Immunotherapy Now Standard for Some Patients With CRC

Immunotherapy Now Standard for Some Patients With CRC

Dr. Michael J. Overman
With recent approvals of pembrolizumab and nivolumab for the treatment of refractory deficient mismatch repair (dMMR) metastatic colorectal cancer (CRC), immunotherapy has made its way into the field of gastrointestinal oncology. Many questions remain, however, including optimal approaches to dMMR cancers and reasons for a lack of efficacy seen in proficient MMR (pMMR) CRC. Experts discussed these issues and the management of immunotherapy toxicities during the Education Session, “Where We Stand With Immunotherapy in Colorectal Cancer,” held June 2.

“dMMR tumors have much better outcomes than pMMR tumor types,” said Michael J. Overman, MD, of The University of Texas MD Anderson Cancer Center, who chaired the session and spoke about optimal approaches to dMMR CRC. The exception to this is in stage IV CRC, where the situation is reversed and dMMR tumors have poorer outcomes than pMMR tumors. Approximately 15% of all CRCs have dMMR, but this decreases as stage increases.1 Only about 4% of stage IV CRCs demonstrate dMMR.

MMR is a mechanism used by cells to repair damaged DNA, particularly during DNA replication. Deficiency in this system leads to an accumulation of mutations, thus CRCs that are deficient in MMR have substantially increased tumor mutation burden. Because the types of errors recognized and repaired by MMR occur at areas of DNA repeats known as microsatellites, dMMR CRC is also termed microsatellite-instability high (MSI-H).

Because of the known differences in outcomes, as well as in effective therapies, guidelines have now shifted with regard to testing for MMR status. “Any patient with a colon or rectal cancer should be tested for deficient MMR,” Dr. Overman said. This is true regardless of family history or other baseline characteristics.

Testing can be done using immunohistochemistry staining, polymerase chain reaction (PCR), or next-generation sequencing. Dr. Overman said that they all function reasonably well, although there are some downsides to each approach. PCR testing has a sensitivity and specificity of 97% and 95%, respectively, although it is known to be less accurate in noncolon cancers. IHC staining has a sensitivity of 92% and a specificity of 99%, but it cannot detect loss-of-function mutations. Next-generation sequencing is effective but there are multiple approaches using different microsatellite targets, with varying sensitivity and specificity.

“I think it is key that everyone understands the universal testing approach is now listed in guidelines,” Dr. Overman said. “Everyone should be tested.”

Checkpoint Inhibitors

It has long been known that dMMR CRC has a unique immune tumor microenvironment consisting of tumor-
infiltrating lymphocytes and a Crohn-like lymphoid reaction.1 The immune activity is likely responsible for the favorable outcomes observed in resected dMMR CRC and led to the initial trials of immune checkpoint inhibitors in this setting.

In May 2017, the U.S. Food and Drug Administration (FDA) approved pembrolizumab for treatment of patients with dMMR CRC after prior treatment with fluoropyrimidine, oxaliplatin, and irinotecan, and who do not have satisfactory alternate treatment options.

Soon afterward, in July 2017, the FDA also approved nivolumab for the treatment of dMMR CRC after the same prior treatment. This was based on results of the CheckMate-142 study, in which 74 patients with dMMR CRC were treated with single-agent nivolumab; that study showed a response rate of 31%, and 69% demonstrated disease control for 12 weeks or longer.1 At 12 months, the progression-free survival (PFS) rate was 50%, and the overall survival (OS) rate was 73%.

The same study also examined the combination of nivolumab and ipilimumab in 119 patients with dMMR CRC.1 The combination appeared effective, with an overall response rate of 55%, a 12-month PFS rate of 71%, and a 12-month OS rate of 85%.

Dr. Overman noted that the responses appear to be extremely durable. Long-term follow-up of several cohorts has revealed a flattening of both PFS and OS curves, suggesting the effect is a lasting one. This is true of nivolumab and pembrolizumab monotherapy, as well as the combination of an anti–PD-1 agent with an anti–CTLA-4 agent; the 12-month PFS rate in a cohort of patients receiving the combination was 77%, compared with 48% in a group receiving nivolumab monotherapy.

With pembrolizumab monotherapy, Dr. Overman noted that in 18 patients who stopped therapy at 2 years per the study protocol (11 with a response and seven with residual disease), the median time off therapy was 8 months and none have yet recurred. The 12-month PFS rates seen with both nivolumab and pembrolizumab are among the highest seen across the variety of tumor types in which these agents have been tested and approved.

Several phase III trials are now ongoing in this field. In one, atezolizumab is being tested alone and in combination with chemotherapy (mFOLFOX6) and bevacizumab in more than 300 patients with MSI-high metastatic CRC. Enrollment has completed for another study in the metastatic setting. A total of 270 patients will be randomly assigned to either mFOLFOX6 and bevacizumab or pembrolizumab monotherapy.

“Standard-of-care therapy for second-line [and after] for dMMR CRC is nivolumab or pembrolizumab,” Dr. Overman said. “I think we do need more understanding of resistance mechanisms,” he added, since not all patients with dMMR CRC respond to these immunotherapy approaches.

Proficient MMR Tumors

Dr. Michael A. Morse
“We know that there is no benefit for anti–PD-1 antibody therapy in pMMR tumors,” said Michael A. Morse, MD, MHS, FACP, of the Duke University Cancer Institute. “It’s night and day.” He spoke about why pMMR tumors do not respond and provided some potential approaches to change this situation.

“It does seem to be that having more T cells in the tumor may be important,” Dr. Morse said. Having neoantigens or something else for the T cells to respond to may improve the response to immunotherapy. It is possible that WNT signaling could play a role in this, as it has been found to be inversely correlated with T-cell infiltration. Targeting WNT signaling is difficult, though, because it is mainly a protein-protein interaction, Dr. Morse said.

Still, some therapeutic approaches have been proposed. For example, in one ongoing clinical trial, the porcupine inhibitor CGX1321 is being tested along with pembrolizumab in patients with advanced gastrointestinal tumors.

There is also potential with approaches that modulate the cytokine environment since this can also induce an immune response. This could be accomplished by inducing inflammation in the tumor. A phase Ib study of talimogene laherparepvec along with atezolizumab is ongoing in patients with triple-negative breast cancer or colorectal cancer with liver metastases.

Dr. Morse said it could also be possible to focus on natural killer (NK) cells rather than on T cells as a way to improve pMMR responses to immunotherapy. Early results from a first-in-human trial of monalizumab, which suppresses inhibitory signaling by tumors on NK cells, plus durvalumab in metastatic pMMR CRC will be presented on Sunday (Abstract 3540). There were three partial responses out of 37 patients, while none would be expected with only the checkpoint inhibitor in this patient population.2

Looking toward the future, Dr. Morse said cancer vaccine approaches may be necessary. “It is likely that there will be patients who do not have adequate T-cell responses,” he said. “If we could activate them with cancer vaccines, then we would have the substrate for checkpoint molecules to work on at the tumor site.”

He stressed that the dramatic differences between dMMR and pMMR outcomes and responses suggest a new approach may be needed. “At some point we’re going to have to diverge; they are biologically different,” Dr. Morse said. “It may be that we have to forge our own path and study microsatellite-stable as a completely separate entity without considering what goes on in MSI-high.”

Managing Toxicity

Dr. Marc S. Ernstoff
Since checkpoint inhibitors are now considered standard in some patients with dMMR CRC, considerations regarding toxicity have become more relevant. “It is important for us to recognize these toxicities and learn how to deal with them,” said Marc S. Ernstoff, MD, of Roswell Park Comprehensive Cancer Center. He noted that a recent survey of oncologists found that only approximately half of respondents said they were somewhat or very comfortable managing
immune-related adverse events (irAEs).

irAEs can be caused by several mechanisms, including the development of autoimmunity, expression of immune checkpoints, and inflammatory cytokine releases. Often, the toxicity can be delayed and it can affect nearly every organ.

Pruritus, rash, and diarrhea are among the most common irAEs. The timing can vary: skin reactions tend to occur early, and gastrointestinal, endocrine, and hepatic toxicities are usually seen within the first 12 weeks of therapy.

“About 10% to 20% of patients will have lasting, unresolved toxicities requiring ongoing management,” Dr. Ernstoff said.

Some of the main principles of irAE management include ruling out other causes such as infection or comorbid diseases and consulting early on with organ-specific specialists when appropriate. Also, several established treatment algorithms have now been published that provide guidance for management. It is important to be aware of life-threatening AEs such as myocarditis, myositis, pneumonitis, and bowel perforation.

“Immune checkpoint inhibition of the PD-1/PD-L1 axis is, in general, well-tolerated and has an excellent safety profile compared to conventional chemotherapy,” Dr. Ernstoff said. “Familiarize yourself and your community with the toxicity profile.”  

–Dave Levitan