In this phase 3 trial conducted across multiple centers, we have demonstrated that the combination of PD1-T cells with XELOX plus bevacizumab leads to a significant improvement in both progression-free survival and overall survival when compared to the use of XELOX plus bevacizumab alone in patients with metastatic colorectal cancer. There were no notable variances observed in terms of ORR and DCR, and the addition of PD1-T cells did not raise any new safety concerns.

At the time of planning the trial, immune checkpoint inhibitors had not yet been approved as the first-line treatment for the dMMR mCRC, so patients were not mandatory to detect the MMR status during screening, and patients with dMMR mCRC were also included in the trial. Based on the available MMR status of the enrolled patients, the distribution of patients with dMMR was balance between the two arms. Besides, no significant statistical interaction was observed between the treatment effect and MMR status for PFS. Thus, a potential confounding effect on treatment outcome associated with the difference in MMR status can be excluded.

In this study, the combination of XELOX and bevacizumab was selected as the initial therapy for both dMMR and pMMR mCRC according to the recommendations from global and regional guidelines.3 After induction and maintenance treatment, few patients went to receive surgery. This finding suggests that patients were correctly selected to participate in the trial, as eligible patients should have been considered to be inappropriate for surgery according to the study design.

In recent years, adoptive cell immunotherapy has mainly focused on gene engineered T cell therapy, such as CAR-T cells, and TCR-T cells. However, the application of gene engineered T cell therapy in mCRC is limited in regards to efficacy and toxicity.20,21 In contrast, previous studies have attempted to investigate the impact of combining chemotherapy with CIK/DC-CIK cell immunotherapy in patients with mCRC, and the findings indicated that the addition of CIK/DC-CIK cells to chemotherapy significantly improved clinical outcomes compared with chemotherapy alone.13,22 Our previous phase II study also found that CIK cell therapy combined with 5-fluorouridine, leucovorin and oxaliplatin (FOLFOX4) chemotherapy significantly prolonged OS when compared to the use of chemotherapy alone as an initial treatment for patients with mCRC; while the PFS was not significantly improved.17 The current study demonstrated that PD1-T cells combined with XELOX plus bevacizumab exhibited superior efficacy in terms of PFS and OS compared to XELOX plus bevacizumab alone. The difference in the impact of adoptive cell immunotherapy on PFS of patients with mCRC observed in our current and previous studies might be attributed to variations in trial design, chemotherapeutic regimes and immune cell type. In our previous study, the primary endpoint focused on 3-year OS rate, and patients received CIK cell treatment combined with FOLFOX4 regimen without bevacizumab.17 However, in the current study, the primary endpoint was PFS, and patients received PD-1 blockade-activated DC-CIK cell treatment combined with XELOX plus bevacizumab. Therefore, the previous study may be not sufficiently powered to detect differences in PFS as the sample size was computed with respect to the primary endpoint of 3-year OS rate. Besides, chemotherapy without bevacizumab and the use of immune cells without anti-PD-1 antibody activation in the previous study may also affect the potential differences in PFS between the treatment groups. Nonetheless, these findings together provide compelling evidence in favor of the effectiveness of combining adoptive cell therapy with first‐line chemotherapy for improved outcomes of patients with mCRC.

One of the obstacles of immune cell therapy is the presence of an immunosuppressive tumor microenvironment.23 PD-1/PD-L1 inhibitory signal is the thoroughly studied immunosuppressive pathway, and antibodies targeting PD-1/PD-L1 have gained approval for treating various types of tumors. However, systemic therapy with anti–PD‐1 antibody has been correlated with immune‐related side effects and may trigger an immunosuppressive response through interacting with different types of immunosuppressive cells present within the tumor microenvironment.24,25,26 In the study, to overcome the PD-1/PD-L1 inhibitory pathway, the autologous DC-CIK cells were in vitro directly incubated with a low-dose of pembrolizumab to generate the PD-1 blockade-activated DC-CIK cells (i.e., PD1-T cells). The median dose of pembrolizumab used in the study was 12 mg per infusion, which is much smaller than what was used in clinical practice. The possible excess pembrolizumab in the PD1-T cell product may also affect the host immune cells. However, this systemic effect may be weak, as low dose (0.3–1 mg/kg) of anti-PD-1 antibody has shown little clinical activity.27,28 Therefore, we believe the clinical activity of PD1-T cells were primarily due to in vitro pre‐activated DC‐CIK cells, rather than the systemic effects of pembrolizumab on host immune cells. In addition, our previous studies have shown that PD1-T cells were active in several kinds of solid tumor, including CRC.18,19 After blocking the PD1/PD-L1 signal pathway with a low dose of pembrolizumab, the cytotoxicity of DC‐CIK cells was enhanced.18,29

Previous studies have shown the conceptually mechanisms on the synergistic effects of immune cells combined with chemotherapy. Oxaliplatin used in the study is demonstrated to have immunogenic effects, including induction of immunogenic cancer cell death and enhancement of effector immune response,30 which may lead to tumor antigens released from tumor tissues and increase the sensitivity of malignant cells to immune-mediated cytotoxic activity.31 On the other hand, similar to the antitumor mechanism exhibited by CIK/DC-CIK cells, PD1-T cells can effectively kill tumor cells through granzyme and perforin-mediated cytotoxic lysis after an MHC-independent tumor recognition,18,32 which is different from the anti-tumor mechanism of chemotherapeutic drugs. Meantime, antiangiogenic therapy with bevacizumab could normalize tumor vasculature to let cytotoxic T-cell circulate into tumor cells.33,34 These findings indicate that the addition of immune cells to XELOX plus bevacizumab can improve the efficacy of chemotherapy.

The clinical issue underlying the trial holds greater clinical relevance within the subgroup of 173 patients with pMMR CRC. Although PD-1/PD-L1 inhibitors can be beneficial for patients with dMMR mCRC, immune checkpoint blockade (ICB) monotherapy is ineffective in patients with pMMR.10 Differing from the ICB therapy that restore the existing immunoreaction by targeting the tumor-induced immune deficiency, adoptive cell immunotherapy directly infiltrated into the pMMR mCRC that is considered a “cold tumor”.9,12,20 In the study, patients with pMMR mCRC had significantly survival benefits from the addition of PD1-T cells to chemotherapy. The immunotherapy group exhibited a 38% lower risk of disease progression compared to the control group. Similarly, two recent clinical trials named AtezoTRIBE and MEDITREME also observed the clinical benefit from the addition of ICB therapy to chemotherapy in pMMR/microsatellite stable (MSS) mCRC patients.35,36 However, the clinical benefit was somewhat small, with only 1.5 month increase in median PFS in the AtezoTRIBE study, while the magnitude of clinical benefit was difficult to determine due to its single-arm design in the MEDITREME study.35,36 Therefore, a phase 3 study was needed to provide additional validation for the therapeutic efficacy of combining ICB therapy with chemotherapy in untreated pMMR/MSS mCRC patients. Nevertheless, the two studies, together with ours, suggest that chemotherapy may promote tumor immunogenicity and enhance the anti-tumoral effect of immunotherapy. Another clinically relevant question is whether there is an association between liver metastases and the effectiveness of ICB therapy in patients with mCRC. Unlike previous studies that reported resistance to ICB therapy among mCRC patients with liver metastases,20,37,38,39 the present study demonstrated that PD1-T cell therapy could provide benefits for patients with liver metastases. This difference may be partly explained by the decreased T-cell infiltration in tumors of patients with liver metastases,39,40 while adoptive cell immunotherapy may increase the T-cell infiltration. It would be intriguing to explore the impact of PD1-T cell therapy on pMMR mCRC patients with liver metastases in future clinical study.

This study demonstrates that combining PD1-T cells with XELOX plus bevacizumab has a similar safety profile to previous studies involving CIK cells combined with FOLFOX4 chemotherapy,17 as well as CIK cells combined with GP (gemcitabine and cisplatin) chemotherapy in non-small-cell lung cancer.41 These findings suggested that the addition of T cell therapy did not further increase the risk of TRAEs compared to the chemotherapy alone. The incidences of TRAEs, grade 3 or above TRAEs, serious adverse events (SAEs), and TRAEs leading to treatment discontinuation were comparable between the two groups. The most frequently occurring grade 3 or above TRAEs in both groups were hematologic toxicity, hand-foot syndrome, diarrhea, and allergy, which is mainly attributed to XELOX plus bevacizumab chemotherapy. There was no significant difference in the incidence of grade 3 or higher hematologic and non-hematologic toxicity between the two groups, and the observed incidence is consistent with previous clinical studies.42

Our study has certain limitations. First, at the beginning of this study, the initial treatment choices for patients with dMMR mCRC were chemotherapy plus bevacizumab or cetuximab; thus, the MMR status was not used as one of the inclusion criteria, and about 12% of patients missed the MMR detection. The trial was not powered for analyzing the effect of PD1-T cell therapy on this dMMR population. Second, PD1-T cells were manufactured in a Good Manufacturing Practice-certified facility of each study centers, which may yield potential variability in the PD1-T cell agent, although standard operating procedures under strict quality control and assurance were trained before beginning this trial. Third, the patient randomization allocation procedure was conducted manually, which is more likely to make allocation error than using a central telephone-in or web-based random system. Fourth, the primary endpoint was not centrally reviewed. Finally, potential efficacy-related biomarker tests remain to be analyzed in future studies.

In conclusion, our data indicate that PD1-T cells in combination with XELOX plus bevacizumab is deemed safe and demonstrates significantly improved PFS and OS when compared to XELOX plus bevacizumab alone for mCRC patients. We will continue our efforts to identify predictive markers by analyzing the blood samples collected during this study. PD1-T cells combined with XELOX plus bevacizumab may represent a promising new alternative therapeutic option for patients with mCRC and could be practice changing.