Significance of postoperative adjuvant immunochemotherapy after curative resection of colorectal cancers: identification of responders incorporating the age factor.

To identify responders when protein-bound polysaccharide (PSK) is used in adjuvant immunochemotherapy after curative resection of colorectal cancers, we examined the host and tumor factors that affect the prognosis incorporating the age factor. A total of 101 patients who had undergone macroscopic curative resection of colorectal cancer were treated with mitomycin C + fluoropyrimidine oral antineoplastics + PSK (MFP therapy) for two years in principle. These cases were divided into two age groups of <65 years [n=55; 54.8 +/- 8.3 years (mean +/- SD)] and > or =65 years (n=46; 69.1 +/- 3.3 years). Host factors including humoral factors (complement C3 and C4), immunosuppressive acidic protein (IAP), lymphocyte transformation (cellular factors) induced by various mitogens [phytohemagglutinin (PHA), pokeweed mitogen (PWM), and PSK], and tumor markers (CEA, CA19-9) were measured. The cases were divided by the cut-off value of each parameter into > or = cut-off value and < cut-off value groups, and the 5-year survival rates were compared. The cut-off values obtained for these parameters and the tumor factor (Dukes class) were subjected to multivariate analysis to identify the markers that affect prognosis. The 5-year mortality rate was 74.5% in the <65 age group and 56.8% in the > or =65 age group, with a tendency of better prognosis in the <65 age group (p=0.1109). Compared to the <65 age group, the > or =65 age group showed higher levels of C3 (2-way ANOVA: p=0.0582), C4 (p=0.0009) and IAP (p=0.0110) over time, but lower PSK-induced stimulation index (SI) as an indicator of cellular immunity) (p=0.0001) and PHA-induced SI (p=0.2650) over time. These results indicated that compared to patients aged <65 years, patients aged > or =65 years were characterized by lowered cellular immunity in addition to augmented complement production and an aggravated immunosuppressive state, suggesting the presence of some differences in host immune function with aging. Using the Cox proportional hazard model, the prognostic determinant was found to be Dukes C in the <65 age group, and CEA level in the > or =65 age group. The present results suggested that analysis of prognostic determinants of this therapy should take into account the age factor. Especially in elderly subjects, responders to PSK may be identified using the preoperative CEA value.