Detection of NRG1 Fusions in Solid Tumors: Rare Gold? Free

In this issue of Clinical Cancer Research, Jonna and colleagues report that they identified NRG1 fusions in approximately 0.2% of over 20,000 patients with diverse tumor types including 0.3% in patients with lung cancer on the basis of a targeted RNA-sequencing assay in tumor specimens (1). NRG1 fusions were originally described in patients with lung adenocarcinoma and negative testing for other common driver mutations. These molecular aberrations are enriched in patients with invasive mucinous adenocarcinoma of the lung. The findings by Jonna and colleagues highlight the potential relevance of this genomic subset in the context of a growing body of literature showing evidence that this oncogenic event may be clinically actionable.

NRG1 produces a protein that belongs to the neuregulin (NRG) family of ligands with specificity for ERBB3 (HER3) and ERBB4 (HER4). HER3 and HER4 belong to the ERBB family of transmembrane receptors, which also includes ERBB1 (EGFR) and ERBB2 (HER2). The ERBB family members are receptors for growth factors and they form homo- and heterodimers upon binding to cognate ligands. The dimers then activate downstream pathways, especially MAPK and PI3K/AKT (2). HER2 does not have a cognate ligand; rather it forms heterodimers preferentially with HER3 when these other ERBB family members are ligand activated. In contrast, HER3 binds to the EGF-like domain of neuregulins but has a much less active kinase compared with EGFR and HER2. Upon binding to neuregulins, to signal, HER3 forms heterodimers with either EGFR or HER2 (2). HER2/HER3 heterodimers exert the strongest signaling activity among the homo- and heterodimers that form between the ERBB receptors. HER2 phosphorylates and activates HER3 within the HER2/HER3 complex. HER3 subsequently binds to p85 and activates the PI3K/AKT pathway. In addition, the HER2/HER3 heterodimer signals through the PI3K/AKT and MAPK pathways (2). There are four neuregulin genes (NRG1–4), and NRG1 the best characterized gene of the group encodes a transmembrane protein, which can exist in several types with over 30 isoforms. The EGF-like domain of NRG1 functions as the key receptor binding region of the ligand. NRG1 fusions exert protumorigenic functions in lung cancer that require activation of HER2/HER3 heterodimers. NRG1 fusions are cancer driver genomic events when detected in newly diagnosed patients, but they can also be subclonal events in patients with other driver mutations at the time of progression to targeted therapy. In the latter scenario, a RALGAPA1–NRG1 fusion has been reported as a putative driver of secondary resistance to alectinib in a patient with ALK fusion–positive lung cancer (3). Several fusion partners have been described, like CD74 and SLC3A2 that are fused with NRG1 type III-β3 isoform (2). NRG1 contributes the EGF-like domain to the resulting novel transmembrane fusion protein. The fusion, just like the wild-type form of NRG1 may drive ligand activation directly in an autocrine or juxtacrine manner through surface expression of the tethered EGF-like domain or this domain may be cleaved and function as a free ligand activating receptors on surrounding cells in a paracrine manner (Fig. 1; ref. 4). The III-β3 isoform is the least prone to undergo cleavage and most of the protein exists in the transmembrane preprotein form. However, some degree of cleavage and EGF-like domain shedding is known to occur for both the wild-type and the fusion mutant (4) The NRG1 fusion product is a transmembrane protein with an extracellular EGF-like domain that binds to HER3 in the same or adjacent cells (Fig. 1). In addition, the EGF-like domain can be cleaved to some extent and activate HER3 in a paracrine fashion. In all cases, binding of the EGF-like domain of NRG1 to HER3 triggers the formation of HER2–HER3 heterodimers and activation of downstream pathways. The presence of NRG1 fusions promotes migration, proliferation, and tumor growth in in vitro and in vivo models.

NRG1 fusion–positive cancer dependence on ERBB family of receptors, indicates possible vulnerabilities that can be explored from a therapeutic perspective. In particular, it suggests the potential for treating these patients with agents that target HER2 or HER3. Although wild-type NRG1 bound to HER3 activates EGFR/HER3 heterodimers, EGFR-specific tyrosine kinase inhibitors with no activity against HER2 do not adequately suppress growth and signaling pathways in NRG1 fusion–driven models (5). In contrast, drugs with dual EGFR/HER2 activity, like lapatinib, afatinib, and neratinib show higher promise (4, 5). A potential explanation is that NRG1 fusion encodes for a protein that directs signaling preferentially through the HER2/HER3 complex, which causes the most robust activation of downstream pathways compared with other ERBB family homo- and heterodimers; thereby suppression of HER2/HER3 signaling is required for adequate response. Clinically, patients with NRG1 fusion–positive lung cancers treated with afatinib have mixed outcomes in case reports: in some reports, patients achieved clinical responses to afatinib that generally lasted for less than 12 months (6) and in others, no such benefit was observed (5). A durable clinical response has also been reported in an individual patient treated with the HER3 inhibitory antibody GSK2849330 (5). There is also preclinical rationale to target both HER3 and HER2 together rather than any component of the heterodimer alone (4). Targeting of HER2 with inhibitory antibodies (e.g., trastuzumab and pertuzumab) or antibody drug conjugates (e.g., TDM1) could also be explored clinically.

So do the current data support incorporating NRG1 fusions in molecular companion diagnostic testing for patients with lung and other cancers? The answer is yes, as NRG1 fusions are cancer driver molecular alterations that appear to follow the classical oncogene addiction model and they may be clinically actionable.

However, in the absence of clinical trial data, it is possible that reported cases are skewed by strong publication bias. It is therefore unknown to what extent all patients with various NRG1 fusions truly benefit from HER2 or HER3 directed agents and how outcomes compare with standard-of-care treatments. Whether any benefit from such treatments is limited to patients with NRG1 fusion–positive lung cancer or it can be expanded to histology agnostic cases with NRG1 fusion is also unknown. As the list of actionable and potentially actionable genetic alterations increases in non–small cell lung cancer and other solid tumors, the use of high throughput next-generation sequencing in complementary DNA and RNA-based assays that test multi-gene panels will reveal NRG1 fusion cases, provided the utilized primers include NRG1 among their analytes. The available early data are intriguing and fully support the design of a basket trial of ERBB2 and/or ERBB3 inhibitors for this molecular group of patients. The robust collection of drug activity together with biomarker analyses to address the potential impact of any underlying molecular heterogeneity on therapeutic outcomes will be vital to fully inform how well we can make this rare gold really shine.

A. Dimou reports receiving honoraria from Roche/Genentech. No potential conflicts of interest were disclosed by the other author.