Validation of bidimensional measurement in nasopharyngeal carcinoma

BACKGROUND: Our previous study showed a close relationship between computed tomography (CT)-derived bidimensional measurement of primary tumor and retropharyngeal nodes (BDMprn) and gross tumor volume of primary tumor and retropharyngeal nodes (GTVprn) in nasopharyngeal carcinoma (NPC) and better prognosis for NPC patients with smaller BDMprn. In this study, we report the results on of a study to validate the use of BDM in a separate cohort of NPC patients. METHODS: We retrospectively reviewed 103 newly diagnosed NPC cases who were treated with radiotherapy/concurrent chemoradiotherapy (CCRT) or CCRT with adjuvant chemotherapy from 2002 to 2009. We used magnetic resonance imaging (MRI) to measure BDMprn. We calculated overall survival, recurrence-free and distant metastasis-free survival curves and set a BDMprn cut off point to categorize patients into a high- or low-risk group. We then used Cox proportional hazard model to evaluate the prognostic influence of BDMprn after correcting age, gender and chemotherapy status. RESULTS: After adjusting for age, gender, and chemotherapy status, BDMprn remained an independent prognostic factor for distant metastasis [Hazard ratio (HR) = 1.046; P = 0.042] and overall survival (HR = 1.012; P = 0.012). Patients with BDMprn < 15 cm(2 )had a greater 3-year overall survival rate than those with BDMprn ≧ 15 cm(2 )(92.3% vs. 73.7%; P = 0.009). They also had a greater 3-year distant metastasis-free survival (94% vs.75%; P = 0.034). CONCLUSION: The predictive ability of BDMprn was validated in a separate NPC cohort. A BDMprn of 15 cm(2 )can be used to separate NPC patients into high- and low-risk groups and predict survival rates and metastasis potential. It can, therefore, be used as a reference to design clinical trials, predict prognosis, and make treatment decisions.