Comprehensive assessment of the feasibility of adoptive cell therapy in colorectal carcinoma

Background

Adoptive cell therapy (ACT) can induce objective clinical responses in solid tumors. However, its potential in colorectal cancer (CRC) remains poorly exploited. To this end, several questions need to be addressed: Can tumor-infiltrating lymphocytes (TIL) be isolated from CRC, including cryopreserved samples? Can the contamination by the gut microbiota be circumvented? Are CRC TIL similar to those expanded in melanoma trials?

Methods

Cryopreserved samples (n = 10) were obtained from primary colon adenocarcinoma tumors (n = 5) or liver metastasis (n = 5). Dissected fragments (n= 10 - 32) were plated and stimulated with IL-2 (6’000 IU/ml) for 21-28 days (pre-REP). A rapid expansion protocol (REP) consisting of a polyclonal stimulation with PHA was performed in cases where the pre-REP yield was low (< 50 x10⁶ TIL). Microbiological testing (BD BACTEC^TM) was performed at Day 0 (D0) and was repeated at D14 of the pre-REP after culturing in antibiotic-containing media. Polychromatic flow cytometry analyses were performed at harvest.

Results

Sufficient TIL were successfully obtained from all patients tested. The yield of TIL at harvest was broad, not linked to fragments numbers and ranged from 10⁶ to 10⁹ cells (198 ± 90.9 x106; mean ± SEM). Only 4/10 patients required a secondary REP to reach >50 x10⁶ TIL. Among the primary tumors, the highest potential for TIL isolation was observed for the right-side colon tumors, as opposed to the left side (478 ± 329 x10⁶ vs 5.3 ± 3.8 x10⁶ cells), while no significant difference was observed between primary and metastatic samples. Bacterial contamination was detected at D0 in all primary tumors (and none in metastatic samples; p = 0.02, χ2), but microbiological tests turned negative at D14. Flow cytometry analyses showed that T cells represented 78 ± 6% of the pre-REP TIL with CD4/CD8 ratios ranging from 0.5 to 180 (median 1.62). CD8+ T cells were 74 ± 9.4% effector memory (CCR7⁻CD45RA⁻) and 5.9 ± 2.3% central memory (CCR7⁺CD45RA⁻), and had an activated phenotype (HLADR⁺CD25⁺PD-1⁺), which is similar to the documented results of melanoma TIL.

Conclusions

Consistent with previous findings in melanoma, we demonstrated the feasibility of TIL expansion in CRC and provide the rationale to move forward with personalized immunotherapy in CRC.

Clinical trial identification

Legal entity responsible for the study

Centre de Therapies expérimentales. CHUV/UNIL

Funding

Centre de Therapies expérimentales. CHUV/UNIL - This Research Project was supported by ESMO with the aid of a grant from Amgen. Any views, opinions, findings, conclusions, or recommendations expressed in this material are those solely of the authors and do not necessarily reflect those of ESMO or Amgen.

Disclosure

All authors have declared no conflicts of interest.