Ratios Derived from an Array of Standard Hematologic Indices Predict Oncologic Outcomes in Colon Cancer

Background: The interaction between inflammation and cancer is well established. Surrogate markers of systemic inflammation such as the neutrophil lymphocyte ratio, may be associated with long-term oncological outcomes. This study aimed to characterize the relationship between several ratios derived from haematologic indices using a classification and regression tree analysis. Methods: Hematologic white cell ratios were established for all patients undergoing colonic cancer resection (N = 436) with curative intent in a regional cancer center. The optimal ratios associated with overall survival were established in a training set (n = 386) using a classification and regression tree technique (CRT). The association between ratios and overall survival was assessed in a separate test set (n = 50). Within the test set, two groups were generated based on each ratio (above and below the cut-off point identified in the training set). The association between ratios and overall survival was assessed using a stepwise cox proportional hazards regression model. Results: The following ratios (identified by CRT) were associated with adverse overall survival in the test set; (a) neutrophil tolymphocyte ratio (NLR) greater than 3.4 (HR 3.4, p\0.001) and (b) white cell count lymphocyte ratio (WLR) greater than 5.28 (HR 4.1, p = 0.03). Conclusions: This is the first study to apply recursive partitioning in determining the relationship between hematological ratios and overall survival in colon cancer. Haematologic ratios were predictive of oncologic outcome.