The Dartmouth Institute for Health Policy and Clinical Practice, Lebanon, NH, USA.
BMC Infect Dis. 2010 Jun 23;10:183. doi: 10.1186/1471-2334-10-183.
HIV-associated lipodystrophy syndrome (HALS) is characterized by insulin resistance, abnormal lipid metabolism and redistribution of body fat. To date, there has been no quantitative summary of the effects of insulin sensitizing-agents for the treatment of this challenging problem.
We searched MEDLINE, the Cochrane Library, clinical trial registries, conference proceedings and references for randomized trials evaluating rosiglitazone, pioglitazone or metformin in patients with evidence of HALS (last update December 2009). Two reviewers independently abstracted data and assessed quality using a standard form. We contacted authors for missing data and calculated weighted mean differences (WMD) and 95% confidence intervals (CI) for each outcome.
Sixteen trials involving 920 patients met inclusion criteria. Rosiglitazone modestly improved fasting insulin (WMD -3.67 mU/L; CI -7.03, -0.31) but worsened triglycerides (WMD 32.5 mg/dL; CI 1.93, 63.1), LDL (WMD 11.33 mg/dL; CI 1.85, 20.82) and HDL (WMD -2.91 mg/dL; CI -4.56, -1.26) when compared to placebo or no treatment in seven trials. Conversely, pioglitazone had no impact on fasting insulin, triglycerides or LDL but improved HDL (WMD 7.60 mg/dL; CI 0.20, 15.0) when compared to placebo in two trials. Neither drug favorably impacted measures of fat redistribution. Based on six trials with placebo or no treatment controls, metformin reduced fasting insulin (WMD -8.94 mU/L; CI -13.0, -4.90), triglycerides (WMD -42.87 mg/dL; CI -73.3, -12.5), body mass index (WMD -0.70 kg/m2; CI -1.09, -0.31) and waist-to-hip ratio (WMD -0.02; CI -0.03, 0.00). Three trials directly compared metformin to rosiglitazone. While effects on insulin were comparable, lipid levels and measures of fat redistribution all favored metformin. Severe adverse events were uncommon in all 16 trials.
Based on our meta-analysis, rosiglitazone should not be used in HALS. While pioglitazone may be safer, any benefits appear small. Metformin was the only insulin-sensitizer to demonstrate beneficial effects on all three components of HALS.
HIV 相关脂肪营养不良综合征（HALS）的特征是胰岛素抵抗、脂代谢异常和体脂重新分布。迄今为止，尚无针对这一棘手问题的胰岛素增敏剂治疗效果的定量总结。
我们检索了 MEDLINE、Cochrane 图书馆、临床试验注册处、会议记录和参考文献，以评估罗格列酮、吡格列酮或二甲双胍治疗有 HALS 证据的患者的随机试验（最后更新日期为 2009 年 12 月）。两位审稿人独立提取数据，并使用标准表格评估质量。我们联系了作者以获取缺失数据，并计算了每个结局的加权均数差值（WMD）和 95%置信区间（CI）。
16 项涉及 920 名患者的试验符合纳入标准。罗格列酮可适度改善空腹胰岛素（WMD-3.67 mU/L；CI-7.03，-0.31），但可使甘油三酯（WMD32.5 mg/dL；CI1.93，63.1）、LDL（WMD11.33 mg/dL；CI1.85，20.82）和 HDL（WMD-2.91 mg/dL；CI-4.56，-1.26）恶化与安慰剂或无治疗相比在 7 项试验中。相反，吡格列酮对空腹胰岛素、甘油三酯或 LDL 没有影响，但与安慰剂相比，在 2 项试验中可改善 HDL（WMD7.60 mg/dL；CI0.20，15.0）。两种药物都没有对脂肪重新分布的指标产生有利影响。基于 6 项安慰剂或无治疗对照试验，二甲双胍可降低空腹胰岛素（WMD-8.94 mU/L；CI-13.0，-4.90）、甘油三酯（WMD-42.87 mg/dL；CI-73.3，-12.5）、体重指数（WMD-0.70 kg/m2；CI-1.09，-0.31）和腰臀比（WMD-0.02；CI-0.03，0.00）。3 项试验直接比较了二甲双胍与罗格列酮。虽然胰岛素作用相当，但血脂水平和脂肪分布指标均有利于二甲双胍。在所有 16 项试验中，严重不良事件均不常见。
根据我们的荟萃分析，罗格列酮不应在 HALS 中使用。虽然吡格列酮可能更安全，但任何益处似乎都很小。二甲双胍是唯一一种对 HALS 的所有三个成分均具有有益作用的胰岛素增敏剂。
