Diverse ERBB2/ERBB3 Activating Alterations and Coalterations Have Implications for HER2/3-Targeted Therapies across Solid Tumors

Abstract

Abstract: Although ERBB2 (HER2) is an established oncogenic driver and therapeutic biomarker in several cancers, current drug approvals do not reflect the diverse spectrum of activating alterations across indications in which HER2-targeted therapies may yield clinical benefit. In most cancer types, HER2 status is defined by HER2 overexpression/amplification assessed by IHC and FISH, which do not provide genomic context. We sought to define the pan-tumor landscape of activating ERBB2 and ERBB3 genomic alterations detected by comprehensive genomic profiling (CGP). We queried institutional databases of solid tumor CGP, including 429,666 patients who underwent Foundation Medicine testing and 83,332 patients whose tumors were profiled using Memorial Sloan Kettering Integrated Mutation Profiling of Actionable Cancer Targets (MSK-IMPACT). We identified activating ERBB2 and ERBB3 alterations across solid tumor types, including many off-label for current HER2 drug approvals. Whereas non–small cell lung cancer represented the highest proportion of ERBB2-mutated (i.e., single-nucleotide variants and short insertions/deletions) cancers (19.0%), breast, colorectal, bladder, and gastroesophageal cancers combined accounted for 50.4% of ERBB2-mutated tumors. Within non–small cell lung cancer, 26% of activating mutations were not included in clinical trials that led to approval of the antibody–drug conjugate trastuzumab deruxtecan. We also present three clinical cases demonstrating clinical benefit from off-label use of HER2-targeted therapies. We identified substantial populations of patients with diverse ERBB2/ERBB3 activating alterations, which represent unmet therapeutic needs. We demonstrate that CGP provides additional genomic information, inclusive of ERBB2 amplification and mutation status together with potential resistance/response-modifying co-alterations, allowing for more nuanced HER2 status interpretation than is possible with IHC/FISH alone.

Significance: CGP provides genomic context for HER2 status beyond the information provided by IHC and FISH, including detection of ERBB2 mutations and co-alterations that may suggest sensitivity/resistance to HER2-directed therapies, and is therefore crucial for guiding treatment choice and understanding individual patient response.

Figure 1

Pan-tumor landscape of ERBB2/ERBB3 activating alterations detected via Foundation Medicine tissue comprehensive genomic profiling. Prevalence of (A) ERBB2 and (B) ERBB3 activating alterations (known or likely functional significance) across solid tumors. “Multiple” includes patients with AMP + MUT, AMP + RE, MUT + RE, and AMP + MUT + RE and patients with >1 MUT. C, Cancer type distribution of ERBB2 (left) and ERBB3 (right) MUT tumors. D, Prevalence of ERBB2 and ERBB3 alteration types in five major ERBB2 or ERBB3 MUT cancer types, respectively. Note that AMP/MUT/RE prevalence across plots are not additive due to tumors with multiple alterations. AMP, amplification; CNS, central nervous system; CRC, colorectal cancer; GEC, gastroesophageal cancer; HCC, hepatocellular carcinoma; MUT, mutation (single-nucleotide variants and insertions/deletions); NSCLC, non-small cell lung cancer; RE, rearrangement.

Figure 2

ERBB2 activating mutations in select cancers. Distribution of ERBB2 mutations across key HER2 protein domains (A) in the combined cohort of five major ERBB2 MUT cancer types and (B) within each cancer type. The number of mutations is indicated. C, Lollipop plot mapping ERBB2 mutations across the HER2 protein in select cancer types. The two most frequent codon hotspots for each cancer type are labeled; % represents the proportion of observed mutations at a particular codon in the respective cancer type. D, Rate of ERBB2 co-amplification with ERBB2 activating mutations in select cancer types. The number of ERBB2 MUT tumors is indicated. E, ERBB2 activating mutations detected in tissue CGP and reported by Foundation Medicine were compared with the list of mutations included in the DESTINYLung-01 trial. Matching mutations were considered “on-label,” and nonmatching mutations were considered “not on-label” for T-DXd in NSCLC. AA, amino acid; CDx, companion diagnostic; CRC, colorectal cancer; ECD, extracellular domain; GEC, gastroesophageal cancer; KD, kinase domain; MUT, mutation (single-nucleotide variants and insertions/deletions); NSCLC, non-small cell lung cancer; TMD, transmembrane domain; T-DXd, trastuzumab deruxtecan.

Figure 3

HER2-directed therapy response-modifying alterations in select cancers. Co-occurrence of alterations that have been reported to impact response to HER2-directed therapies (mAb, TKI, and ADC) in tumors with ERBB2 amplification or ERBB2 activating mutations, including TP53 MUT, PIK3CA MUT, CCNE1 AMP, RB1 MUT/DEL, PTEN MUT/DEL, ERBB3 AMP/MUT, EGFR MUT, MET AMP/Ex14 MUT, and IGF1R AMP. Patients with multiple ERBB2 alteration types (e.g., AMP + MUT) were excluded from the analysis. ADC, antibody-drug conjugate; AMP, amplification; DEL, deep deletion; CRC, colorectal cancer; GEC, gastroesophageal cancer; mAB, monoclonal antibody; MUT, mutation (single-nucleotide variants and insertions/deletions); NSCLC, non-small cell lung cancer; TKI, tyrosine kinase inhibitor.

Figure 4

Detection of ERBB2 activating alterations in select cancers via Foundation Medicine liquid comprehensive genomic profiling. A, Prevalence comparison of ERBB2 amplification (top) and ERBB2 mutations (bottom) detected in patient-matched tissue and liquid CGP across select cancer types in all matched liquid samples (N = 1,922) and only in matched liquid samples with elevated ctDNA TF (≥1%; n = 1,047). The number of samples in each group is indicated. PPA and NPV for liquid biopsy detection of (B) ERBB2 amplification and (C) ERBB2 mutations in select cancer types combined. Ninety-five percent confidence interval (CI; Wilson Continuity Corrected Method) are shown. AMP, amplification; CRC, colorectal cancer; GEC, gastroesophageal cancer; LBx, liquid biopsy; MUT, mutation (single-nucleotide variants and insertions/deletions); NSCLC, non-small cell lung cancer; NPV, negative predictive value; PPA, positive percent agreement; TBx, tissue biopsy.

Figure 5

Clinical experience targeting ERBB2 mutations. A, Flow diagram of MSKCC patients with ERBB2 MUT tumors treated with HER2-directed therapies. The distributions of ERBB2 mutation type and class of HER2-targeted therapy received are shown for response evaluable patients (N = 176) and for patients with clinical response (N = 75). “Other” ERBB2 mutations in the clinical response group included an R678Q juxtamembrane domain mutation and a V697L mutation. B, Distribution of treated ERBB2 MUT tumors across cancer types stratified by response. C, Representative tumor imaging from three patients with ERBB2 MUT tumors. C1–C2: 59-year-old female with cervical cancer harboring ERBB2 S310F mutation. PET scans of the retrocaval node performed at baseline and 8 weeks into treatment with neratinib are shown. C3–C4: 53-year-old female with metastatic breast cancer harboring ERBB2 L755S mutation. PET scans of the liver at baseline and 12 months into treatment with neratinib plus trastuzumab are shown. C5–C6: 48-year-old male with lung cancer harboring ERBB2 V659E mutation. CT scans of the liver at baseline and 8 weeks into trastuzumab emtansine (T-DM1) treatment are shown. ADC, antibody-drug conjugate; CR, complete response; CRC, colorectal cancer; CUP, cancer of unknown primary; ECD, extracellular domain; KD, kinase domain; mAB, monoclonal antibody; MUT, mutation (single-nucleotide variants and insertions/deletions); NSCLC, non-small cell lung cancer; NR, no response; PR, partial response; TKI, tyrosine kinase inhibitor; TMD, transmembrane domain.