The traditional Chinese medicine Qiliqiangxin in heart failure with reduced ejection fraction: a randomized, double-blind, placebo-controlled trial

Study design and setting

The executive committee designed and oversaw the conduct and analysis of the trial in collaboration with the sponsor, Shijiazhuang Yiling Pharmaceutical. The safety of patients in the trial was overseen by an independent data safety monitoring and clinical event adjudication committee (for a complete list of the committee members see Supplementary Information, ‘QUEST committees and investigators’).

This was a randomized, double-blind, placebo-controlled, parallel-group, event-driven, multicenter clinical study. The trial design was conducted and is reported in accordance with the protocol and statistical analysis plan (SAP), which are available with the full text in the supplementary Information (Supplementary Information, ‘Protocol and statistical analysis plan’)^30,31. The study adhered to the CONSORT guidelines. The protocol was reviewed and approved by the independent ethics committee of the First Affiliated Hospital of Nanjing Medical University (approved no. of ethics committee: 2018-SR-275) and the ethics committee of each participating study center. The trial was registered at registration no. ChiCTR1900021929 (registration date: 16 March 2019).

Data were collected using an electronic data capture application and Epidata v.3.1 software and managed by the independent statistics committee in strict accordance with a predefined SAP. The analyses were conducted by independent statisticians from the Peking University Clinical Research Institute. The first draft of the manuscript was prepared by the first author, who had unrestricted access to the data, and was reviewed and edited by all authors. All authors made the decision to submit the manuscript for publication and assume responsibility for the accuracy and completeness of the analysis.

Participants

The enrollment period was from 24 May 2019 to 24 May 2021. The eligibility requirements at screening included an age of at least 18 years, left ventricular ejection fraction (LVEF) of 40% or less, NYHA functional grading of II to III and stable clinical symptoms. Patients diagnosed as grade IV within 2 weeks before enrollment were also included in the study. Patients were required to have a plasma NT-proBNP level of ≥450 pg ml⁻¹.

All patients (in both the QXQL and placebo groups) were required to receive standard heart failure drug therapy following the guidelines for the diagnosis and treatment of heart failure in China³², including an ACEi, ARB or ARNi, a β-blocker and an MCA. The optimal therapeutic dose of these drugs was required, except in the case of contraindication or intolerance.

The detailed inclusion and exclusion criteria were as follows. Inclusion criteria: (1) signed informed consent; (2) age of ≥18 years at the time of consent; (3) established documented diagnosis of heart failure for at least 3 months according to the Chinese heart failure diagnosis and treatment guidelines issued by the Chinese Medical Association Cardiovascular Branch; (4) LVEF of ≤ 40% (by echocardiogram, radionuclide imaging, ventriculogram, contrast angiography or cardiac magnetic resonance imaging); (5) NYHA cardiac functional grading of II to III, with stable clinical symptoms, or diagnosis as grade IV within 2 weeks before enrollment; (6) serum NT-proBNP level of ≥450 pg ml⁻¹; (7) receipt of a standardized baseline treatment regimen without dose adjustment given intravenously for at least 2 weeks before enrollment; and (8) no use of another TCM medicineor herbs having the same contents as QLQX, such as Danshen and Tongxinluo capsules. Exclusion criteria: (1) heart failure caused by valvular disease, congenital heart disease, pericardial disease, arrhythmia or noncardiaogenic disease or caused by vital organ failure (such as renal failure, hepatic failure, etc.), right-sided heart failure caused by pulmonary or other definite causes or acute heart failure; (2) plans to undergo coronary revascularization (percutaneous coronary intervention or coronary artery bypass grafting) or cardiac synchronization therapy after randomization or receipt of cardiac resynchronization therapy before enrollment; (3) any condition other than a cardiovascular disease, including but not limited to, malignant tumor, severe mental illness, hematopoietic diseases, neuroendocrine system disease, liver transaminase and alkaline phosphatase levels more than three times the upper limit of normal, abnormal renal function, a serum creatinine level of >2 mg dl⁻¹ (176.82 μmol l⁻¹) and a potassium level of >5.5 mmol l⁻¹; (4) left ventricular outflow tract obstruction, myocarditis, aortic aneurysm, aortic dissection or obvious hemodynamic changes caused by an unrepaired valve; (5) cardiogenic shock, uncontrollable malignant arrhythmia, sinus or atrioventricular block at second degree, type II or above, without pacemaker treatment, progressive unstable angina pectoris or acute myocardial infarction; (6) uncontrolled hypertension, defined as a systolic blood pressure of ≥180 mmHg and/or a diastolic blood pressure of ≥110 mmHg, or a systolic blood pressure of <90 mmHg and/or a diastolic blood pressure of <60 mmHg; (7) participation in another clinical study with an investigative product during the month before enrollment; (8) women of child-bearing potential (that is, those who were not chemically or surgically sterilized or who were not postmenopausal) who were not willing to use a medically accepted method of contraception that was considered reliable in the judgment of the investigator, from the time of signing the informed consent to the end of the study and four weeks thereafter, women who had a positive pregnancy test at enrollment or randomization, or women who were breast-feeding; (9) an allergic constitution (known to be allergic to the research drug); and (10) inability of the patient, in the opinion of the investigator, to understand and/or comply with study medications or procedures or any conditions that might render the patient unable to complete the study.

Investigational products

QLQX has 11 components. The proportions of these when preparing 1,000 capsules are as follows: astragali radix, 450 g; ginseng radix et rhizoma, 225 g; aconiti lateralis radix praeparata, 112.5 g; Salvia miltiorrhiza radix et rhizoma, 225 g; Descurainiae semen, 150 g; alismatis rhizoma, 225 g; polygonati odorati rhizoma, 75 g; cinnamomi ramulus, 90 g; carthami flos, 90 g; Periplocae cortex, 180 g; citri reticulatae pericarpium, 75 g.

UPLC fingerprint analysis combined with chemometric methods was applied to evaluate the differences and similarities in the chemical constituents of QLQX capsules from the ten batches, which showed good consistency in preparation (Extended Data Fig. 2 and Extended Data Table 5).

To mimic the experimental drug, scorch-fried medicated leaven and caramel coloring were used in the placebo. The matching placebo capsules had identical color, weight, size, smell, specification, property of contents, labels and packaging as the QLQX capsules to ensure blinding during the study. Shijiazhuang Yiling Pharmaceutical (Shijazhuang, People’s Republic of China) provided the investigative products (QLQX capsule and the matching placebo capsules) for this research. An independent third party conducted blind coding of the experimental drug and placebo, rendering them virtually indistinguishable to participants and investigators.

The investigative product was manufactured consistently to ensure the quality and safety of the drug and strictly adhered to the good manufacturing practices of Chinese national drug production. Details of the manufacturing process, product controls and composition are presented in chemistry manufacturing and control (CMC) reports (Supplementary Information, ‘CMC report’).

Randomization and masking

Participants were randomly assigned (1:1) to receive QLQX or placebo in addition to the established heart failure medication regimen. This study adopted a block randomization method, and the participants were randomly grouped and managed through the randomization and trial supply management (RTSM) system. Randomization and drug numbers were generated by an independent statistician from the Peking University Clinical Institute using SAS software and were integrated into the RTSM system. The allocation list was stored in the RTSM system and was not available to any member of the research team. At each participating hospital, patients who provided written informed consent and met the study criteria were randomized by investigators who obtained the assigned treatment and code number from the RTSM system. Each package was labeled with a unique code number that was used to assign treatment to the participants but did not indicate treatment allocation to the investigator. After the database had been locked and the statistical analysis plan had been finalized, unblinding was conducted via the RTSM system.

Study procedures

All patients provided written informed consent and entered a 14-day screening period, during which the trial inclusion and exclusion criteria were checked and baseline information was gathered. After screening, patients were randomly assigned to receive either QLQX (0.3-g capsules, four capsules three times daily) or matching placebo (0.3-g capsules, four capsules three times daily). A web-based response system was used to determine treatment assignment. After randomization, eligible patients were evaluated at 1 and 3 months and every other 3 months after that. Patients were free to discontinue treatment for any reason, and the reasons for withdrawal were recorded in the case report form. Dose reduction (to 2−3 capsules three times daily of QLQX or placebo) or temporary discontinuation was permitted in cases of adverse complications, with a subsequent increase in dose or resumption of treatment, if possible.

The recruitment period lasted for 24 months. The shortest follow-up period was 12 months, and the median duration of follow-up duration was 18.3 months (IQR, 14.3 to 23.5 months).

Study outcomes

The primary outcome was the time to the first major adverse cardiovascular event, which was defined as either cardiovascular death or first HHF occurring within the follow-up period. An episode of worsening heart failure was either an unplanned hospitalization with objective evidence of an exacerbation of heart failure (as determined through clinical examination and/or laboratory evidence) or an urgent visit resulting in intravenous therapy for heart failure.

Secondary outcome measures included all-cause mortality, secondary endpoint events (treatment termination due to worsening heart failure, successful resuscitation after cardiac arrest, malignant arrhythmia, nonfatal stroke), cardiovascular death and first HHF in patients with ischemic etiology within the follow-up period and serum NT-proBNP levels at 3 months. All outcomes were adjudicated by the Clinical Event Adjudication Committee, which was unaware of trial group assignment, according to prespecified criteria. The definitions of endpoints and adverse events, as well as the procedures applicable to the Clinical Event Adjudication Committee, are described in the supplementary protocol (Supplementary Information, ‘Protocol’).

The prespecified safety analyses included serious adverse events, adverse events associated with discontinuation of a trial drug, adverse events of interest and laboratory findings of note. Data on other adverse events were routinely collected at each follow-up examination or unexpected visit.

Statistical analysis

The sample size was determined to provide adequate power to assess the outcomes of cardiovascular death and first HHF. We estimated that the annual rate of the composite endpoint would be 25% in patients receiving placebo and 20% in the QLQX group within the 12−36 months of follow-up.

The error was controlled at an overall two-sided α level of 0.05 and β level of 0.20 with two scheduled internal interim efficacy analyses, and the statistical stopping guideline for compelling benefit required a one-sided nominal P value <0.0001 at the first analysis and <0.00605 at the second analysis in favor of QLQX as the primary endpoint. On the basis of these calculations, it was estimated that approximately 3,080 patients would need to be included to provide the required number of 620 expected composite primary end point events, with an anticipated recruitment period of 24 months.

Only one interim analysis was performed when two-thirds of the primary endpoint events had occurred due to the pandemic. Termination for futility was set to be triggered if the conditional power was below 20%, whereas termination for efficacy would be triggered if the interim analysis yielded a P value of <0.00605. The significance of the two-sided α level for the multiple comparisons across the primary outcomes was 0.04628, with one interim efficacy analysis taken into account (with consumed two-sided α₁ = 0.012096).

Full analysis set (FAS) refers to the dataset that includes all patients who underwent randomization and received at least one dose of the study drug, with minimal and reasonable exclusion of individuals adhering to the intention-to-treat principle. The PPS was a subset of the FAS in which individuals had optimal adherence to the protocol with drug administration of at least 80% and no significant protocol violations. Protocol violation refers to (1) important violations of inclusion criteria; (2) individuals who did not receive the study drug; and (3) lack of post-randomization observation data. The safety set (SS) included all randomized participants who received at least one treatment and underwent safety evaluations.

Baseline characteristics are summarized as the mean + s.d., median and IQR or percentages. The baseline covariates were subject to a small amount of missing data that were assumed to be randomly missing. The comparability of characteristics between study groups was assessed using Student’s t-test for continuous variables and the Wilcoxon test for categorical variables. Intergroup comparisons were made using the chi-square test for categorical variables and Student’s t-test or the Wilcoxon test for continuous variables. Intragroup comparisons were conducted using the Wilcoxon signed-rank test for categorical variables. Time-to-event data were evaluated using Kaplan–Meier estimates and Cox proportional-hazards models, with the trial site as a random effect, to estimate the HR and 95% CI for the primary and secondary outcomes. Cumulative primary endpoint curves were constructed using Kaplan−Meier methods, and differences between curves were tested using the log-rank method.

We employed post hoc analysis using the Fine−Gray and cause-specific hazards competing risk-adjusted model, thereby enhancing the robustness and reliability of our findings. The model accounted for the competing risk of noncardiovascular mortality. We also assessed the consistency of the treatment effect among the prespecified subgroup variables of interest, which was verified using the Instrument for assessing the Credibility of Effect Modification Analyses (ICEMAN)³³. Sensitivity analyses were not prespecified and were conducted to assess the robustness of the primary findings by (1) censoring at follow-up of 12-months; (2) complete case analysis; and (3) using the PPS outcomes.

Statistical analyses were conducted using SAS v.9.4 software. Two-sided P < 0.05 was considered statistically significant.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.