LAUSR

TheAmericanSocietyforClinicalPathologytogetherwith the College of American Pathologists, the Association for Molecular Pathology, and the American Society of Clinical Oncology have recently published a guideline for biomarkers in the evaluation of colorectal cancer (CRC).1 In this article the authors support testing for RAS mutations as a negative predictor of response to EGFR-targeted monoclonal antibodies, whereas BRAF mutations and mismatched repair (MMR) status are recommended for their prognostic value and as heralds of Lynch syndrome. Unexpectedly, human epidermal growth factor receptor 2 (HER2, also referred to as ERBB2) amplification is not considered.1 In metastatic CRC as in other tumors there is a compelling need to discriminate validated prognostic or predictive biomarkers from others with insufficient evidence. However, the landscape of molecular biomarkers is rapidly evolving, and new paradigms of precision medicine imply therapeutic strategies based on enrichment designsthatallowcapturingefficacyofuncommononcogenic molecular targets with prominent actionability. HER2 amplification is a clinically relevant genetic alteration in metastatic CRC as documented in the HER2 Amplification for Colorectal Cancer Enhanced Stratification (HERACLES)2,3 and MyPathway4 clinical studies. We propose to consider HER2 as an emerging predictive biomarker in this setting for the following reasons: (1) it can be screened for with established diagnostic tools,5 (2) it is detected in a sizable 5% of patients with KRAS codon 12/13 wild-type metastatic CRC as we2,5 and others6 have established on more than 2500 stage 4 patients, (3) it can be acted on at the therapeutic level,2-4 and (4) it is suggested to be a negative predictor of benefit to anti-EGFR monoclonal antibodies.2,7 The HERACLES studies are academic precision oncology efforts that started by discovering the actionability of HER2 in pertinent in vivo models7 strengthened by the identification of histology-specific diagnostic criteria for reliable assessment of HER2 positivity in 830 patients.5 We reported a 3% to 5% incidence for this biomarker,2,5 that has been recently confirmed in subsequent studies showing a rate of positivity, defined as 2+/3+ HER2 overexpression assessed by immunohistochemical analysis corresponding to HER2 amplification detected by in situ hybridization, ranging from 1.3% to 2.2% in all stages, molecularly unselected, up to 5.2% of patients in the metastatic setting, KRAS wild-type.6 Importantly, in our case series HER2 protein expression, as defined above, and gene amplification tally quite accurately.5 In the HERACLES-A trial, 33 KRAS exon 2 wild-type, HER2-positive metastatic CRC patients refractory to standard treatments were treated with a combination of lapatinibandtrastuzumab.Theoverallresponserate(ORR) was 30.3%, with 2 patients achieving complete response. The complete response is still maintained in one of these two at 36 months.2,3 These data should be regarded as an extraordinary achievement for a chemotherapy-free regimeninsuchaheavilypretreatedpopulation.Similarresults were recently observed in the MyPathway trial,4 which reported a 38.2% ORR with trastuzumab and pertuzumab in 34 HER2-positive patients. Importantly, patients with concomitantHER2amplificationandKRASmutationsfailed torespond.4 Treatmentswerewelltoleratedwithnograde 4 or 5 adverse events observed nor withdrawals owing to patient request.2,3 Finally, several experimental and clinical studies demonstrated that HER2 activation substitutes for EGFR dependence in a fraction of patients with CRC, making this biomarker a potential negative predictor of benefit to cetuximab or panitumumab.2,7 These emerging data support consideration of HER2 as a biomarker in metastatic CRC. With a 3% incidence, HER2-positive metastatic CRC represents an uncommon but clinically relevant fraction of patients, with about 1500 new cases per year in the United States (calculated from estimated annual deaths)—a figure that is similar to other genetic alterations against which licensed drugs are effective,suchaspembrolizumabforMMR-deficientmetastatic CRC.8 Results of the HERACLES-A and MyPathway studies consistently document the therapeutic actionability of HER2,andthereforewereasonitshouldberecommended for testing in KRAS exon 2 wild-type, metastatic CRC. The lack of prospective phase 3 data validation should be considered as a caveat. However, randomized clinical trials will take a long time to achieve results in such a selected population, although there are both a strong underlying biologic rationale and a compelling clinical actionability for considering HER2 as a predictive biomarker. This is already recognized by expert opinion, and there has been some integration for biomarker-directed maintenance treatment in ongoing precision oncology trials (NCT02291289), and inclusion of HER2 in clinical guidelines, such as the ones by the European Society of Medical Oncology and National Comprehensive Cancer Network.