Breast carcinomas with low amplified/equivocal HER2 by Ish: potential supporting role of multiplex ligation-dependent probe amplification

Cristiana Ercolani, Caterina Marchiò, Anna Di Benedetto, Alessandra Fabi, Letizia Perracchio, Patrizia Vici, Francesca Sperati, Simonetta Buglioni, Vincenzo Arena, Edoardo Pescarmona, Anna Sapino, Irene Terrenato, Marcella Mottolese

This is a retrospective cross sectional study aimed to verify whether Multiplex Ligation-dependent Probe Amplification (MLPA), a quantitative molecular assay, may represent a valuable reflex test in breast cancer with equivocal HER2 expression by immunohistochemistry and HER2 gene signals/nucleus (s/n) ranging between 4.0 and 5.9 by in situ hybridization. A series of 170 breast carcinomas scored as 2+ for HER2 expression by immunohistochemistry, were selected from our files and analyzed in parallel by silver in situ hybridization and by MLPA. According to ASCO-CAP 2013 guidelines, 54/170 tumors, displaying 4.0-5.9 HER2 gene s/n, were defined as low amplified (ratio ≥ 2) or equivocal (ratio < 2) on the basis of centromere enumeration probe 17 (CEP17) status. An independent set of 108 score 2+ breast cancers represented the external validation set. Concordance between the two techniques was assessed through the use of Cohen's K statistic. A concordance rate of 78.2% (Cohen's K statistic: 0,548 95% CI:[0,419-0,677]) between in situ hybridization and MLPA was found in the whole series of 170 cases and of 55.5% (Cohen's K statistic: -0,043 95% CI:[-0,271-0,184]) in the 54 tumors presenting 4.0-5.9 HER2 gene s/n. By MLPA, we found HER2 amplification or gain in 14% of the 21 BC presenting a disomic status and in 18% of the 33 BC presenting a CEP17 > 2.0. These data were further confirmed in the external validation set. Interestingly, the 54 low amplified/equivocal breast carcinomas presented a frequency of hormonal receptor positivity significantly higher than that observed in the amplified tumors and similar to the non-amplified one (p = 0.016 for estrogen receptor and p = 0.001 for progesterone receptor). To avoid to offer patients an ineffective therapy, HER2 status should be studied more thoroughly in low amplified and equivocal cases which can have lower response rates and shorter time to progression to trastuzumab. In this context, our data indicate that MLPA may be a reliable, objective supporting test in selecting HER2 positive breast cancer patients.