Pre-treatment Haemoglobin and Peripheral Blood Lymphocyte Count as Independent Predictors of Outcome in Carcinoma of Cervix

Aims: To evaluate pre-treatment haemoglobin and peripheral blood lymphocyte (PBL) counts as predictors of treatment outcome in cervix carcinoma treated with radical chemoradiation. Materials and methods: Pre-treatment PBL counts and haemoglobin concentrations were retrieved from full blood count examinations from 111 patients who received concurrent chemoradiotherapy. Overall survival and relapse-free survival were obtained using the Kaplan-Meier method by ranking the data by median haemoglobin and PBL, singly and then in association. Their independence and significance as predictors of outcome were analysed using the Cox proportional hazard model. Results: Survival rates were significantly higher in patients whose haemoglobin level or PBL counts were at or above the corresponding median value. At 5 years, rates of overall survival were 77% versus 41% (P = 0.0003) and 75% versus 42% (P = 0.002), when dichotomised around median haemoglobin and PBL, respectively. In multivariate and univariate analyses, both PBL and haemoglobin were independent and significant predictors for risk of death and relapse. Their predictive power was dramatically enhanced when the data were stratified into four groups by associating patients with haemoglobin >= median or < median with those whose PBL was >= or < median. Conclusion: Baseline PBL and haemoglobin seem to be strong, independent predictors of treatment outcome in carcinoma of the cervix, particularly if patient response is ranked using the predictors simultaneously. The hypothesis needs to be tested and, if confirmed, the markers should be used in combination to identify those at greater risk of failure who may benefit from additional therapy, with further validation in prospective trials offering treatment modification. (C) 2013 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.