Longitudinal analysis of peripheral blood immune status and prognosis dynamic prediction for advanced colon cancer with first-line chemotherapy.

[BACKGROUND] The prognostic evaluation of advanced CRC patients places increased importance on longitudinal peripheral blood immune status. This study aimed to identify prognosis associated longitudinal immune markers and construct dynamic prognostic models for advanced CRC patients with first-line chemotherapy.

[METHODS] Metastatic CRC patients treated with standard first-line palliative chemotherapy were retrospectively collected at Shanghai General Hospital from May 2013 to May 2020. Lymphocyte subsets, inflammatory indices, and tumor markers in peripheral blood were repeatedly assessed before each chemotherapy cycle. Joint models were used to identify significant longitudinal prognostic markers. A dynamic prognostic model was established using random forests for time-dependent predictors, and internally validated using tenfold cross-validation.

[RESULTS] Increased levels of CRP, CEA, CA199, and IL-6, as well as the CD4 + CD29 + cell proportion and the CD4 + CD45RO + /CD4 + ratio were identified as significant risk factors for overall survival (OS) in metastatic CRC patients. Conversely, the increased levels of CD3-CD19 + cell proportion and the CD4 + CD45RA + /CD4 + ratio were identified as favorable factors for OS. A dynamic prognostic model demonstrated good discriminative ability, with AUC values of 0.827, 0.787, 0.726, and 0.693 for 2-, 3-, 4-, and 5-year predictions, respectively. A high ratio of CD4 + CD45RA + /CD4 + before the third to fourth chemotherapy cycle was associated with significantly better OS. Normal CRP and IL-6 levels in the early phase of first-line chemotherapy indicated a good prognosis.

[CONCLUSION] This study highlights the prognostic significance of measuring longitudinal immune status in advanced CRC patients and develops an internally validated dynamic prediction model. External validation is needed before clinical implementation.