Pembrolizumab Plus Bevacizumab/CAPOX Yields Responses in pMMR/MSS mCRC With High TILs


Key Takeaways

  • Pembrolizumab, bevacizumab, and CAPOX achieved a 96% disease control rate in pMMR/MSS mCRC with high TILs.
  • Complete and partial response rates were 21% and 54%, respectively, with a median follow-up of 19 months.
David Tougeron, MD, PhD

David Tougeron, MD, PhD

The combination of pembrolizumab (Keytruda), bevacizumab (Avastin), and CAPOX (capecitabine and oxaliplatin) led to responses and disease control in patients with previously untreated mismatch repair–proficient (pMMR)/microsatellite stable (MSS) metastatic colorectal cancer (mCRC) who had high tumor-infiltrating lymphocytes (TILs), according to data from the phase 2 POCHI trial (NCT04262687).1

Findings presented at the 2024 ESMO Gastrointestinal Cancers Congress showed that at a median follow-up of 19 months, patients treated with the combination (n = 28) experienced a complete response (CR) rate of 21% and a partial response rate of 54%. The stable disease and progressive disease rates were 21% and 4%, respectively. The disease control rate was 96%. The median duration of response was 10 months, and at data cutoff, 46% of patients remained on treatment.

Additional data showed the 12-month progression-free survival (PFS) rate was 68%, and the 24-month overall survival (OS) rate was 67%.

“These results justify the evaluation of this combination [pembrolizumab plus bevacizumab and CAPOX] in patients with [pMMR/MSS mCRC] with high immune [infiltration] in a phase 3 [study],” lead study author David Tougeron, MD, PhD, said during the presentation. Tougeron is a digestive oncologist at Poitiers University Hospital and a professor, Faculty of Medicine and Pharmacology, at the University of Poitiers in France.

Immune checkpoint inhibitors have generally been considered ineffective treatment options for patients with pMMR/MSS mCRC. However, high TIL levels have been associated with good prognosis and potential benefit with the use of immune checkpoint inhibitors, and approximately 15% of patients with mCRC have high TILs.

POCHI was a single-arm, open-label, multicenter, proof-of-concept trial that enrolled patients at least 18 years of age with histologically proven pMMR/MSS mCRC who were naive to systemic therapy in the metastatic setting. Patients were required to have an available primary tumor containing a tumor-free margin, and they also needed to have at least 1 positive immune score.1,2

Notably, neoadjuvant or adjuvant chemotherapy or radiotherapy for the treatment of the primary tumor or resected metastatic disease was allowed if recurrence occurred more than 6 months after the end of treatment. Other key inclusion criteria consisted of a World Health Organization performance status of 0 or 1; a life expectancy of at least 3 months; and adequate hematological, liver, and renal function.2

Patients were excluded if they had disease that was mismatch repair deficient or microsatellite instable. Other exclusion criteria included active infection requiring antibiotics at the start of treatment; active or untreated central nervous system metastases; previous allogeneic stem cell transplant or solid organ transplant; and a history of idiopathic pulmonary fibrosis, medicinal product–related pneumonia; or proof of active pneumonia or pneumonitis.

All patients received pembrolizumab at 200 mg, bevacizumab at 7.5 mg/kg, and standard CAPOX once every 3 weeks.1 CAPOX consisted of oxaliplatin at 130 mg/m2 once every 3 weeks plus pembrolizumab 2000 mg/m2 of capecitabine per day on days 1 to 14 of each cycle.2 Treatment continued until disease progression, unacceptable toxicity, patient refusal, consent withdrawal, pregnancy, or investigator decision.

The trial’s primary end point was 10-month PFS rate per RECIST 1.1 criteria.1 Secondary end points included OS and histological response in the case of secondary resection.2

Investigators screened 182 patients across 40 cancer centers between April 2021, and May 2024, and 15% (n = 28) had at least 1 positive immune score, allowing them to enroll in the trial.1 Among patients enrolled, the median age was 66 years, and 61% were men. Eighty-six percent of patients had an ECOG performance status of 0, and the remaining 14% had a status of 1. Primary tumor location was either right sided (43%), left sided (46%), or in the rectum (11%). Sixty-five percent of patients harbored RAS mutations, and 46% had liver metastases at baseline.

Regarding safety, 64% of patients experienced at least 1 grade 3 or 4 treatment-related adverse effect; however, no toxic deaths were reported.

“Biomarker analyses are already ongoing to identify predictors of CR [for pembrolizumab plus bevacizumab and CAPOX in this patient population,” Tougeron concluded.

Disclosures: Dr Tougeron reported consultancy, advisory fees, and honoraria from Servier, Pierre Fabre, Merck Serono, MSD, Bristol Myers Squibb, AstraZeneca, Roche, Sanofi, and Takeda; receiving research funding from Sandoz, AstraZeneca, Servier, and MSD; and receiving travel grants from Pierre Fabre, MSD, Servier, and Roche.

POCHI was funded in part by MSD and Veracyte.


  1. Tougeron D, Emile JF, Bodere A, et al. Pembrolizumab in combination with xelox and bevacizumab in patients with microsatellite stable (pMMR/MSS) metastatic colorectal cancer (mCRC) and a high immune infiltrate: A proof of concept study - preliminary results of FFCD 1703 POCHI trial. Ann Oncol. 2024;35(suppl 1):S212. doi:10.1016/j.annonc.2024.06.005
  2. Chemotherapy and immunotherapy as treatment for MSS metastatic colorectal cancer with high immune infiltrate (POCHI). Updated August 18, 2023. Accessed July 1, 2024.
Related Videos
Karl Semaan, MD, MSc
Bradley McGregor, MD, discusses findings from a phase 1b study of abemaciclib  in clear cell renal cell carcinoma.
Toni K. Choueiri, MD
Neil J. Shah, MBBS
Chun Chao, PhD, MS
Mary Philip, MD, PhD
Rom S. Leidner, MD
Sarwish Rafiq, PhD
Efficacy and Safety of Fruquintinib in Patients With Metastatic Colorectal Cancer According to Prior Treatment Sequence in the Refractory Setting: Results From FRESCO and FRESCO-2
Richard Kim, MD