Abstract

Background Recipients of allogeneic HSCT frequently require support with total parenteral nutrition (TPN), in expense of increased risk of infections associated with hyperglycemia, particularly in neutropenic period. Previously, van den Berghe et al. showed that intensive insulin therapy reduced the morbidity including infections and mortality in patients cared in the ICU. Here we assessed the clinical impact of hyperglycemia in patients undergoing myeloablative HSCT.

Methods A retrospective cohort of consecutive 112 adults treated between January 2002 and June 2006 was reviewed, and 21 patients were excluded due to coexisting infectious diseases, preexisting neutropenia or graft failure. The remaining 91 patients with various hematological malignancies were categorized according to mean blood glucose (BG) level, which developed in neutropenic period, to 1) “normoglycemia” (BG≤110 mg/dl, n=28), 2) “mild hyperglycemia” (110<BG≤150 mg/dl, n=49) and 3) “moderate hyperglycemia” (150 mg/dl<BG, n=14). Conditioning regimens included BU/CY (n=45), CY/TBI (n=43) and CA/CY/TBI (n=3). GVHD prophylaxis included cyclosporine- (n=62) and tacrolimus-based regimen (n=29). Stem cell sources included bone marrow (n=46), peripheral blood (n=41) and cord blood cells (n=4). Infection prophylaxis was oral ciprofloxacin, acyclovir and fluconazole. The primary endpoint of this study was the occurrence of febrile neutropenia (FN) and infectious episodes including bacteremia, pneumonia and central venous catheter infection. The secondary endpoint was abnormal laboratory data as parameters for organ dysfunction, which were used in the study by van den Berghe, including elevation of serum creatinine ≥2.0 mg/dl or more than twice of the baseline, serum total bilirubin ≥2.0 mg/dl and serum C-reactive protein (CRP) ≥15 mg/dl. For statistical analysis, student T, chi-square, and Wilcoxon rank-sum tests were used.

Results There was no essential difference between the 3 groups in the average caloric intake, occurrence of FN and infectious episodes. However, hyperglycemia was significantly associated with higher number of febrile days (normoglycemia 3.3±3.7 days; mild hyperglycemia 5.5±4.1days p=0.017; moderate hyperglycemia 7.4±5.0days: p=0.004), hypercreatininemia (normoglycemia 3.6 %; moderate hyperglycemia 28.6% OR 10.8, p=0.018), hyperbilirubinemia (normoglycemia 10.7 %; moderate hyperglycemia 42.9% OR 6.3, p=0.017), CRP ≥15 mg/dl (normoglycemia 14.3 %; moderate hyperglycemia 64.3% OR 10.8, p<0.001), longer hospital stay (normoglycemia 43.0±20.5 days; mild hyperglycemia 60.7±31.3 days: p=0.003, moderate hyperglycemia 77.8±51.3 days; p=0.004) and in-hospital mortality (normoglycemia 0%, mild hyperglycemia 10.2%; p=0.08, moderate hyperglycemia 21.4%; p<0.001).

Conclusion Hyperglycemia during neutropenia was associated with increased risk of infection and organ dysfunction, which further lead to vicious cycle of infectious complications. The results may support the possibility that intensive glucose control reduces the morbidity including infectious complications and organ dysfunction after HSCT.

Disclosure: No relevant conflicts of interest to declare.

Author notes

*

Corresponding author