Case history

Here, we report a patient with a history of LS-type cancers. The patient was first diagnosed with stage IC endometrial cancer at the age of 44 years and underwent a total abdominal hysterectomy and bilateral salpingo-oophorectomy, followed by radiation therapy and adjuvant chemotherapy. The patient developed a colorectal adenocarcinoma (rectosigmoid) aged 64 and underwent a low-anterior resection. Routine IHC screening of this tumour for MMR deficiency identified dual loss of MSH2 and MSH6 expression, suggesting a primary defect in MSH2. Germline testing at that time identified a pathogenic deletion of exons 4–6 in MSH2, which was predicted to result in truncation of the MSH2 protein. When a urothelial carcinoma of the bladder was found at age 66, it was treated definitively with fulguration. In 2017, after presenting with abdominal pain, the patient aged 67 years was subsequently found to have a locally advanced pancreatic adenocarcinoma. CT of the abdomen showed extension into the retroperitoneum and involvement of the mesenteric vasculature, duodenum and mesenteric lymph nodes. Endoscopic ultrasound revealed severe stricture of the duodenum, and biopsy showed ductal adenocarcinoma. Unfortunately, the biopsy volume was inadequate for IHC testing of the MMR proteins or PCR-based MSI testing. After starting chemotherapy, the patient underwent another biopsy of the pancreatic cancer and IHC of all four MMR proteins was performed successfully. Surprisingly, the pancreatic cancer showed intact protein staining of all four MMR proteins, including MSH2 and MSH6 (figure 1). This finding suggested a sporadic occurrence of the tumour apparently unrelated to the germline MSH2 mutation. Unfortunately, attempts to determine MSI status and mutational burden via NGS failed due to insufficient tumour material in the PDAC biopsy sample, hence we were unable to corroborate the IHC findings with other biomarkers of MMR status. Given the unexpected IHC result in the PDAC, we undertook retrospective IHC testing of the MMR proteins in each of the prior malignancies (endometrial carcinoma, colorectal adenocarcinoma and bladder carcinoma) from the patient. All three other tumours had dual loss of MSH2 and MSH6 staining, consistent with MSH2-LS-associated tumorigenesis in them.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Immunohistochemical staining of the mismatch repair proteins in the patient's tumours. (A–E) Endometrial adenocarcinoma: H&E staining (A), with intact MLH1 (B) and pms2 (C), and dual loss of MSH2 (D) and MSH6 (E) immunostaining in tumour cell nuclei. (F–J) Colonic adenocarcinoma: H&E staining (F), with intact MLH1 (G) and pms2 (H), and dual loss of MSH2 (I) and MSH6 (J) immunostaining in tumour cell nuclei. (K–O) High-grade papillary urothelial carcinoma: H&E staining (K), with intact MLH1 (L) and pms2 (M), and dual loss of MSH2 (N) and MSH6 (O) immunostaining in tumour cell nuclei. (P–T) Pancreatic ductal adenocarcinoma: H&E staining (P) with intact MLH1 (Q), pms2 (R), MSH2 (S) and MSH6 (T) immunostaining in tumour cell nuclei. All photomicrographs taken at ×400 magnification.

An alternative hypothesis is that haploinsufficiency for MSH2 may have contributed to tumorigenesis of the PDAC. In support of this, reduced expression of MSH2 mRNA has been shown to significantly decrease MMR efficiency using in vitro assays,13 to promote cell-cycle progression and decrease the rate of apoptosis in the proliferation of pituitary adenomas,14 and to cause low-level MSI in patients with LS patients without a cancer diagnosis.15

Finally, we considered the possibility that intact IHC staining of the MMR proteins in the PDAC may have represented a false-negative IHC test result (lack of sensitivity to detect true MMR deficiency). Rare instances of intact IHC staining of a MMR protein have been observed in LS-associated tumours despite loss of function of both genetic alleles, if one of the pathogenic mutations was a missense mutation.16 This presumably allowed for antibody binding in the IHC test, even if the encoded protein was malfunctional.16 However, we considered this unlikely in the PDAC of our patient, given IHC staining of MSH6 was also intact. MSH2 and MSH6 are binding partners in the functional MutSα heterodimer, and retention of the MSH6 protein is dependent on stable binding with functional MSH2. If MSH2 function was compromised, loss of staining of its dependent binding-partner MSH6 would be anticipated. Our finding raises the question of which biomarker is optimal for determining MMR status in rarer and extracolonic tumours. In this case scenario, given the PDAC arose in the context of known MSH2-LS, and only a limited biopsy sample was available, we considered expression status of MSH2 and its binding partner MSH6 by IHC would be the most informative for determining MMR status.

Irrespective of the mechanistic basis for tumour development, or which test may prove optimal for determining MMR status in limited biopsy samples, the finding of intact MMR protein expression in the PDAC at the time of clinical treatment decision-making deemed the patient ineligible for immune therapy with pembrolizumab. This is, to our knowledge, the first documented case of an MMR-intact pancreatic ductal adenocarcinoma in the setting of MSH2-LS, which is associated with high lifetime risks for the full spectrum of Lynch-type cancers. Interestingly, another study has described MMR-intact PDAC in two patients with PMS2 germline mutations, although one of these patients also carried a germline mutation of BRCA2.8 Germline PMS2 mutations are associated with lower lifetime risks and later age of cancer onset than MSH2 mutations, hence sporadic occurrences of cancer in PMS2 mutation carriers may be less surprising. Based on our experience, we recommend that each tumour in patients with LS be tested for MMR protein expression, as discordant results may have profound effects on treatment opportunities.