Our Pathology Department analyzes HER2 status in more than 1000 BC per year. In particular, between 2009 and 2016, 3066 consecutive cases showed IHC HER2 equivocal status (score 2+) and underwent ISH analysis to definitely identify HER2 positive BC. Even though in the majority of cases, ISH testing is able to accurately discriminate amplified from non-amplified tumors, 11% of our BC routine series presented a mean of HER2 gene s/n ranging between 4.0 and 5.9. According to the 2013 ASCO-CAP guidelines [
11], 8% were amplified due to a HER2/CEP17 ratio ≥ 2.0 and 3% were equivocal since CEP17 signals produced a ratio < 2.0. The two categories, low amplified and equivocal, as highlighted in our routine series, are in line with those reported by Ballard et al. [
14] in a recent and elegant study concerning the non classical HER2 FISH results. It is of paramount clinical importance to carefully define the HER2 status in these uncommon subsets of BC in which the mean CEP17 copy number drives in the final classification based on the ISH algorithm proposed by the guidelines. To this end, we evaluated whether MLPA, a quantitative molecular method [
26,
28,
29,
40] may be a helpful reflex test to define the HER2 status in 54 BC presenting a non classical HER2 ISH results (4.0–5.9 HER2 gene s/n). One hundred and sixteen BC including 55 HER2 clearly amplified (>6.0 gene s/n) and 61 clearly non amplified (less than 4.0 gene s/n, ratio < 2.0) BC represented our control series. Moreover, an independent set of 108 IHC 2+ BC, of which 10 had a non classical HER2 gene pattern, represented our external validation set. Several studies already demonstrated that MLPA is an easy method to perform and interpret HER2 status with a good correlation with IHC and ISH [
28‐
31]. Nevertheless, to the best of our knowledge, this is the first study which investigated whether MLPA may provide useful diagnostic information in two independent BC series presenting a mean of 4.0–5.9 HER2 genes/n which can be often difficult to define as amplified or not. The concordance rate between DDISH and MLPA obtained in this series of 54 BC was 55.5%, a percentage significantly lower than that obtained in the 116 BC made up by clearly amplified and non amplified BC (concordance 88.8%). This observation further indicates the peculiar biological characteristics of these low amplified or equivocal cases. As already reported by Ballard [
14], we observed, in fact, that tumors harboring 4.0–5.9 HER2 gene s/n had a frequency of ER positivity similar to the non amplified category. These data were confirmed in the external validation set supporting the recommendation to consider all the clinico-pathological features on a case-by-case basis when planning HER2-targeted therapies in these critical cases. Several recent studies which compared MLPA and ISH techniques, were broadly concordant with our results. The data reported so far have demonstrated that there are specific scenarios in which MLPA may be of particular value in HER2 testing when selecting patients to undergo anti HER2 treatment [
26‐
28,
40]. In this context, BC patients displaying a mean of the HER2 gene s/n ranging between 4.0–5.9, including both low amplified and equivocal tumors, could benefit most from MLPA. We have previously reported a small cohort of HER2 equivocal BC where 75% presented a normal HER2 status whereas 25% showed HER2 gain [
27]. Here we now report data on a series of 54 low amplified and equivocal cases where MLPA uncovers HER2 amplification in about 17% of cases and seems to be an optimal reflex test mainly in the group of 33 cases defined equivocal by DDISH. The external validation series had a strikingly similar frequency. Whether patients with HER2 equivocal tumors should receive targeted therapy remains a challenging question for oncologists, particularly as reporting guidelines evolve. The 2007 ASCO-CAP guidelines [
41] did not recommend anti HER2 treatment for patients with an equivocal test. On the other hand, the 2013 ASCO/CAP guidelines [
11] opened the possibility to oncologists to consider HER2-targeted therapy also for patients with equivocal HER2 test results, even after reflex testing with an alternative assay. Clearly, there is a need to further examine this critical subset of tumors, whose relative frequency has increased following the application of the 2013 ASCO-CAP guidelines [
11]. A recent re-evaluation of HERA trial FISH results according to the 2013 ASCO-CAP guidelines on 6018 BC, showed an increase of equivocal cases from 0.7% to 1.9% [
9]. In another recent study [
42], the authors reviewed a consecutive series of 904 BC observing a switch from HER2 FISH negative in HER2 FISH equivocal in 7,3% of cases. The majority of these reclassified cases presented a gained CEP17. The lack of an overall agreement between ISH and MLPA assays, mainly in tumors with 4.0–5.9 HER2 gene s/n, is due to both genetic heterogeneity and tumor cellularity of the sample [
27]. In line with other authors [
28], we observed that the best correlation between the two techniques is found when there is more than 30% of cancer cells in the test sample. Therefore, in order to eliminate any potential bias in our results due to low tumor cell percentage, the 170 BC included in the present study had to have at least 30% of tumor cells in the entire section. Furthermore, a tumor macrodissection was performed before MLPA testing. All cases belonging to the second group of 54 BC had the same range of mean HER2 gene s/n (4.0–5.9), but varied in the mean CEP17 s/n. According to 2013 ASCO-CAP, all the disomic tumors were HER2 amplified by ISH whereas MLPA found HER2 amplification or gain in 14% of the latter cases. In the group of equivocal tumors MLPA may objectively identify HER2 amplification in 18% of BC. This data raises the question whether the lack of amplification by MLPA is due to a bias induced by the method itself or vice-versa MLPA was able to exactly identify only the fraction of the true amplified cases. Concerning the first issue, the lack of amplification by MLPA may be due to intermixed genetic heterogeneity, an important biological mechanism that correlates with a low level of amplification and equivocal HER2 status. Lee et al. [
43] analyzed 443 HER2 positive BC and found HER2 regional and genetic heterogeneity in 6.2% and 6.8%, respectively. These authors demonstrated that both types of heterogeneity were significantly associated with low levels of HER2 gene amplification. The objective results provided by MLPA in identifying the HER2 status may be of particular clinical importance. Recent studies indicated, in fact, that patients with overall low-level or equivocal HER2 amplification had significantly lower response rate to neo-adjuvant trastuzumab [
44] and shorter time to progression and overall survival in metastatic BC treated with trastuzumab based chemotherapy than did those with high-level amplification [
18,
19]. Qian et al. [
45] compared the HER2 FISH status based on the 2007 and 2013 ASCO/CAP guidelines in 1931 BC cases. The authors noted that the last guidelines, although improving the identification of amplified cases, provided an increase in the equivocal cases. Furthermore, the observation that FISH equivocal cases do not always reflect HER2 overexpression might suggest the growing need to carefully select patients who can benefit from HER2 targeted treatment. These concerns were widely debated during the 15th St Gallen International Breast Cancer Conference [
46]. A randomised phase III trial, NSABP B47 is ongoing and will probably elucidate the best treatment approach for these patients.