Mismatch repair deficiency in upper GI cancers needs methylation, MSI and germline testing, not IHC alone

Design

Retrospective single-centre review

Population

91 MMR-deficient upper GI and pancreatobiliary cancers in 90 patients

Primary outcome

IHC patterns and results of ancillary molecular tests

Effect

Unusual IHC 12%; IHC/MSI discordance 3/19; MLH1 methylation 37/46; Lynch 6/36

This single-centre review of 91 mismatch repair-deficient upper GI and pancreatobiliary cancers (90 patients) examined IHC patterns and ancillary tests in real practice. Stomach (44%), oesophagus (16%) and small bowel (12%) were the commonest sites.

Paired MLH1/PMS2 loss was the usual pattern (78%), but 12% showed unusual patterns of loss and 10% paired MSH2/MSH6 loss. Where MSI was tested, IHC and MSI disagreed in 3 of 19 cases (16%). MLH1 promoter methylation explained 80% of tested MLH1/PMS2-deficient tumours. Lynch syndrome was confirmed in 6 of 36 patients who had germline testing (17%), and somatic MMR variants were found in 9 of 21 sequenced tumours.

The practical conclusion: MMR deficiency outside the colon arises by several routes, and a single IHC result neither explains the mechanism nor rules Lynch syndrome in or out. Pathology reports should recommend the next step — methylation for MLH1/PMS2 loss, germline referral for MSH2/MSH6 or unexplained loss, and MSI or NGS when patterns are unusual.

• For MLH1/PMS2 loss in an upper GI cancer, recommend MLH1 promoter methylation testing
• For MSH2/MSH6 loss or unmethylated MLH1 loss, recommend germline Lynch testing
• For unusual patterns, recommend MSI-PCR or tumour NGS
• State in the report that IHC alone does not exclude Lynch syndrome
• Make sure the treating team knows MMR status also bears on immunotherapy eligibility

Why it matters

Lynch syndrome was found in one in six tested patients with upper GI MMR deficiency, a group where it is rarely sought.

Don't overread it

A single-centre series of 91 cancers with incomplete ancillary testing; the percentages are from small tested subsets.

The statistics, in plain English

Several figures come from subsets — 19 with MSI, 36 with germline tests, 21 with sequencing — so they are imprecise. One in six with Lynch among those tested may overestimate the rate if testing targeted high-risk patients.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.