Complete Responses in 100% of dMMR Rectal Cancer With Neoadjuvant PD-1 Inhibition

CHICAGO -- Single-agent dostarlimab (Jemperli) led to complete responses in 100% of a small group of patients with locally advanced mismatch repair-deficient (dMMR) rectal cancer, allowing them to avoid surgery, chemotherapy, and radiation, at least for the time being.

All 14 patients followed for at least 6 months had clinical complete responses with no evidence of tumor on follow-up MRI. Follow-up in the cohort range from 6 to 25 months, and none of the patients have received additional therapy. Four other patients with limited follow-up have preliminary evidence of response, including one clinical complete response.

The 6-month course of the PD-1 inhibitor was well tolerated, and no patient developed grade ≥3 adverse events, reported Andrea Cercek, MD, of Memorial Sloan Kettering Cancer Center (MSKCC) in New York City, at the American Society of Clinical Oncology (ASCO) annual meeting. The study was published simultaneously in the New England Journal of Medicine.

"We observed 100% complete response in the first 14 consecutive patients," said Cercek. "We noted no grade 3 or 4 adverse events. No patients have required chemotherapy, radiation, or surgery. There has been no disease recurrence during the follow-up period. Longer follow-up is certainly required to establish the durability of this treatment."

"This data provides the framework for immunoablative therapies and highlights the clinical impact of biomarker-drive therapy in early-stage disease," she said. "The tumor-agnostic mismatch repair-deficiency population of early-stage disease has the potential to eliminate the need for chemotherapy, radiation, and surgery in 3%-4% of all cancers. This has the potential to be translated rapidly into areas around the world without access to modern chemotherapy, radiation, and surgery."

Not Yet Practice Changing

The results are clinically meaningful and scientifically plausible but cannot be considered practice changing at this point, said ASCO invited discussant Kimmie Ng, MD, MPH, of Dana-Farber Cancer Institute in Boston.

"My answer is 'not yet' for several reasons," said Ng. "The sample size is still relatively small. The median follow-up is still extremely short with a median of 6.8 months. We know from the OPRA trial that 88% of tumor regrowth can happen up to 2 years after completion of TNT [total neoadjuvant therapy]. All patients were enrolled at a single institution [MSKCC], which arguably has the most extensive expertise in nonoperative management of rectal cancer."

"Finally, the only endpoint available right now is overall response, with no data on survival or other clinically relevant outcomes," she noted.

Ideally, unresolved issues about the treatment would be addressed in a randomized clinical trial (RCT), comparing neoadjuvant dostarlimab with standard TNT. However, such a trial is unlikely, said Ng. The condition is rare, reflected in the 29 months required to accrue 18 patients. Off-protocol use of neoadjuvant immunotherapy is likely to occur following publication of the dostarlimab results. As news of the results circulate, patients' willingness to be randomized in a clinical trial is questionable.

In the absence of data from a RCT, more patients and longer follow-up on other clinically relevant endpoints, such as 3-year disease-free survival, overall survival (OS), and organ preservation, are needed.

"Importantly, we need multi-institutional participation," said Ng. "We need to confirm the high clinical complete response rates and that the complex nonoperative management of rectal cancer can be replicated in all cancer care settings."

Study Background, Results

Standard of care for locally advanced rectal cancer consists of chemotherapy, radiation, and surgery. Recent studies have provided support for neoadjuvant chemotherapy followed by chemoradiation and surgery, said Cercek. The strategy produces pathologic complete response in as many as a fourth of patients but is associated with potentially severe complications and toxicity.

Given the life-altering impact of surgical resection of the rectum -- which often requires a permanent colostomy -- interest in organ sparing, nonoperative treatment has increased, Cercek continued. Clinical complete response to neoadjuvant treatment as a surrogate for pathologic complete response offers patients a nonoperative option that leads to survival benefits similar to those observed with surgery.

From 5%-10% of rectal adenocarcinomas are dMMR and typically respond poorly to standard chemotherapy. Single-agent checkpoint inhibition has produced response rates of 33%-55% in metastatic dMMR colorectal cancer, and responses have often been durable, leading to prolonged OS. On the basis of those results, investigators hypothesized that single-agent anti-PD1 treatment might be beneficial in dMMR locally advanced rectal cancer.

Dostarlimab initially received FDA approval for recurrent or advanced dMMR endometrial cancer. More recently, the FDA granted a tissue-agnostic approval to the PD-1 inhibitor for recurrent/advanced dMMR solid tumors that have progressed on prior therapy and have no suitable alternatives.

Cercek reported updated findings from an ongoing phase II trial to evaluate the frequency and durability of responses to neoadjuvant dostarlimab. A preliminary report from the study showed that the first 11 patients treated had complete responses. Investigators plan to enroll a total of 30 patients with stage II/III dMMR rectal cancer.

The primary objectives are overall response rate with single-agent dostarlimab with or without chemoradiation and pathologic complete response or clinical complete response at 12 months after PD-1 inhibition with or without chemoradiation. Rectal MRI and endoscopy determine stable disease, and response (partial, near complete, and complete). Clinical complete response is defined by endoscopy, digital rectal exam, and rectal MRI.

The first 18 patients had a median age of 54, and women accounted for 12 of the patients. Fourteen patients have T3/4 tumors, and all but one patient had one or more involved lymph nodes. Ten of 17 evaluable patients had germline mutations, and all 18 patients had BRAF V600E wild-type tumors.

In the first 14 patients, clinical complete response occurred at the 6-month follow-up in 12 patients, and two met criteria for complete response by 3 months. The four remaining patients had not reached 6 months of follow-up, but one patient had already met criteria for clinical complete response.