The pipeline of new agents in gastrointestinal cancers is robust, with the emergence of several antibody-drug conjugates, KRAS G12C inhibitors, and novel TKI/immune checkpoint inhibitor combinations.
Wells Messersmith, MD, FASCO, FACP
The pipeline of new agents in gastrointestinal (GI) cancers is robust, with the emergence of several antibody-drug conjugates (ADCs), KRAS G12C inhibitors, and novel TKI/immune checkpoint inhibitor combinations, according to Wells Messersmith, MD, FASCO, FACP.
Several new antibody-drug conjugates (ADCs) have shown promising activity in small subgroups of patients, although disappointing results have been reported thus far with regard to their use in patients with RAS-mutant disease, outside of the KRAS G12C–mutated subgroup, according to Messersmith. However, at least 3 studies are underway to evaluate approaches in patients with other RAS mutations, Messersmith added.
The heterogeneity of tumors in GI cancers continues to pose a challenge, according to Messersmith. As such, understanding how circulating tumor DNA will play into the management of patients as RAS inhibitors emerge in the clinic will be an area of interest going forward, Messersmith added.
“There are some interesting signals with checkpoint inhibitors and TKIs but we just do not know who is going to respond,” Messersmith, a professor of medicine and medical oncology at the University of Colorado Cancer Center, said during a presentation at the 6th Annual School of Gastrointestinal Oncology. “Most of these [efforts] have at least a biomarker hypothesis and how we find these rare subsets is going to be a challenge for the field.”
ADCs were first proposed as far back as the 1980s, but it took until the early 2000s for the first one to be approved for use, according to Messersmith. In 2013, ado-trastuzumab emtansine (T-DM1; Kadcyla) was approved for use in patients with HER2-positive breast cancer. Since then, many of the ADCs to receive regulatory approval have emerged in the breast cancer paradigm, up until recently, Messersmith noted.
The ADCs approved for use in the treatment of patients with solid tumors include trastuzumab emtansine and fam-trastuzumab deruxtecan-nxki (Enhertu) for HER2-positive breast cancer and HER2-positive gastric cancer, enfortumab vedotin-ejfv (Padcev) for locally advanced and metastatic urothelial carcinoma, and sacituzumab govitecan-hziy (Trodelvy) for triple-negative breast cancer.2
“Several steps have to happen [with] ADCs; they need to bind to an antigen, it has to get internalized, and then through endosomal and lysosomal biology, which is as of yet not all that understood in terms of the predictors of efficacy, you then release the payload,” Messersmith said. “That’s where the magic bullet part of this comes in, where it is really only supposed to release the payload at sites of tumor. Then, the payload does its thing in terms of killing tumor cells. For each of these steps, you can see where’d you get mechanisms of resistance.”3
ADCs are made up of an antigen, which is the target of the antibody, along with a cytotoxic payload. The most important advance in the past decade, however, has been in terms of linkers, according to Messersmith. Having a linker that is strong enough to hold the payload in circulation is critical, but it also must release the cytotoxic payload to the tumor site, Messersmith explained. “It took many, many years to get the right linkers in place to actually accomplish these goals,” Messersmith said.
In the phase 2 NCI-MATCH trial (NCT02465060), which examined T-DM1 in patients with HER2-amplified tumors, the rate of HER2 amplification was found to be low in patients with GI cancers; however, those numbers added up, Messersmith noted. Moreover, in those patients with GI cancers, T-DM1 elicited some responses, although the duration of benefit proved to be brief.4
In January 2021,trastuzumab deruxtecan was FDA approved for the treatment of adult patients with locally advanced or metastatic HER2-positive gastric or gastroesophageal junction (GEJ) adenocarcinoma who have received a previous trastuzumab (Herceptin)-based regimen. This ADC consists of an anti-HER2 antibody with a cleavable tetrapeptide-based linker, and a cytotoxic topoisomerase I inhibitor.5
“That has actually been a more common payload [to emerge] over the past several years,” Messersmith said.
The regulatory decision was based on data from the phase 2 DESTINY-Gastric01 trial (NCT03329690), which included patients with HER2-expressing advanced gastric or GEJ adenocarcinoma who had received at least 2 prior treatments, including fluoropyrimidine and a platinum agent. Eighty-six percent of patients had received prior treatment with taxanes, and 33% had prior treatment with checkpoint inhibitors.
A total of 187 participants were randomized 2:1 to receive either trastuzumab deruxtecan at a dose of 6.4 mg/kg in a 3-week cycle (n = 125), or physician’s choice of either irinotecan or paclitaxel (n = 62). The primary end point of the study was overall response rate (ORR), while secondary endpoints included overall survival (OS), duration of response (DOR), progression-free survival (PFS), confirmed ORR, and safety.
The results showed an ORR of 51.3% (95% CI, 41.9%-60.5%; P <.0001) for trastuzumab deruxtecan. “In the third-line setting, to have response rate that high is actually pretty impressive,” Messersmith noted. “You can see that 10 patients had a complete response, while 41 patients had a partial response, and the disease control rate was 85.7%. [The agent] was quite active, so you could see why the FDA would approve it, despite the fairly limited patient numbers.”
Additionally, the median OS with the ADC was 12.5 months (95% CI, 9.6-14.3) vs 8.4 months (95% CI, 6.9-10.7) with physician’s choice (HR, 0.59; 95% CI, 0.39-0.88; P = .0097). The median PFS in the investigative and control arms was 5.6 months (95% CI, 4.3-6.9) vs 3.5 months (95% CI, 2.0-4.3), respectively (HR, 0.47; 95% CI, 0.31-0.71).
“[These survival rates] go down gradually; they are not leveling like they do with immunotherapy,” noted Messersmith. “However, [these are still] fairly impressive results when you compare them with [what has been seen with] trifluridine and tipiracil [TAS-102; Lonsurf], regorafenib [Stivarga], or some of these other drugs that we use in the third-line setting for other GI cancers.”
In the phase 2 DESTINY-CRC01 trial (NCT03384940), trastuzumab deruxtecan was examined in patients with unresectable and/or metastatic, HER2-expressing colorectal cancer (CRC). The primary end point for this trial was ORR by independent central review in cohort A. Notably, previous HER2-targeted treatment was permitted.
Results showed an ORR of 45.3% (95% CI, 31.6%-59.6%), and tumor shrinkage over time was well sustained, according to Messersmith. The median PFS was 6.9 months (95% CI, 4.1–not evaluable), while the median OS was not reached.6 “These data are immature, but still are pretty impressive for the third-line setting and beyond,” noted Messersmith.
A majority of the treatment-emergent adverse effects (TEAEs) reported in this study were grade 1 or 2, and many of them were GI-related toxicities. The most common TEAEs included nausea, anemia, and decreased neutrophil count.
“It’s not surprising given that you have an irinotecan-type of drug,” Messersmith said. “It shows that some of the payload is clearly getting into the circulation…The idea is that there is a much higher amount of drug being delivered to the tumor site.”
Additionally, 5 incidents of interstitial lung disease were observed during this study, 2 of which were grade 5. The median time to investigator-reported onset of these incidents was 80 days (range, 22-132), and all 5 patients were treated with corticosteroids. Two of the patients with interstitial lung disease recovered, 1 did not recover and later died from disease progression, and 2 died as a result of the TEAE.
In the 2 patients who died, onset of the interstitial lung disease ranged from 40 days to 126 days, both received steroids as treatment, and death occurred 6 to 18 days after diagnosis. As a result, new protocol recommendations encourage monitoring for symptoms, and holding treatment of the ADC and starting steroids immediately if interstitial lung disease is suspected.
“[This agent showed] promising and durable activity. For a third-line response rate to be around 45% is pretty impressive,” Messersmith said. “[Also, we saw] a median PFS of around 7 months and we’re used to seeing a median PFS of less than 2 months in this population. This could be an option down the line as further data come out.”
Other HER2-targeted ADCs in development include A166, XMT-1522, MEDI-4276, ARX788, RC48-ADC, BAT8001, and PF-06804103.
“Nowadays, [these ADCs] are able to dial in the exact number of warheads per antibody, and as you do that, you seem to get a much better efficacy to toxicity ratio,” Messersmith said.
However, overexpression does not always equate to benefit with ADCs in GI cancers. For example, there have been many challenges with Trop2 targeting. Trop2 is a glycoprotein that was initially identified as a surface marker of trophoblast cells and has subsequently been found to be overexpressed in many solid tumors, including most GI cancers.7
Sacituzumab govitecan was examined in patients with GI cancers and only yielded 1 response; as such, the agent is not moving forward, noted Messersmith. Another agent, PF-06664178, also failed to elicit responses in this patient population. As such, further development of the agent for this purpose was stopped.
KRAS signaling can lead to multiple cellular-proliferation pathways, and this complexity and redundancy has made pathway targeting very complicated, according to Messersmith. As such, many unsuccessful strategies have been reported over the past 15 years.8
“It is a very redundant and complex pathway to try to inhibit,” Messersmith said. “There was this idea that maybe we could just inhibit downstream effector signals, such as a MEK inhibitor…but now, we have more modern agents.”
One of those agents is adagrasib (MRTX849), a covalent inhibitor of KRAS G12C that irreversibly and selectively binds KRAS G12C in its inactive, GDP-bound state. The agent was optimized for desired properties of KRAS G12C inhibition, has high selectivity for the mutant KRAS G12C proteins vs wild-type KRAS, and has favorable pharmacokinetic properties, including oral bioavailability, a long half-life, and extensive tissue distribution.9
Adagrasib was examined in patients with CRC as part of the phase 1 KRYSTAL-1 trial (NCT03785249). Previously reported data from the phase 1 portion of the research showed that adagrasib had clinical activity in patients with pretreated KRAS G12C–mutant CRC. For the trial, patients were treated on a standard dose escalation, where they received up to 600 mg/kg of the agent twice daily.
The median follow-up for patients with CRC was 4.3 months. The confirmed ORR with the agent in patients with CRC was 17% (n = 3/18), while the disease control rate was 94% (n = 17/18). The agent was fairly well tolerated, with AEs that included nausea, diarrhea and vomiting. These data, Messersmith noted, are still very early; however, all patients enrolled to the study were currently still on treatment, as of the August 2020 data cutoff.10
As a follow-up to this study, the phase 3 KRYSTAL-10 trial (NCT04793958), is a randomized, open-label trial that will compare second-line adagrasib in combination with cetuximab (Erbitux) with chemotherapy in patients with metastatic, KRAS G12C–mutant CRC.
The primary outcome measures for this study are PFS and OS, while the secondary outcome measures are safety, ORR per RECIST v1.1 criteria, DOR, and patient-reported outcomes. To be eligible for enrollment, patients must have confirmed metastatic CRC, confirmed KRAS G12C mutation, and have progressed on a first-line fluoropyrimidine-, oxaliplatin-, or irinotecan-based regimen.
Combination regimens comprised of TKIs and immune checkpoint inhibitors has represented an interesting area of research in the GI cancer treatment paradigm, according to Messersmith.
Initially, promising results were reported from a phase 1 trial (NCT01988896) examining the combination of atezolizumab (Tecentriq) and cobimetinib (Cotellic) in patients with solid tumors, including those with metastatic CRC.11 However, data from the phase 3 IMblaze 370 trial (NCT02788279), which looked at atezolizumab with or without cobimetinib vs regorafenib in patients with previously treated metastatic CRC, showed that the response rate achieved with the combination had dropped.12
The phase 2 LEAP-005 trial (NCT03797326) examined the combination of lenvatinib (Lenvima) plus pembrolizumab (Keytruda) in 30 patients with previously treated advanced solid tumors. Primary end points for this study were ORR and safety. The GI cohorts in the trial included patients with third-line gastric cancer, third-line CRC, and second-line biliary tract cancer.13
Patients in the CRC cohort experienced an ORR of 21.9% (95% CI, 9.3-40.0) with the combination, while those in the gastric cancer cohort had an ORR of 9.7% (95% CI, 2.0-25.8); those in the biliary tract cancer cohort also had an ORR of 9.7% (95% CI, 2.0-25.8).
“I have seen these patients in the clinic, and we have had multiple of these studies and investigator-initiated studies opened, and you definitely have patients who get these incredible responses,” Messersmith said. “Unfortunately, however, it is not the majority, so trying to figure out which of these patients is going to benefit [from this approach] is a key challenge.”
The median PFS was 2.5 months (range, 1.8-4.2) in the gastric cancer cohort, 2.3 months (range, 2.0-5.2) in the CRC cohort, and 6.1 months (range, 2.1-6.4) in the biliary tract cancer cohort. “Two-thirds of the [patients with third-line CRC] are progressing at the 3-month interval, but if you make it out to 33 or 4 months, your chance of making it out to 9 months is actually pretty good,” Messersmith noted.
One of the challenges in the field is to determine how to preclinically evaluate novel combinations, but different models are under exploration, Messersmith concluded.