Hyperglycemia During Chemotherapy for Hematologic and Solid Tumors Is Correlated With Increased Toxicity

Abstract

Few data are available concerning how glycemic control can affect outcomes in cancer patients treated with chemotherapy. Charts of non-Hodgkin lymphoma (NHL) and prostate cancer (PC) patients treated at Moffitt Cancer Center between January 1999 and September 2006 were reviewed, and patients who received cyclophosphamide, doxoruicin, vincristine, prednisone, rituximab or Docetaxel plus steroids were eligible. Demographics, vitals, comorbidity, laboratory parameters including baseline and average glucose level during chemotherapy, G4 hematological toxicity (HemT), and G3-G4 non-hematological toxicity (NHemT), progression, and death dates were recorded. A total of 349 patients were eligible (NHL/PC: 162/187). G4 HemT was experienced by 76 (47%) NHL and 9 (5%) PC patients. Seventy-nine NHL and 90 PC patients had G3-G4 NHemT. The most frequent NHemT were as follows: neuropathy (25.3%), fever (non-neutropenic, 18.9%), fatigue (15.2%), for NHL; and fatigue (22.1%), thromboembolic events (11.6%), and diarrhea (9.3%) for PC. For NHL patients, G3-G4 NHemT was associated with baseline hyperglycemia (P = 0.0384, Wilcoxon Rank-Sum test) as well as with average glycemia during chemotherapy (P = 0.0048), whereas there was no significant correlation for HemT. For PC patients, a positive correlation was found between baseline hyperglycemia and G4 HemT (P = 0.0241), while univariate correlations between average glycemia during chemotherapy and G4 HemT and between both baseline and average glycemia with NHemT were not significant, multivariate correlation between average glycemia during chemotherapy and overall severe toxicity was significant at 0.05 level. In NHL patients, hyperglycemia correlates with NHemT, and a similar although less clear pattern is suggested in PC patients. Prospective studies are needed to assess whether a better glycemic control during chemotherapy can improve toxicity and outcomes.