OBJECTIVES: To examine whether lactate dehydrogenase (LDH) can predict the prognosis of oral cavity squamous cell carcinoma (OSCC) and to determine the optimal cut-off values for LDH. METHODS: This retrospective study included 184 patients with OSCC, treated with surgery between 2006 and 2014. The association between LDH and T, N classification was investigated using the Mann-Whitney test. Cut-off values for LDH were determined with a recursive partitioning analysis (RPA). Survival rates were estimated using the Kaplan-Meier method. A Cox hazard model was used to assess the prognostic capability of LDH. RESULTS: There was no association between LDH and T or N classification (p = .657, .619, respectively). RPA determined the cut-off values for LDH as 160 and 220 IU/L. The five year survival for low-, moderate-, and high-LDH groups were 87.7, 73.7, and 50.9%, respectively (p < .001). The hazard ratios (HRs) for death in moderate- and high-LDH groups were 2.92 (95%CI =1.02-12.30, p = .001) and 7.36 (95%CI =2.54-31.20, p < .001), respectively. The model including LDH-based stratification (Akaike's information criterion (AIC) = 516) was better than the model including clinical stage (AIC =528). CONCLUSION: Pretreatment serum LDH is an independent prognostic factor for overall survival in patients with OSCC.
OBJECTIVES: To examine whether lactate dehydrogenase (LDH) can predict the prognosis of oral cavity squamous cell carcinoma (OSCC) and to determine the optimal cut-off values for LDH. METHODS: This retrospective study included 184 patients with OSCC, treated with surgery between 2006 and 2014. The association between LDH and T, N classification was investigated using the Mann-Whitney test. Cut-off values for LDH were determined with a recursive partitioning analysis (RPA). Survival rates were estimated using the Kaplan-Meier method. A Cox hazard model was used to assess the prognostic capability of LDH. RESULTS: There was no association between LDH and T or N classification (p = .657, .619, respectively). RPA determined the cut-off values for LDH as 160 and 220 IU/L. The five year survival for low-, moderate-, and high-LDH groups were 87.7, 73.7, and 50.9%, respectively (p < .001). The hazard ratios (HRs) for death in moderate- and high-LDH groups were 2.92 (95%CI =1.02-12.30, p = .001) and 7.36 (95%CI =2.54-31.20, p < .001), respectively. The model including LDH-based stratification (Akaike's information criterion (AIC) = 516) was better than the model including clinical stage (AIC =528). CONCLUSION: Pretreatment serum LDH is an independent prognostic factor for overall survival in patients with OSCC.