AIMS: To improve prediction efficiency by incorporating complete blood count (CBC) into the TNM system on 5 year disease-specific survival (DSS) for patients with nasopharyngeal carcinoma (NPC). MATERIALS AND METHODS: The CBC of 3237 patients undergoing radiotherapy was retrospectively evaluated. In total, 2820 patients treated with non-intensity-modulated radiotherapy (IMRT) were randomly divided into development (1895 patients) and validation cohorts (925 patients). The association of potential risk factors with 5 year DSS was tested by Cox proportional hazards analysis and a prognostic index was created by assigning weighted scores proportional to a regression coefficient to each factor. Each cohort was divided into low, intermediate and high prognostic index. The prognostic index was validated in the validation cohort and compared with the TNM system on prediction of 5 year DSS. Validation was repeated in another independent group of 417 patients treated with IMRT. RESULTS: Eight independent prognostic factors were identified: gender, age, T or N stages, anaemia or thrombocytosis during radiotherapy, continuous reduction in haemoglobin, high neutrophil-lymphocyte ratio before radiotherapy. Each was assigned a number of points. The area under curve (AUC) of the prognostic index was larger than that of Union Internationale Contre le Cancer/American Joint Cancer Committee TNM system 2009 (0.697 versus 0.619, P < 0.001). CONCLUSION: A CBC-based prognostic index was developed and had a higher prediction efficiency on 5 year DSS in NPC than the TNM system alone.
AIMS: To improve prediction efficiency by incorporating complete blood count (CBC) into the TNM system on 5 year disease-specific survival (DSS) for patients with nasopharyngeal carcinoma (NPC). MATERIALS AND METHODS: The CBC of 3237 patients undergoing radiotherapy was retrospectively evaluated. In total, 2820 patients treated with non-intensity-modulated radiotherapy (IMRT) were randomly divided into development (1895 patients) and validation cohorts (925 patients). The association of potential risk factors with 5 year DSS was tested by Cox proportional hazards analysis and a prognostic index was created by assigning weighted scores proportional to a regression coefficient to each factor. Each cohort was divided into low, intermediate and high prognostic index. The prognostic index was validated in the validation cohort and compared with the TNM system on prediction of 5 year DSS. Validation was repeated in another independent group of 417 patients treated with IMRT. RESULTS: Eight independent prognostic factors were identified: gender, age, T or N stages, anaemia or thrombocytosis during radiotherapy, continuous reduction in haemoglobin, high neutrophil-lymphocyte ratio before radiotherapy. Each was assigned a number of points. The area under curve (AUC) of the prognostic index was larger than that of Union Internationale Contre le Cancer/American Joint Cancer Committee TNM system 2009 (0.697 versus 0.619, P < 0.001). CONCLUSION: A CBC-based prognostic index was developed and had a higher prediction efficiency on 5 year DSS in NPC than the TNM system alone.
Authors: Jian Ji Pan; Wai Tong Ng; Jing Feng Zong; Sarah W M Lee; Horace C W Choi; Lucy L K Chan; Shao Jun Lin; Qiao Juan Guo; Henry C K Sze; Yun Bin Chen; You Ping Xiao; Wai Kuen Kan; Brian O'Sullivan; Wei Xu; Quynh Thu Le; Christine M Glastonbury; A Dimitrios Colevas; Randal S Weber; William Lydiatt; Jatin P Shah; Anne W M Lee Journal: Cancer Date: 2016-07-19 Impact factor: 6.860
Authors: R Kumar; E Geuna; V Michalarea; M Guardascione; U Naumann; D Lorente; S B Kaye; J S de Bono Journal: Br J Cancer Date: 2015-03-31 Impact factor: 7.640