HER2 amplification in gastroesophageal adenocarcinoma: correlation of two antibodies using gastric cancer scoring criteria, H score, and digital image analysis with fluorescence in situ hybridization.

We assessed 103 resected gastroesophageal adenocarcinomas for HER2 amplification by fluorescence in situ hybridization (FISH) and 2 commercial immunohistochemical assays. Of 103, 30 (29%) were FISH-amplified. Both immunohistochemical assays had greater than 95% concordance with FISH. However, as a screening test for FISH amplification, the Ventana Medical Systems (Tucson, AZ) 4B5 antibody demonstrated superior sensitivity (87%) compared with the DAKO (Carpinteria, CA) A0485 (70%). Of the cases, 28 were immunohistochemically 3+ or immunohistochemically 2+/FISH-amplified with the 4B5 assay compared with only 22 cases with the A0485 assay, representing a large potential difference in patient eligibility for anti-HER2 therapy. Cases with low-level FISH amplification (HER2/CEP17, 2.2-4.0) express lower levels of HER2 protein compared with cases with high-level amplification (HER2/CEP17, ≥4.0), raising the possibility of a differential response to anti-HER2 therapy. The H score and digital image analysis may have a limited role in improving HER2 test performance.

Challenging cases encountered in colorectal cancer screening for Lynch syndrome reveal novel findings: nucleolar MSH6 staining and impact of prior chemoradiation therapy.

Many pathology laboratories have developed specific screening protocols to detect patients with Lynch syndrome. With recent recommendations to test all patients with newly diagnosed colorectal cancer for Lynch syndrome, the volume of testing will increase, and the most economic and reliable screening test will prevail. Although the detection of microsatellite instability by polymerase chain reaction and the detection of loss of the mismatch repair proteins by immunohistochemistry can each be used as a screening tool, each methodology has its strengths and weaknesses. During the time of our study, we used both polymerase chain reaction and immunohistochemistry to screen for Lynch syndrome in colorectal cancer specimens. We encountered 21 cases that posed significant interpretive challenges. A previously unpublished pattern of nucleolar MSH6 staining and potential spurious results induced by chemoradiation therapy are described. We feel that it is important to report these cases so that potential pitfalls in screening for Lynch syndrome can be avoided.