After decades of unfulfilled promise, immunotherapy now plays a role in the management of many patients with advanced cancers. From leukemia and multiple myeloma (chimeric antigen receptor-T cells) to hepatocellular carcinoma and lung cancer (checkpoint inhibitors) to uveal melanoma (tebentafusp-tebn—a bispecific gp100 peptide HLA-directed CD3 T-cell engager targeting a specific HLA genotype),[1] the armamentarium and indications keep growing, but not so for patients with colorectal cancer (CRC).
Decades ago, a vaccine-based adjuvant treatment showed promise in patients with colon cancer,[2] and the immune-modulator levamisole was approved in combination with fluororuracil in patients with stage III CRC.[3] However, the adjuvant vaccine did not pan out, and levamisole was pulled from the market because it caused a demyelinating syndrome.
In the current era, the field was set back by a definitively negative study in patients with unselected metastatic CRC[4] and, even in a biomarker-enriched group such as microsatellite instable-high (MSI-H) patients, one of every three progresses immediately through checkpoint inhibitor therapy.[5] We now know that patients with mostly extrahepatic metastases are likelier to benefit from immunotherapy,[6] but designing trials for that population excludes the majority of patients with metastsatic CRC. In other words, we have hit a roadblock in the development of immunotherapy in patients with metastatic CRC. Therefore, it seemed appropriate to break with drug development orthodoxy and look away from advanced disease.
In the companion to this article, Shahin et al[7] takes us to the nonmetastatic disease setting. In the precheckpoint inhibitor era (just a few years ago), MSI-H status in patients with stage II colon cancer argued against fluoropyrimidine chemotherapy,[8] and more recently, we have seen that one fourth of patients with MSI-H localized rectal cancer fail to respond to conventional neoadjuvant chemoradiation.[9] These observations justified studies of neoadjuvant immunotherapy for localized colon or rectal primary cancers.
The neoadjuvant setting is a tricky platform for drug development because one must be certain not to compromise the curability of the cancer. However, that risk can be managed, as has been demonstrated by the I-SPY (Investigation of Serial Studies to Predict Therapeutic Response Through Imaging and Molecular Analysis) series initiated in breast cancer more than a decade ago. With an adaptive trial design, new agents were added to standard therapies, and pathologic response was the intermediate end point.
A study of neoadjuvant checkpoint inhibitor therapy was initiated in Europe in patients with nonmetastatic MSI-H colon cancer in 2017. The NICHE trial was a safety study, wherein two cycles of checkpoint inhibitor therapy were administered before curative surgery.[10] All 21 patients underwent uneventful resection at 6 weeks. Surprisingly, despite receiving no more than two doses of ipilimumab and nivolumab and having <6 weeks to show its effects, all tumors had substantial regression.
Cercek et al[11] targeted patients with MSI-H nonmetastatic rectal cancer with single-agent dostarlimab every 3 weeks for up to six months, withholding standard multimodality therapy until evidence of disease progression. Being less common than MSI-H colon cancer (just 2%-3% of rectal cancers are MSI-H), patients accrued slowly but a report on the first 12 patients on-study for at least six months showed that all were in complete clinical remission. In follow-up, with more patients accrued and more time, no patient has recurred (A. Cercek, personal communication, January, 2023).
Although it could be argued that these results should be filed in the too good to be true category, they also meet the too good to ignore threshold. Shahin et al[7] propose a new paradigm wherein 10%-15% of patients with CRC that are MSI-H may need nothing other than checkpoint inhibitor therapy. In their review, the authors spend little time on toxicities and complications for a good reason—there were almost none seen. It is hard to argue against a change in paradigm for patients with rectal cancer right now.
Next up? Although we hope it does not happen but suspect it will, patients whose MSI-H rectal cancers progress on checkpoint therapy would provide an opportunity to identify mechanisms of resistance that could inform the CRC—immunotherapy interaction. Meanwhile, NICHE included 20 patients with microsatellite stable colon cancer, and two showed substantial tumor regression on pathology. Studies in these patients could identify cofactors that give insight into the mechanisms of response. Such information could inform future studies in advanced CRC.
J Oncol Pract. 2023;19(5):263-264. © 2023 American Society of Clinical Oncology
