https://orcid.org/0000-0001-5440-9643
Abstract
In advanced HER2-positive (HER2+) breast cancer, the new antibody-drug conjugate trastuzumab deruxtecan is more effective compared with trastuzumab emtansine (T-DM1). However, trastuzumab deruxtecan can have considerable toxicities, and the right treatment sequence is unknown. Biomarkers to guide the use of anti-HER2 therapies beyond HER2 status are needed. Here, we evaluated if preestablished levels of ERBB2 mRNA expression according to the HER2DX standardized assay are associated with response and survival following T-DM1. In ERBB2 low, medium, and high groups, the overall response rate was 0%, 29%, and 56%, respectively (P < .001). ERBB2 mRNA was statistically significantly associated with better progression-free survival (P = .002) and overall survival (OS; P = .02). These findings were independent of HER2 immunohistochemistry (IHC) levels, hormone receptor, age, brain metastasis, and line of therapy. The HER2DX risk score (P = .04) and immunoglobulin signature (P = .04) were statistically significantly associated with overall survival since diagnosis. HER2DX provides prognostic and predictive information following T-DM1 in advanced HER2+ breast cancer.
Antibody drug-conjugates targeting HER2 have changed the treatment landscape of HER2-positive (HER2+) advanced breast cancer (BC) (1-5). Among them, trastuzumab-emtansine (T-DM1) improves progression-free survival (PFS) and overall survival (OS) in patients with HER2+ metastatic BC previously treated with trastuzumab and a taxane (3,5). Recently, trastuzumab-deruxtecan (T-DXd) has shown superiority to T-DM1 in the second-line setting (2). However, the toxicity profile of T-DXd is not trivial. In addition, T-DXd is highly efficacious after T-DM1 (1), but no data exist about the activity of T-DM1 after T-DXd and there is uncertainty regarding the best treatment sequence (6).
To date, no biomarker of prognosis and/or treatment benefit has been implemented in advanced HER2+ BC. In early-stage HER2+ BC, the HER2DX assay is prognostic and predictive (7). HER2DX provides an ERBB2 mRNA score with specific cutoffs to identify HER2+ from HER2-negative BC according to the American Society of Clinical Oncology/College of American Pathologists (ASCO/CAP) guidelines (8) and 2 different expression levels within HER2+ BC (medium and high). ERBB2 mRNA might be a potential predictive biomarker of T-DM1 response (9-12). Here we evaluated the HER2DX variables in patients with advanced HER2+ BC treated with T-DM1 (Supplementary Methods, available online).
Eighty-seven consecutive patients diagnosed with HER2+ advanced BC and treated with T-DM1 were evaluated (Figure 1, A). Baseline patient characteristics are reported in Supplementary Table 1 (available online). Median follow-up since T-DM1 initiation was 35.8 months. Overall response rate (ORR), median PFS, and median OS were 45% (6 complete and 33 partial responses), 5.8 months, and 24.3 months, respectively (Figure 1, B and C).
Study design. A) HER2DX standardized assay was performed in archival formalin-fixed paraffin embedded (FFPE) tumor biopsies from patients with advanced HER-positive (HER2+) breast cancer (BC) treated with trastuzumab emtansine (T-DM1). HER2DX was evaluated in archival FFPE tumor samples. B) Progression-free survival (PFS) in all patients. C) Overall survival (OS) in all patients. CI = confidence interval; ORR = overall response rate.
The ERBB2 mRNA range (5.1-fold difference between lowest and highest quartiles) varied according to centrally reviewed HER2 immunohistochemistry (IHC) (Figure 2, A). According to preestablished cutoffs, ERBB2 mRNA high, medium, and low groups represented 70.2%, 19.5%, and 10.3%, respectively. ERBB2 mRNA was statistically significantly associated with ORR as a continuous variable (Figure 2, B; Supplementary Table 2, available online) and according to prespecified cutoffs (odds ratio = 5.29, P = .003).
Association of ERBB2 mRNA expression with efficacy following trastuzumab emtansine (T-DM1). A) ERBB2 mRNA expression across the centrally reviewed HER2 immunohistochemistry (IHC) groups. B) ERBB2 mRNA expression in patients with stable disease (SD)/progressive disease (PD) vs partial or complete response (PR/CR). C) Progression-free survival (PFS) according to ERBB2 mRNA expression (preestablished cutoffs). D) Overall survival (OS) according to ERBB2 mRNA expression (preestablished cutoffs). E) PFS according to ERBB2 mRNA expression (preestablished cutoffs) in patients treated with T-DM1 in the first to third line setting. F) OS according to ERBB2 mRNA expression (preestablished cutoffs) in patients treated with T-DM1 in the first to third line setting. ORR = overall response rate.
High ERBB2 expression was statistically significantly associated with better PFS and OS as a continuous (Supplementary Table 2, available online) and categorical variable according to prespecified cutoffs (Figure 2, C and D). ERBB2 remained statistically significantly associated with ORR and PFS when adjusted by the other clinical-pathological variables (Supplementary Table 3, available online). Notably, HER2 IHC was statistically significantly associated with better PFS and OS in univariate analyses, but it was not when ERBB2 expression was included in multivariable analyses (Supplementary Table 3, available online). In the patient subset treated with T-DM1 in the first to third lines, ERBB2 was statistically significantly associated with better PFS (hazard ratio [HR] = 0.70, 95% confidence interval [CI] = 0.58 to 0.86, P < .001) and OS (HR = 0.75, 95% CI = 0.61 to 0.92, P = .005) as a continuous and categorical variable (Figure 2, E and F).
To further validate the value of ERBB2 mRNA in advanced HER2+ BC treated with anti-HER2 therapies, we interrogated tumor samples of 91 patients treated with trastuzumab and lapatinib in the EGF104900 phase III trial (13). ERBB2 mRNA was associated with better PFS (HR = 0.81, 95% CI = 0.72 to 0.91, P < .001) and OS (HR = 0.85, 95% CI = 0.75 to 0.95, P = .006) as a continuous variable and as group categories (Supplementary Figure 1, available online). Of note, ERBB2-low disease in this study represented 19.8% of all cases.
In the T-DM1 dataset, we also explored the impact of tissue type (primary vs metastasis) in the ability of HER2DX ERBB2 mRNA to predict prognosis. In a univariate analysis, tissue type was not statistically significantly associated with PFS and OS (Supplementary Table 3, available online) and did not affect the association of ERBB2 mRNA with PFS (adjusted HR = 0.74, 95% CI = 0.62 to 0.88, P < .001) and OS (adjusted HR = 0.81, 95% CI = 0.68 to 0.96, P = .01). Additionally, when we evaluated 24 HER2+ paired primary and metastatic samples of an internal dataset (14), ERBB2 expression did not show a statistically significant difference between supplementary tissue types, suggesting that it is overall stable during tumor evolution (Supplementary Figure 2, available online).
Finally, we explored other variables provided by the HER2DX assay. The HER2 amplicon signature was statistically significantly associated with ORR, PFS, and OS as a continuous (Supplementary Table 2, available online) and categorical variable (Supplementary Figure 3, A-C, available online). HER2 amplicon score and ERBB2 expression were moderately correlated (Pearson coefficient = 0.59, P < .001). HER2DX pathological complete response (pCR) score was statistically significantly associated with ORR and OS but not PFS (Supplementary Table 2, available online).
HER2DX risk and immunoglobulin G signature scores, as continuous variables, were significantly associated with OS from diagnosis (HR = 1.36, 95% CI = 1.02 to 1.83, P = .04 and HR = 0.73, 95% CI = 0.54 to 0.98, P = .04, respectively). Immunoglobulin G signature as a categorical variable was also associated with OS from diagnosis (Supplementary Figure 3, D, available online). Forty-eight patients had a prior diagnosis of early-stage HER2+ BC; in 21 of these cases, HER2DX was performed in the primary tumor, including tumor and nodal staging. HER2DX high-risk disease was identified in 20 (95%) of 21 patients who had a prior diagnosis of early-stage HER2+ BC. The only patient with HER2DX low-risk disease had been diagnosed in 2003 with a pT1cN0 hormone receptor-positive disease and presented distant metastasis in 2014. Finally, we observed a significant association of HER2DX risk score with OS (HR = 1.27, 95% CI = 1.05 to 1.52, P = .01) in 125 patients with HER2+ BC who relapsed at a distant site from the publicly available METABRIC dataset (15).
Here, we show that the standardized HER2DX genomic assay provides potential predictive and prognostic value in advanced HER2+ BC treated with T-DM1. ERBB2 mRNA was previously associated with T-DM1 benefit in retrospective analyses of 3 trials (10-12) in HER2+ metastatic BC. However, these research-based determinations of ERBB2 mRNA did not evaluate specific cutoffs.
The HER2DX ERBB2-low group, which represents 10%-20% of HER2+ tumors, has an extremely poor response to T-DM1 and survival outcome. This group might benefit from other antibody drug-conjugates such as T-DXd (2) and might be spared T-DM1, reducing unnecessary toxicities and relatively high costs. In contrast, the HER2DX ERBB2-high group might be good candidates to indicate T-DM1 because of high efficacy of T-DM1, lower cost than T-DXd, less toxicity than T-DXd, and high efficacy of T-DXd at progression from T-DM1 (1). In contrast, no data exist regarding the activity of T-DM1 after progression to T-DXd.
Limitations of this study are the retrospective nature design along with the limited number of patients involved. Patients were treated according to everyday clinical practice and were heterogeneous with respect to the previous treatments received and tissue type available.
To conclude, the introduction of new anti-HER2 drugs is changing the treatment landscape and improving outcomes, with median OS exceeding 5 years. Questions remain unanswered regarding the optimal therapies and sequencing strategies for each patient. To guide these decisions, implementation of prognostic and predictive biomarkers will be needed.
Funding
This study was funded by Hospital Clinic, Dipartimento di Scienze Chirurgiche Oncologiche e Gastroenterologiche (DiSCOG)—University of Padova DOR 1721185/17 and DOR 1830512/18 (to MV.D.), and Reveal Genomics (Barcelona, Spain). F. B-M received funding from Fundación Científica Asociación Española Contra el Cáncer (Ayudas Investigador AECC 2021- INVES21943BRAS). F.S. received funding from the European Society for Medical Oncology (ESMO Fellowship—Translational) and BBVA Foundation/Hospital Clinic of Barcelona (Joan Rodés—Jose Baselga Advanced Research Contract in Oncology), O.M-S. received funding from Sociedad Española de Oncología Médica (SEOM visiting fellow 2022).
Notes
Role of the funder: The funders, except for ESMO, BBVA Foundation and AECC, had a role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Disclosures: Potential conflicts of interest are the following: A.P. reports advisory and consulting fees from Roche, Pfizer, Novartis, Amgen, BMS, Puma, Oncolytics Biotech, MSD, Guardan Health, Peptomyc and Lilly, lecture fees from Roche, Pfizer, Novartis, Amgen, BMS, Nanostring Technologies and Daiichi Sankyo, institutional financial interests from Boehringer, Novartis, Roche, Nanostring, Sysmex Europa GmbH, Medica Scientia Innovation Research, SL, Celgene, Astellas and Pfizer; stockholder and hold a salary from Reveal Genomics, SL and patents from HER2DX. C.M.P is an equity stockholder and consultant of BioClassifier LLC, and for Reveal Genomics. C.M.P is also listed as an inventor on patent applications for the Breast PAM50 assay. J.S.P is an equity stockholder and consultant for Reveal Genomics and is also listed as an inventor on patent applications for the Breast PAM50 assay, and a patent application on DNA-based predictors of breast tumor phenotypes. F.B-M. has a patent application EP21383165. L.P is listed as an inventor on patent PCT/EP2021/070788. G.G reports fees for invited speaker from EliLilly and Novartis, and Consulting Fees from Gilead. M-V.D reports fees for Consulting Fees and Fees for Non-CME Services from Astrazeneca. Daiichi Sankyo, EliLilly, Exact Sciences, Gilead, MSD, Novartis, Pfizer, Seagen. FM reports consulting fees from Novartis and Roche. V.G. reports Consulting Fees and Fees for Non-CME Services from Amgen, Exact Science, Gilead, GSK, Lilly, MerkSerono, MSD, Pfizer and Sanofi.
Author contributions: Conceptualization: AP, AV, CMP and JSP. Data curation: FBM, GG, NC, TP, MVD, FM, TG, OMS, MMA, FS, BC, LA, MV, BA and MM. Formal Analysis: AP, FBM, GG. Methodology: AP, FBM, GG, LP, PG, ES and BG. Supervision: AV, CMP, JSP, PFC, AP, VG. Writing—original draft: AP and FBM. Writing—review and editing: FBM, GG, NC, TP, LP, JM, PG, MVD, FM, TG, OMS, MMA, FS, BC, LA, MV, BA, MM, ES, BG, AV, PV, JSP, CMP, PFC, AP, VG.
Data availability
Due to the nature of this research, participants of this study did not agree for their data to be shared publicly. However, data can be made available under a data transfer agreement and upon Ethics Committee approval and we encourage investigators interested in data access and collaboration to request them using the following link: https://www.clinicbarcelona.org/en/idibaps/research-areas/oncology-and-haematology/translational-genomics-and-targeted-therapies-in-solid-tumours/tools.
References
Author notes
Fara Brasó-Maristany and Gaia Griguolo contributed equally to this work.
Aleix Prat and Valentina Guarneri contributed equally to this work.
