Autologous atrial appendage micrografts transplanted during coronary artery bypass surgery: design of the AAMS2 randomized, double-blinded, and placebo-controlled trial

Abstract

Background

The AAMS open-label clinical study demonstrated the safety and feasibility of epicardial transplantation of autologous right atrial appendage micrografts (AAMs) during coronary artery bypass grafting (CABG) surgery. The study also provided the first indications of therapeutic efficacy of the AAMs, as delivered within an extracellular matrix patch, to reduce ischemic scar and increase viable ventricular wall thickness. To further evaluate the initial beneficial effects observed in the AAMS study, we designed the randomized, double-blinded, and placebo-controlled AAMS2 trial. Focusing on patients with ischemic heart disease (IHD) and myocardial scar, the AAMS2 trial aims to generate state-of-the-art structural and functional imaging data of the myocardium treated with an AAMs-patch during CABG.

Methods

The AAMS2 trial recruits IHD patients who are set to undergo non-urgent CABG and present with an ischemic myocardial scar in preoperative cardiac magnetic resonance imaging (CMRI) with late gadolinium enhancement. Patients are randomized (1:1) to receive a collagen-based matrix patch (Hemopatch®), with or without AAMs, epicardially onto the scar border. The primary endpoint, assessed by CMRI preoperatively and at 6 months post-operative follow-up, focuses on the left ventricle scar mass. The secondary endpoints center on the change in scar mass by the AAMs-patch site and evaluation of therapy safety and feasibility as well as its effects on myocardial structure and function by echocardiography. Change in blood N-terminal-pro-BNP levels in the timeframe is the co-primary endpoint.

Discussion

Data from the AAMS2 trial provides the first randomized, blinded, and placebo-controlled evaluation of efficacy on epicardial AAMs transplantation for ischemic myocardial scar. This data will pave the road towards rational design of larger AAMs therapeutic efficacy-addressing trial(s).

Trial registration

ClinicalTrials.gov, NCT05632432, registered 30 November 2022, https://clinicaltrials.gov/study/NCT05632432.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s13063-025-09379-4.

Introduction

During myocardial infarction (MI), timely revascularization limits the injury and recovers the stunned myocardium [1]. However, the activation of innate responses to damage, i.e., scarring and imbalanced compensatory mechanisms, still unfortunately often drives the disease towards heart failure (HF) [2, 3]. Alarmingly, on top of its rising prevalence in the aging population, ischemic HF is increasingly affecting individuals with obesity, causing a snowballing effect on cardiovascular disease burden that is especially dominant in low- and middle-income countries [4, 5].

Unfortunately, no therapy, excluding heart transplantation, exists to cure the ischemic myocardial damages. These damages and secondary adverse remodelling sustain the disease and its progression in many of those who resist even the best contemporary care, that is, robust revascularization and modern pharmacotherapy [6]. To bridge this gap in therapy, reactivation of myocardial regenerative processes has been explored for decades [7]. This rationale stems from findings that zebrafish and axolotls regenerate myocardium throughout their lifespan [8], while such endogenous healing in mice is limited to the first postnatal week [9]. Notably, however, by inhibiting fatty acid oxidation in cardiomyocyte mitochondria, the dormant regenerative healing program was recently reactivated for a marked repair of adult mice’s heart [10]. Evidence on existing cardiac regenerative capacity has been reported also in humans. Namely, Bergmann et al. measured the incorporation of ¹⁴C, mostly derived much from the Cold War nuclear tests, in human myocardia and found the cardiomyocytes to undergo slow, yet lifelong, turnover [11]. In 2016, Haubner et al. reported a full cardiac recovery of a human neonate suffering a massive perinatal MI with cardiogenic shock [12]. These data highlight human myocardial regeneration to be possible but endogenously tightly restricted. A recent review by Koopmans and van Rooij explores the gatekeeping molecular pathways and mechanisms of cardiomyocyte proliferation [13].

Massive efforts have been dedicated to translating the phenomenon of myocardial regeneration from experimental research to clinical application. Unfortunately, along this translational path, successes have remained scarce [7]. The causes for such results span the hardships of modeling the clinical disease [8], pharmacokinetic challenges of injection-based therapy delivery [14], and the allogenic origin of the cells investigated [7], stressing the need to evaluate alternative therapy approaches and more straightforward protocols [15].

Cells and micrografts of the atrial appendages have demonstrated regenerative properties to drive cardiac repair [16–19]. For autologous therapy, parts of the right atrial appendage (RAA) can be readily and safely harvested during coronary artery bypass grafting (CABG) surgery while ensuring RAA closure at the end of the operation [20, 21]. We have already demonstrated that both RAA processing to atrial appendage micrografts (AAMs) and the epicardial transplantation of AAMs are safe and feasible to perform during CABG surgery (Fig. 1) [20, 21]. After processing of a RAA biopsy to AAMs, over 90% of the isolated cells were viable at a yield of 9.76 × 10⁶ ± 0.53 cells for each gram of tissue (n = 11) [21]. The viability of human AAMs was further demonstrated in ex vivo cultures [21]. In mice, 2 months after MI and AAMs transplantation, histological analysis demonstrated their epicardial persistence [16]. In pigs, their immunosuppressive potency was noted [22]. Interestingly, as measured 6 months after CABG by cardiac magnetic resonance imaging (CMRI) with late gadolinium enhancement, our clinical open-label study demonstrated a significant increase in the viable myocardium thickness at the documented AAMs-patch transplantation site (+ 1.0 mm [0.2–1.3 mm]) as compared to controls (− 1.4 mm [− 1.7 to 0.0 mm]) [20, 21]. Notably, by this time, the AAMs-patch was intractable with the CMRI. An indicative trend for scar mass reduction was also obtained. Together, these results support the premise that similar cardioreparative processes, as identified in a mice model of ischemic HF [16], are active when AAMs therapy is delivered in the clinical setting.

Fig. 1.

The results of the clinical AAMs-patch open-label pilot study. The preclinical results of an epicardial AAMs-patch transplantation are presented in Fig. 4. It prompted an open-label study, focusing on ischemic HF patients with a myocardial scar in preoperative CMRI, which supported the feasibility and safety of AAMs-patch transplantation as an intraoperative therapy adjuvant to CABG in a clinical setting [20]. This study revealed the median time for the AAMs-patch setup to be 33 min, while only minutes were needed for the attachment onto the epicardium. The 6-month follow-up found good tolerability of the AAMs-patch. Moreover, in the 6-month follow-up CMRI records, the AAMs-patch was intractable; however, the transplant area showed significant increase in the live ventricular wall thickness as compared to the baseline [20]. An indicative trend for reduction in scar mass by the site was also found. The results support a larger randomized, double-blinded, and placebo-controlled clinical trial described herein. AAMs, atrial appendage micrografts; AAMs-patch, collagen-based matrix patch encasing atrial appendage micrografts; CABG, coronary artery bypass grafting; CMRI, cardiac magnetic resonance imaging with late gadolinium enhancement; LVEF, left ventricular ejection fraction. The CMRI results are reprinted as minutely modified from Frontiers in Cardiovascular Medicine, 2021 Nummi A, Mulari S, Stewart JA, et al. Epicardial Transplantation of Autologous Cardiac Micrografts During Coronary Artery Bypass Surgery (2021) [20] with permission from publisher under the Creative Commons Attribution License (CC BY 4.0)

This AAMS2 randomized, double-blinded, and placebo-controlled trial is designed to provide the first clinical evaluation of AAMs’ therapeutic efficacy. The trial recruits a total of 50 patients scheduled for elective CABG surgery and presenting with an ischemic myocardial scar as identified in preoperative CMRI. The patients are randomized to AAMS2 (collagen-based patch + AAMs) and control (only patch) groups. To minimize variation in the myocardial tissue state, patients with an MI in the last 30 days prior to CABG will be excluded. The primary endpoints are the changes in the scar area and mass within the left ventricle, as determined by CMRI with late gadolinium enhancement prior to and 6 months after CABG. Change in the scar mass of the ventricle at the AAMs-patch site is the secondary endpoint. The change in N-terminal pro-brain natriuretic peptide (NT-pro-BNP) levels in the study timeframe is a trial co-primary endpoint.

Methods

The AAMS2 randomized double-blinded, and placebo-controlled trial addresses the effect of AAMs-patch on myocardial ischemic scar. It assesses cardiac function by CMRI when delivered epicardially on the scar border at the end of CABG surgery. The trial’s estimated final enrolment is 50 participants in a 1:1 group allocation ratio. The trial protocol has been approved by the ethics review board at Helsinki University Hospital (HUS; Dnr. HUS/12322/2022) and the Finnish Medicines Agency Fimea (FIMEA; Dnr. FIMEA/2023/004090). The trial is registered at ClinicalTrials.gov (NCT05632432), and its outline is shown in Fig. 2. The trial conforms to the SPIRIT checklist for randomized clinical trials. At the publication of this article, the trial has recruited 10 participants.

Fig. 2.

Outline of the AAMS2 trial. The main inclusion criteria (“ECHO”), screening failure criteria (“CMRI”), and the primary (“CMRI”) and the co-primary (“NT-proBNP”) endpoint of the trial are shown in red. In addition to transthoracic echocardiography (“ECHO”), which is performed: (i) at recruitment, (ii) preoperatively, and postoperatively at (iii) hospital discharge and (iv) at 3-month follow-up, transesophageal echocardiography is also done by the perfusion anesthesiologist at the beginning of CABG surgery to assess RAA anatomy and any presence of sludge. Clinical metadata (“OTHER”) include the recording of Framingham cardiovascular risk factors, electrocardiogram, medication with changes, MACCE, quality-of-life numeration with the SF-36 questionnaire, and a survey of dyspnea and angina pectoris symptoms with NYHA and CCS classifications. The RNA-stabilized study blood samples, focusing on adenosine-based epitranscriptomic profile characterization, are collected as previously published [23]. Abbreviations: AAMs, atrial appendage micrografts; AAMs-patch, collagen-based matrix patch encasing atrial appendage micrografts; CABG, coronary artery bypass grafting; CCS, Canadian Cardiovascular Society (angina pectoris grading classification); CMRI, cardiac magnetic resonance imaging with late gadolinium enhancement; LVEF, left ventricular ejection fraction; NT-proBNP, N-terminal pro-B-type natriuretic peptide; NYHA, New York Heart Association (dyspnea grading classification); MACCE, major adverse cardiac and cerebrovascular events; RAA, right atrial appendage; SF-36, 36-item short form survey (standardized questionnaire for assessment of overall quality of life); 6MWT, 6-min walking test

Power analysis and endpoints

The trial’s primary and secondary endpoints are listed in Table 1. They are based on power calculations utilizing the only data available on the method from our previous open-label study [20]. The primary endpoint is the change in myocardial scar tissue mass (g) and area (%) within the whole left ventricle, while the analogous effects on the scar at the AAMs-patch transplantation site from preoperative to 6-month postoperative CMRI recordings are, along with cardiac systolic function, and feasibility and safety of the AAMs-patch transplantation, the secondary endpoints. The change in the NT-proBNP levels across the trial timeframe is a co-primary endpoint.

Table 1.

The AAMS2 trial endpoints

Primary endpoint
	- Change in the myocardial scar tissue within the left ventricle
	• Preoperative vs 6-month-follow-up 5SD CMRI for mass (g) and area (%)
Co-primary endpoint
	- Change in the plasma NT-proBNP levels across the trial period
Secondary endpoints
Efficacy	- Change in the myocardial scar tissue by the patch transplantation site
	• Preoperative vs 6-month-follow-up 5SD CMRI for mass (g) and area (%)
	- MACCE* during the whole trial period
	- Deaths and deaths due to primary cardiovascular cause
	• Death and cause (ICD-10) of death during the whole trial period
	- Measured by CMRI preoperatively and at 6-month-follow-up by the patch site
	• Change in the live left ventricular wall thickness
	• Change in viable left ventricular myocardium
	• Change in the local ventricular wall systolic and diastolic function
	• Change in left ventricular ejection fraction
	- Change in NYHA class at the 3- and 6-month vs. preoperative NYHA
	- Change in 6-min walk test (preoperative vs. 6-month-follow-up)
	- Changes in the quality of life (RAND36 preoperative vs. 6-month-follow-up)
Exploratory secondary endpoints
Safety	-Telemetric rhythm postoperatively
	• Incidence of VT, VF, atrial fibrillation/flutter, or other arrythmias
	- Days in need for invasive vasoactive medication postoperatively
	- In-hospital infections (transplant-related = 1; non-related = 2; no = 0)
	- Days in hospital
	- Anticipated SADE* (serious adverse device effects) during the trial - Other serious adverse events* and unanticipated SADE* during the trial
Efficacy	- Measured by TTE preoperatively and at 3-month-follow-up by the patch site
	• Changes in systolic and diastolic function of the ventricular wall • Changes in local myocardial strains
Feasibility	- Success in completing the delivery of the patch to the epicardium
	• 0 = success, 1 = no success
	- Waiting time for the ready micrograft transplant
	• Time from the ready AAMs-gel to the AAMs-patch transplantation
	- Waiting time in minutes for the heart
	• After all the anastomoses finished and before the patch transplanted
	- Closing the right atrial appendage
	• Closing the RAA after biopsy for preparing the AAMs-patch
	0 = no additional suturing needed, 1 = additional suturing needed

CMRI cardiac magnetic resonance imaging with late gadolinium enhancement, LVEF left ventricular ejection fraction, MACCE major adverse cardiac and cerebrovascular events, RAA right atrial appendage, RAND36 questionnaire for quality of life, TTE transthoracic echocardiography, VT/VF ventricular tachycardia/fibrillation, 5SD 5-standard deviation

^*See the “Adverse events” section for the adverse event definitions, including MACCE

The power analysis was carried out using SAS 9.4 TS Level 1M4 software (SAS Institute Inc., Cary, NC, USA), the POWER Procedure Wilcoxon-Mann–Whitney Test with the fixed scenario elements O’Brien-Castelloe approximation method and two-sided statistical evaluation. With a total sample size of 50 (two groups, group size 25, distribution 1:1), these parameters yield a power greater than 80% at an α-error level of 0.05. Figure 3 presents the power analysis output graphs. A detailed consideration of the nature of the power analysis is provided at the end of Discussion.

Fig. 3.

Power analysis for the AAMS2 trial. The power analysis is based on the unblinded and non-randomized data from the open-label safety and feasibility study of the perioperative AAMs-patch method (n = 6, AAMs-patch-treated; n = 5 controls without patch) [20]. As calculated by the POWER Procedure Wilcoxon-Mann–Whitney Test with fixed scenario elements O’Brien-Castelloe approximation method and two-sided statistical evaluation, the change (Δ) in infarction scar area (%), and mass (g), as evaluated from the 5SD CMRI imaging data preoperatively and 6 months postoperatively at the AAMs-patch transplantation site, the total sample size of 50 (two groups, group size 25, distribution 1:1), yields a power greater than 80% at an α level of 0.05. Analysis was carried out with SAS 9.4 TS Level 1M4 software (SAS Institute Inc., Cary, NC, USA). The indicative effect size on fibrosis is shown in Fig. 1 [20]. A trend for a decrease in the change of NT-proBNP levels was found in those patients receiving an AAMs-patch [20]. A Evaluation using change in infarction area percentage (%). B Evaluation using change in infarction area mass (g). C Evaluation using change in NT-pro-BNP circulatory concentrations (plasma sample analysis preoperative vs. 6 months postoperatively). The red line represents 80% power at a total sample size of 50. A detailed consideration of the nature of the power analysis is provided at the end of the Discussion. α, level of type I error (α-error level representing the proportional level for false-positive result assessment); CMRI, cardiac magnetic resonance imaging with late gadolinium enhancement; NT-proBNP, N-terminal pro-B-type natriuretic peptide; 5SD, 5-standard deviation

Patient selection and enrolment

The AAMS2 trial will recruit patients with ischemic heart disease (IHD) requiring surgical revascularization, and with a visible myocardial scar in preoperative CMRI. To standardize the myocardial tissue state, patients with an AMI within the last 30 days will be excluded.

Similar to our AAMs-patch feasibility pilot [21], and the AAMs-patch open-label study [20], for this AAMS2 trial the recruitment of patients, regardless of their gender, is carried out by an academic hospital cardiologist at either Helsinki University Central Hospital (Helsinki, Finland) or Oulu University Hospital (Oulu, Finland). The patients’ usual waiting time on the hospital’s elective surgery list ranges between 2 and 8 weeks, thus allowing any changes to medication, as made by the cardiologist on clinical grounds unrelated to the trial, to take effect before surgery. After CABG, the recruited patients are called for a clinical control visit (denoted as the 3-month follow-up) and a dedicated trial control visit (at 6–8 months postoperatively, denoted as the 6-month follow-up).

Patients meeting both the inclusion and exclusion criteria (Table 2) will be provided a consent form describing the trial and are provided with sufficient time and information to make an informed decision on participation. Before a subject undergoes any study procedure, an informed consent discussion will be conducted, and written informed consent will be obtained. After optimization of medications, if in preoperative CMRI visible scar cannot be identified or its anatomy is unsuitable for AAMs-patch transplantation, patients already recruited are excluded due to screening failure (Table 2). This also applies in the case that the left ventricular ejection fraction (LVEF) is < 25% in CMRI or logistical issues prevent it, typically done just days prior to planned CABG, to be performed before the CABG in adequate time. The trial will be conducted following the Declaration of Helsinki on Ethical Principles for Medical Research Involving Human Subjects [24]. The data collected in the trial will fulfil EU regulations for personal health data protection, including the General Data Protection Regulation (GDPR).

Table 2.

The AAMS2 trial criteria

Eligibility	Age: 18–75 years
	Sexes: All
	Healthy volunteers: not accepted
Inclusion	Informed consent obtained
	Ischemic heart disease requiring CABG for revascularization
	Clinical evidence of ischemic myocardial scar (e.g., history of LVEF between ≥ 15% and ≤ 40% at recruitment by echocardiography, OR known prior STEMI, OR pathological Q-waves on the electrocardiogram)
	NYHA Class II–IV heart failure symptoms
Exclusion	Heart failure due to left ventricular outflow tract obstruction
	Expected life expectancy < 1 year
	Acute myocardial infarction (AMI) within last 30 days
	History of life-threatening and likely repeating ventricular arrhythmia or resuscitation, or an implantable cardioverter defibrillator
	Stroke or other disabling condition in 3 months before screening
	Severe valve disease or scheduled valve surgery
	Renal dysfunction (GFR < 45 ml/min/1.73 m)
	Other major disease limiting life expectancy
	Contraindications for coronary angiogram or CMRI
	Allergy or hypersensitivity to the fibrin glue or to the Hemopatch®
	A part of any special patient group*
	Participation in some other clinical trial
	Screening failure in preoperative study CMRI (No visible scar, OR infarct scar position unsuitable for patch transplantation [i.e., scar within the interventricular septum], OR LVEF < 25%, OR the preoperative CMRI has not been performed prior to the scheduled CABG

CABG coronary artery bypass grafting, CMRI cardiac magnetic resonance imaging with late gadolinium enhancement, LVEF left ventricular ejection fraction

^*Pregnant, nursing, handicapped persons in an especially vulnerable position, or those in emergency situations

Baseline morbidity

A general baseline morbidity assessment will be carried out and involves a patient information system search for significant cardiovascular and non-cardiovascular comorbidities, electrocardiogram, structured patient interview and a case report form fill-out comprising dyspnea and angina pectoris symptom scaling, numeration of quality of life, traditional Framingham cardiovascular risk factors [25], supplemented with body mass index (BMI), personal history of MI, or family history of IHD, and medication list review. Data on patient medication will be curated and analyzed based on anatomical therapeutic chemical classification using the defined daily dose values enabling dosage comparisons. The CRF and trial checklists are provided as online-only supplementary files (Supplementary Files S1–9).

Randomization and blinding

Participants passing the screening failure are randomized to receive either standard CABG and collagen-based patch without the AAMs (control group) or CABG with an epicardial transplantation of an AAMs-patch (collagen-based patch + AAMs) after all the coronary anastomoses have been completed (the AAMs-patch group). Patients are randomized to the groups using sex-stratified block randomization via an openly available online tool at www.sealedenvelope.com with block sizes 2 and 4, and stratification according to sex (female, male). Randomization is carried out by the study nurse. In this trial, we randomize 50 participants that are treated during CABG either with Hemopatch® or the AAMs-patch. Specifically, the research nurse texts the Sealed Envelope service phone number “AAMS2” (the trial abbreviation), followed by an execution command “randomise”, and completed by the detailing patient pseudonym. Then, the research nurse receives the allocation information by a text message. This is done a day prior to each patient’s surgery to grant adequate time to organize practicalities required for the perioperative AAMs-patch assembly. The study nurse oversees the allocation in a double-blind manner, where the patient and the evaluating cardiologist(s), radiologist(s), and academic researchers remain blinded to the allocation. Given the nature of the treatment intervention (intraoperative AAMs preparation vs. no preparation), the operating cardiac surgeon or the study nurse cannot be blinded intraoperatively. However, the operating cardiac surgeon is blinded when planning and deciding the revascularization strategy. Moreover, all the CMRI and transthoracic echocardiography (TTE) measurements as well as laboratory analyses are done by persons blinded to the patients’ study group allocations. In this trial, if considered necessary to remove the epicardial patch material, it is done without consideration of the randomization. Hence, no emergency unblinding is required in this trial.

Preparation and administration of atrial appendage micrografts

A piece of the RAA is harvested at the beginning of cardiac surgery upon right atrial cannulation as a part of the heart-and-lung-machine setup. For the AAMs-patch group, the RAA tissue is then weighed and mechanically grinded to micrografts in the operating room by using the Rigeneracon blade (Rigenera-system, HBW s.r.l., Turin, Italy) as previously described [16, 20, 21]. As in the open-label safety and feasibility study [20], the targeted weight of the sample tissue for AAMs preparation is approximately 0.6–1.0 g. For the CABG control group, the tissue piece is stored as a sample (see the “RAA tissue samples” section). Dedicated CE-marked instrumentation kits to support tissue processing in the operating room are obtained from EpiHeart Oy (EpiHeart CMT Core kit, Helsinki, Finland).

After grinding the RAA with a Rigeneracon blade and subsequent centrifugation (5 min, 400 g) to pellet the AAMs, the cold cardioplegia supernatant is removed. The AAMs-pellet is collected with 0.4 mL of Tisseel (Baxter AG, Vienna, Austria) fibrinogen solution (diluted 1:1 in 0.9% NaCl). Then, the AAMs in fibrinogen are spread onto a cooled sterile metallic dish. The spread AAMs in fibrinogen are mixed in situ with 0.2 mL of Tisseel thrombin solution (diluted 1:30 in 0.9% NaCl). The AAMs–fibrinogen–thrombin mixture is allowed to undergo spontaneous gelling for at least 10–15 min. Then, the gelled AAMs–fibrin mixture is maintained cooled (+ 6– + 8 °C), covered, and sterile while waiting for transplantation. When all the anastomoses are ready, the AAMs–fibrin gel is gently lifted with a spatula and transplanted onto the dry matrix sheet (Hemopatch® Sealing Hemostat 45 mm × 45 mm; catalog ref. 1506256; Baxter International Inc., IL, USA), thus forming the AAMs-patch (Supplementary Video S1), just prior to transplantation by the surgeon. Then, the Hemopatch® edges are moistened with sodium bicarbonate solution (4.2–8.4%) to activate the polyethylene glycol-coating of the patch. Finally, to achieve proper epicardial adherence, the AAMs-patch with the moistened patch edges is transplanted onto the epicardium of the scarred border zone of the myocardium by using a dry gauze with uniform pressure applied for 2 min. Based on the CMRI data, the exact transplantation site is preoperatively designated to the ischemia-induced scar. In this trial, the AAMs patch shall not wait longer than 6 h prior to transplantation.

To standardize the effects of the patch material, the control group receives the patch without the AAMs. A photograph is taken to record the location of the transplanted patch. The patients are carefully monitored according to clinical routine after the operation in the intensive care unit (Table 2, Supplementary File S4). Tissue processing-supporting instrumentation (EpiHeart Oy, Helsinki, Finland) usability is recorded in detail by the research nurse, and the usability outcome is included as one of the AAMS2 trial’s secondary outcomes (Table 1).

Data collection

Clinical, laboratory, and drug treatment data are collected in the hospital electronic health records. Any visit related to the operation or their cardiovascular system condition, as well as drug treatment changes, are collected by the study investigators. This data is stored pseudonymized with the other data from the patient. In addition, the cardiovascular-related changes in medication with Anatomical Therapeutic Classification (ATC) codes and defined daily doses (DDDs) will be recorded for analysis to serve as source data for surrogate modelling of an improved (i.e., reduced medication or dosage) or worsened (added medication or increased dosage) disease state.

Concomitant care, possible harm, and trial adherence

No concomitant care or interventions are prohibited. In this trial, based on regulatory body and company evaluations, the use of Hemopatch® is on-label (CE-marked) in the AAMs-patch-treated cohort, whilst it is an off-label use in the control group, where no AAMs are given. Hence, any product liability for possible participant harm(s), which can be causally verified to the use of Hemopatch® sealing the AAMs micrografts epicardially, is on the manufacturer, Baxter International Inc., IL, USA. Also in the control cohort, where no AAMs are being used, the liability for compensation for any participant harm(s) that can be causally linked to the Hemopatch® product itself is on the manufacturer. If the possible participant harms cannot be causally linked to the product, the liability is on the respective hospital trial investigators, all of whom have extensive insurance covering compensation for any damages in accordance with the national Patient Injury Act of Finland during the treatment and examination of patients. In Finland, these processes are centrally governed and handled by the Patient Insurance Center.

The participant’s adherence is monitored by the research nurse by continuously updating an anonymized stepwise trial progression file. In case of participant- or investigator-derived deviation from the trial protocol, the participant is followed and examined per the trial protocol as extensively as possible after the deviation. In case of participant discontinuation, the study data accumulated until the date of discontinuation are included in the trial datasets and reported separately.

Adverse events

In this trial, adverse events are divided into (1) major adverse cardiac and cerebrovascular events (MACCE), (2) anticipated SADE (serious adverse device effects), (3) other SAE (serious adverse events), and (4) unanticipated SADE. All the events are continuously monitored and reported to regulatory bodies. In this trial, the MACCE definition comprises death (all-cause), MI, any acute coronary revascularization, or stroke. MI is defined according to the fourth universal definition of MI as either (1) perioperative myocardial injury (i.e., type 5 MI ≤ 48 h post-CABG, creatinine kinase muscle-brain isoenzymes [CK-MB] ≥ 10 times the upper reference) or (2) postoperative MI (i.e., an increase in the CK-MB or troponin concentration above the upper reference limit with ischemic symptoms or signs). Stroke is indicated by neurological deficits and confirmed by a neurologist on the basis of imaging modalities with lesion(s) concordant with the clinical presentation. New revascularizations span any postoperative coronary intervention.

As a part of our overall comprehensive safety evaluation of the method, anticipated SADE are events identified by us with a possible causal relationship to the AAMs-patch therapy. These include the following: (1) mediastinitis, (2) postoperative pericardial effusion requiring subxiphoidal drainage or resternotomy, (3) major bleeding (BARC classes 4–5) [26] from the RAA biopsy site, or (4) major postoperative arrhythmia (ventricular fibrillation or ventricular tachycardia over 30 s). Mediastinitis is diagnosed according to Centers for Disease Control and Prevention (CDC) guidelines [27]. The gathered data in mice [16], pigs [22], and humans [20] on the AAMs-patch method do not indicate heightened risk for these events.

Other SAE comprise any adverse event that has led to either death, life-threatening illness, (prolongation of) hospitalization, medical intervention to prevent life-threatening illness, or chronic disease. According to our risk assessment of the AAMs-patch therapy, these could include the following: myocarditis, pericardial effusion, HF exacerbation, resternotomy, atrial tachycardia or fibrillation, atrial flutter, transient ischemic attack, major bleeding (BARC 3–5) [26], acute kidney injury, or other hospitalization due to ischemic cause.

Blood samples

The AAMS2 trial assesses blood, plasma, and RAA tissue (see the “RAA tissue samples” section) samples for their RNA with a focus on their contained post-transcriptional, epitranscriptomic modifications, as described in the IHD-EPITRAN study design (www.ihd-epitran.com) [23]. These modifications, as recently reviewed comprehensively by us in the field of cardiovascular diseases [28], are emerging regulators of both cardiac disease and regeneration. The schedule of sample collection is summarized in Fig. 2.

Briefly, the study blood sample set contains tubes for RNA-stabilized TEMPUS™ whole blood (3 mL × 8), as well as EDTA-stabilized blood (9 mL × 3), from which RNA-stabilized plasma (900 µL × 10) and standard plasma (500 µL × 4–10) are separated and aliquoted for storage [23]. In addition, a sample for N-terminal pro-brain natriuretic peptide (NT-pro-BNP) measurement by the clinical laboratory (HUSLAB, Helsinki, Finland; NordLab, Oulu, Finland) is collected in the trial preoperatively and at both 3-month and 6-month clinical and study follow-ups, respectively.

Right atrial appendage (RAA) tissue samples

Since the RAA tissue sample removed during CABG in the control group is not used for producing AAMs, it is collected for analyses. The initial piece of RAA tissue will be divided into two pieces and stored in (1) RNAlater solution (AM7021, ThermoFisher Scientific Inc., Waltham, MA, USA) with an overnight incubation (4 °C) prior to storage (< − 70 °C), and (2) formalin (4%) for 2 weeks and then in ethanol (70%) to prevent overfixation (at 4 °C).

The RNA-stabilized RAA piece is used to profile the epitranscriptome as in the IHD-EPITRAN study [23]. Briefly, the RNAlater-stored (AM7021, ThermoFisher Scientific Inc., Waltham, MA, USA) RAA piece will be subjected to RNA extraction, fractionation, and sequencing, and bioinformatics. Overall, these produced epitranscriptomes are correlated with changes in repeated CMRI, TTE, and blood sampling results.

Echocardiography

For functional and anatomical insight into the participants’ cardiac status, the participants are assessed with TTE both pre- and postoperatively (Fig. 2). Postoperative TTE recordings are done at hospital discharge (approximately 1 week postoperatively) and at 3 months of follow-up. The recordings are performed with prespecified acquisition methods and imaging windows by a few designated cardiologists and include both anatomical and functional assessments of atria, valves, and ventricles. The presence or absence of pericardial effusion, thrombus, and aneurysm is recorded. The detailed protocol is provided as a Supplementary File S10. Also, a perfusion anesthesiologist will perform transesophageal echocardiography in the operating room during anesthesia to evaluate both left and right atrial appendages and atria for blood flow velocities, possible sludge, thrombus, and anatomy before CABG. The raw data will also be exported and stored for further state-of-the-art functional analyses (myocardial imaging), such as strain and strain rate measurements.

Late gadolinium enhancement cardiac magnetic resonance imaging (CMRI)

CMRI is performed preoperatively to evaluate cardiac function and myocardial anatomy, including ischemic and fibrotic areas. The preoperative CMRI excludes patients without fibrotic scar suitable for the patch transplantation or those with severe systolic left ventricular impairment and prone for postoperative complications (Table 2). Further, it guides the AAMs-patch transplantation site at the end of CABG surgery (see the “Randomization and blinding” section). Its repetition at the 6-month follow-up enables assessment of change in scar or vital myocardial tissue mass, firstly, in the left ventricle and, secondly, by the epicardial AAMs-patch transplantation site (see the “Power analysis and endpoints” section). Specifically, the whole body 1.5-T MRI scanner (In Helsinki, Siemens Sola or Avanto-fit, Siemens AG, Erlangen, Germany; in Oulu, Siemens Sola or Sola-fit, Siemens AG, Erlangen, Germany) is used for image acquisition. Cardiac structure and function are evaluated with a standardized CMRI protocol using electrocardiogram and respiratory gating. Short-axis cine images are used for left and right ventricular volumetric measurements. Myocardial contractility is evaluated using longitudinal, circumferential, and radial strain measurements from short- and long-axis cine images. Late gadolinium enhancement is used to measure infarction volume and mass using a 5-SD semiautomatic gain estimate, as previously suggested for semiautomatic thresholding for infarction detection [29]. Image post-processing is performed with Medis Suite software (Medis Medical Imaging Systems, Leiden, The Netherlands) with QMass and QStrain applications.

Quality of life assessment

Health-related quality of life (HRQoL) is assessed using the RAND36 36-Item Short Form Health Survey (SF-36) questionnaire [30]. The questionnaire is standardized with specified mean and standard deviation values for eight dimensions that range from physical functioning and subjective feelings of vitality and health to bodily pain. The obtained scores are compared for a Finnish cohort without any chronic disease. Also, a subjective symptom evaluation is performed for the two cardinal symptoms of IHD and HF, angina pectoris and exertional dyspnea, with standard classification systems as developed by the Canadian Cardiovascular Society (CCS) and the New York Heart Association (NYHA), respectively [31, 32].

Six-minute walking test (6MWT)

To measure the general physical capacity, a 6MWT is performed for all participants preoperatively and postoperatively at the 6-month follow-up [33]. The parameters included are the following: the actualized walking distance in meters, the predicted walking distance, oxygen saturation (baseline at rest and lowest during the test), and subjective related symptoms according to Borg’s scale (0–10). The test is scheduled just before CABG (Fig. 2) to minimize the effects of preoperatively initiated pharmacotherapies on exertional capacity prior to CABG and AAMs-patch transplantation.

Data confidentiality

In this trial, all collection of personal or other study information collected is carried out by healthcare professionals committed in verbatim to the utmost standards of data safety and confidentiality. The produced data are stored in the research registry in the University Hospitals’ and University of Helsinki’s safe network hard drives and on the servers of the Finnish IT Center for Science designed for sensitive data storage, all with an automated backup. The tailored data storage services—SD-Connect, SD-Desktop, and SD-Publish—provided by Finnish CSC–IT Center for Science (https://research.csc.fi/sensitive-data) are financially supported by Finland’s Ministry of Education and Culture. These services are designed to comply with the EU General Data Protection Regulation (GDPR), which is followed throughout the trial lifetime. Access to these registries is controlled via role-based access rights. Only those specified in the registry description approved by the HUS Ethics Committee can access the data therein.

The participants’ TTE and CMRI data are stored and accessed via software fulfilling the Hospitals’ data security guidelines. Case report formats are stored both physically in the safe Hospital premises with access control and electronically in the research registry. The pseudonymized sequencing datasets with metadata will be made available upon publication principally via the long-term sensitive data storage service SD-Publish, which appoints an independent data monitoring committee to evaluate and process all applications (in writing) by other scientists interested in re-using the deposited pseudonymized research data. The principal investigators of the trial will have access to the final trial dataset. Contractual agreements will not limit this right. The identificatory data and pseudonymized data will be stored for 15 years after the trial completion. Then, the pseudonymization codes and keys will be erased, and the anonymized data will be principally stored in SD-Apply or in another public repository. The anonymized group-level datasets of the trial results will be made available by submission(s) for publication(s) in esteemed journals of the field of cardiology and cardiac surgery.

Coordinating center and trial monitoring

University of Helsinki (UH) together with the Helsinki University Hospital (HUS) coordinate the trial and form the coordinating center. The responsibilities of the coordinating center are to secure data protection, hold periodic meetings on the trial progression and matters arising among the trial investigators, ensure timely communication of any possible changes in the criteria or trial status, coordinate reporting with the trial’s external regulatory bodies, realize the up-to-date documentation requirements to transfer the results from Oulu University Hospital to perform the final endpoint analyses. The coordinating center is also responsible for data synthesis, organizing the results data meeting discussions, and finally the publication of the results in the peer-reviewed journal(s) as open science articles. The scientists of the coordinating center meet every 2 months (i.e., UH and HUS) and weekly (i.e., HUS internal meetings, UH internal meetings). The AAMS2 trial is externally monitored by the Clinical Research Institute HUCH to ensure trial subjects’ rights, safety, and well-being. A detailed monitoring plan has been formulated and is in effect with the monitor and the research group. The trial monitor regularly audits the trial conduct and data registries for appropriateness and completeness twice a year. The researchers constituting the coordinating center meet twice a year with the trial monitor, thus constituting the trial steering committee activity in the trial. In addition, the AAMS2 trial has a Trial Monitoring Committee with national medical professionals from the field of cardiology, cardiac surgery, and nursing to evaluate the appropriateness of the trial in a periodical manner, as each 10 patients are recruited, from a focused medical viewpoint. As per local standards, trial size, and evaluated risk, this trial does not have a separate data monitoring committee nor the endpoint adjudication committee. The trial has no Stakeholder and Public Involvement Group (SPIG) meetings.

Interim analysis

If considered appropriate, an interim analysis can be performed when a significant proportion of the total recruited participants (i.e., 20–30 patients) have undergone the full protocol. This consideration is undertaken jointly with the Trial Monitoring Committee. All the trial investigators can access the interim analysis results. The principal investigator makes the final decision to terminate the trial.

Statistical analyses

Comparisons between groups will be performed with the Mann Whitney U test. Ordinal variables are tested with the Chi Square test. Multiple comparisons are corrected with the Bonferroni method; significant findings are further tested groupwise using the Mann Whitney U test. Quality of life data is presented as mean and analyzed with the independent samples t-test (two-sided). Analyses are performed with the IBM SPSS Statistics 27 program (IBM Corp., Armonk, NY) or equivalent. Only those participants are included in the main endpoint analyses (Table 1) who have both undergone the treatment (AAMs-patch or Hemopatch®-only) and completed the 6-month follow-up CMRI imaging with late gadolinium enhancement.

Public or patient involvement

There were no public or patient involvement in the design of this trial.

Protocol amendments

Any important protocol amendment (i.e., changes to eligibility criteria, endpoints, or analyses) is communicated to trial investigators, clinical trial registries, and regulators as soon as possible by the principal investigator. Any deviation from the trial protocol will be documented with a breach report form.

Discussion

Once ischemic cardiac damages develop, there is no effective cure for them. Whether due to prolonged subclinical course, microvascular dominance, missed timing, repeated insults, suboptimal care, or just decades of exposure to its injurious mechanisms, chronic ischemia, by its prime damages, necrosis and fibrosis, exposes the heart to adverse remodeling and HF. Despite myriad breakthroughs to its care, including LVADs, the advanced disease remains incurable, its mortality high [34], and without near game-changing horizons to reliably reverse its manifest remodeling or to restore the once lost tissue. Globally, 64.3 million people are estimated to suffer from HF [34], while largely due to ageing and westernization of life habits, its prevalence has been consistently estimated to increase. In the USA alone, it has been estimated to exceed 8.0 million people by 2030 [35].

Fortunately, with the powerful modalities of contemporary care, clinicians are well-equipped to relieve the ischemic failing heart from its most severe burdens; that is, the overt neurohumoral and hemodynamic loads. These are primarily caused by the renin–angiotensin–aldosterone system overactivation and circulatory volume and pressure overloads. Further advancements are highlighted by the most recent entry of sodium-glucose transport protein 2 (SGLT2) antagonists to the HF therapy guidelines [36], thus likely optimizing the autophagy of the burdened heart [37]. Despite the availability of such comprehensive downstream care, the disease progression and treatment resistance still often reveal those upstream, myocardial structural deficits compromising the functional reserves. Here, the fact that nearly all mammalian hearts lose their regenerative power after birth [38] is of paramount importance.

For the ischemic failing human heart, a massive body of research over the last decades has evaluated the injections of stem and more differentiated cells as myocardial regeneration-inducing strategies. Overall, as reviewed by Bolli et al., they have emerged with mixed results [7]. While all the larger trials have failed to demonstrate an increase in LVEF [7], three have reported benefits on MACE in ischemic HF with reduced ejection fraction (HFrEF) by utilizing ixmyelocel-T cells or bone-marrow-derived mesenchymal stem cells (BM-MSCs) [39–41], both of which are immunomodulatory [42]. Since 1990, kickstarting from the discovery of high tumor necrosis factor-α (TNF-α) levels in patients with cachectic HFrEF [43], investigations have verified cardiac inflammation, along with angiogenesis defects, ischemia, and maladaptive remodeling, to play a pivotal role in the progression of ischemic HFrEF. Indeed, several cytokines have been identified as regulators in the progression from early adaptive to late maladaptive remodeling in ischemic HFrEF [44]. Preclinically, the cardiac tissue macrophages (CTMs) have been shown to orchestrate adaptive remodeling by sensing stretch to produce growth factors, while the maladaptive remodeling is hallmarked by inflammatory cytokine production and a higher fraction of recruited, bone marrow-derived macrophages [45]. Clinically, this “cytokine hypothesis” on the HF progression has proved challenging to harness [46]. The recent DREAM-HF trial, the largest HF cell therapy trial with 537 patients treated by transendocardial injections of BM-MSCs (TEi-BM-MSCs), found a slight reduction in 3-point MACE (MI, stroke, cardiovascular death) and an increase in LVEF in its median ~ 30 months follow-up only in those with hsCRP > 2 mg/mL [47]. However, the trial failed to meet its primary endpoint of recurrent HF hospitalizations or secondary endpoints on its congestive complications. These investigations stress the instrumentality of modulating inflammation and the need for novel avenues for myocardial tissue reparation.

In our approach, we cover the epicardially transplanted AAMs with a matrix sheet to concentrate and direct the transplant’s paracrine effects specifically towards the myocardium. Mechanistically, the revisited view on the cardiac cell therapy mechanisms suggests that paracrine factors released by the transplanted cells are instrumental for the activation of myocardial responses towards structural and functional restoration [17–19]. As the atrial tissue is endogenously active in the secretion of regulatory factors, such as natriuretic peptides and extracellular vesicles with cardioreparative effects [48], we hypothesize that epicardial AAMs provide a targeted, lasting, and locally sufficient set of factors for regeneration and reparation-inducing myocardial therapy. We have found that, associated with both structural and functional improvement after MI, CTMs migrate to the myocardium from epicardially transplanted atrial tissue [49]. These self-renewing and functionally pleiotropic macrophages are migratory and respond to and modulate inflammation. They are key elements for cardiac regenerative potency [50, 51]. They are of early embryonic origin and are initially formed in the yolk sac and the fetal liver [52]. They first populate the heart subepicardially until epicardial signals [53–55] induce them to invade deeper to the myocardium. Remarkably, recent work by Connor Lantz and colleagues showed that deleting apoptotic recognition receptor MERTK on these macrophages was sufficient to halt regeneration in pup hearts [56]. Mechanistically, as supported by the results of Li et al. from the Thomas Braun laboratory at Max Planck Institute, showing that reorganization of cardiomyocyte lipidomes can rewire the mouse heart to a regenerative state [10], they specified the regenerative block to a lack of thromboxane A2 (TxA2) synthesis by the MERTK-deleted macrophages. In turn, the eicosanoid lipid mediator, TxA2, produced by the neonatal CTMs, promoted cardiomyocyte proliferation by shifting their metabolism towards glycolysis, while the macrophages in adult hearts failed to do so [56]. This finding, supported by the previous reports [52, 57–60], can be seen to converge on the fact that cardiac macrophages are replaced in the steady state over time with monocyte-derived macrophages from bone marrow. Moreover, the CTM subset abundance correlates with the benefit from LVAD [61], and their conditioned media [62], secreted legumain [63], and the many other immunoregulators [64] have regenerative properties in vivo. Taken together with our data on the CTM migration from epicardially transplanted atrial tissue [49], these results support the hypothesis that the therapeutic effect of AAMs is, at least partially, driven by the CTMs. To further highlight the importance of AAMs transplantation, other cells of the appendages have also been assigned roles for improved cardiac healing after ischemic damage [17–19].

Preclinically, the epicardial AAMs-patch transplantation during LAD-ligation-induced MI and HF demonstrated myocardial tissue protection, attenuated scarring, and retained cardiac function (Fig. 4) [16]. Untargeted mass-spectrometric proteomics revealed the widespread cardioprotective effects right under the AAMs-patch, which also reached the remote areas, as assessed from the samples of the interventricular septum [16]. After treatment with AAMs, this subtransplant myocardium demonstrated significant negative associations for oxidative stress and mitochondrial electron transport chain activity, while a positive association for an increase in antioxidant glutathione metabolism was noted [16]. These changes suggest AAMs-induced changes in myocardial metabolism towards glycolysis, a process pervasively linked already before Lantz et al. [56] to increased cardiac regenerative capacity [65–68]. Together with the recent literature on cardiac regeneration and the instrumentality of both the epicardium and CTMs in the phenomenon [69–71], our results indicate the epicardial AAMs-patch, containing CTMs, with key effects on the processes that drive mammalian scar resolution and reparation of the once lost myocardium. Based on these, a synthesis hypothesis is shown in Fig. 5 for the AAMs-patch mechanism.

Fig. 4.

Preclinical AAMs-patch therapy effects in a myocardial ischemia model. A In our preclinical mouse study with LAD-ligation-induced MI and ischemic HFrEF. B Even transplantation of only an epicardial patch without the AAMs attenuated scarring and persistently salvaged myocardial function, which was further enhanced when AAMs were included (AAMs-patch) [16]. C The site-specific untargeted proteomics revealed the molecular-level blueprints and AAMs’ putative mechanisms of action. This analysis revealed that below the AAMs-patch transplantation, in comparison with the patch-only transplantation, more than 200 proteins were expressed differentially in the injured heart, which were associated with upregulated cell viability, protein synthesis, muscle formation, angiogenesis, and glycolysis (a metabolic shift associated with an enhanced regenerative ability [65–68]), while the attenuation of inflammation, oxidative stress, and cell death were noted in tandem [16]. The interventricular septal areas also showed significant associations for cell viability. AAMs, atrial appendage micrografts; LAD, left anterior descending (coronary artery); MI, myocardial infarction. Subplot (B) is reprinted as minutely modified from the Journal of Heart and Lung Transplantation, 39/7, Xie Y, et al. Epicardial transplantation of atrial appendage micrograft patch salvages myocardium after infarction, 707–718 (2020), with permission from the publisher (Elsevier) under the Creative Commons Attribution–NonCommercial–NoDerivs (CC BY-NC-ND 4.0 DEED) license

Fig. 5.

Possible mechanism of epicardial AAMs–patch transplantation. Upper panel. Studies with species that regenerate their hearts endogenously well have recently shown the instrumentality of cardiac tissue macrophages (CTMs) in the process. They are of early embryonic origin, while the bone marrow, monocyte-derived macrophages command the early healing phases from the abrupt inflammation to rapid fibrosis. Classically, the phenotypes governing the phases have been denoted as proinflammatory (M1) and anti-inflammatory (M2). They operate while the CTMs activate upon necrosis to traverse the subendothelial space to form a niche with the injury-activated epicardium [70, 71]. Once the fibrosis is established, the CTMs and epicardium activate regeneration, a process experimentally linked to the metabolic rewiring of the cardiomyocytes by the scar border [72], which has been modelled to be dictated by the CTMs [73]. Several factors have been described for this sequence (see Discussion) [64]. Lower panel. In humans, like in most adult mammals, myocardial healing after its necrosis is virtually fully halted at fibrosis. Studies on CTM biology have shown them to deplete by age, especially from the subepicardial compartment [69], which might be one reason for the halted cardiac regeneration in adult mammals, although it is seen in neonate humans and mice pups. Zebrafish hearts also start to fail to regenerate after repeated injuries, which might reflect CTM depletion [74]. The preclinical data (Fig. 4) [16], and the first indication of increased live myocardium, with a trend for a reduced scar mass by the AAMs-patch site clinically (Fig. 1) [20], support a model of AAMs-patch transplantation to provide a reparative niche for the scarred and aged adult human myocardium. Here, the inevitable necrosis of some AAMs can be argued to activate the adjacent epicardium alike the CTMs within them. This design awards the molecular and cellular pathophysiological rationale to assess the AAMs-patch therapy effects on the scar in this trial. AAMs, atrial appendage micrografts; ANGPTL4, Angiopoietin-like 4; IL-6, interleukin 6; CLCF1, Cardiotrophin-like cytokine factor 1; Csf1a, (Macrophage) colony stimulating factor 1a; CTM, cardiac tissue macrophage; CXCL12, CXC motif chemokine 12, a.k.a. stromal cell-derived factor 1 (SDF-1); FSTL1, Follistatin-related protein 1; OSM, oncostatin M; RAA, right atrial appendage; RA, retinoic acid; TXA2, Thromboxane A2

Clinically, the surgical procedure of epicardial administration of AAMs was feasible and safe during CABG (Fig. 1) [20]. Intraoperative processing time of 7–8 min for micrograft isolation, together with the total time of approximately 30 min for the full perioperative AAMs-patch assembly, provided confidence and the clinical proof-of-concept for the utilization and further development of the AAMs-patch therapy [20].

The AAMS2 trial has notable advantages to the field of cardiac cell therapy. First, the straightforward perioperative protocol with minimal mechanical processing avoids many of the hindrances arising from the integration of most of the currently investigated cardiac cell therapy protocols to the everyday clinical practice. Most of the approaches are costly, time-requiring, depend on specialized infrastructure, and span cell harvesting, sorting, expansion, storage, and transplantation [15]. Furthermore, the transplantation often necessitates an invasive procedure. Second, the autologous approach circumvents alloreactivity, yet induces inflammation crucial for regeneration; this has been shown even in the extreme context of intramyocardial injection of cryokilled cells [75]. Third, the on-site transplantation of the AAMs, encased in a collagen-based sheet, is inherently powered to concentrate the cells and paracrine effects to the site—a major hardship in the previous trials [76, 77].

Although this researcher-initiated AAMS2 trial is limited by its size to assess the secondary endpoints exhaustively, the power calculations support the evaluation of its primary endpoints (Fig. 3) [20]. Several trials in the field have proven the general scale of the recruitment goal rational, including the phase II trial assessing TEi-BM-MSCs in ischemic HFrEF patients (n = 60), which achieved its endpoints [78]. However, the qualitative aspects of the power calculations (Fig. 3) warrant discussion. The power analysis source data, as derived from the only such available study, is from our open-label safety and feasibility study [20], which is both non-randomized and unblinded. Also, the controls did not receive any study intervention (i.e., an empty patch) during CABG. In this trial, the patch material has been adapted to the routinely clinically used product (i.e., Hemopatch®) in both centers, and the controls receive an empty patch. Finally, to conform to the standards to most reliably measure the scar mass with CMRI, that is, at a whole ventricle level, it stands to note the power calculations to represent local ventricle scar mass measurements.

With these decisions, while the evaluation of the AAMs’ effects is robust, the direct applicability of the power analysis is reduced. However, we consider the power analysis to be well indicative, while we consider the AAMS2 trial’s marked strengths to lie in its randomized, double-blinded, and placebo-controlled design, an imperative step on the road to evaluate the method’s therapeutic efficacy. Further, therapeutically, the trial’s primary endpoint on change in scar mass is both sound and relevant as well as evaluated by state-of-the-art objective and feasible modality. Finally, by considering its intraoperative and autologous nature against the major gap in IHD therapeutics, lacking myocardium-repair-inducing remedies, the method can be envisioned to harbour a particularly clear complementary applicability to the cardiac clinics. On this road, the results from this researcher-initiated trial will pave the road towards rational design of larger AAMs therapeutic efficacy-addressing trials.

Trial status

At the time of publication of this article, 10 patients have been recruited to the trial. The completion of patient recruitment is estimated to be finished by late 2027. The full completion date of the trial is set for mid-2028. The participants are recruited in two national academic hospital centers in Finland: Helsinki University Hospital in Helsinki (estimated recruitment of 25–30 participants), and Oulu University Hospital in Oulu (estimated recruitment of 20–25 participants). Together, these hospitals perform yearly more than 500 isolated CABG surgeries.

Protocol date and version

11.4.2024, version 1.

3.6.2025, version 2.

Abbreviations

AAMs

Atrial appendage micrografts

ANGPTL4

Angiopoietin-like 4

ATC

Anatomical Therapeutic Classification

BARC

Bleeding Academic Research Consortium

BMI

Body mass index

CABG

Coronary artery bypass grafting

CCS

Canadian Cardiovascular Society (classification of angina pectoris)

CK-MB

Creatinine kinase muscle-brain isoenzyme

CMRI

Cardiac magnetic resonance imaging

CXCL12CXC

motif chemokine 12, Aka Stromal cell-derived factor 1 (SDF-1)

DDD

Defined daily dose

ECM

Extracellular matrix

GDPR

General Data Protection Regulation

IHD

Ischemic heart disease

HF

Heart failure

HRQoL

Health-related quality of life

HFrEF

Heart failure with reduced ejection fraction

hsCRP

High sensitivity complement reactive protein

LAD

Left anterior descending (coronary artery)

LLMs

Large language models

LVEF

Left ventricular ejection fraction

MACE

Major adverse cardiovascular event

MACCE

Major adverse cardiac and cerebrovascular event

MI

Myocardial infarction

MERTK

Proto-oncogene tyrosine-protein kinase MER

NT-pro-BNP

N-terminal pro-brain natriuretic peptide

NYHA

New York Heart Association (classification of dyspnea)

RAA

Right atrial appendage

RA

Retinoic acid

SADE

Serious adverse device effects

SF36

Short form questionnaire RAND36 (for HRQoL)

SGLT2

Sodium-glucose transport protein 2

TEi-BM-MSCs

Transendocardial injection of bone-marrow-derived mesenchymal stem cells

TNF-α

Tumor necrosis factor-α

TTE

Transthoracic echocardiography

6MWT

6-Minute walking test

Authors’ contributions

A.Ny., A.V., E.K., K.O., R.K., S.K., and V.S. conceived the study design and coordination. V.S. wrote the manuscript, collected the literature, and designed the illustrations; E.K. and A.Ny. wrote and revised the manuscript as well as provided supervision. All authors read, provided comments, and approved the final manuscript. A.K. and E.K. recorded the Supplementary Video S1. Authorship requirements for the trial result publication(s) will follow the International Committee of Medical Journal Editors (ICMJE) guidelines.

Funding

Open Access funding provided by University of Helsinki (including Helsinki University Central Hospital). This work was supported by the Finnish Foundation for Cardiovascular Research (V.S. 200174, A.V. 200196 and 220094), Government-allocated block grants to specialty area (A.V. TYH2023339, TYH2021346), Aarne Koskelo Foundation (V.S.), The Finnish Medical Foundation (V.S. 3857), The Ida Montin Foundation (V.S. 20210362), and the Emil Aaltonen Foundation (V.S. 210212 K). Open access is funded by Helsinki University Library. The funders have had no contribution in the design of the trial nor in the writing of the manuscript. The funders will not be involved in the collection, management, analysis, and interpretation of data; writing of the subsequent report(s); or the decision to publish the subsequent report(s).

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The trial protocol has been approved by the ethics review board at Helsinki University Hospital (HUS; Dnr. HUS/12322/2022), and the Finnish Medicines Agency Fimea (FIMEA; Dnr. FIMEA/2023/004090). The trial is registered at ClinicalTrials.gov (NCT05632432). The trial will be conducted following the Declaration of Helsinki on Ethical Principles for Medical Research Involving Human Subjects [24]. The cardiac surgeon part of the trial obtains the informed consent from the participants after sufficient information orally and in verbatim has been provided with adequate time for making this consent. No trial procedure is performed prior to obtaining the consent from the participant.

Consent for publication

Not applicable.

Competing interests

E.K. and A.Nu. are stakeholders in EpiHeart Oy. A.K. is the Chief Engineer at EpiHeart Oy, which provides dedicated CE-marked instrumentation kits to support tissue processing in the operating room for AAMs’ patch assembly. The other authors have no competing interests to disclose. Large language models (LLMs) or other artificial intelligence (AI)-based tools were not utilized in any phase of the preparation of this manuscript.