icon star paper   Articles  
Back grey_arrow_rt.gif
Changes in Metabolic Profile Among Antiretroviral-Naïve Patients Initiating Protease Inhibitor Versus Non-Protease Inhibitor Containing HAART Regimens
  AIDS Journal of Acquired Immune Deficiency Syndromes 2003; 33(5):653-655
*Fehmida Visnegarwala; *Jorge Darcourt; *Prasuna Sajja; ‡Vandana Menon; †Oliver Ong; †Mario Maldonado; *A. Clinton White, Jr. *Department of Medicine, Section of Infectious Disease and †Section of Endocrinology , Baylor College of Medicine, Houston, Texas and ‡Division of Clinical Care Research Tufts-New England Medical Center Boston, Massachussetts
To the Editor:
Lipodystrophic syndrome, insulin resistance, and frank diabetes are emerging as important metabolic consequences of protease inhibitor (PI) therapy in patients with HIV. These metabolic abnormalities may be due to direct effects of PI on rates of whole body lipolysis, differentiation and apoptosis of adipocytes, and inhibition of glucose transporter activity. There are few prospective studies examining metabolic profiles in antiretroviral (ARV)-naive patients, in a clinical setting.
The study protocol was approved by the Baylor College of Medicine Institutional Review Board. Eligible patients were ARV naive, able to give informed consent, and initiated on a highly active antiretroviral therapy (HAART) regimen. Exclusion criteria included history of type 1 or 2 diabetes, endocrine disorders, acute illnesses, and use of drugs known to interfere with insulin secretion or action. Initially, 131 patients screened in for the study. Seventy-nine of these patients were subsequently initiated on HAART and therefore enrolled into the study. Of these, 41 were initiated on a PI-based regimen and 38 were on PI-sparing regimens. In the PI group, 4% of patients were on saquinavir/ritonavir, 16% were on indinavir/ritonavir, and 80% were on nelfinavir. In the PI-sparing group, 22% were on abacavir, 34% were on nevirapine, and 44% were on efavirenz.
Demographic and medical history data were collected by medical record review. Levels of low-density lipoprotein (LDL) cholesterol were calculated using the Friedewald formula. Serum insulin levels were measured by radioimmunoassay and insulin resistance was assessed by the homeostasis model of assessment. Anthropometry and laboratory measurements were repeated 6 months after initiation of HAART therapy. Baseline characteristics were compared between patients started on PI versus PI-sparing therapy using the Student t test, [chi]2, or Fisher exact test as appropriate. A paired t test was used to test for differences within each group in values of variables from baseline to 6 months. Between-group differences in the change score for each variable in the PI versus PI-sparing groups were tested using Student t test.
Thirty-seven percent of the initial 131 patients had family history of premature coronary heart disease (CHD), 63% were current smokers, 8% had diabetes or existing CHD, almost half had high-density lipoprotein (HDL) cholesterol <40 mg/dL, a quarter had hypertension, 11% had Framingham risk score >10%, and 16% met the Adult Treatment Panel III definition for metabolic syndrome. Forty-six percent of the initial cohort had >2 risk factors and 27% had 1 risk factor for CHD.
At baseline, there was no difference between groups in age (median, 38 years), sex (35% female), race (67% African-American), viral load (median, 4.8 log10 HIV RNA copies/mL), and CD4 counts (median, 140 cells/mm3). The 2 groups were similar in terms of metabolic variables such as levels of fasting serum glucose, insulin, total cholesterol, triglycerides, and insulin resistance. Patients on PI had significantly lower baseline body mass index (BMI) (22.7 ± 4.0) as compared with those on PI-sparing regimens (25.2 ± 5.5) and were less likely to be placed on d4T (54%) as compared with those started on PI-sparing regimens (82%). In both groups, after 6 months of therapy, there was a median of 5.1 log reduction in HIV RNA levels, and CD4 count increased by 134 cells/mm3. Change in metabolic parameters from baseline to 6 months of HAART are reported. In the PI group (n = 19), BMI, waist:hip ratio (WHR), fasting insulin levels, and insulin resistance increased significantly after 6 months of therapy. In the PI-sparing group, there was a significant increase in BMI but not in WHR, fasting insulin, or insulin resistance. Levels of fasting glucose were similar at baseline and 6 months in both groups. Fasting triglyceride levels increased from baseline values in the PI group (127 to 194 mg/dL) but did not change in the PI-sparing group (125 to 125 mg/dL). HDL-C (good cholesterol) went from 37 to 32 mg/dL in PI group and 37 to 49 mg/dL in PI-sparing group. Fasting LDL-C (bad cholesterol) went from 37 to 49 mg/dL in PI group and from 37 to 32 in PI-sparing group as reported in Tables.
In this cohort of ARV-naive patients, there is a high prevalence of traditional cardiovascular risk factors including modifiable risk factors such as low HDL cholesterol, hypertension, and smoking. As clinicians are confronted with an aging HIV-infected population, there needs to be an increased awareness of the risk of other chronic diseases in these patients, especially CHD. Consistent with data from the DEXA substudy of the ACTG 384, our patients in the PI and PI-sparing groups had a significant increase in their BMI in the first few months after HAART initiation. This may reflect an improvement in nutritional status secondary to control of HIV infection. In contrast to our findings, Mulligan et al.9 did not find significant weight changes in their prospective study of patients who were started on PI or in their control groups of patients on lamivudine, not on ARV or on stable regimens without PI. PI use, in this study, was associated with an increase in WHR. This observed increase in WHR with PI therapy may reflect overall increase in BMI and increased visceral adiposity.
There was an increase in fasting insulin, with no change in fasting glucose, among the PI-treated patients. The hyperinsulinemia observed in our study may be a marker for the development of insulin resistance among PI-treated patients. In our study, while there was an equivalent increase in BMI in both groups, this was associated with hyperinsulinemia and insulin resistance only in patients on PI regimens. This is in accordance with previous studies that have shown a direct effect of PI on insulin resistance. Treatment with PI-containing regimens was associated with a significant increase in triglyceride levels from baseline, consistent with data from other studies. The latter was despite significantly lower baseline BMI and lesser use of d4T in the PI group. There also was a trend in reduction in HDL cholesterol.
We were able to obtain complete follow-up on only 50% of the original cohort. Thus the findings of this study mainly reflect the clinical course in patients who are compliant with HAART, which yields important information on the adverse effects of these medications in general clinical practice. In conclusion, PIs appear to exert early deleterious effects on metabolic parameters related to cardiovascular risk. These findings have clinical significance for selection of HAART regimen, monitoring, and long-term implications of therapy.
  icon paper stack View Older Articles   Back to Top   www.natap.org