Faculty Bibliography

OBJECTIVES/OBJECTIVE:ERBB2 (human epidermal growth factor receptor 2 [HER2]) amplification/overexpression in colorectal carcinomas (CRCs) may predict response to HER2 inhibitors. We correlated ERBB2 amplification by next-generation sequencing (NGS) with HER2 overexpression by immunohistochemistry. METHODS:NGS was performed on specimens containing 20% or more tumor. HER2 immunohistochemistry (clone SP3) was scored semiquantitatively by H-score. ERBB2 fluorescence in situ hybridization (FISH) was performed to examine copy alterations in one HER2-heterogeneous tumor. RESULTS:ERBB2 amplification was detected in 2% of 1,300 CRCs analyzed by NGS. HER2 immunohistochemistry was examined in 15 cases with ERBB2 amplification (six or more copies), 10 with low gain (three to five copies), and 77 copy neutral. ERBB2 amplification and HER2 immunohistochemistry showed perfect concordance at an H-score of 105 or more. FISH confirmed homogeneous ERBB2 amplification in a tumor showing HER2 protein expression heterogeneity. ERBB2 amplification anticorrelated with RAS/RAF mutations (P = .0001). No ERBB2-amplified cases showed mismatch repair deficiency. CONCLUSIONS:NGS-detected ERBB2 amplification highly correlates with HER2 overexpression in CRC, but immunohistochemistry is required to capture protein-level heterogeneity.

AIMS/OBJECTIVE:While dysplasia in colonic sessile serrated adenoma/polyps (SSAs) often shows loss of MLH1 on immunohistochemistry, the significance of loss of MLH1 expression in non-dysplastic crypts in these polyps is less well studied. The purpose of this study was to evaluate the prevalence of MLH1 loss in non-dysplastic crypts in SSAs, and to evaluate its significance with regard to progression of these polyps. METHODS AND RESULTS/RESULTS:400 SSAs including 158 SSAs without dysplasia, 219 SSAs with dysplasia (SSAD), and 23 SSAs with invasive adenocarcinoma (SSAC) were evaluated immunohistochemically for MLH1 loss in both non-dysplastic and dysplastic portions of the polyps. 71 of 400 (18%) SSAs showed loss of MLH1 in non-dysplastic crypts. The prevalence of MLH1-deficient non-dysplastic crypts was higher in polyps with dysplasia or carcinoma (7%, 22%, and 52% in SSAs, SSADs, and SSACs, respectively; p<0.0001). When SSAs with MLH1-deficient vs. MLH-1-proficient dysplasia were compared, those with MLH1-deficient dysplasia were more likely to have MLH1-deficient non-dysplastic crypts (66% vs. 8.1%, p<0.0001) and a greater number of discrete foci (3.6 vs 1.1 foci, p=0.008). Also, non-dysplastic crypts with MLH1 loss were more likely to be contiguous to dysplasia when the dysplasia also showed MLH1 loss (26% vs 0%, p=0.02). CONCLUSIONS:Our results suggest that loss of MLH1 in non-dysplastic crypts in SSAs precedes the development of MLH1-deficient dysplasia and adenocarcinoma and may be a biomarker of an advanced serrated polyp even in the absence of dysplasia.

It is believed that sessile serrated adenomas/polyps lead to the development of microsatellite unstable cancer via a dysplasia-carcinoma sequence. Little is known regarding the morphologic and biologic features, and outcome of sessile serrated adenomas/polyps with dysplasia, or of its specific dysplasia subtypes (intestinal versus serrated). The aims of this study were to analyze and compare the clinical, pathologic, and outcome characteristics of sessile serrated adenomas/polyps with serrated versus intestinal dysplasia. The study included 86 patients with sessile serrated adenomas/polyps with dysplasia (50 serrated dysplasia, 22 intestinal dysplasia, 14 mixed serrated and intestinal dysplasia). The clinical and pathologic features, and the prevalence rate of prior, concurrent, and future neoplastic lesions, were compared between sessile serrated adenomas/polyps with intestinal versus serrated dysplasia and with matched control patients with ≥1 conventional adenoma. The mean age of the patients, polyp size, and prevalence of adenocarcinoma within the polyps were significantly higher in sessile serrated adenomas/polyps with high versus low-grade dysplasia. Sessile serrated adenomas/polyps with intestinal dysplasia showed a significantly higher rate of adenocarcinoma (23%) compared with those with serrated dysplasia (6%, P=0.05), and the high-grade lesions occurred at a significantly younger age in the former compared with the latter (65 versus 76 years, P=0.05). Compared with patients with conventional adenomas, patients with sessile serrated adenomas/polyps with dysplasia showed a significantly higher rate of invasive carcinoma within the polyps (12 versus 0%, P=0.01) and a significantly lower association with prior or future conventional adenomas. Sessile serrated adenomas/polyps with dysplasia should be considered high-risk neoplastic precursor lesions, particularly those with intestinal dysplasia. Cancer may develop from sessile serrated adenomas/polyps with either type of dysplasia.