Comparisons of the metabolic effects of treatment regimens based on protease inhibitors, non-nucleoside analogues, and both in combination, have been drawn from a clinical trial of five years duration. Greater increases in triglycerides and LDL cholesterol were observed with triple-class regimens. Effects on insulin, insulin resistance, and body composition were not found to vary significantly between the three antiretroviral treatment strategies. The study is published in the April 15th issue of the Journal of Acquired Immune Deficiency Syndromes.
The FIRST study (Flexible Initial Retrovirus Suppressive Therapies – CPCRA 058) was a long-term, prospective clinical trial that compared three first-line antiretroviral strategies. Participants (all antiretroviral-naïve) were provided with a nucleoside backbone and were then randomised to receive either a protease inhibitor (PI), non-nucleoside reverse transcriptase inhibitor (NNRTI), or both. Specific treatments within each class were chosen by the participant and their treating physician. Published results from FIRST (MacArthur 2006) showed that the three-class strategy was not more effective than either two-class treatment, and was more toxic overall.
In this report, a distinct substudy (CPCRA 061) of the larger trial compared the metabolic outcomes of the three strategies. A total of 422 participants (from 1,397 in the FIRST trial) were enrolled: 141 on PIs, 141 on NNRTIs, and 140 on both. Characteristics were as follows (and very similar between the treatment arms): mean age 38 years, 22% female, 60% black, 27% white, 10% Latino, mean CD4 count 215 cells/mm3, mean viral load 5 log10 copies/ml. Very few participants were already on lipid-lowering or diabetic treatment.
Patients were followed between August 1999 and September 2005 (median, 62 months). The treatments reflect what was in use during that period: nelfinavir (Viracept) was the most commonly used PI (61%), followed by ritonavir-boosted PIs (24%) and indinavir (Crixivan) (12%). Efavirenz (Sustiva) was the most common (63%) in the NNRTI-only arm; in the NNRTI+PI arm, nevirapine (Viramune) was used as frequently. d4T (Zerit, stavudine) was also in common use, in about 36% of all participants. Treatment switching was also fairly common: 51%, 38% and 78% of people in the PI, NNRTI, and dual arms (respectively) switched medications at least once.
Changes in lipids: HDL cholesterol
HDL cholesterol also rose quickly at first, levelling off between four and eight months on therapy. Overall increases were (in mg/dl): 5.58 (PI), 10.34 (NNRTI), 9.18 (PI+NNRTI). The HDL increases in the PI group were significantly lower than the NNRTI group (p
The researchers note that this “corroborates the previously reported finding of a favourable effect on HDL … with use of NNRTI-containing regimens compared with PI-containing regimens.”
Changes in lipids: triglycerides and LDL cholesterol
In people on PI or NNRTI strategies, triglyceride levels increased slowly over the course of the study (PI: 0.50 mg/dl/month, p = O.01; NNRTI: 0.68 mg/dl/month, p
LDL cholesterol increased dramatically in the first month in all groups: 14.29 mg/dl (PI), 11.55 (NNRTI), and 25.51 (PI+NNRTI). This was followed by a slow subsequent decline, which still resulted in net overall increases over the study period: 8.46 mg/dL (PI), and 7.85 (NNRTI), and 12.84 (PI+NNRTI). Increases in the PI and NNRTI groups were not significantly different from each other, but were both significantly smaller than the three-class group (p
The researchers note that “the PI + NNRTI strategy was associated with greater increases in triglycerides and LDL cholesterol … although no significant differences were noted in these parameters between the PI and NNRTI strategies. For triglycerides, these differences by strategy were more prominent among those participants who initiated the study at higher triglyceride levels.” The report notes that there are considerable differences between individual protease inhibitors as far as their effect on triglycerides.
Changes in body composition
Body composition was evaluated by overall body mass, body mass index and bioelectric impedance analysis (BIA), plus measures of tissue area estimated from skin fold measurements at the waist, mid-arm and mid-thigh.
During the first four months on all strategies, increases were seen in total body weight, body mass index (BMI), total body fat (TBF) and fat-free mass (FFM). All of these increases were sustained throughout the study period, and no significant differences were seen between the treatment strategies.
Subcutaneous areas associated with fatty tissue (mid-arm, mid-thigh and waist) also increased in the first four months for all treatments. After these initial increases, estimated subcutaneous fatty tissue declined back down to below baseline levels after eight months (in the mid-arm), and back to baseline levels after twelve months (in the mid-thigh) and 20 months (in the waist).
Overall, NNRTI- and PI-based therapy did not show significantly different results for LDL cholesterol levels (which spiked early on, followed by return to baseline levels by 40 months), for insulin, insulin resistance and glucose levels (all of which increased over the study period), or for changes in regional body composition. As the report states, the “central adiposity and loss of subcutaneous fat and evidence of insulin resistance that occurred irrespective of strategy suggest that the NRTIs may be responsible for the development of lipoatrophy and insulin resistance, as described in other studies.”
Changes in glucose and insulin
Glucose levels, insulin levels and insulin resistance increased significantly for all three treatment arms; unlike lipid levels, these changes occurred gradually over the course of the study. There were no significant differences between the three strategies.
Glucose increases (in mg/dl) were: 3.27 (PI, p= 0.05), 4.88 (NNRTI, p
Conclusions
Noted study limitations include “the fact that … nelfinavir and indinavir were the primary PIs available; thus, the study does not provide information on the effects of newer PIs,” and that switches from initial therapies may have affected outcomes – however, this reflects the real-world situation where “changes in treatment are likely to occur”.
Overall, the study confirms that “the 3 ART strategies assessed were associated with substantial changes in metabolic parameters and body composition” in the long term; the three-class strategy including PIs, NRTIs and NNRTIs “is associated with some of the least favourable metabolic effects.” As “some effects were ubiquitous across ART strategies,” rather than implicating NNRTIs or PIs as a class, “this supports the possible fundamental effects of NRTIs on changes in body composition and lipid and glucose metabolism.”
Shlay J et al. Long-term body composition and metabolic changes in antiretroviral naïve persons randomized to protease inhibitor-, nonnucleoside reverse transcriptase inhibitor-, or protease inhibitor plus nonnucleoside reverse transcriptase inhibitor-based strategy. J Acquir Immune Defic Syndr 44: 506-517, 2007.