ASCO highlights: neoadjuvant immunotherapy in mismatch repair deficient colorectal cancer

Potential advantages of neoadjuvant systemic therapy for localized colorectal cancer are to increase response, reduce the risk of distant metastases, prolong overall survival, and in rectal cancer to improve quality of life through organ-preserving strategies (sphincter-preserving surgery, “watch & wait” in the case of clinically complete response). Patients with T4 and node-positive tumors have a 3-year disease-free survival (DFS) of only 60% and those with T4N0 disease have a 5-year DFS of only 74% despite adjuvant fluoropyrimidine and oxaliplatin chemotherapy and may benefit most from neoadjuvant approaches [1, 2]. Patients with dMMR have a better overall prognosis, but are less sensitive to adjuvant chemotherapy. In the metastatic setting, immunotherapy is already part of standard therapy for this subgroup. Neoadjuvant immunotherapy is of great interest for patients with dMMR locally advanced colorectal cancer given the remarkable pathologic and clinical responses reported in several randomized trials [3‐7] and the high organ preservation rate in locally advanced rectal cancer [4].

At ASCO 2022, impressive 100% complete clinical response rate with 6‑months of immunotherapy with dostarlimab in locally advanced dMMR rectal cancer in the first 14 patients were presented (Fig. 1; [7]). None of these patients required chemoradiotherapy or surgery and no recurrence occurred during the observation period. At ASCO meeting 2024, Andrea Cercek presented an update of the ongoing study [8]. Currently, 48 patients are included. Lynch syndrome was diagnosed in approximately 50% of patients. The median TMB was 53.6 Mut/MB. In all, 42 patients have already completed immunotherapy and a clinical complete response rate was still reported in 100% of patients and none of the patient required radiochemotherapy or surgery. The median follow-up was 17.9 months and no recurrences were detected. The therapy was well tolerated and no grade 3–4 immunotherapy-associated adverse events were documented. Among the patients, there were 6 “late responder” who showed complete clinical response after 8 months of immunotherapy. Immunotherapy with dostarlimab for locally advanced dMMR rectal cancer is an unprecedented success story and will change the clinical routine of this subgroup of patients. Despite the ongoing study many clinics already offer dostarlimab to patients with locally advanced dMMR rectal cancer.

Recently, the NICHE‑1 [9] and NICHE‑2 [10] studies demonstrated high pathological response rates of 100 and 98% with complete pathological response in about two thirds of patients with combined immunotherapy with one dose of ipilimumab (1 mg/kg) on day 1 and two doses of nivolumab (3 mg/kg) on days 1 and 15 followed by surgery 6 weeks after study inclusion in patients with localized dMMR colon cancer. The combined immunotherapy was well tolerated and none of the patients had a recurrence during follow-up. In the NICHE-1 study also patients with proficient MMR (pMMR) were included with pathological response in one third of patients (9/30) and complete pathological response in 3 patients. In the NICHE-3 study [11], two doses of nivolumab plus relatlimab (anti-LAG3) resulted in 100% overall pathologic response rate among 19 patients with 79% complete pathological response and 89% major pathological response. The therapy was well tolerated with grade 3 immunotherapy-related adverse events in only 1 patient. At ASCO 2024, three further studies investigating neoadjuvant immunotherapy in dMMR colon cancer were presented.

The NEOPRISM-CRC study [12] investigated an interesting approach, namely whether all dMMR colon cancers should be treated with immunotherapy or only those that are particularly immunogenic with a high TMB (Fig. 2). In all, 32 patients with high-risk stage II or stage III dMMR colon cancer were included. All patients received one dose of immunotherapy with pembrolizumab. Then patients were randomized according to the amount of TMB. Patients with TMB high (< 20 Mut/MB) or TMB medium (5–19 Mut/MB) received two further cycles of pembrolizumab followed by surgery and patients with TMB low (< 5 Mut/MB) underwent immediate surgery. The administration of adjuvant chemotherapy with FOLFOX or CAPOX was physician choice. Primary endpoint was complete pathological response rate. Of the patients, 82% had stage III and 62% had a T4 tumor. All patients (n = 33) except 1 patient had a high TMB; therefore, the hypotheses of the study could not be answered. However, 59% of patients had a complete pathological response. A grade 3–4 immunotherapy-associated adverse event was documented in 6.2% of patients. One patient died due to immunotherapy-associated pneumonitis. The study will be now expanded to 70 patients. Secondary endpoints have not yet been reported.

A further phase II study investigated in 33 patients with stage II–III dMMR colon cancer 3 cycles of neoadjuvant tislelizumab, an PD‑1 antibody, followed by surgery (Fig. 3; [13]). Primary endpoint was major pathological response. The median age of the participating patients was 52 years, 30% had rectal cancer and 70% had colon cancer; furthermore, 75% of the tumors were classified as T4 tumor and 82% were node positive. About 90% of patients had a major pathological response (≤ 10% residual viable vital tumor) and 62% of patients had a complete pathological remission. Surgery was not performed in 4 patients, 3 of whom were managed nonoperatively and 1 patient had disease progression. The immunotherapy was well tolerated. Grade 1–2 immunotherapy-associated adverse events occurred in 30% of patients.

In the third presented phase Ib study [14], the question of whether immunomonotherapy is sufficient or a whether combined immunotherapy in the neoadjuvant setting is more effective in dMMR colon cancer was investigated (Fig. 4). In all, 101 patients with stage IIb or stage III dMMR colon cancer were included. In the experimental arm patients received combined immunotherapy with one cycle of IBI310, a CTLA‑4 antibody, and 2 cycles of sintilimab, an PD‑1 antibody. In the control group patients were treated with 2 cycles of sintilimab monotherapy. Primary endpoint of the study was pathologic complete remission rate. The results showed that complete pathological remission was achieved in 80% of patients with combined immunotherapy and in 47.7% of patients complete pathological remission was achieved with immunomonotherapy. One patient died due to immunotherapy-related myocarditis. In China, a phase 3 study is currently investigating this combined immunotherapy versus adjuvant chemotherapy in dMMR colon cancer.

Neoadjuvant immunotherapy in dMMR colorectal cancer resulted in high response rate. In an ongoing clinical trial with neoadjuvant immunotherapy with dostarlimab in dMMR locally advanced rectal cancer, all patient had a clinical complete response and radiochemotherapy and surgery could be avoided in all patients. None of the patients had a recurrence. Nevertheless, a longer follow-up is required to assess the local recurrence rate, development of distant metastases, and the impact of overall survival. The ongoing AZUR-1 trial, a multicenter, open-label phase II registrational clinical trial investigating the efficacy and safety of dostarlimab as monotherapy for treatment-naïve patients with dMMR locally advanced rectal cancer and aims to confirm the findings of the study led by Andrea Cercek [7, 8]. In my opinion dostarlimab is a groundbreaking therapy and should be administered to all patients with dMMR locally advanced rectal cancer in the clinical setting. Dostarlimab, which is well tolerated, can spare chemotherapy, radiotherapy and surgery and another treatment is difficult to justify.

Whether neoadjuvant immunotherapy in dMMR colon cancer is as clinically meaningful as in rectal cancer is not yet clear. It is important to recognize that surgical resection alone is curative in the majority of stage II dMMR colon cancers and adjuvant therapy applies predominantly to stage III disease. Moreover, staging of dMMR localized colon cancer is specially challenging because microsatellite instability is often associated with falsely positive enlarged lymph nodes owing to immune infiltration. Thus, a substantial proportion of patients with dMMR colon cancer would not need any treatment beyond surgical resection. In contrast to rectal resection, surgery of colon cancer is much less complicated without long-term consequences—raising the question whether avoidance of surgery is a goal in all localized colon cancer patients. For some patients with a higher morbidity risk with surgery due to comorbidities, a nonoperative management in patients with a clinical complete response after neoadjuvant immunotherapy would be a reasonable goal. Another goal of neoadjuvant immunotherapy could be the possibility to avoid adjuvant chemotherapy in dMMR colon cancer, which is less effective as in pMMR colon cancer. But so far, we have no direct comparison between neoadjuvant immunotherapy and adjuvant chemotherapy. However, immunotherapy is generally well tolerated, but can lead to long-term side effects such as endocrinological dysfunction. Furthermore, we have no data whether neoadjuvant immunotherapy prolongs DFS or overall survival (OS). Despite the high pathological response rate with a few cycles of neoadjuvant immunotherapy, we need long-term outcomes to evaluate the impact of DFS and OS [15]. In my opinion the high proportion of patients with a pathological response observed after immunotherapy, together with the safety profile, may provide sufficient justification to provide immunotherapy to radiographically high-risk disease (e.g., T4-tumors) in the clinical setting.