Expert Insights on NRG1 Fusions as Driver Alterations and Therapeutic Targets

Evolving Paradigms, The Expanding Continuum of Care in NRG1 Fusion-Positive Cancers,

Targeted OncologyTM: Considering the rarity of NRG1 fusions, to what extent could the development of NRG1 fusion–targeting therapies fill unmet treatment needs?

TANIOS BEKAII-SAAB, MD: In pancreatic cancer, about 7% or less of patients will have KRAS wild-type cancer. Out of these KRAS wild-type tumors, [up to about] 20% could have an NRG1 fusion. We always find that these retrospective reports overestimate the incidence of rare alterations because of a number of factors, including selection bias. Pancreatic cancer remains a disease that has few options outside of chemotherapy for the overwhelming majority of patients. We frankly have less than 10% of patients with a targetable alteration with less commonly directed approaches such as the uncommon BRCA mutation, the rare G12C, and the extremely rare MSI-H [high levels of microsatelliteinstability] and/or NTRK [tropomyosin receptor kinase A] fusion. This is why we want to fish for every little piece.

NRG1 fusions are likely to account for about 1% of all alterations. Today, the unmet need is knowing the 0.5% to 1% of a subset in a certain disease that we can go after with a targeted approach. It is pretty amazing that 20 years ago, any target that affected a low percentage of patients with cancer was seen as not worthy of developing. Moving the field forward with targeting these rare alterations remain logistically challenging but worth aggressively pursuing.

D. ROSS CAMIDGE, MD, PHD: In thoracic oncology, we have become comfortable with chasing down rare but actionable slices of the oncogene pie; NTRK fusions are a great example of this. If you think your NGS [next-generation sequencing] panel should include NTRK, then it should also include NRG1; they are largely mutually exclusive with other oncogenes so they are a true driver. There is evidence of actionability from trials with HER3 and pan-HER directed therapies. Based on results of the few studies published to date, NRG1 fusion‒positive tumors are under-responsive to standard therapies, such as chemotherapy and immunotherapy.

STEPHEN LIU, MD: One of the surprising results in the eNRGy1 registry was how poor the outcomes were in NRG1-fusion–positive non-small cell lung cancer [NSCLC] with standard therapy. Immunochemotherapy, for example, had a response rate of 0% and a median progression-free survival of only 3.3 months.

There really is an unmet need for novel therapies in this space, and several agents have shown promise. The major challenge from a drug development perspective is the rarity of NRG1 fusions, compounded by the trend that most NGS platforms used now are DNA-based, while an RNA-based approach is more sensitive. Fortunately, there are now multiple RNA-based testing platforms that are commercially available and able to detect these events regularly.

MICHAEL JON PISHVAIAN, MD, PHD: Pancreatic cancer is the deadliest solid cancer, and it is expected to be the second leading cause of cancer-related death in the United States by 2030. Despite some improvements over the last 30 years, chemotherapy has really been the only treatment. As a clinician who primarily treats pancreatic cancer, I know that an extra 3 to 6 months [of survival] is clinically meaningful, and I am happy to tell patients that they are likely to live a year rather than 4 to 6 months. But in truth, we really have a lot of room for improvement in helping these patients live significantly longer.

There are a lot of studies exploring why pancreatic cancer is so resistant to chemotherapy. One of the well-established reasons is its underlying genetic profile. The KRAS gene mutation, which is the primary driver of most pancreatic cancers, makes them very resistant to chemotherapy. In addition, when you actually look at these tumors under a microscope, there aren't a lot of cancer cells; it’s mostly connective tissue that is being stimulated by cancer cells, so just physically getting a drug to kill the cancer cells is difficult. Also, there isn’t much of an immune reaction because these cancer cells have really learned how to hide from the immune system. These are just some of the key mechanisms of resistance to standard chemotherapy.

Chemotherapy generically targets cell division. Even PARP [poly-ADP ribose polymerase] inhibitors that work so powerfully in BRCA-mutated cancers are still causing DNA damage the way chemotherapy does. Targeting of gene fusions is a very active area of study in pancreatic cancer, but currently, we don’t have a lot of therapies that can target downstream intracellular pathways.

Any opportunity to improve survival by specifically targeting what is driving pancreatic cancer cells would be highly welcome. The biggest challenge to making the most of therapies that target NRG1 fusions is quite simply that people need to know that they should look for NRG1 fusions. This is true across the board for all fusion genes, including FGFR [fibroblast growth factor receptor], NTRK, RET, and NRG1, wherein the part of the gene that’s fused is hyperactive, which is often a driver of cancer cell growth. Shutting down that driver pathway can make a very profound impact on these cancer cells. In diseases like pancreatic cancer, any therapy that is going to be particularly profound will make a massive impact because our treatment options aren't that good.

TO: Which characteristics of NRG1 fusions do you think contribute to their correlation with lack of response to standard-of-care treatments, development of resistance to existing therapies, and poor overall and disease-free survival rates in patients whose tumors present with NRG1 fusions?

BEKAII-SAAB: There are not much data. This is a rare subset of tumor, so it's very difficult to understand its natural history. I go back for examples to the cancers I treat. In pancreatic or colon cancer, if a patient's tumor lacks a RAS mutation, he or she is already in a better prognostic group. Overall, these cancers behave less aggressively because of the absence of a RAS mutation. That has been my observation and my experience with these particular patients; their cancer tends to progress slower than cancers with RAS mutations.

We need more data to understand a little bit more about the prognostic indications of NRG1 fusions. We know this fusion is a driver, meaning it is going to drive the cancer to the point where it spreads and metastasizes, but we need to know if does it as aggressively as other alterations.

LIU: The biology of NRG1 fusion–positive cancers has been under-explored. In NSCLC, we know that these cancers tend to have a low expression of PD-L1 [programmed death ligand 1] and low tumor mutational burden, which may predict the lack of response to immunotherapy-based options. It is not yet clear how activation of these signaling pathways would lead to poor outcomes with chemotherapy.

PISHVAIAN: I don’t think there is a clear answer for that yet. We know that fusion genes are aggressively driving cancers; they are constitutively sending signals that cause these cancers to grow, which is part of why chemotherapy isn’t doing much. The silver lining is that once you shut down those pathways, it will often make a profound impact on treating the cancer. 4

TO: What is the rationale behind targeting the HER3 signaling pathway in tumors with NRG1 fusions?


LIU: NRG1 fusions are a very unique genomic alteration. Many of the actionable fusions we are familiar with, including ALK [anaplastic lymphoma kinase], ROS1 and NTRK, lead to a constitutively active kinase domain. NRG1 fusions work in a different way. NRG1 is a ligand for the HER3 receptor. When NRG1 binds to HER3, the HER3 receptor forms a heterodimer, often with HER2, and that triggers downstream signaling. NRG1 fusions often result in a pathologic tethering of the NRG1 ligand to the cell membrane, keeping NRG1 in close proximity to HER3, which leads to increased signaling. Agents targeting that HER3/HER2 pathway have shown some early efficacy and are being explored in ongoing clinical trials.


PISHVAIAN: The way NRG1 fusions function is that the gene alteration fuses together an EGF [epidermal growth factor]-like domain to a cell surface receptor. That EGF domain will then activate any of the HER proteins, such as HER1, which is EGFR, and HER2, HER3, HER4. When you have these fusion genes, you can activate the HER pathway (in particular, the HER3 pathway) and cause dimerization with either HER2 or EGFR. Therefore, if you have an antibody or a small molecule inhibitor that blocks HER3, you can prevent that activation in a key driving pathway.HER3-targeted drugs have been available for a while, but it was almost happenstance that existing HER3-targeted therapies were able to work to control cancer cells driven by NRG1 fusions. Now even more specific therapies are in development and clinical testing.


TO: What are some of the key questions regarding NRG1 fusions that you would like to see addressed in preclinical studies and clinical trials?

BEKAII-SAAB: There were data presented at the American Society of Clinical Oncology annual meeting, which showed very meaningful responses with zenocutuzmab in patients with NRG1 fusion–positive pancreatic cancer. We would like to see these data confirmed through expanded patient data sets. The other key question relates to whether we should consider a combination approach. A concern is that most responses [to monotherapy] tend to be limited in duration. Combination strategies with agents that target the MAPK [mitogen-activated protein kinase] pathway [and] the HER pathway should be prioritized with combinations. The preclinical strategy has to look at combinations while the clinical strategy is to continue to expand upon the promising findings that were found primarily in lung and pancreatic cancer.

I think from the preclinical standpoint, [investigators] are focused on areas where we can combine HER3-targeted therapies or NRG1 fusion–targeted agents that target that specifically target signaling that is part of this whole pathway. The preclinical strategy has to look at combinations. The clinical strategy is to continue to expand on the promising findings that we found primarily in lung and pancreatic cancer. These are the diseases where we are seeing most of the initial responses, but it seems the NRG1 fusions may be a little bit more prevalent in those 2 pathologies.

LIU: We are in the early stages of understanding these events. There is much more heterogeneity with the fusions and we do not know how this influences behavior or how to properly leverage these differences. The primary question is how to best target these fusions. We have seen some efficacy targeting the HER2 pathway, but it does not work for everyone, and in cases where we do see responses, we often see resistance emerge. At the moment, we know very little about that process as well.


PISHVAIAN: My colleague Rachna Shroff, MD, [chief of Gastrointestinal Medical Oncology at the University of Arizona Cancer Center] is an expert in cholangiocarcinoma and has really studied FGFR fusions. I was surprised to learn from her that despite the fact that there are a few dozen fusion partners, it doesn’t really seem to matter which fusion partners with FGFR; the activation is profound.

I would like to know if there are any subtle differentiations between NRG1 fusion partners and whether some are more responsive than others. I would like to know this for FGFR, NTRK, and RET, too.

I would also like to better understand mechanisms of resistance to therapies targeting NRG1-driven cancers. It’s a little more challenging with some of the antibody drugs, which are the leading contenders for efficacy just because mechanisms of resistance to small molecule inhibitors tend to be more targetable from a biochemical pathway. If a cancer becomes resistant to an antibody, then the entire pathway will no longer be relevant to that cancer.

TO: Why was the CRESTONE trial of seribantumab designed as a “tumor agnostic” study?

BEKAII-SAAB: It has to do with the rarity of the target and the lack of a clear understanding about the relevance of the target in various malignancies. This follows a similar path to alterations such as NTRK fusions and microsatellite instability high–driven cancers. For now, it appears that lung and pancreatic cancers are driving the pack. In lung cancer, we know that slightly less than 1% of lung cancers have NRG1 fusions. In pancreatic cancer, up to 1% may have NRG1 fusions. It is important to note that NRG1 fusions are mutually exclusive with other oncogenic alterations such as those in EGFR, KRAS, ALK, ROS1, and RET.


CAMIDGE: As with NTRK fusions, driver oncogenes often largely function independent of the histology. For rare cancer, this approach is also attractive from a trial standpoint to maximize accrual.

LIU:NRG1 fusions have been best described in NSCLC but also exist in pancreatic adenocarcinoma and many other tumor types. As best as we can tell, responses seem to occur independent of tissue of origin. This is a similar paradigm to NTRK fusions, which are seen across tumor types; NTRK inhibitors have demonstrated efficacy across tumor types, leading to tumor agnostic approvals. I suspect we will see a similar path forward with NRG1 fusions.

PISHVAIAN: There has been a paradigm shift in the past 10 years; precision medicine took a little while to get off the ground. Twenty years ago, pharmaceutical manufacturers were focused on developing drugs that could treat as many people as possible. A drug that could target 0.2% of tumors that harbor a specific mutation did not seem financially sound. But after we learned how powerful these drugs are, it became feasible to pursue these niche therapeutic categories.

TO: In which settings (first line, second line, etc) might seribantumab therapeutic have a role? Are there any FDA-approved agents with which you are interested in learning about potential synergistic activity with seribantumab as part of a combination therapy regimen?

BEKAII-SAAB: As I mentioned, I think agents targeting NRG1 fusions are being looked at in the refractory setting, which is likely the setting that may be submitted for approval by the FDA at some point. Is it crazy to think that you can take those to the first line? No, absolutely not. In pancreatic cancer, KRAS wild-type tumors tend to do very well. You have to be able to separate biology from treatment effect (ie, prognostic from predictive effect). Moving biologic targets to earlier lines of therapy make sense and may have a more profound effect.

CAMIDGE: The line of therapy will depend on the ease of detecting these changes and the efficacy of the therapy relative to other standard-of-care treatments. A modest response rate and a slow turnaround time for testing will make this a second-line or later approach. A fast, dramatic benefit in quickly identifiable cases would support a frontline or even chemotherapy-HER3 combination approach in the front line.

LIU: At this time, I would be comfortable enrolling patients who have had at least 1 prior line of therapy (second line and beyond), depending on the specific clinical circumstance. Seribantumab is a HER3 monoclonal antibody. In other settings, we have seen efficacy combining monoclonal antibodies with tyrosine kinase inhibitors, and I would like to see a similar approach explored with NRG1 because there are so many agents already available.

Our priority right now is to demonstrate monotherapy efficacy. Future combinations should really be based on an understanding of primary and acquired resistance, and how to best overcome and then prevent it. Hopefully, the ongoing studies will help guide forthcoming combination strategies.

PISHVAIAN: I think it’s going to be challenging to get approval in the frontline setting from the outset. The trials that are ongoing right now, which are mostly for the second-line setting and beyond, will hopefully lead to the approval of some of these drugs. After that, first-line therapy indications are likely to be pursued, just like the FGFR trials that are ongoing right now.

The only challenge with first-line therapy (and this is really true in my area of pancreatic cancer) is that when patients initially present, they are often very ill and symptomatic; they just don’t have time to wait 4 weeks to get tested
for gene fusions and need to start chemotherapy right away. Fortunately, we have 2 chemotherapy regimens that
have been shown to improve or at least maintain quality of life
for most patients.

My approach is to test patients for gene fusions as soon as I meet them. If they happen to have a fusion, I can get them treated with a fusion-targeted drug in the maintenance setting or in the second-line setting.