-up in April 2025, a chest CT scan revealed the incidental finding of a left upper lobe lung mass. Biopsy was reported as an adenocarcinoma. Immunohistochemical (IHC) staining showed diffuse positive staining for CK 8/18, CDX2 and EMA, patchy positive staining with WT1 and negative staining for PAX8, ER, PR, TTF1, CK7, CK20, CK5/6, p53 and Bcl-2. Attempts were made to determine whether the lung metastasis was from the previous ovarian endometrioid carcinoma. Re-examination and additional IHC studies of the ovarian tumour revealed positivity for CDX2, CK8/18 and villin. The tumour was deemed negative for WT1, PAX8 and CK20. CDX2-positive staining led to concern that there could be a second primary gastrointestinal carcinoma present. A Cancer TYPE ID© test (Biotheranostics, San Diego) gave an 89% prediction that the tumour originated from the intestinal tract and ruled out with 95% confidence numerous other entities, including an ovarian carcinoma. This test is based on the real-time quantitative RNA expression profile of 92 genes. Initial carboplatin-paclitaxel chemotherapy was changed to FOLFOX. At this point, the patient contacted family members, who recalled that her maternal cousin had been diagnosed with follicular nodular thyroid disease and an ovarian Sertoli–Leydig cell tumour more than a decade before, and a GPV in DICER1—c.1525C>T, (p.R509*) had been identified as part of a research project conducted at McGill University.1 Further testing of blood DNA and RNA, including 77 genes associated with hereditary cancer (Ambry, Aliso Viejo), identified this same DICER1 GPV in this subject. At this point, the subject of this report contacted one of us (W.D.F.), to ask whether we had any insights into her case. Following signed informed consent, key medical records and pathology material were requested from the treating hospitals and the patient herself. Given that rare non-epithelial ovarian tumours, but not ovarian carcinomas, have an established association with DICER1 GPVs,2 our initial thought was that another diagnosis was possible. Moreover, we recently reported on an ovarian yolk sac tumour (YST) in a 28-year-old woman with a DICER1 GPV who was originally diagnosed with a grade 2 primary ovarian mucinous adenocarcinoma, FIGO stage I. The tumour contained a DICER1 somatic hotspot variant: c.5437G>A, p.E1813K.3 Hotspot missense variants in the RNase IIIb domain of DICER1 nearly always accompany a GPV in DICER1 in DICER1-related tumour predisposition (DRTP).2 Review of the initial laboratory investigations taken before the 2022 surgery revealed CEA of 27.7 ng/ml, CA125 of 150 U/ml and an AFP of 2125.0 ng/ml. All three tests quickly normalized post-surgery and have remained within normal limits. Upon pathology review, the tumour consisted of cells with vacuoles mimicking secretory endometrium arranged in papillary and cribriform architecture. This morphological appearance is one of the less known patterns of YST (Figure 1A,B). On IHC studies, the tumour cells show diffuse positive reactions to SALL4 and AFP and negative reactions to ER, which helps to confirm the diagnosis of YST (Figure 1C,D). Although we could not perform IHC study for SALL4 in the lung biopsy (tissue exhausted), the results of the IHC panel performed at the original hospital were consistent with metastatic YST. DICER1 sequencing of the YST demonstrated the known GPV c.1525C>T (p.R509*) (Figure 1E) and in addition, a tumour-confined missense variant, c.5126A>G (p.Asp1709Gly) (Figure 1F), typical for DRTP, confirming that the DICER1 mutations are the key events in the occurrence of this YST. It is known that DICER1 mutations promote a fetal-type phenotype, and as there are ongoing efforts to find the most accurate descriptive nomenclature for these tumours, it is possible that some tumours currently classified as YST may eventually be reclassified as examples of DICER1 associated primitive polyphenotypic tumours.4 There were a few ‘red flags’ in this case. The patient had an elevated serum AFP level and a close relative had DRTP. However, the association between YST and DRTP is not widely recognized—the first case was only reported in 2022.3 In addition, many pathologists may be unfamiliar with the full morphologic spectrum of YST. Its morphology can mimic other tumours and can be misdiagnosed unless specifically considered. In this patient's tumour, the YST component mimicked secretory endometrium—a recognized but less common pattern of YST. The patient's age (30s) is also at the upper end of the typical range for ovarian YST.5 Furthermore, the presence of endometriosis can bias the differential toward endometrioid carcinoma, a far more common malignancy arising in that setting. How can similar misdiagnoses be prevented? Awareness of the link between YST and DRTP would be helpful. Pathologists could be encouraged to incorporate clinical–pathologic correlation, including review of pertinent laboratory data. Familiarity with the less common morphologic patterns of YST should also be fostered. In non–subspecialty-based practices, early consultation with gynaecologic pathology might be advisable.6 Lili Fu: pathological diagnosis including immunohistochemistry, preparation of figures, manuscript editing. Afrida Ahmed: collection of all relevant clinical information, manuscript editing. Celine Domecq: Preparation of samples and DNA sequencing. Elizabeth N. Skinner: clinical information, literature review. William D Foulkes: project management, manuscript writing. All authors approved the final version of the manuscript. WDF was funded by the Canadian Institutes for Health Research (FDN—148380). The authors declare that they have no conflicts of interest with respect to this submission. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
Fu et al. (Thu,) studied this question.