坎地沙坦对服用血管紧张素转换酶抑制剂的左室收缩功能下降的慢性心力衰竭 患者的疗效：CHARM合用研究

摘要

背景  

在血管紧张素转换酶（angiotensin-converting-enzyme，ACE）抑制剂基础上加用血管紧张素Ⅱ 1 型受体拮抗剂可改善慢性心力衰竭（chronic heart failure，CHF）患者的血流动力学指标、神经体液活性和左室重构。本研究旨在探讨这些药物是否有助于改善临床结局。

 

方法

1999 年 3 月至 1999 年 11 月入组 2548 例正接受ACE 抑制剂治疗的纽约心脏协会心功能分级Ⅱ ~ Ⅳ级 CHF 且左室射血分数≤ 40% 的患者，随机分成坎地沙坦组（n = 1276，目标剂量 32 mg，1 次 / 天）和安慰剂组（n = 1272）。基线时， 55% 的患者还使用 β 受体阻滞剂，17% 使用螺内酯。主要终点是心血管死亡或因 CHF 入院的复合终点事件。根据意向性治疗进行分析。

 

结果

中位随访时间为41个月。坎地沙坦组483例（38%）患者和安慰剂组 538 例（42%）患者发生主要终点事件 [ 未校正风险比 ：0.85（95% CI ：0.75 ~ 0.96），P = 0.011 ；协变量校正后，P = 0.010]。坎地沙坦治疗各种主要终点事件均显著降低，因 CHF 入院的总例数也显著降低。在所有预先定义的亚组（包括基线使用 β 受体阻滞剂的患者）中观察到相似的坎地沙坦获益。

 

解释

在 ACE 抑制剂和其他治疗基础上加用坎地沙坦可使左室射血分数下降的 CHF 患者的相关心血管事件出现具有临床意义的进一步降低。

 

引言

对于左室射血分数下降的慢性心力衰竭（chronic heart failure，CHF）患者，尽管应用血管紧张素转换酶（angiotensin-converting-enzyme，ACE）抑制剂、β 受体阻滞剂和螺内酯等全面的常规治疗，但死亡率和发病率仍较高。理论上，在 ACE 抑制剂基础上加用血管紧张素Ⅱ 1型受体拮抗剂是治疗 CHF 的较佳方案。血管紧张素Ⅱ可通过人体心脏组织和血管的非 ACE 酶解通路产生，且即使在长期应用大剂量 ACE 抑制剂治疗 CHF 期间，血管紧张素Ⅱ仍可持续产生 [1-5]。因此，血管紧张素受体拮抗剂应可更全面地抑制血管紧张素Ⅱ的作用。相反，ACE 抑制剂也可阻断与 ACE 相同的激肽酶Ⅱ介导的缓激肽的降解。缓激肽具有直接和间接的血管扩张、抗有丝分裂和抗血栓作用，这些作用对 CHF 有益 [6,7]。因此，ACE 抑制剂 / 血管紧张素受体拮抗剂联合治疗优于ACE 抑制剂单药治疗。

 

包括左室功能障碍治疗策略的随机评估（Randomized Evaluation of Strategies for Left Ventricular Dysfunction，RESOLVD）先导研究 [8]在内的一些研究发现，ACE 抑制剂 / 血管紧张素受体拮抗剂联合治疗 CHF 可改善血流动力学指标、左室重构和神经体液活性 [8,9]。这种联合治疗也可增加 CHF 患者的运动耐量，改善 CHF 患者的纽约心脏协会（New York Heart Association，NYHA）心功能分级 [10]。

 

作为坎地沙坦治疗心力衰竭降低死亡率和发病率的评估（Candesartan in Heart failure: Assessment of Reduction in Mortality and morbidity，CHARM）研究的一部分，本项前瞻性 CHARM 合用研究 [11-13]旨在探讨联合应用血管紧张素受体拮抗剂坎地沙坦和 ACE 抑制剂是否也可改善临床结局。本研究在左室射血分数下降的CHF 患者中比较了坎地沙坦与安慰剂治疗的效果。

 

方法

CHARM 研究方案的设计（包括随机分组、监测和随访）已在既往文献中予以详述 [11-13]。

 

患者

1999 年 3 月至 1999 年 11 月，从 26 个国家的 618个中心纳入符合以下标准的患者 ：年龄≥ 18 岁，过去 6个月内测定的左室射血分数≤ 40%，NYHA 心功能分级Ⅱ ~ Ⅳ级（如心功能为Ⅱ级，应为过去 6 个月内因心脏事件入院者），接受稳定剂量 ACE 抑制剂治疗≥ 30 天。告知研究者已知的可降低 CHF 患者发病率和死亡率的ACE 抑制剂剂量，并要求研究者陈述自己认为每例患者接受的 ACE 抑制剂剂量是否为最佳剂量。本研究获得所有参与中心的伦理委员会或机构审查委员会批准，并且所有患者均签署知情同意书。

 

方法

按双盲方式将患者随机分入坎地沙坦组和匹配的安慰剂组，初始剂量为 4 mg 或 8 mg，每日一次（图 1）。分配代码由数据安全监测委员会独立保存。根据强制滴定方案，在可耐受情况下，每 2 周治疗剂量加倍，并建议监测血压、血清肌酐和血钾。从 6 周起，目标剂量为 32 mg，每日一次。在随机分组后 2 周、4 周、6 周和 6 个月对患者进行随访，此后每 4 个月随访 1 次，直至研究结束。对于在北美纳入的患者亚群，在基线、6 周和此后每年进行1次常规实验室检查，以确定有无安全性问题。

主要终点为心血管死亡或因 CHF 加重而入院治疗。预先定义的次要终点为 ：心血管死亡、因 CHF 入院、或非致死性心肌梗死；心血管死亡、因 CHF 入院、非致死性心肌梗死或非致死性卒中；心血管死亡、因 CHF 入院、非致死性心肌梗死、非致死性卒中或冠状动脉血运重建；全因死亡或因 CHF 入院 ；新发糖尿病。

 

除非观察到明确的非心血管死因，否则所有死亡均记录为心血管死亡。CHF 入院定义为因心力衰竭而需入院治疗，即出现心力衰竭加重的体征和症状并需要静脉利尿剂治疗。心力衰竭加重的表现需包括以下至少一项 ：劳力性呼吸困难加重、端坐呼吸、夜间呼吸困难、肺水肿、外周水肿加重、疲劳增加或运动耐量降低、肾灌注不足（即肾功能恶化）、颈静脉压升高和CHF 的影像学征象。

 

心肌梗死的诊断应满足如下标准 ：肌酸激酶或肌酸激酶 -MB 升至正常上限的 2 倍以上 ；或如果肌酸激酶或肌酸激酶 -MB 测定值不详，肌钙蛋白 I 或 T 升至正常上限的 2 倍以上 ；或经皮腔内冠状动脉血管成形术 24 h 内上述指标升至正常上限的 3 倍 ；或冠状动脉旁路移植术24 h 内上述指标升至正常上限的 5 倍。除了上述实验室指标标准外，患者还需有如下心电图改变，即两个或两个以上相邻导联的心电图出现新的 Q 波（或 V1 或 V2导联出现R波），左束支传导阻滞，或缺血性ST-T波改变；或有符合心肌梗死的典型临床表现，定义为以下其中一种 ：心肌缺血型疼痛持续 20 分钟以上，肺水肿，或原因不明的心源性休克。

 

统计学方法

计划纳入的样本量为 2300 例患者，此样本量发现主要终点事件发生率相对降低 16%（假设安慰剂组事件年发生率为 18%）的效能为 80%。对意向治疗人群进行分析，包括所有随机分组的患者。通过至发生首次事件的时间来分析所有主要终点事件。在初步分析中，采用logrank 检验法比较时间 - 事件的分布情况。计算风险比（hazard ratio， HR）及其 95% 可信区间（confidenceinterval，CI）。此外，采用含有治疗和其他前瞻性定义的协变量（表 1）的 Cox 回归模型来校正这些预设基线因素的风险比，因为这些因素可能会改变事件发生率。采用双侧P 值，P ＜ 0.05 被认为具有统计学意义。

赞助方角色

研究赞助者负责管理数据，其代表参与数据分析和数据解读。所有最终数据分析均由赞助者进行，并由英国伦敦卫生与热带医学院的统计学中心独立进行验证。

 

结果

在入组的 2548 例患者中，坎地沙坦组 1276 例，安慰剂组 1272 例（图 1）。2003 年 3 月 31 日结束随访，中位随访时间为 41 个月。

 

表 1 所列为包括背景药物治疗详情的基线特征。依那普利、赖诺普利、卡托普利和雷米普利是最常用的 ACE 抑制剂，这些药物占所用 ACE 抑制剂的 74%。在坎地沙坦组患者中，这些药物的日平均剂量分别为16.8 mg、17.7 mg、82.2 mg 和 6.8 mg ；在安慰剂组患者中分别为 17.2 mg、17.7 mg、82.7 mg 和 7.3 mg。研究者认为随机分组时各组有 96% 的患者接受了最佳剂量的 ACE 抑制剂。55% 的患者在基线时使用 β 受体阻滞剂，17% 使用螺内酯 [12]。研究结束时，坎地沙坦组和安慰剂组分别有 64% 和 68% 的患者正在使用β 受体阻滞剂。坎地沙坦组和安慰剂组使用螺内酯的患者比例分别增至 20% 和 25%。研究结束时，坎地沙坦组和安慰剂组分别有 2.3% 和 5.0% 的患者正在接受开放标签血管紧张素受体拮抗剂治疗。

 

坎地沙坦组 483 例（38%）患者和安慰剂组 538 例（42%）患者出现主要终点事件，即心血管死亡或因CHF入院[未校正HR： 0.85（95% CI： 0.75 ~ 0.96）， P = 0.011；协变量校正后P = 0.010 ；图 2]。坎地沙坦组和安慰剂组的年事件发生率分别为 14.1% 和 16.6%。

其他终点事件见表 2。坎地沙坦除了可降低心血管死亡率和因 CHF 入院风险之外，还可降低各种次要复合终点事件的风险。坎地沙坦组的心血管死亡例数为 302例（24%），而安慰剂组为 347 例（27%）（未校正 HR ：0.84，95% CI ：0.72 ~ 0.98，P = 0.029 ；协变量校正后P = 0.021）。此外，坎地沙坦组随机分组后因 CHF 首次入院的患者比例、多次因 CHF 入院的患者比例和因 CHF入院总例数均降低（表 3）。坎地沙坦组和安慰剂组出现心肌梗死的总例数分别为 44 例和 69 例（P = 0.012）；卒中总例数分别为 47 例和 41 例（P = 0.62）；冠状动脉血运重建总例数分别为 69 例和 75 例（P = 0.46）。

 

坎地沙坦组的全因死亡例数为 377 例（30%），而安慰剂组为 412 例（32%）（未校正 HR： 0.89，95% CI： 0.77 ~ 1.02，P = 0.086 ；协变量校正后P = 0.105）。坎地沙坦组全因死亡或 CHF 入院例数为 539 例（42%），而安慰剂组为 587 例（46%）（未校正 HR： 0.87， 95% CI： 0.78 ~ 0.98，P = 0.021）。坎地沙坦组 852 例患者因任何原因入院 2462 次，安慰剂组 858 例入院 2798 次（患者例数比较，P = 0.7 ；入院次数比较，P = 0.023）。坎地沙坦组和安慰剂组各有 72 例（6%）患者新发糖尿病（未校正 HR ：0.98，95% CI ：0.70 ~ 1.35，P = 0.88）。

 

在所有预设亚组中，使用坎地沙坦均降低心血管死亡或因 CHF 入院的发生风险，但无证据表明其治疗效果存在异质性 [13]。特别是对于基线时联用 ACE 抑制剂和β 受体阻滞剂的患者，坎地沙坦降低其发生风险（图 3）。在坎地沙坦组和安慰剂组中，这些患者的死亡率分别为25% （175 例 /702 例）和 27% （195 例 /711 例）（HR： 0.88，95% CI ：0.72 ~ 1.08，P = 0.22）。在坎地沙坦组和安慰剂组中，基线时未使用 β 受体阻滞剂的患者的死亡率分别为 35% （202 例 /574 例）和 39% （217 例 /561 例）（HR：0.88，95% CI ：0.73 ~ 1.07，P = 0.20）。在使用推荐剂量ACE 抑制剂的患者中观察到的坎地沙坦疗效与在使用较低剂量的患者中观察到的相同（图 3）。

86% 患者的坎地沙坦或安慰剂初始剂量为 4 mg，每日一次，14% 患者的初始剂量为 8 mg，每日一次。6 个月时，坎地沙坦组和安慰剂组患者使用研究药物的日平均剂量分别为 24 mg 和 27 mg。坎地沙坦组和安慰剂组分别有 61% 和 73% 的患者在随机分组后 6 个月内达到目标剂量 32 mg。

 

末次研究随访时，坎地沙坦组和安慰剂组分别有220 例（25%）和 155 例（18%）存活患者不再使用研究药物（任何原因）。总体上，坎地沙坦组和安慰剂组分别有 309 例（24%）和 233 例（18%）患者因不良事件或异常实验室检查值而长期停用研究药物（P=0.0003，表 4）。

坎地沙坦组和安慰剂分别有7%（32例/436例）和6%（27例/447例）患者的血清肌酐水平相对基线升高至少1倍（P = 0.5）。在坎地沙坦组和安慰剂组中，分别有11%（8 例 /73例）和 4%（3 例 /71 例）基线时使用螺内酯的患者的血清肌酐水平相对基线升高至少 1 倍（P = 0.21）。

 

坎地沙坦组和安慰剂组分别有 3%（12 例 /447 例）和 1%（5 例 /459 例）的患者的血钾浓度≥ 6 mmol/L（P=0.089）。坎地沙坦组和安慰剂组分别有4% （3例/74例）和 1%（1 例 /71 例）基线时使用螺内酯的患者的血钾浓度≥ 6 mmol/L。

 

6 个月时，坎地沙坦组收缩压（4.6 mmHg，P=0.007）和舒张压（3.0 mmHg，P=0.004）相对基线的降幅大于安慰剂组。在坎地沙坦组中，基线时使用 β 受体阻滞剂的患者的血压降幅不大于未使用 β 受体阻滞剂的患者。

 

坎地沙坦组和安慰剂组分别有 2 例和 3 例患者发生血管性水肿，所受累患者均在使用 ACE 抑制剂，有 2 例患者（每组 1 例）需入院治疗。坎地沙坦组 1 例患者停用研究药物。

 

讨论

对于左室射血分数低的 CHF 患者，在 ACE 抑制剂基础上加用坎地沙坦可降低心血管死亡和因 CHF 入院的发生风险。这种坎地沙坦的治疗获益见于所有预设亚组患者（包括接受 β 受体阻滞剂和其他治疗的患者），并且不存在治疗异质性。

 

研究发现，血管紧张素Ⅱ在长期 ACE 抑制剂治疗期间仍会持续产生 [1-5]，并且机制研究显示在 ACE 抑制剂治疗的基础上加用血管紧张素受体拮抗剂有助于改善患者的神经体液活性、血流动力学和左室重构 [8,9]，本研究结果与这些既往研究结果一致。这些坎地沙坦的治疗获益也见于接受 β 受体阻滞剂和 ACE 抑制剂的患者。例如，在RESOLVD 先导研究中 [14]，在联用卡托普利、美托洛尔和坎地沙坦的情况下，观察到的逆转左室重构的作用最明显。本研究结果可扩大观察结果至主要临床结局的改善。

 

尽管难以对不同研究进行直接比较，但从表面上来看，本研究结果似乎与缬沙坦心力衰竭试验（Valsartan Heart Failure Trial，Val-HeFT）[15]的结果不同。Val-HeFT研究显示，在患者接受 ACE 抑制剂（93% 的患者）、β受体阻滞剂（35%）和螺内酯（5%）等常规治疗的基础上加用缬沙坦使主要复合终点事件 [ 即死亡或心血管发病率（因 CHF 入院，因 CHF 而在门诊接受≥ 4 小时的静脉治疗，或心脏骤停复苏）] 的发生风险降低 13.2%。该复合终点结果主要源于因CHF入院率降低27.5%，因为缬沙坦对心血管死亡率和总死亡率均无影响。令人意外的是，在基线时同时使用 ACE 抑制剂和 β 受体阻滞剂的 1610例（35%）患者中观察到缬沙坦与一个不良终点指标相关，这引发了对神经内分泌抑制过度的担忧。一些指南因此不建议同时使用三种神经体液抑制剂进行治疗 [16-19]。研究者认为本研究结果可以消除这些担忧。

 

本研究人群非常适合与 Val-HeFT 的整体人群（几乎所有患者均接受 ACE 抑制剂治疗）比较。两项研究的结果一致表明，在常规治疗基础上加用血管紧张素受体拮抗剂可改善临床获益。本研究与 Val-HeFT 研究的明显差异可能源于所用血管紧张素受体拮抗剂的特定类型或剂量。另外，Val-HeFT 研究中对小样本量的亚组分析的效能不足，这可能也是产生差异的原因。

 

本研究在接受 ACE 抑制剂推荐剂量治疗的患者中观察到坎地沙坦获益。本研究中基线时患者所用依那普利的日平均剂量为17.0 mg，与左室功能障碍研究[20]（Studies Of Left Ventricular Dysfunction） 所 用 剂 量 16.6 mg 和Val-HeFT 研究 [15]所用剂量 17.0 mg 相似。本研究还明确指出这种获益具有临床意义。在平均 3.0 年的研究期间，坎地沙坦组和安慰剂组分别有 37.9% 和 42.3% 的患者发生心血管死亡或因 CHF 首次入院。每治疗 100 例患者，就可绝对减少 4.4 例患者发生事件，这与为了预防一起首发心血管死亡或因 CHF 入院事件而需治疗的患者数量（23 例）相符。这不仅降低首发事件的发生率，而且也降低常见的多次 CHF 入院的发生率，从而减少患者痛苦、降低治疗费用[21,22]。这些获益源于肾素-血管紧张素-醛固酮系统抑制药物的疗效，而其特有的不良反应却比较少见。坎地沙坦组因肾功能不全和高钾血症而停药的患者比例较高，这提示使用时需密切监测肾功能和血钾。综上所述，对于左室射血分数低的 CHF 患者，在 ACE抑制剂和其他治疗（包括 β 受体阻滞剂）基础上加用坎地沙坦可使心血管死亡率和发病率出现具有临床意义的降低，并且这种联合治疗方案的耐受性通常较好。

 

---

Summary

Background

---

Introduction

Mortality and morbidity among patients with chronic heart failure (CHF) and reduced left-ventricular ejection fraction remain high, despite the use of full conventional treatment, including angiotensin-converting-enzyme (ACE) inhibitors, β blockers, and spironolactone. The addition of an angiotensin II type 1 receptor blocker to an ACE inhibitor is a theoretically attractive treatment strategy in CHF. Angiotensin II can be produced by non-ACE enzymatic pathways in human cardiac tissue and blood vessels, and its generation seems to continue even during chronic, high-dose, ACE-inhibitor treatment in CHF.^{1, 2, 3, 4 and 5} Angiotensin-receptor blockers should, therefore, provide more complete inhibition of the actions of angiotensin II. Conversely, ACE inhibitors also block the breakdown of bradykinin, mediated by kininase II, which is identical to ACE. Bradykinin has direct and indirect vasodilator, antimitotic, and antithrombotic actions that could be of benefit in CHF.^6 and 7 Consequently, treatment with combined ACE inhibitors and angiotensin-receptor blockers might have advantages over ACE-inhibitor monotherapy.

In several studies, including the Randomized Evaluation of Strategies for Left Ventricular Dysfunction pilot study,⁸ favourable effects on haemodynamic indices, left-ventricular remodelling, and neurohumoral activity in CHF have been reported with combined ACE inhibitors and angiotensin-receptor blockers.^8 and 9 This combination of treatment also increases exercise capacity and improves New York Heart Association functional class.¹⁰

In the prospective Candesartan in Heart failure: Assessment of Reduction in Mortality and morbidity (CHARM)-Added trial, part of the CHARM programme,^{11, 12 and 13} we investigated whether combining an angiotensin-receptor blocker, candesartan, with ACE inhibitors also improves clinical outcome. We compared the effect of candesartan with that of placebo among patients with CHF and reduced left-ventricular ejection fraction.

Methods

The design of the CHARM programme has been described in detail elsewhere, including randomisation, monitoring, and follow-up.^{11, 12 and 13}

Patients

Eligible patients were aged 18 years or older, had left-ventricular ejection fraction 40% or lower measured within the past 6 months, New York Heart Association functional class II–IV (if class II, patients had to have admission to hospital for a cardiac reason in the previous 6 months), and treatment with an ACE inhibitor at a constant dose for 30 days or longer. We enrolled patients between March, 1999, and November, 1999 in 618 centres in 26 countries. Investigators were advised of the doses of ACE inhibitors known to reduce morbidity and mortality in CHF and also asked to state whether each patient was, in their opinion, on an optimum dose of ACE inhibitor, judged individually. The study was approved by ethics committees or institutional review boards in all participating centres and all patients gave written, informed consent.

Methods

We randomly assigned patients, in a double-blind way, candesartan or matching placebo, which could be started at 4 or 8 mg once daily (figure 1), the assignment code being held at an independent centre and by the data safety monitoring board. The treatment dose was doubled every 2 weeks, as tolerated, according to a forced titration protocol, with recommended monitoring of blood pressure, serum creatinine, and potassium. The target dose was 32 mg once daily from 6 weeks onwards. After randomisation, patients were seen at 2, 4, and 6 weeks, at 6 months and, thereafter, at every 4 months until the end of the trial. In a subset of patients enrolled in North America, routine laboratory assessments were done at baseline, 6 weeks, and yearly thereafter for safety reasons.

 

The primary outcome was cardiovascular death or unplanned admission to hospital for the management of worsening CHF. Prespecified secondary outcomes were: cardiovascular death, admission to hospital for CHF, or non-fatal myocardial infarction; cardiovascular death, admission to hospital for CHF, non-fatal myocardial infarction, or non-fatal stroke; cardiovascular death, admission to hospital for CHF, non-fatal myocardial infarction, non-fatal stroke, or coronary revascularisation; death (any cause) or admission to hospital for CHF; and development of new diabetes.

We classified all deaths as cardiovascular unless an unequivocal non-cardiovascular cause was established. A CHF hospital admission was defined as admission to hospital necessitated by heart failure and primarily for its treatment. A patient admitted for this reason had to show signs and symptoms of worsening heart failure and require treatment with intravenous diuretics. Evidence of worsening heart failure had to include at least one of the following items: increasing dyspnoea on exertion, orthopnoea, nocturnal dyspnoea, pulmonary oedema, increasing peripheral oedema, increasing fatigue or decreasing exercise tolerance, renal hypoperfusion (ie, worsening renal function), raised jugular venous pressure, and radiological signs of CHF.

A diagnosis of myocardial infarction was made if the following conditions were met: creatine kinase or creatine kinase-MB more than twice the upper limit of normal, or troponin I or T more than twice the upper limit of normal if neither creatine kinase or creatine kinase-MB were available; or three times the upper limit of normal for the same markers within 24 h of percutaneous transluminal coronary angioplasty; or five times the upper limit of normal for the same markers within 24 h of coronary artery bypass grafting surgery. In addition to these marker criteria, a patient had to have experienced electrocardiographic changes in two or more contiguous leads showing new Q waves (or R waves in V1 or V2), left-bundle-branch block, or ischaemic ST-T wave changes, or typical clinical presentation consistent with myocardial infarction defined as one of the following: cardiac ischaemic type pain lasting more than 20 min, pulmonary oedema, or cardiogenic shock not otherwise explained.

Statistical methods

The planned sample size of 2300 patients was designed to provide around 80% power to detect a 16% relative reduction in the primary outcome, assuming an annual placebo event rate of 18%. The analysis was done on an intention-to-treat basis and included all randomised patients. We analysed all major outcomes by time to first event. For the primary analysis we used the logrank test to compare the time-to-event distributions. The hazard ratios were estimated together with 95% CI. In addition, we used a Cox's regression model with treatment and other prospectively defined covariates (table 1) to adjust the hazard ratio for these prespecified baseline factors, which might alter the event rates. We used two-sided p values and took p<0·05 to be significant.

Role of the funding source

The sponsor of the study managed the data, and its representatives were involved in the data analysis and data interpretation. All final data analyses were done by the sponsor and verified independently by the statistical centre at the London School of Hygiene and Tropical Medicine, London, UK.

Results

Of 2548 patients enrolled, 1276 were assigned candesartan and 1272 placebo (figure 1). Follow-up was concluded on March 31, 2003. The median duration of follow up was 41 months.

The baseline characteristics, including details of background medical treatment, are given in table 1. Enalapril, lisinopril, captopril, and ramipril were the most commonly used ACE inhibitors, together accounting for 74% of all ACE inhibitors used. The mean daily doses of these drugs in the candesartan group were 16·8, 17·7, 82·2, and 6·8 mg, respectively, and in the placebo group were 17·2, 17·7, 82·7, and 7·3 mg, respectively. Investigators stated that they thought 96% of patients in each group were receiving optimum doses of ACE inhibitor at randomisation. 55% of patients were treated with β blockers at baseline and 17% with spironolactone.¹² By the end of the study, 64% of patients in the candesartan group and 68% in the placebo group were taking β blockers. The proportion of patients taking spironolactone had risen to 20% in the candesartan group and to 25% in the placebo group. Open-label angiotensin-receptor-blocker treatment was being used in 2·3% of the candesartan group and 5·0% of the placebo group by the end of the trial.

483 (38%) patients in the candesartan group and 538 (42%) in the placebo group experienced the primary outcome of cardiovascular death or admission to hospital for CHF (unadjusted hazard ratio 0·85 [95% CI 0·75–0·96], p=0·011; covariate adjusted p=0·010; figure 2). The annual event rates were 14·1% in the candesartan group and 16·6% in the placebo group.

Other outcomes are shown in table 2. Candesartan reduced cardiovascular mortality and the risk of admission to hospital for CHF individually, as well as the risk of each of the secondary composite outcomes. There were 302 (24%) cardiovascular deaths in the candesartan group compared with 347 (27%) in the placebo group (unadjusted 0·84 [0·72–·98], p=0·029; covariate adjusted p=0·021). Candesartan also reduced the proportion of patients experiencing a first hospital admission for CHF after randomisation, the proportion of patients with multiple admissions for CHF, and the total number of hospital admissions for CHF (table 3). The total number of patients who had myocardial infaraction was candesartan 44, placebo 69 (p=0·012); stroke: candesartan 47, placebo 41 (p=0·62); and coronary revascularisation procedures: candesartan 69, placebo 75 (p=0·46).

 

The number of deaths from any cause in the candesartan group was 377 (30%) compared with 412 (32%) in the placebo group (unadjusted 0·89 [0·77–1·02], p=0·086; covariate adjusted p=0·105). 539 (42%) patients treated with candesartan and 587 (46%) with placebo died from any cause or were admitted for CHF (unadjusted 0·87 [0·78–0·98], p=0·021). In the candesartan group, 852 patients had 2462 hospital admissions for any reason and 858 placebo patients had 2798 admissions (p=0·7 for patients and p=0·023 for admissions). 72 (6%) patients in the candesartan group and 72 (6%) in the placebo group developed new diabetes (unadjusted 0·98 [0·70–1·35], p=0·88).

Candesartan reduced the risk of cardiovascular death or admission to hospital for CHF in all predefined subgroups, with no evidence of heterogeneity of treatment effect.¹³ In particular, candesartan reduced this risk in patients treated with β blockers in addition to an ACE inhibitor at baseline (figure 3). Among these patients, 175 (25%) of 702 died in the candesartan group and 195 (27%) of 711 died in the placebo group (0·88 [0·72–1·08], p=0·22). The numbers of deaths in patients not taking a β blocker at baseline were 202 (35%) of 574 in the candesartan group and 217 (39%) of 561 in the placebo group (0·88 [0·73–1·07], p=0·20). Candesartan was as effective among patients taking a recommended dose of ACE inhibitor as in those taking lower doses (figure 3).

86% of patients started on 4 mg and 14% on 8 mg of candesartan or placebo daily. The mean daily doses for patients taking study drug at 6 months were 24 mg in the candesartan and 27 mg in the placebo group. 61% of the candesartan and 73% of the placebo group reached the target dose of 32 mg within 6 months of randomisation.

At the final study visit, 220 (25%) survivors in the candesartan group and 155 (18%) in the placebo group were no longer taking study medication for any reason. Overall, 309 (24%) patients in the candesartan group and 233 (18%) patients in the placebo group permanently discontinued study medication because of an adverse event or an abnormal laboratory value (p=0·0003, table 4).

In 32 (7%) of 436 in the candesartan group, creatinine at least doubled from baseline, compared with in 27 (6%) of 447 in the placebo group (p=0·5). Among patients taking spironolactone at baseline, serum creatinine at least doubled from baseline in eight (11%) of 73 patients in the candesartan group and three (4%) of 71 in the placebo group (p=0·21).

In the candesartan group, 12 (3%) of 447 patients developed potassium concentrations 6 mmol/L or higher compared with five (1%) of 459 in the placebo group (p=0·089). For patients taking spironolactone at baseline, three (4%) of 74 in the candesartan group developed potassium concentrations of 6 mmol/L or higher compared with one (1%) of 71 in the placebo group.

By 6 months, blood pressure was lowered from baseline by 4·6 mm Hg systolic (p=0·007) and 3·0 mm Hg diastolic (p=0·004) more in the candesartan group than in the placebo group. The reduction in blood pressure with candesartan was not greater among patients treated with β blockers at baseline than among those not treated with β blockers.

There were two cases of angioedema in the candesartan group and three in the placebo group. All affected patients were taking an ACE inhibitor at the time and two required hospital admission (one placebo and one candesartan). One patient taking candesartan had study medication discontinued.

Discussion

Among patients with CHF and a low left-ventricular ejection fraction, the addition of candesartan to an ACE inhibitor decreased the risk of cardiovascular death, and admission to hospital for CHF. This beneficial effect of candesartan was seen in all prespecified subgroups of patients, including those treated with β blockers and other treatments, with no evidence of treatment heterogeneity.

Our findings are consistent with the evidence that angiotensin II continues to be produced despite chronic ACE-inhibitor treatment,^{1, 2, 3, 4 and 5} and mechanistic studies showing favourable neurohumoral, haemodynamic, and left-ventricular remodelling effects from adding an angiotensin-receptor blockers in patients already treated with an ACE inhibitor.^8 and 9 These potentially beneficial effects are also seen in patients treated with β blockers and ACE inhibitors. For example, in the Randomized Evaluation of Strategies for Left Ventricular Dysfunction pilot study,¹⁴ the greatest left-ventricular reverse remodelling was seen with the combination of enalapril, metoprolol, and candesartan. Our results extend those observations to improvements in important clinical outcomes.

Our findings may superficially seem to be in conflict with those of Valsartan Heart Failure Trial (Val-HeFT),¹⁵although direct comparisons between trials are difficult to make. In Val-HeFT. the addition of valsartan to conventional treatment, including ACE inhibitors in 93% of patients, β blockers in 35%, and spironolactone in 5%, reduced the risk of the composite co-primary outcome of death or cardiovascular morbidity (admission for CHF, ≥4 h intravenous treamtent for CHF without admission, or cardiac arrest with resuscitation) by 13·2%. This effect on the composite outcome was principally explained by a 27·5% reduction in CHF hospital admission, since valsartan had no effect on cardiovascular mortality or total mortality. Unexpectedly, in the 1610 (35%) patients treated with both ACE inhibitors and β blockers at baseline, valsartan was associated with a worse outcome. This latter finding has caused concern about excessive neuroendocrine inhibition and led guidelines to discourage triple neurohumoral blockade.^{16, 17, 18 and 19} We believe our results can remove these concerns.

Comparison of the overall Val-HeFT population— nearly all of whom were treated with ACE inhibitors— with our population is most approriate. The findings of the two trials show consistently that adding an angiotensin-receptor blocker to conventional treatment has incremental clinical benefit. The apparent differences between our trial and Val-HeFT might be explained by the particular type or dose of angiotensin-receptor blocker used. Alternatively, underpowered analyses of small subgroups in Val-HeFT might have led to the difference.

The benefits of candesartan were evident in our study among patients treated with recommended doses of ACE inhibitors. For example, the mean daily dose of enalapril taken at baseline was 17·0 mg, which compares favourably with 16·6 mg in those taking the drugs in the treatment group of the Studies Of Left Ventricular Dysfunction²⁰ and 17·0 mg in Val-HeFT.¹⁵ We also show clearly that this benefit is clinically important. Over the mean 3·0 years duration of the trial, 37·9% of patients in the candesartan group experienced a cardiovascular death or first admission to hospital for CHF compared with 42·3% in the placebo group. This absolute reduction of 4·4 patients with events per 100 patients treated corresponds to a number needed to treat of 23 to prevent one first event of cardiovascular death or CHF admission. It is not only first events that are reduced. Multiple CHF admssions, which are common, distressing, and costly, are also reduced.^21 and 22 These benefits are obtained at the expense of infrequent adverse effects that are characteristic of drugs inhibiting the renin-angiotensin-aldosterone system. The higher rates of withdrawals for renal dysfunction and hyperkalaemia in the candesartan group indicate the need for careful monitoring of renal function and serum potassium. In conclusion, the addition of candesartan to an ACE inhibitor and other treatments, including a β blocker, is generally well tolerated in patients with CHF and a low left-ventricular ejection fraction and leads to a clinically important reduction in cardiovascular mortality and morbidity.

Conflict of interest statement

M A Pfeffer, K Swedberg, C B Granger, JJ V McMurray, and S Yusuf have served as consultants to or received research grants from AstraZeneca and other major cardiovascular pharmaceutical companies. J Östergren has served as a consultant and received research grants from AstraZeneca. P Held, E L Michelson, and B Olofsson are employees of AstraZeneca.