<International Circulation>: The HPS2-THRIVE study published this year is a study on the effects of niacin/laropiprant combination in patients at high risk of occlusive vascular events; however， there was no difference in the primary end point events compared with the control group， and the increase of HDL-C was not clear. In your opinion， how should we select patients for HDL-C elevation therapy? Is it feasible to increase HDL-C for the treatment of atherosclerosis in the clinical practice?

　　Prof. Rosenson: First of all， the HPS2-THRIVE trial had a flawed design in that they selected individuals who had LDL cholesterol levels that were at target on simvastatin with or without ezetimide. So the average LDL cholesterol was 60 mg/dl， each non-HDL cholesterol and HDLcholesterol levels were average at the start of the trial. We know that niacin has many side affects. It is challenging to prescribe because of the time and effort to inform， educate the patient about potential side effects in order to improve the care – it’s difficult for patients to take. It appears that laropiprant did not diminish many of these niacin-associated side affects as might have been expected.

　　At the end of the trial， the results were non-significant， and it’s not unsurprising that there was a lack of benefit when one considers the baseline lipid values. They excluded individuals who had LDL cholesterol levels greater than or equal to a certain level on simvastatin and ezetimibe， and those are the individuals that we would consider using niacin for. So the trial doesn’t address an important clinical question， and this was discussed in an editorial that I wrote with Doctor Antonio Gotto and published in Current Atherosclerosis Reports earlier this year.

　　It is important to discuss the HDL hypothesis in more detail. Epidemiologists have utilized HDL cholesterol， which is the cholesterol cargo in the HDL particle， as a biomarker of cardiovascular risk. The question is: is the HDL cholesterola target of therapy or should we consider other components of the HDL particle as more appropriate targets for intervention?

　　First of all， low levels of basal HDL cholesterol are typically associated with elevated LDL particle concentrations， or its surrogate measure， apolipoprotein B. The Framingham Offspring Study reported that the lower the HDL cholesterol， the higher the concentration of small particles， and the higher the concentration of the total LDL particles. In the AFCAPS/Tex-CAPS study of men and women with low or below average HDL cholesterol levels， the apolipoprotein protein B increase at the lowest HDL cholesterol levels and cardiovascular risk were associated most strongly with high apolipoprotein B levels.

　　And these two studies are cited because it may not be the low level of HDL cholesterol， but maybe the associated high levels of circulating anthropogenic lipoproteins that occur in people of low levels of HDL cholesterol. This is a situation that is in large part mediated by cholesterol transfer proteins. We know that in the Veterans Administration HDL Intervention trial， a study of men with higher myocardial infarction who have low levels of HDL cholesterol， (less than 40 mg per deciliter)， that low levels of HDL particles were the most important predictor of recurrent cardiovascular events. In that study， gemfibrozil increased the HDL particle concentration and that was the most important biomarker that explained the reduction in cardiovascular events in that trial.

　　Turning back to niacin， niacin raises HDL cholesterol by 31 to 37 percent， but increases HDL particle concentration by only one to two percent. So you load up the particles with cholesterol， but not necessary increase the number of HDL particles. Similarly， Torcetrapib， which was evaluated in the ILLUMINATE trial， was an agent that increased HDL cholesterol by 60 percent but increased HDL particles by 1 only percent.

　　So the strategy that we load up the particles with cholesterol without an increase in number of HDL particles is a failed strategy， and one can ask the question， “Why?” HDL particles have a very rich proteome. Different protein constituents that are involved in HDL’s effect on improving endothelial function， reducing oxidative modification of LDL， reducing inflammation， and reducing apoptosis. So， if you increase the cholesterol load in the core of the HDL particle， what happens to the proteome? It changes. And that’s been shown in work by Chapman -- that large ratio particles are less effective as antioxidants and anti-inflammatory particles than the smaller protein rich HDL particles. We now need to refocus our energies to develop new therapies that increase HDL particle concentration and preserve its important anti-atherosclerotic functions.

　　Of the strategies that are being tested， delipidation of HDL as shown by Bryan Brewer is an effective approach， the administration of apolipoprotein A1 mimetics will increase new HDL particles that interact with the critical ABCA1 transporter and do have important anti-atherosclerotic effects. So I’m hoping that the investigation of these novel HDL therapies will reduce atherosclerosis and cardiovascular events.

　　《国际循环》：今年公布的HPS2-THRIVE研究是关于烟酸和拉罗皮兰联合治疗对降低高危血管闭塞事件的研究，但该研究并未显示出与对照组相比主要终点事件的差异，HDL-C的升高并不是很明显。在您看来，实施升高HDL治疗时如何进行合适的患者选择？升高HDL治疗动脉粥样硬化在临床实践中是否可行？

　　Rosenson教授：首先，HPS2-THRIVE试验的设计存在缺陷。该研究入选的是使用辛伐他汀加或不加依折麦布降LDL-C治疗达标后的患者，其平均LDL-C为60 mg/dl，在试验开始时，非HDL-C及HDL-C均处于平均水平。我们知道烟酸具有很多不良反应。鉴于将其潜在不良反应告知患者并进行相关教育需耗费大量时间和努力，故临床上烟酸的处方应用尚存在极大挑战。而研究结果提示，联用拉罗皮兰似乎并未出现预期的减弱烟酸不良反应的作用。

　　试验结束时，结果也并不显着。故得出与基线相比血脂水平并无获益的结论也就并不令人吃惊。研究者排除了那些应用辛伐他汀加和依折麦布后LDL-C水平≥特定水平的患者，这些患者正是我们通常考虑要应用烟酸的患者。因此，该试验并未解决这一重要临床问题，我与Antonio Gotto医生在今年早些时候发表在Current Atherosclerosis Reports的相关社论中就此进行了讨论。

　　对HDL假说进行更详细的讨论十分重要。流行病学家曾利用HDL-C即HDL颗粒中的胆固醇成分作为心血管风险的标志物。现在的问题是，HDL-C是否能作为治疗标靶？HDL颗粒的其他组分是否更合适作为干预靶标？

　　首先，较低的HDL-C水平通常与LDL颗粒浓度或其替代指标载脂蛋白B浓度增加具有相关性。弗雷明汉后代研究显示，HDL-C水平越低，小颗粒及总LDL颗粒浓度越高。在AFCAPS/Tex- CAPS研究中，HDL-C水平低于平均水平的男性及女性中，HDL-C水平最低组患者的载脂蛋白B浓度显着增加，其心血管风险与高载脂蛋白B水平具有较强的相关性。

　　这两项研究均被广为引用，因为HDL-C水平较低的患者中，不是HDL-C的低水平而可能是循环中较高的人为脂蛋白才是风险增加的原因所在。HDL-C水平较低在很大程度上是由胆固醇转运蛋白介导。退伍军人管理局HDL干预试验对HDL-C水平较低（<40 mg/dl）的高危心肌梗死患者进行了研究，其结果显示，低HDL颗粒水平是复发性心血管事件最重要的预测因素。在这项研究中，吉非贝齐能增加HDL颗粒浓度，是试验中能解释该药降低心血管事件最重要的生物标志物。

　　而烟酸虽能使HDL-C增加31%~37%，但却仅能使HDL颗粒浓度增加1%或2%。因此，尽管其能够增加颗粒中的胆固醇含量，但却未能增加HDL颗粒数量。ILLUMINATE试验评估了降脂药托彻普的作用，得到了相似的结果。托彻普能使HDL-C增加60%，却仅能使HDL颗粒增加1%。

　　因此，在不增加HDL-C颗粒的情况下增加颗粒中胆固醇含量的治疗策略是一种失败的策略。人们会问这是为什么？HDL颗粒拥有丰富的蛋白质组，而不同的蛋白成分分别参与到HDL改善内皮细胞功能、降低LDL氧化修饰、减轻炎症反应及减少凋亡等效应中。所以，如果你增加的是HDL颗粒核心中的胆固醇之后，会对蛋白质组有怎样的影响呢？此时，蛋白质组会发生改变。Chapman等的研究证实了这一点。其结果显示，与富含小蛋白的HDL颗粒相比，富含大蛋白的HDL颗粒的抗氧化及抗炎作用较弱。因此，我们现在需要重新调整我们的研究方向以研发能够增加HDL颗粒浓度并保留其抗动脉粥样硬化功能的新型治疗策略。

　　在目前正在被检验的相关策略中，Bryan Brewer的研究显示，HDL去脂化策略是一种非常有效的治疗方法。应用载脂蛋白A1类似物可增加新HDL颗粒的水平，后者与ABCA1转运体相互作用，具有非常重要的抗动脉粥样硬化效应。因此，我希望这些新的以HDL为靶标的治疗策略能降低动脉粥样硬化及心血管事件的发生。