Regression analysis of prognostic factors in colorectal cancer after curative resections.

The clinical, laboratory, and pathologic data of 310 patients who had curative resections were prospectively collected and analyzed in a multiple stepwise regression model. Although several factors (i.e., venous invasion) were of importance in univariate analysis, the following conclusions reflect the outcome and relative importance of the regression analysis only. Blood loss as an initial symptom and duration of symptoms were associated with a better prognosis. Location of the primary tumor, age, and sex did not appear to have prognostic value. Observations during operation such as palpable lymph nodes, fixity to adjacent organs, and tumor spill were related to a diminished tumor-free survival. Laboratory data (hemoglobin, leukocytes, ESR, GGTP, SGOT, SGPT, LDH, total protein, CEA) were tested for their potential prognostic values. Only a preoperative low protein level or an elevated CEA level were associated with an increased risk of death due to recurrent tumor. The histopathologic features (stage and grade), with the exception of venous invasion, were of relative importance in the determination of prognosis. The aforementioned variables can be included in a prognostic index on the base of which high-risk groups suitable for adjuvant studies can be identified.