Effect of Tight Control of Blood Glucose during Hyper-CVAD Chemotherapy for Acute Lymphocytic Leukemia.

Weiser et al discovered that patients (pts) with acute lymphocytic leukemia (ALL) who developed hyperglycemia (HG) during induction chemotherapy with hyper-CVAD (fractionated cyclophosphamide, vincristine, doxorubicin, dexamethasone alternating with methotrexate and high-dose cytarabine) had shorter median overall survival (OS) as well as shorter complete remission duration (CRD). We hypothesized that normalization of HG would reverse this adverse prognosis. Therefore, a prospective study was performed to determine if intensive management of HG with insulin during hyper-CVAD consolidation therapy would improve outcome.

Between May 2004 and July 2008, 52 pts with newly diagnosed ALL, Burkitt’s lymphoma, or lymphoblastic lymphoma treated with hyper-CVAD and who developed HG (glucose level of > 180 mg/dL on at least two random measurements) were enrolled. Pts were randomized to the standard care arm or the intervention care arm of HG control. Pts randomized to the standard arm were treated according to conventional care; whereas those in the intervention arm received insulin therapy with insulin glargine and aspart which were titrated to keep fasting blood glucose (Glc) levels <120 mg/dL and postprandial Glc levels <180 mg/dL without hypoglycemia.

An interim analysis for both superiority and futility was planned after one half of the total number of expected deaths were observed. In the standard and intervention arms, the median age was 46.5 yrs (range 17–70) and 57.5 yrs (range 18–85) , respectively (p=0.053). Median follow-up among all patients was 20 mos (range 2 – 47). Based on a repeated measures model with terms for treatment arm, cycle, and the interaction between the two, mean serum Glc levels were significantly lower in the intervention arm as compared to the standard arm (p = 0.009). Eighteen pts died, 8 in the standard arm and 10 in the intensive arm. OS was not statistically different between the two study arms (p = 0.386). OS at one yr was 80.8% (CI 67–97.4%) and 63.5% (CI 47–85.8%) for the standard and intervention arms, respectively. Five of the deaths in the intervention arm were in pts aged 70 years or older. When adjusted for age, the intervention arm was associated with a 13% reduction in the hazard of death compared to the standard arm (hazard ratio = 0.87, 95% CI 0.32, 2.39, p = 0.79). Achievement of CR, number of pts with an ICU admission, number of hospitalizations, and number of infections were not statistically different between the two arms. CRD was not statistically different between the two arms (p = 0.503).

In conclusion, intensive insulin management of glucocorticoid-induced HG as practiced did not improve OS or CRD. Since glycemic control improved without impact on survival, either the improvement was not sufficient enough to be clinically significant or insulin had a deleterious effect on the malignancy that negates the benefits of glycemic control. Future research will address the latter possibility as we have preliminary in vitro data demonstrating that elevated insulin concentrations stimulate cell growth and resistance to doxorubicin-induced apoptosis in leukemia cell lines and primary ALL samples.