ABSTRACT
BACKGROUND: HER2 is overexpressed in 10 to 20% of gastro-esophageal adenocarcinoma (GE-ADK), and is a target for trastuzumab in metastatic patients. We conducted a study to compare HER2 expression between diagnostic biopsies (DBs) and surgical specimens (SSs) of GE-ADK, and to determine the influence of non-trastuzumab containing neoadjuvant chemotherapy (NAC) on this expression. PATIENTS AND METHODS: Pathological specimens from biopsies of 228 patients operated on between 2004 and 2011 were collected. Two cohorts treated (n = 141) or not (n = 87) with a NAC were constituted. Two blind independent pathological HER2 analyses on DB and on SS were carried out using immunohistochemistry (IHC) and colorimetric in situ hybridization (CISH). HER-2 overexpression (HER2+) was defined by a score 3+ in IHC, or 2+ with a positive CISH test, according to the specific HER2 scoring guidelines for GE-ADK. RESULTS: Paired HER2 status could be determined for 218 out of the 228 patients (95.6%). HER2+ rates were 13.3% on DB (29/218) and 14.7% on SS (32/218). HER2+ tumors were mainly cardial or esophageal adenocarcinomas, with a well-differentiated, intestinal histological type. HER2 status differed between DB and SS in 6% of cases. When DB analyses were added to SS analyses, the relative increase in HER2+ cases was 13.5% (17.1% for patients with NAC and 23.5% for patients with histological response to NAC, versus 7.1% for patients without NAC, P = 0.4, NS). Differences between DB and SS HER2 expression could be explained by intratumoral heterogeneity and by a HER2 expression decrease in SS after NAC in responding patients possibly due to a higher chemosensitivity of HER2-positive clones. CONCLUSION: The determination of HER2 status on DB provides results that complete those obtained with SS. Combining the analysis of DB and of SS enables to optimize the selection of trastuzumab-eligible patients in case of metastatic relapse, and particularly in previously NAC-responding patients.
