The hypothesis that sacubitril/valsartan is superior to ramipril in reducing time-to-first cardiovascular death or the development of heart failure (hospitalization or outpatient) in patients with acute myocardial infarction was tested in PARADISE-MI (Prospective ARNI versus ACE Inhibitor Trial to Determine Superiority in Reducing Heart Failure Events after MI).¹ The risk of this primary composite outcome using the prespecified time-to-first event analysis of the clinical end point committee (CEC)–adjudicated events was not significantly reduced in the sacubitril/valsartan group.² As such, all secondary analyses are considered exploratory.

Because the comparator arm renin-angiotensin system inhibitor has previously proven effective in reducing mortality in this population, exploring additional prespecified end points could offer relevant information regarding the effects of angiotensin receptor neprilysin inhibition in high-risk patients after myocardial infarction. This research letter provides analyses regarding more expansive evaluations of clinical outcomes focusing on total (first and recurrent) CEC-adjudicated, and investigator-reported events, as well.

The design, baseline characteristics, and primary results of this institutional review board–approved trial are published.^1,2 In brief, PARADISE-MI was a double-blind, active-controlled randomized, clinical trial that compared sacubitril/valsartan with ramipril in 5661 patients with written informed consent, an acute myocardial infarction (within 0.5–7 days of presentation) with either left ventricular ejection fraction ≤40% or transient pulmonary congestion. Previous heart failure or clinical instability (requiring intravenous diuretics, inotropes, or blood pressure support within the 24 hours before randomization) were major exclusions, and all patients provided written informed consent.

Time-to-first event data were analyzed using Cox models, whereas timing and occurrence of recurrent events (hospitalizations for heart failure, outpatient heart failure, or cardiovascular death) were analyzed using a negative binominal regression model with a Weibull baseline intensity function with treatment, type of myocardial infarction, percutaneous coronary intervention, and region as factors.² Analyses are exploratory and P values were not adjusted for multiplicity. Data are available on request from the authors.

In the previously reported time-to-first analysis of CEC-adjudicated primary events, there were 338 patients (11.9%) with a first event among the 2830 randomly assigned to sacubitril/valsartan and 373 patients (13.2%) with a first event among the 2831 randomly assigned to ramipril (hazard ratio, 0.90 [95% CI, 0.78–1.04], P=0.17).² Comparisons between the components of the end points were previously reported.²

Among the patients randomly assigned to sacubitril/valsartan, there were 443 patients (15.7%) with a first investigator-reported cardiovascular death, hospitalization for heart failure, or outpatient heart failure, and among the patients randomly assigned to ramipril there were 516 patients (18.2%) with one of these events (hazard ratio, 0.85 [95% CI, 0.75–0.96], P=0.01; Figure [A]).

Figure. Clinical outcomes. Shown are Kaplan-Meier curves for the investigator-reported primary composite outcome, analyzed as time to first event (A), and Nelson-Aalen event curves for total (first and recurrent) clinical end point committee (CEC)–adjudicated (B) and investigator-reported (C) primary events of total hospitalizations for heart failure, total outpatient episodes of symptomatic heart failure, and cardiovascular death. A, Investigator-reported time-to-first event: 443 patients (15.7%) in the sacubitril/valsartan group (cardiovascular death, 115; hospitalization, 231; and outpatient, 97) and 516 patients (18.2%) in the ramipril group (cardiovascular death, 111; hospitalization, 264; and outpatient, 141). B, CEC adjudicated total primary events: 452 primary events in the sacubitril/valsartan group (240 hospitalizations, 44 outpatient episodes, and 168 cardiovascular deaths) and 539 events in the ramipril group (286 hospitalizations, 62 outpatient episodes, and 191 cardiovascular deaths). C, Investigator-reported total primary events: 672 primary events in the sacubitril/valsartan group (395 hospitalizations, 122 outpatient episodes, and 155 cardiovascular deaths) and 802 events in the ramipril group (447 hospitalizations, 176 outpatient episodes, and 179 cardiovascular deaths).

In the sacubitril/valsartan group, there were a total of 452 CEC-adjudicated primary events among 338 patients with at least one of these events, whereas there were 539 total events among 373 patients in the ramipril group (rate ratio, 0.79 [95% CI, 0.65–0.97], P=0.02; Figure [B]). The use of open label sacubitril/valsartan after a CEC-confirmed heart failure event (hospitalization or outpatient episode) was 17 of 199 (8.5%) in the sacubitril/valsartan group and 16 of 234 (6.8%) in the ramipril group.

In the sacubitril/valsartan group, there were 672 investigator-reported total events among 443 patients with at least one of these events, whereas in the ramipril group, there were 802 total events among 516 patients (rate ratio, 0.79 [95% CI, 0.67–0.93], P=0.004; Figure [C]). The use of open label sacubitril/valsartan following an investigator report of heart failure was 23 of 327 (7.0%) in the sacubitril/valsartan group and 28 of 400 (7.0%) in the ramipril group.

When the results of a major randomized trial fail to achieve the level of significance to declare a difference for its primary objective, additional analyses may provide useful information. Questions have been proposed to better understand, although not to change, the neutral result.³ With adequate statistical power and no major deficiencies in trial design or conduct, this retrospective probing of PARADISE-MI focused on choice of the primary outcome. Although recurrent event analyses capture a more complete assessment of the adverse patient and economic impact of heart failure, time-to-first event analysis was chosen because once a patient in PARADISE-MI developed heart failure, open label sacubitril/valsartan use was anticipated to confound the recurrent event analysis. In retrospect, with the minimal use of open label sacubitril/valsartan (<10%) in patients developing symptomatic heart failure during the conduct of our trial, adopting the more expansive outcome of total events would have been a more appropriate primary end point to assess the influence of sacubitril/valsartan relative to ramipril on the full burden of heart failure.

Although these exploratory analyses do not alter the primary neutral findings for use after acute myocardial infarction, the consistent findings of reductions in recurrent heart failure events using both CEC adjudications and investigator reports provides more supportive information for the already indicated replacement of an angiotensin-converting enzyme inhibitor with sacubitril/valsartan once the clinical transition to symptomatic heart failure has occurred.^4,5

PARADISE-MI (Prospective ARNI vs ACE Inhibitor Trial to Determine Superiority in Reducing Heart Failure Events After MI) was sponsored by Novartis.

Disclosures Dr Pfeffer reports research grant support through Brigham and Women’s Hospital from Novartis; and consulting fees from AstraZeneca, Boehringer Ingelheim and Eli Lilly Alliance, Corvidia, DalCor, GlaxoSmithKline, National Heart, Lung, and Blood Institute (NHLBI) CONNECTs (Master Protocol Committee), Novartis, Novo Nordisk, Peerbridge, and Sanofi; and has equity in DalCor. Dr Claggett reports consultancy fees from Amgen Inc, Beren Therapeutics, Myokardia, Novartis, and Peerbridge Health. Dr Lewis reports research grant support from Novartis; consulting fees from Amgen, Dal-Cor, and Merck. Dr Granger reports consultancy fees and/or research support from Novartis, AstraZeneca, Bayer Healthcare Pharmaceuticals, Boehringer Ingelheim, Bristol-Myers Squibb, Correvio, Daiichi Sankyo Co., Janssen Global Services, and Pfizer Inc. Dr Køber reports consultancy fees from Boehringer Ingelheim. Dr Maggioni reports consultancy fees from Novartis. Dr Mann reports consultancy fees from Bristol-Myers Squibb, Cardurion, LivaNova USA Inc, Novartis, Novo Nordisk, and Renovacor. Dr McMurray reports research grant support through his employer, Glasgow University, from Novartis. Dr Rouleau reports consultancy fees from Novartis Pharma, AstraZeneca, Bayer, Bristol-Myers Squibb and Sanofi-Aventis US, LLC. Dr Solomon reports research grant support through Brigham and Women’s Hospital from Actelion Pharmaceuticals, Alnylam Pharmaceuticals, Amgen, AstraZeneca, Bayer, Bellepheron, Bristol-Myers Squibb, Celladon Corporation, Cytokinetics, Eidos, Eli Lilly and Company, GlaxoSmithKline, Ionis, Mesoblast, Myokardia, Neurotronik, Novartis, Novo Nordisk, Respicardia Inc, Sanofi Pasteur Inc, and Theracos; consulting fees from Abbott Vascular, Action Medical Research, Akros, Alnylam Pharmaceuticals, American Regent, Amgen, Arena, AstraZeneca, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb Eli Lilly and Company, Cardiac Dimensions, Cardior, Cardurion, CellProThera, Cytokinetics, Daiichi Sankyo Company, Genomics, GlaxoSmithKline, Johnson and Johnson, Merck, Moderna, Myokardia, Novartis, Quantum, Roche Diagnostics Corporation, Sanofi Pasteur Inc, Sarepta Therapeutics Inc, Tenaya, Theracos, Tremeau and has equity in Dinaqor; other: Corvia Medical. Dr Steg reports research grant support from Bayer; consulting fees from Amarin Pharma Inc, Amgen, AstraZeneca, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Janssen Pharmaceuticals, Novartis, Novo Nordisk, Regeneron Pharmaceuticals, Sanofi US Services Inc, and Servier. Dr Berwanger reports consultancy fees from Amgen, AstraZeneca, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Novartis Pharma, and Pfizer. Dr Cikes reports research support from Abbott and Novartis; and Institutional, Organizational, or Other Financial Benefit from: Abbott, GE Healthcare, Pfizer, and Teva Pharmaceutical Industries. Dr De Pasquale reports consultancy fees from Novartis and Institutional, Organizational, or Other Financial Benefit from Novartis. Dr Landmesser reports research support from Amgen, Bayer, and Novartis; consultancy fees from Bayer, Boehringer Ingelheim, Daiichi Sankyo Company, Novartis, Pfizer, and Sanofi Pasteur Inc. Dr Petrie reports research support from AstraZeneca, Boehringer Ingelheim, Novartis, and SQ Innovation; consulting fees from AstraZeneca, Boehringer Ingelheim, Novartis, Novo Nordisk, SQ Innovation, and Vifor Pharma. Dr Senni reports consulting fees from Abbott Fund, Bayer, Bayer Healthcare, Novartis, Merck, and Vifor Pharma. Dr van d reports research support from AstraZeneca, Pfizer, and Vifor Pharma. Drs Lefkowitz, Zhou, and Wang are full-time employees of Novartis. Dr Braunwald reports research grant support through Brigham and Women’s Hospital from AstraZeneca, Daiichi-Sankyo, Merck, and Novartis; consulting for Amgen, Boehringer-Ingelheim/Lilly, Cardurion, MyoKardia, NovoNordisk, and Verve. The other authors report no conflicts.

This manuscript was sent to Faiez Zannad, Guest Editor, for review by expert referees, editorial decision, and final disposition.

For Sources of Funding and Disclosures, see page 89.

Registration: URL: https://www.clinicaltrials.gov; Unique identifier: NCT02924727.

在 PARADISE-MI（前瞻性 ARNI 与 ACE 抑制剂试验，确定在减少 MI 后心力衰竭事件方面的优势）。¹在沙库巴曲/缬沙坦组中，使用临床终点委员会 (CEC) 裁定事件的预设首次事件发生时间分析得出的这一主要复合结局的风险并未显着降低。²因此，所有二次分析都被认为是探索性的。

由于比较组肾素-血管紧张素系统抑制剂先前已被证明可有效降低该人群的死亡率，因此探索其他预先指定的终点可以提供有关心肌梗死后高危患者血管紧张素受体脑啡肽酶抑制作用的相关信息。这封研究信提供了关于更广泛的临床结果评估的分析，重点是总的（首次和复发的）CEC 裁定事件和研究者报告的事件。

该机构审查委员会批准的试验的设计、基线特征和主要结果已公布。^1,2简而言之，PARADISE-MI 是一项双盲、主动对照的随机临床试验，在 5661 名急性心肌梗死（出现后 0.5-7 天内）患者中进行了书面知情同意，比较了沙库巴曲/缬沙坦与雷米普利左心室射血分数≤40%或一过性肺充血。既往心力衰竭或临床不稳定（随机分组前 24 小时内需要静脉利尿剂、正性肌力药或血压支持）被排除在外，所有患者均提供了书面知情同意书。

首次事件发生时间数据使用 Cox 模型进行分析，而复发事件（心力衰竭住院、门诊心力衰竭或心血管死亡）的发生时间和发生时间使用负二项式回归模型进行分析，该模型具有 Weibull 基线强度函数和治疗、心肌梗死类型、经皮冠状动脉介入治疗和地区作为因素。²分析是探索性的，P值未针对多重性进行调整。数据可应作者要求提供。

在先前报告的 CEC 裁定主要事件的首次发生时间分析中，随机分配到沙库巴曲/缬沙坦组的 2830 名患者中有 338 名 (11.9%) 发生了首次事件，373 名患者 (13.2%) 发生了首次事件2831 人随机分配给雷米普利（风险比，0.90 [95% CI，0.78–1.04]，P = 0.17）。²之前报告了终点组成部分之间的比较。^2个

在随机分配到沙库巴曲/缬沙坦组的患者中，有 443 名患者 (15.7%) 首次报告研究者心血管死亡、因心力衰竭住院或门诊心力衰竭，而在随机分配到雷米普利组的患者中，有 516 名患者（ 18.2%）与这些事件之一（风险比，0.85 [95% CI，0.75–0.96]，P =0.01；图 [A]）。

数字。 临床结果。显示的是研究者报告的主要复合结果的 Kaplan-Meier 曲线，分析为首次事件的时间 ( A )，以及总（首次和复发）临床终点委员会 (CEC) 裁定的 Nelson-Aalen 事件曲线 ( B )和研究者报告的 ( C ) 主要事件包括心力衰竭住院总次数、症状性心力衰竭门诊总发作次数和心血管死亡次数。一种，研究者报告的首次事件发生时间：沙库巴曲/缬沙坦组 443 名患者 (15.7%)（心血管死亡，115 名；住院，231 名；门诊，97 名）和雷米普利组 516 名患者 (18.2%)（心血管死亡，111 人；住院，264 人；门诊，141 人）。B，CEC 判定的主要事件总数：沙库巴曲/缬沙坦组 452 起主要事件（240 起住院、44 起门诊和 168 起心血管死亡）和雷米普利组 539 起事件（286 起住院、62 起门诊和 191 起心血管死亡） . C, 研究者报告的主要事件总数：沙库巴曲/缬沙坦组 672 起主要事件（395 次住院、122 次门诊和 155 例心血管死亡）和雷米普利组 802 次事件（447 次住院、176 次门诊和 179 例心血管死亡） .

在沙库巴曲/缬沙坦组中，338 名患者共发生 452 起经 CEC 判定的主要事件，至少有 1 起此类事件，而雷米普利组的 373 名患者共发生 539 起事件（率比，0.79 [95%] CI, 0.65–0.97], P =0.02；图 [B])。沙库巴曲缬沙坦组 199 人中有 17 人 (8.5%) 和雷米普利组 234 人中有 16 人 (6.8%) 在 CEC 确认的心力衰竭事件（住院或门诊发作）后使用开放标签沙库巴曲/缬沙坦。

在沙库巴曲/缬沙坦组中，443 名患者中有 672 起研究者报告的总事件，其中至少有 1 起事件，而在雷米普利组中，516 名患者中有 802 起总事件（率比，0.79 [95% CI， 0.67–0.93]，P = 0.004；图 [C])。沙库巴曲缬沙坦组 327 人中有 23 人 (7.0%) 和雷米普利组 400 人中有 28 人 (7.0%) 在研究者报告心力衰竭后使用开放标签沙库巴曲/缬沙坦。

当一项主要随机试验的结果未能达到显着性水平以声明其主要目标存在差异时，额外的分析可能会提供有用的信息。提出问题是为了更好地理解（尽管不会改变）中性结果。^3个凭借足够的统计功效并且在试验设计或实施方面没有重大缺陷，PARADISE-MI 的这项回顾性探索侧重于主要结果的选择。尽管复发事件分析可以更全面地评估不良患者和心力衰竭的经济影响，但仍选择首次事件发生时间分析，因为一旦 PARADISE-MI 中的患者出现心力衰竭，开放标签沙库巴曲/缬沙坦的使用预计会混淆重复事件分析。回顾过去，在我们的试验进行期间，在出现症状性心力衰竭的患者中，开放标签沙库巴曲/缬沙坦的使用最少（<10%），

虽然这些探索性分析不会改变急性心肌梗死后使用的主要中性发现，但使用 CEC 裁决和研究者报告得出的复发性心力衰竭事件减少的一致发现为已经表明的血管紧张素转换酶替代提供了更多支持信息一旦临床转变为有症状的心力衰竭，就使用 sacubitril/valsartan 抑制剂。^4,5

PARADISE-MI（前瞻性 ARNI 与 ACE 抑制剂试验以确定在减少 MI 后心力衰竭事件方面的优势）由诺华公司赞助。

披露Pfeffer 博士报告了诺华公司通过布莱根妇女医院提供的研究经费支持；来自阿斯利康、勃林格殷格翰和礼来联盟、Corvidia、DalCor、葛兰素史克、国家心脏、肺和血液研究所 (NHLBI) CONNECT（主协议委员会）、诺华、诺和诺德、Peerbridge 和赛诺菲的咨询费；并拥有 DalCor 的股权。Claggett 博士报告了 Amgen Inc、Beren Therapeutics、Myokardia、Novartis 和 Peerbridge Health 的咨询费。刘易斯博士报告了诺华公司的研究资助支持；Amgen、Dal-Cor 和 Merck 的咨询费。Granger 博士报告了来自 Novartis、AstraZeneca、Bayer Healthcare Pharmaceuticals、Boehringer Ingelheim、Bristol-Myers Squibb、Correvio、Daiichi Sankyo Co.、Janssen Global Services 和 Pfizer Inc. 的咨询费和/或研究支持。Køber 博士报告了勃林格殷格翰的咨询费。Maggioni 博士报告了来自诺华公司的咨询费。Mann 博士报告了来自 Bristol-Myers Squibb、Cardurion、LivaNova USA Inc、Novartis、Novo Nordisk 和 Renovacor 的咨询费。McMurray 博士报告了他的雇主格拉斯哥大学从诺华公司提供的研究资助支持。Rouleau 博士报告了来自 Novartis Pharma、AstraZeneca、Bayer、Bristol-Myers Squibb 和 Sanofi-Aventis US, LLC 的咨询费。Solomon 博士报告了来自 Actelion Pharmaceuticals、Alnylam Pharmaceuticals、Amgen、AstraZeneca、Bayer、Bellepheron、Bristol-Myers Squibb、Celladon Corporation、Cytokinetics、Eidos、Eli Lilly and Company、GlaxoSmithKline、Ionis、Mesoblast、Myokardia 的布莱根妇女医院的研究资助支持, Neurotronik, Novartis, Novo Nordisk, Respicardia Inc, Sanofi Pasteur Inc, 和 Theracos；Abbott Vascular、Action Medical Research、Akros、Alnylam Pharmaceuticals、American Regent、Amgen、Arena、AstraZeneca、Bayer、Boehringer Ingelheim、Bristol-Myers Squibb Eli Lilly and Company、Cardiac Dimensions、Cardior、Cardurion、CellProThera、Cytokinetics、Daiichi 的咨询费Sankyo Company、Genomics、GlaxoSmithKline、Johnson and Johnson、Merck、Moderna、Myokardia、Novartis、Quantum、Roche Diagnostics Corporation、Sanofi Pasteur Inc、Sarepta Therapeutics Inc、Tenaya、Theracos、Tremeau以及Dinaqor的股权；其他：科维亚医疗。Steg 博士报告了拜耳的研究资助；Amarin Pharma Inc、Amgen、AstraZeneca、Bayer、Boehringer Ingelheim、Bristol-Myers Squibb、Janssen Pharmaceuticals、Novartis、Novo Nordisk、Regeneron Pharmaceuticals、Sanofi US Services Inc 和 Servier 的咨询费。Berwanger 博士报告了来自 Amgen、AstraZeneca、Bayer、Boehringer Ingelheim、Bristol-Myers Squibb、Novartis Pharma 和 Pfizer 的咨询费。Cikes 博士报告了雅培和诺华的研究支持；机构、组织或其他经济利益来自：雅培、GE Healthcare、辉瑞和 Teva Pharmaceutical Industries。De Pasquale 博士报告了来自诺华公司的咨询费以及来自诺华公司的机构、组织或其他经济利益。Landmesser 博士报告了安进、拜耳和诺华的研究支持；Bayer、Boehringer Ingelheim、Daiichi Sankyo Company、Novartis、Pfizer 和 Sanofi Pasteur Inc. 的咨询费。阿斯利康、勃林格殷格翰、诺华、诺和诺德、SQ Innovation 的咨询费，和 Vifor 制药公司。Senni 博士报告了来自雅培基金、拜耳、拜耳医疗保健、诺华、默克和 Vifor Pharma 的咨询费。van d 博士报告了阿斯利康、辉瑞和 Vifor Pharma 的研究支持。Lefkowitz、Zhou 和 Wang 博士是诺华公司的全职员工。Braunwald 博士报告了阿斯利康、第一三共、默克和诺华通过布莱根妇女医院提供的研究资助支持；为 Amgen、Boehringer-Ingelheim/Lilly、Cardurion、MyoKardia、NovoNordisk 和 Verve 提供咨询服务。其他作者报告没有冲突。Braunwald 博士报告了阿斯利康、第一三共、默克和诺华通过布莱根妇女医院提供的研究资助支持；为 Amgen、Boehringer-Ingelheim/Lilly、Cardurion、MyoKardia、NovoNordisk 和 Verve 提供咨询服务。其他作者报告没有冲突。Braunwald 博士报告了阿斯利康、第一三共、默克和诺华通过布莱根妇女医院提供的研究资助支持；为 Amgen、Boehringer-Ingelheim/Lilly、Cardurion、MyoKardia、NovoNordisk 和 Verve 提供咨询服务。其他作者报告没有冲突。

这份手稿已发送给客座编辑 Faiez Zannad，以供专家审阅、编辑决定和最终处理。

有关资金来源和披露，请参见第 89 页。

注册：网址：https://www.clinicaltrials.gov；唯一标识符：NCT02924727。