Bekaii-Saab Breaks Down Significant Developments in the Realm of Gastrointestinal Cancers

Tanios Bekaii-Saab, MD

The care of patients with gastrointestinal cancers continues to rapidly shift, with novel frontline combinations under exploration in hepatocellular carcinoma (HCC), new insights helping to inform optimal sequencing strategies in colorectal cancer (CRC), and innovative chemotherapy regimens generating discussion in pancreatic cancer, according to Tanios Bekaii-Saab, MD.

“It’s an exciting time. Today, [we have] all these options; targeted options, intense chemotherapy [approaches], and less-intense chemotherapy [regimens],” Bekkai-Saab said. “The landscape is changing quite significantly. Immunotherapy, which at the beginning had a rocky start in gastrointestinal cancers, is now becoming, in some cancers such as HCC, the predominant option…The future looks incredibly bright for our patients with gastrointestinal cancers.”

In an interview with OncLive® during a 2021 Institutional Perspectives in Cancer webinar on gastrointestinal cancers, Bekaii-Saab, a medical oncologist, the medical director of the Cancer Clinical Research Office, and the vice chair and section chief of the Medical Oncology, Department of Internal Medicine, at Mayo Clinic, spotlighted the recent developments made in gastrointestinal cancers.

OncLive®: Atezolizumab (Tecentriq) plus bevacizumab (Avastin) has become the preferred frontline option for patients with HCC. What is your approach for the second-line setting?

Bekaii-Saab: Overall, in the landscape of advanced HCC, treatment options have exploded over the past few years. A lot has changed in the past decade, but more recently, [things have] gotten complicated in a good way. We have atezolizumab [Tecentriq] plus bevacizumab [Avastin] as first-line options for most of our patients. Some patients may [receive] lenvatinib [Lenvima] or sorafenib [Nexavar]. However, most of the complicated decisions have transferred over to the second-line [setting].

What do you do after atezolizumab/bevacizumab? The answer is likely TKIs. We have 2 that are [FDA] approved [for use] in the second line: regorafenib [Stivarga] and cabozantinib [Cabometyx]. Some [may] argue that lenvatinib and/or sorafenib may still have a role, [but] I strongly argue against that [approach]. The best options would be either regorafenib or cabozantinib, and I lean more toward the use of cabozantinib in my clinic.

What are some of the benefits of using cabozantinib over other TKIs? What efforts are being made to examine the agent in the first-line setting?

There are some theoretical advantages to utilizing cabozantinib over other TKIs, at least in the post-bevacizumab setting. [The agent] appears to hit MET and AXL, which both seem to drive resistance to VEGF inhibition; this has been shown in renal cell carcinoma [RCC]. Regorafenib is also a good option for these patients. I typically use [this drug] in the third-line [setting] after cabozantinib; however, it is exchangeable in that setting.

The data that that led to the approval of cabozantinib come from the phase 3 CELESTIAL trial [NCT01908426], which randomized patients to receive cabozantinib or placebo; the trial reported out a few years ago. Patients in the second- and third-line [setting experienced a] meaningful overall survival [OS] and progression-free survival [PFS] improvement. This was done in the setting of patients [in whom] sorafenib [has failed], so the decision-making process when it comes to atezolizumab/bevacizumab gets a little bit more complex. We must draw from what we understand about the drug’s mechanism of action, [as well as] experience [that we have] in other diseases like RCC. HCC, at some level, simulates what goes on in RCC.

Attempts have also been made to move cabozantinib into the first line. The phase 3 COSMIC-312 trial [NCT03755791] is comparing cabozantinib and atezolizumab with sorafenib. A third arm is looking at cabozantinib as a single agent, as we do not have any first-line cabozantinib data available to us. This study is accruing, and if positive, may provide us with yet another option in the first line. [Regardless, it will] give us some valuable information on [the efficacy of] cabozantinib as a single agent.

Beyond cabozantinib, what are other novel combinations are under investigation in the frontline setting?

Other ongoing studies are [examining] lenvatinib and pembrolizumab [Keytruda] and they will be reporting out in the next year or so. The phase 3 LEAP-002 trial [NCT03713593] is examining the VEGF inhibitor lenvatinib plus the immunotherapy pembrolizumab vs lenvatinib alone.

Other [efforts are examining] dual immunotherapies, so PD-1 or PD-L1 [agents] plus CTLA-4 [agents]. We have the phase 3 HIMALAYA trial [NCT03298451], which has completed accrual and will fully report out sometime this year. [The trial is] examining the combination of tremelimumab and durvalumab [Imfinzi] in the frontline setting. Then, we have the phase 3 CheckMate 9DW trial [NCT04039607] which is looking at [first-line] ipilimumab [Yervoy] plus nivolumab [Opdivo] vs sorafenib or lenvatinib; that trial is accruing.

In the next 2 to 3 years, we will see quite a few novel combinations. Depending on how they perform, they may become part of the landscape in the first line and continue to complicate our HCC landscape in a positive way.

If these studies yield positive results, how do you anticipate these approaches to impact the paradigm?

When we look at the phase 1b data coming out from these studies, they all look like [they are] within the same range as [what had been seen with] atezolizumab/bevacizumab. It is going to be very difficult, for example, to shift to a TKI plus a PD-1 or PD-L1 inhibitor unless the data are [significant]. I do not expect that they will be, but we will have to see the results of the study to decide. Overall, there is no advantage to giving an oral agent plus an intravenous (IV) agent. There is an advantage to give an oral vs an IV; however, when it is an oral plus an IV, there are added toxicities and no significant added benefits historically. [As such,] I may be reluctant to switch. We will see what the data [look like]. That is one consideration.

Secondly, nivolumab plus ipilimumab or durvalumab plus tremelimumab, depending on if those studies end up being positive, they offer a non-VEGF platform. That is an added benefit, because many of our patients have risk of bleeding, or significant varices. For those patients, the dual immunotherapy combination may be preferred; it may be a useful alternative to atezolizumab/bevacizumab…We will see what the data show. At this point, atezolizumab plus bevacizumab remains our [best] option for first line, and our second-line options include cabozantinib and regorafenib.

Shifting to CRC, what are some of the challenges faced with regorafenib and what was learned from the phase 2 ReDOS trial (NCT02368886) regarding dose optimization?

In CRC, we are moving more and more toward personalized medicine. HER2 and KRAS G12C have started to emerge as targets. We are seeing a lot of activity that is moving us into a positive direction. Unfortunately, the reality for most of our patients is that the best [approach] continues to primarily be cytotoxic therapy for the first few lines of treatment and then moving on to oral agents. In the refractory setting, we have TAS-102 [Lonsurf] and regorafenib; both options appear to be similar [on paper], but they have never really been compared head-to-head.

We find challenges with regorafenib being started at 160 mg, which is the approved dose. That [agent faced] challenges making it into the landscape of treating CRC because of the high level of toxicity [observed]. The toxicity [was found to] hit early and hit hard. Within 2 to 3 weeks, patients had significant hand–foot syndrome reactions and significant fatigue; therefore, [patients] have issues continuing with the treatment. Many patients drop off early, and do not see the benefit, although benefits [are observed in] the overall patient population. We believe that several patients do not see the full benefit because they dropped early. In the community, meaning both academic and community practice, there has been a lot of [thoughts on] how to dose regorafenib. There was no meaningful way to essentially figure out the right dose for the right patient.

This is [how the] design of the ReDOS trial [was developed]. We took patients and randomized them to 2 arms; 1 arm was the dose-optimization arm, where we started at 80 mg, went to 120 mg, and then 160 mg, as tolerated; the other arm started at 160 mg. The primary end point of the study was a composite one, and it looked at patients who made it to cycle 3 [of treatment], as well as [those who] made it beyond cycle 3. [The trial] had to achieve 2 things: the drug had to be beneficial [enough for patients] to go beyond cycle 3 and the drug needed to be tolerable enough to make it to cycle 3. The primary end point was met, meaning more patients on the dose-optimization arm ended up with a benefit.

Interestingly, we found that the survival of patients on the dose-optimization arm was higher than those who just received the 160 mg dose. [The latter] performed as we would expect, [with an OS of] approximately 6 months vs [10 months on the dose-optimization arm]. This has shifted our practice in the United States and beyond, to utilize the dose-optimization strategy for regorafenib. This [approach was also] included in the National Comprehensive Cancer Network guidelines and has since become widely adopted. This changed the landscape.

[Additionally,] another randomization was integrated into the study. [Investigators examined the] use of preemptive vs reactive clobetasol to reduce the risk of hand–foot syndrome. Essentially, we found that in the first cycle or 2, we cut down on the risk of significant hand–foot syndrome, regardless of what strategy we used. Now we recommend most patients, especially those who we believe to be at risk, [to apply] steroid cream on their hands and feet to decrease the risk of hand–foot syndrome.

How do you choose between regorafenib and TAS-102 in the third-line setting?

That’s [arguably] the most important question. When patients go through 2 lines of therapy, and now require either regorafenib or TAS-102, how do we decide on one vs the other? For most patients who can get through treatment, my first choice is regorafenib. The reason for this is because data suggest [that regorafenib may] be a bit more effective If you use it earlier rather than later. TAS-102 seems to preserve that level of activity.

We have also done a network meta-analysis which suggested that the dose-optimization [strategy with regorafenib] outperforms both TAS-102 and regorafenib given at 160 mg, which is not surprising. This essentially [cemented] our choice of [using] regorafenib first and TAS-102 second, except for patients for whom we believe regorafenib may not be as well indicated, such as those with significant liver dysfunction.

Moving to updates made in the pancreatic cancer paradigm, the phase 3 NAPOLI 3 trial (NCT04083235) has generated excitement. Could you expand on this effort and the hope for this approach?

We recently published this paper with the first results from the phase 1/2 study with NALIRIFOX. The thought behind developing NALIRIFOX is not merely just to simulate FOLFIRINOX. There was certainly some preclinical basis that suggested that the liposomal formulation may interact a little bit better with the other chemotherapy components, and so we conducted this study in pancreatic cancer.

Essentially, we have shown that [this approach] was safe enough. The preliminary results [showed that the] PFS was 9.2 months, which is one of the highest reported in pancreatic cancer—even in early-phase studies—and the OS was close to 13 months. These were encouraging results, and the safety profile was established [enough to bring this] to a phase 3 study. [NAPOLI-3] is looking at NALIRIFOX vs gemcitabine and nab-paclitaxel [Abraxane], which is the standard regimen right now. Looking at the preliminary data, we have high hopes that this study will show superiority for the triplet [regimen] vs the doublet across patient subgroups.

This is the first study to compare a triplet regimen with the doublet regimen. We have never had data on [this] compared head-to-head. I do not want to simplify this to the fact that NALIRIFOX is just another FOLFIRINOX; it is not. The basis of its development is a little bit different. The study itself, if positive, will hopefully provide us with another option for our patients with pancreatic cancer in the first line. The study is accruing very well. I am hoping that we will see some results by next year.