Should you repeat mismatch repair testing in cases of tumour recurrence? An evaluation of repeat mismatch repair testing by the use of immunohistochemistry in recurrent tumours of the gastrointestinal and gynaecological tracts

Comparison of PD-L1 expression and MMR status between primary and matched metastatic lesions in patients with cervical cancer

PURPOSE

Programmed death-ligand 1 (PD-L1) and DNA mismatch repair (MMR) are considered predictive biomarkers for immunotherapy in cervical cancer. However, their expression in primary tumors and metastases does not always match affecting the course of treatment. We investigated the consistency of their expression in primary and matched recurrent/metastatic lesions from patients with cervical cancer.

METHODS

Primary and matched recurrent/metastatic specimens from patients with recurrent cervical cancer (n = 194) were stained for PD-L1 and MMR (MLHI, MSH6, MSH2, and PMS2) using immunohistochemistry. The degree of consistency of PD-L1 and MMR expression in these lesions was analyzed.

RESULTS

The inconsistency rate of PD-L1 expression in primary and recurrent/metastatic lesions was 33.0%, and it varied between the recurrence sites. Positive PD-L1 rate in primary lesions was lower (15.4%) than that in recurrent/metastatic lesions (30.4%). The discordance rate of MMR expression between primary and recurrent/metastatic lesions was 4.1%.

CONCLUSION

We conclude that to use PD-L1 as a predictive biomarker for immunotherapy, analysis of both metastatic and primary lesions may be required. High consistency rate of MMR expression between primary and metastatic lesions suggests that testing primary lesions alone can be sufficient for guiding the course of therapy, thereby solving the difficulty of obtaining recurrent/metastatic specimens in clinic.

Patterns of DNA mismatch repair protein expression for primary and recurrent colorectal cancer at an advanced surgical unit: A retrospective audit

AIM

Lynch syndrome is an inherited cancer syndrome associated with an increased lifetime risk of colorectal cancer (CRC) and characterized by germline mutations to one of four DNA mismatch repair (MMR) genes. Immunohistochemical (IHC) testing is used to screen for Lynch syndrome; however, despite routine completion following resection of primary CRC, it is only variably completed following resection of recurrent disease. This may be significant, as MMR protein expression can change from primary to recurrent CRC. The primary aim of this study is to investigate how MMR profiles change from primary to recurrent CRC; the secondary aim is to assess rates of MMR testing of primary and recurrent disease.

METHOD

We conducted a retrospective analysis of patients undergoing surgery for recurrent CRC from 2018-19 at a high-volume institution. MMR profiles were obtained following both primary and recurrent resection of CRC, and MMR protein expression was evaluated from both time points.

RESULTS

A total of 107 patients met the inclusion criteria and IHC results were obtained for both primary and recurrent resections in 85 cases. MMR profiles changed in nine patients (10.6%), with a loss of staining from primary to recurrent disease in six (7.1%) and a gain of staining in three (3.5%). IHC testing was completed following 88.7% of primary and 39.3% of recurrent resections.

CONCLUSION

MMR profiles can change from primary to recurrent CRC and repeat MMR testing for recurrent CRC is completed in only a minority of cases.

Clinical and pathological analysis of companion diagnostic testing of microsatellite instability-high for pembrolizumab in gynaecologic malignancy

BACKGROUND

Microsatellite instability-high is a known biomarker for anti-PD-1/PD-L1 immune checkpoint therapy. It is also a known tumour feature of Lynch syndrome, detected most frequently in endometrial cancer. However, it remains unclear how microsatellite instability testing is carried out in the clinical field.

METHODS

Ninety-nine patients with gynaecological malignant tumours who underwent microsatellite instability testing as a companion diagnosis for pembrolizumab and 16 patients who previously underwent microsatellite instability testing as a screening for Lynch syndrome were recruited. Clinical information, microsatellite instability status, outcomes, genetic assessments and information about cancer tissue were retrospectively analysed.

RESULTS

Ninety-nine patients had 101 gynaecologic malignant tumours including 26 endometrial, 38 ovarian and 28 cervical cancers, 9 with other tumours including 2 synchronous endometrial and ovarian cancers. All tissue samples were successfully tested, even though some were ≥10-year-old samples. Three cases (3.0%, 3/99) showed microsatellite instability-high; all cases were endometrial cancers with one case of synchronous endometrial and ovarian cancer [11.5% (3/26) in endometrial cancer, 2.6% (1/38) in ovarian cancer], and there was no microsatellite instability-high in cervical and other cancers. One of the endometrial cancer patients received pembrolizumab treatment, but finally died of cancer. Two other cases underwent genetic testing; both were diagnosed as Lynch syndrome. Six cases (37.5%) showed microsatellite instability-high in screening for Lynch syndrome.

CONCLUSIONS

Microsatellite instability-high was less commonly detected as a companion diagnosis for pembrolizumab in unselected gynaecologic patients. Genetic counselling should be always provided along with treatment selection.

p53 Immunohistochemical patterns in HPV-related neoplasms of the female lower genital tract can be mistaken for TP53 null or missense mutational patterns

We have recently encountered p53 immunohistochemical (IHC) patterns in human papillomavirus (HPV)-associated carcinomas of the gynecologic tract, which were confused with absent (null) or overexpression TP53 mutational staining. We therefore evaluated p53 and p16 IHC in 25 squamous cell carcinomas (SCC) (16 vulva, 4 Bartholin's gland, and 5 cervix), 20 endocervical adenocarcinomas (EDAC), 14 high-grade squamous intraepithelial lesions (HSIL), 2 adenocarcinoma in situ (AIS), all of which exhibited morphologic features of HPV. Only cases showing diffuse/strong block-like p16 staining were included for further study. All EDACs underwent TP53 sequencing and HPV in situ hybridization (ISH) was performed in selected cases. p53 IHC staining fell into two main patterns. The most common was designated as "markedly reduced (null-like)" (absence or significantly attenuated staining in >70% of cells), which could be confused with true null mutational pattern. This was present in 14/25 (56%) SCCs, 7/14 (50%) HSILs, and 18/20 (90%) EDACs. The second notable pattern was "mid-epithelial (basal sparing)" (distinct absence of staining in basal cells juxtaposed with strong staining in parabasal cells), seen in 10/25 (40%) SCC, 7/14 (50%) HSIL, and none of the EDACs. There was scattered weak to moderate p53 staining (conventional wild type) in 1/25 (4%) SCC and 2/20 (10%) EDAC. No cases showed strong/diffuse overexpression. One EDAC had a TP53 missense mutation and exhibited "markedly reduced (null-like)" staining. HPV ISH revealed an inverse relationship with p53, cells positive for HPV mRNA were negative for p53. Knowledge of these patterns can help pathologists avoid misinterpreting p53 status in the setting of HPVA cancers.