Immunohistochemical staining for DNA mismatch repair (MMR) proteins is commonly used to screen for Lynch syndrome. Several laboratories have noticed a discrete punctate nuclear staining pattern for MLH1 that caused confusion in interpretation. This study was designed to investigate whether this particular staining pattern represents intact nuclear expression of MLH1. MMR proteins immunostaining and follow-up testing in 161 consecutive colorectal adenocarcinoma cases (86 biopsies, 75 resections) were retrospectively reviewed. Both discrete punctate nuclear staining and diffuse nuclear staining patterns for MLH1 were observed in internal control cells in 76 biopsies and 27 resections. Only diffuse nuclear staining was seen in the remaining 10 biopsies and 48 resections ( P < .0001). Pure discrete punctate nuclear staining pattern for MLH1 was observed in 11 tumors (9 biopsies, 2 resections), and completely negative staining was seen in 13 tumors (2 biopsies, 11 resections; P = .003). Those 24 tumors (21 patients) invariably showed loss of PMS2. Three patients whose biopsies showed pure punctate staining for MLH1 underwent repeat testing on resections: 1 retained the punctate staining and 2 showed complete loss of MLH1. Nine patients who showed loss of PMS2 and pure punctate MLH1 staining underwent molecular testing: 4 had BRAF V600E mutations and 1 had MLH1 gene mutation. Our data showed that discrete punctate nuclear staining for MLH1 is more commonly seen in biopsy specimens. Pure discrete punctate staining pattern is paired with loss of PMS2 expression and may be associated with BRAF or MLH1 gene mutation, thus it should not be interpreted as intact nuclear expression.
