Predicting the duration of chemotherapy-induced neutropenia: New scores and validation

Abstract

Background: The objective of this study was to develop predictive models to classify febrile neutropenic patients into two groups, according to a prediction of the duration of the chemotherapy-induced neutropenia episode. Patients and methods: For this retrospective analysis, 106 patients with solid tumours and an episode of febrile neutropenia (FN) were eligible. A score was attributed to each chemotherapy treatment drug according to its expected toxicity. Three new scores were proposed based only on this classification. Two of them are a combination of the individual drug scores and the third one was built using statistical techniques such as cluster analysis and classification trees. Results: Statistical techniques produced the best score, distinguishing two groups of patients with statistically different neutropenia durations, with median durations until haematological recovery of absolute neutrophil count 2 × 109/l of 4 versus 2 days (P < 0.001). Conclusions: Our methodological approach based on statistical techniques identifies the patients who will need the longest times to recover from FN. The input of this predictive system is only the aggressiveness of the cytotoxic agents in a chemotherapy regimen. Our proposal succeeded in distinguishing two groups of patients and the results show better performance than other scores in previous studies. © The Author 2010. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved.