Michael J. Overman, MD, highlights new combination agents under investigation for the treatment of patients with GI cancers—specifically CRC—and challenges that still need to be addressed.
Michael J. Overman, MD
While PD-1/PD-L1 inhibitors have demonstrated encouraging activity in a subset of patients with gastrointestinal (GI) cancers, there are other drugs in the space that are currently in development, said Michael J. Overman, MD. The challenge lies in figuring out how to optimize the activity seen with this class of agents and bring them into the frontline setting.
“Immuno-oncology is currently a big area of development in GI cancers,” said Overman, who is a professor of gastrointestinal medical oncology at The University of Texas MD Anderson Cancer Center. “It is revolutionary compared with our classic chemotherapy-based approaches.”
For example, the combination regimen comprised of the PD-1 inhibitor nivolumab (Opdivo) and the CTLA-4 inhibitor ipilimumab (Yervoy) continues to be explored in colorectal cancer (CRC).
The phase II CheckMate-142 trial examined the combination in a cohort of 119 patients with microsatellite instability-high (MSI-H) or mismatch repair deficient (dMMR) metastatic CRC. In this cohort, 82 patients received prior treatment with a fluoropyrimidine, oxaliplatin, and irinotecan. Results showed an overall response rate (ORR) or 46% (95% CI, 35%-58%).1 These data led to the FDA’s accelerated approval of the combination for the treatment of adult and pediatric patients 12 years and older with MSI-H mCRC.
Updated data, which were presented at the 2018 ESMO Congress, showed that the combination improves survival in treatment-naïve patients for MSI-high mCRC, highlighting its potential as a new first-line treatment in this population. In 45 patients with no prior treatment, the ORR with the combination was 60%, while the disease control rate was 84%. The complete response rate was 7%.2
In an interview with OncLive during the 2018 ISGIO Meeting on Immune-oncology in Gastrointestinal Cancers, Overman highlighted new combination agents under investigation for the treatment of patients with GI cancers—specifically CRC—and challenges that still need to be addressed.Overman: Immune-oncology is a current big area of development in cancer. There are many different drugs under development in this space, but the drugs that have had the most success have been these drugs targeting PD-1 or PD-L1. There are a number of different drugs that are FDA-approved for several different cancer types. The question is, “What other drugs are coming after the initial wave of multiple PD-1/PD-L1 agents?” There are a lot in development, and [we need to figure out] some of the bigger picture of how to decide which other targets are relevant.One of the big ones is a target called CTLA-4, which is an immune checkpoint on T cells that inhibit an immune response; if you block this target, you can stimulate an immune response. There have been a number of studies looking at combinations targeting PD-1 and CTLA-4.
The classic combination that has been FDA approved is nivolumab and ipilimumab, a combination that has been approved for a while in melanoma. Recently, we had that combination approved in GI cancer, specifically for [patients with] MSI-high or dMMR CRC. We have this approval for patients after they had at least 1 or 2 lines of therapy. This combination really demonstrated pretty dramatic activity and appears to be potentially more active than just single-agent therapy, although it hasn't been compared in a randomized fashion.
However, when you look at the data, there is a refractory data set of patients with MSI-H CRC treated with nivolumab/ipilimumab where you see a 12-month PFS of around 80%, which is just really outstanding. At 1 year, only 20% of patients have progressed, which is a dramatic finding for us and a very high response rate.
There were some updated data reported at the 2018 ESMO Congress looking at that same combination but in a frontline setting for metastatic MSI-high CRC. It's early in the follow-up, but it demonstrates dramatic activity, even when you first use it in patients. Therefore, the study was of 45 patients in the frontline setting, and it does [suggest] that we should be considering immunotherapy agents alone and in combination—especially in this dMMR population—earlier.In the CRC space, we have a lot of interest in recent changes to molecular understanding of CRC and identifying various different molecular subsets and treating them differently. We're definitely in an age where CRC is not an entity but many different subsets within. One example is MSI-H. That is a subset within RC that is about 4% of all metastatic CRCs, and this type should be treated with immunotherapy. There's another subset that has this BRAF V600E mutation; we now have National Comprehensive Cancer Network (NCCN) guidelines to treat that alteration.
There is a regimen of vemurafenib (Zelboraf), irinotecan, and cetuximab (Erbitux) that is now in the NCCN guidelines to treat patients with BRAF-mutated metastatic CRC targeting that mutation. There are a number of trials as well that are targeting that mutation, but BRAF is a valid proven target.
HER2 amplification is another target within CRC for which there are a number of different early trials with various combination-based therapies demonstrating fantastic activity when you target HER2. There are several trials looking at targeting HER2 in CRC. If you have that HER2 amplification, combination therapy should have a very high chance of being active. Therefore, identifying these different targets is critical. Then, the classic one we've had for a long period of time has been KRAS and NRAS status; we've broadened our understanding of molecular subtypes. Now, we recommend testing for NRAS, KRAS, BRAF, MSI, and HER2—5 different molecular alterations that need to be investigated in every patient with metastatic CRC.We've had the most success with MSI-high or dMMR tumors, which is a type of DNA repair error that happens in a rare subset of tumors. When you have that error, then you lead to a lot of mutations, and those tumors tend to be immunogenic. That subset of cancers—it can be colon cancer, it can be any cancer; there is an approval for pembrolizumab (Keytruda) across all tumor types that are dMMR—have good activity to immunotherapy. It’s really the [question] of how do we optimize this good activity and how do we bring this earlier to patients?
There have been efforts to conduct large randomized phase III studies in frontline therapy in dMMR CRC; there have been efforts in stage III disease as adjuvant therapy, and there are also neoadjuvant efforts. [The question of whether] immunotherapy [can be used] earlier is an ongoing active research question.
Another question would be, “What other combinations to use?” There are other combinations. These are immune-responsive subtypes, so we do have a feel that other immune checkpoints will be active. There are other combinations [to explore] going forward. In particular, there's a combination that should be going forward soon through the National Cancer Institute (NCI)-MATCH trial, in which we are looking at the combination of a PD-1 inhibitor with a LAG-3 inhibitor, which is a new checkpoint in this space.
In regard to other GI cancers, which tend to be microsatellite stable (MSS), we have a much lower limit of clinical activity. There, it is much more of a fundamental need to understand the individual tumors and the tumor microenvironment that's at play. A lot of these tumors do have T cells, but they're not functional T cells. One big aspect is understanding not just the quantity of T cells, but the quality of the T-cell function. There is a lot more research that needs to be done in that space to understand what's keeping us from having success in MSS CRC, pancreatic cancer, and in PD-L1—negative gastric cancer, for example.In a broad-spectrum sense, immune-oncology is going to be the main focus for drug development and therapies going forward. We have a number of potential targets that are going to be explored. The challenge with immune therapy is that we don't have great model systems to exactly figure out what the best targets are. The model systems are fairly limited for various different reasons, so there are a lot of options. Which one is the best combination or target is the issue that we are going to understand and better appreciate over the next 5-year time span. I'm very hopeful that within that 5-year time span, we'll have identified several other valid clinically successful targets. We’ll be discussing multiagent immunotherapy combinations that are going to start allowing immune activity across a much broader range of tumor types than what we currently see right now with just single-agent PD-1-based therapy.