Neoadjuvant immune checkpoint blockade has shown strong activity in mismatch repair-deficient or microsatellite instability-high locally advanced colorectal cancer. The PICC-2 trial evaluated whether adding the COX-2 inhibitor celecoxib to neoadjuvant toripalimab could further improve pathological complete response in this setting.
The article, titled “Neoadjuvant toripalimab plus celecoxib versus toripalimab monotherapy for mismatch repair-deficient or microsatellite instability-high, locally advanced colorectal cancer (PICC-2): an open-label, multicentre, randomised, phase 2 trial,” was published online on July 1, 2026, in The Lancet Oncology.
Authors: Huabin Hu, Xutao Shen, Yunfeng Li, Jiaming Zhou, Ping Liu, Jianwei Zhang, Yujie Hou, Xiaozhong Wang, Jiaye Deng, Zhuoxin Zheng, Jianxia Li, Ping Lan, Xiaojian Wu, Liang Kang, Meijin Huang, Zhen He, Xiaosheng He, Zuli Yang, Liang Huang, Hui Wang, Huaiming Wang, Shuangling Luo, Dianke Chen, Xiaoyu Xie, Yan Zhang, Xiaohui Zhai, Shanshan Li, Weiwei Li, Jiancong Hu, Tao Yang, Chao Wang, Weihao Deng, Yan Huang, Wuteng Cao, Fangqian Li, Lishuo Shi, Li Ling, and Yanhong Deng.
Background
Patients with dMMR or MSI-H colorectal cancer are highly responsive to immune checkpoint blockade, while conventional chemotherapy has limited efficacy in this molecular subtype.
Preclinical data suggest that COX-2–prostaglandin E2 signalling can contribute to an inflammatory and immunosuppressive tumour microenvironment. COX-2 inhibition may therefore enhance antitumour immune activity and improve response to PD-1 blockade.
PICC-2 was designed to test whether adding celecoxib to neoadjuvant toripalimab could increase pathological complete response compared with toripalimab alone.
How Toripalimab Works
Toripalimab is a PD-1 immune checkpoint inhibitor designed to restore antitumour T-cell activity. PD-1 is a checkpoint receptor found on T cells. When PD-1 binds to its ligands, PD-L1 or PD-L2, which can be expressed by tumour cells and cells in the tumour microenvironment, it sends an inhibitory signal that reduces T-cell activity. This allows cancer cells to avoid immune attack.
Toripalimab blocks the interaction between PD-1 and PD-L1/PD-L2. By preventing this inhibitory signal, toripalimab helps T cells remain active and continue recognizing and attacking cancer cells. In dMMR/MSI-H colorectal cancer, where tumours often carry many mutations and are more visible to the immune system, PD-1 blockade can produce strong antitumour responses.
Read more about Toripalimab (Loqtorzi) on OncoDaily.
Study Design
PICC-2 was an investigator-initiated, multicentre, open-label, randomised, controlled phase 2 trial conducted at three academic hospitals in China.
Eligible patients were 18–75 years old and had histologically confirmed dMMR or MSI-H colorectal cancer with clinical stage T3–T4 disease, or any clinical T stage with node-positive disease. Patients were required to have ECOG performance status 0 or 1 and adequate haematological, hepatic, and renal function.
Key exclusion criteria included distant metastases, unresectable or recurrent colorectal cancer, obstruction, perforation, or bleeding requiring urgent intervention, previous systemic therapy for colorectal cancer, long-term NSAID use including COX-2 inhibitors or aspirin, active or previous autoimmune disease, and systemic immunosuppression.
Treatment Approach
Patients were randomly assigned 1:1 to receive:
- Toripalimab plus celecoxib
- Toripalimab monotherapy
Toripalimab was given at 3 mg/kg intravenously on day 1 of each 14-day cycle for 12 cycles in both groups. In the combination group, celecoxib was given at 200 mg orally twice daily on days 1–14 of each 14-day cycle. Surgery with curative intent was planned within 4 weeks after the final neoadjuvant toripalimab dose, regardless of tumour response.
Patient Population
Between May 5, 2022, and January 20, 2025, 110 patients were randomly assigned: 55 to toripalimab plus celecoxib and 55 to toripalimab monotherapy.
The median age was 50 years. Overall, 65 patients were male, all patients were Chinese, 88 had colon cancer, and 22 had rectal cancer. A total of 76 patients had cT4 tumours, including 25 with cT4b disease. Clinically node-positive disease was present in 105 patients, and Lynch syndrome was identified in 37 patients.
At the data cutoff of October 10, 2025, median follow-up was 18.1 months.
Key Efficacy Findings
The primary endpoint was pathological complete response, defined as no residual viable tumour in the primary tumour and all sampled lymph nodes at surgery, assessed by central blinded independent pathological review in the intention-to-treat population.
Pathological complete response was reported in 49 of 55 patients receiving toripalimab plus celecoxib and 38 of 55 patients receiving toripalimab monotherapy. This corresponded to a pCR rate of 89% with the combination versus 69% with toripalimab alone. The between-group difference was 19 percentage points, and the result was statistically significant (p=0.014).
In both groups, 49 of 55 patients completed all 12 planned cycles of neoadjuvant therapy. Surgery was performed in 53 patients in the combination group and 51 patients in the monotherapy group. All patients who underwent surgery achieved R0 resection.
Major pathological response was reported in 93% of patients in the combination group and 85% in the monotherapy group. Pathological response in the primary tumour occurred in 96% and 91% of patients, respectively. Tumour regression grade 0 was observed in 91% versus 71%.
In the per-protocol population, pathological complete response was reported in 92% of patients in the combination group and 75% in the monotherapy group. A post-hoc analysis showed clinical complete response in 60% and 45% of patients, respectively.
Subgroup Findings
The benefit of adding celecoxib appeared consistent across some prespecified subgroups. Among patients with cT4b disease, pathological complete response occurred in 92% of patients in the toripalimab plus celecoxib group compared with 50% in the monotherapy group.
Among patients with sporadic colorectal cancer, pathological complete response occurred in 92% and 67% of patients, respectively. An exploratory next-generation sequencing analysis using pretreatment biopsy samples did not identify a significant association between major driver alterations and pathological complete response.
Safety Profile
Treatment-related adverse events of any grade occurred in 38% of patients receiving toripalimab plus celecoxib and 33% receiving toripalimab monotherapy.
Grade 3 treatment-related adverse events occurred in 3 patients in the combination group and 4 patients in the monotherapy group. In the combination group, grade 3 events included tumour perforation and bowel obstruction. In the monotherapy group, grade 3 events included rash, tumour perforation, increased aminotransferase, and hypothyroidism.
No grade 4 or grade 5 treatment-related adverse events were reported. Treatment-related adverse events led to discontinuation in 7% of patients in the combination group and 5% in the monotherapy group. No adverse event led to cancellation of surgery.
Among patients who underwent surgery, surgery-related adverse events of any grade occurred in 11% of patients in the combination group and 8% in the monotherapy group. One patient in the combination group died from postoperative sepsis, which investigators considered unrelated to study treatment.
Why the COX Pathway Matters in CRC
Interest in COX pathway inhibition in colorectal cancer has also been strengthened by recent aspirin data. Aspirin irreversibly inhibits COX enzymes and reduces prostaglandin-mediated inflammatory signalling, a pathway linked to tumour growth, immune regulation, and recurrence risk.
In the ALASCCA trial, published in The New England Journal of Medicine in 2025, low-dose aspirin reduced recurrence in patients with localized colorectal cancer and PI3K pathway alterations. Among patients with PIK3CA hotspot mutations, the three-year cumulative incidence of recurrence was 7.7% with aspirin versus 14.1% with placebo. Among patients with other PI3K pathway alterations, recurrence was 7.7% versus 16.8%, respectively.
PICC-2 is the first randomised trial to show that adding COX-2 inhibition to neoadjuvant PD-1 blockade can significantly improve pathological complete response in dMMR or MSI-H locally advanced colorectal cancer.
The addition of celecoxib to toripalimab increased pCR from 69% to 89%, with a similar safety profile. The findings support further investigation of toripalimab plus celecoxib in larger phase 3 studies and suggest that COX-2 inhibition may be a rational strategy to enhance neoadjuvant immunotherapy response in this population.
Although ALASCCA and PICC-2 studied different drugs, settings, and molecular populations, both add to the clinical interest in inflammation-related pathways in selected patients with colorectal cancer.
Limitations
The study did not include an upfront-surgery control group, so it cannot determine whether neoadjuvant immunotherapy is superior to immediate surgery in all eligible patients.
Radiological staging may have led to overtreatment in some cases, particularly because inflammatory features in dMMR/MSI-H tumours can complicate assessment of T stage and nodal involvement. Follow-up remains immature for survival endpoints, and pathological complete response has not yet been validated as a surrogate endpoint in this setting.
The trial was conducted only in China, included an entirely Chinese population, and had a relatively young median age, which may limit generalisability. Patient-reported outcomes and quality-of-life data were not included.
Takeaway
Neoadjuvant toripalimab plus celecoxib significantly improved pathological complete response compared with toripalimab monotherapy in patients with dMMR or MSI-H locally advanced colorectal cancer.
The findings support further evaluation of toripalimab plus celecoxib in larger phase 3 trials and suggest that COX-2 inhibition may help enhance neoadjuvant PD-1 blockade in dMMR/MSI-H locally advanced colorectal cancer.