Atrial Appendage Micrograft Transplants to Assist Heart Repair After Cardiac Surgery

BACKGROUND AND SIGNIFICANCE

Globally, each year 17.9 million people die of cardiovascular diseases. Ischemic heart disease (IHD) is the cause in half of these cases, thus making it the global leading single cause of death. While 126.5 million patients suffer from IHD worldwide, in Europe 30.3 million patients are afflicted.

IHD is hallmarked by progressively enlarging atherosclerotic coronary plaques. These disease hotspots not only drive myocardial hypoxia, cardiomyocyte hibernation, apoptosis and interstitial fibrosis but are prone for erosion and rupture. Plaque rupture forcefully activates the hemostatic system resulting in thrombotic coronary occlusion, myocardial infarction (MI), and death of cardiac tissue. Due to improved acute care, the patients increasingly survive the acute phase, and the site of injury eventually gets replaced by a scar that typically restricts the filling and pumping of the heart. Depending on the extent of injury and the resulting scar, eventually the increased workload leads to adverse remodeling and emergence of heart failure (HF), an irreversible and incapacitating clinical syndrome with poor prognosis. HF due to an ischemic etiology has been reported to vary from 29% to 45%. For instance, a recent meta-analysis suggests the "all-type" HF prevalence, including the previously unrecognized cases via population-based echocardiographic screening, to be as high as 11.8% among general population aged above 65 years.

CABG surgery is the preferred revascularization method for patients with severe progressed IHD. In Europe, more than 245,000 CABG surgeries were carried out in 2016. Regardless of age, CABG surgery has been shown to have an overall beneficial effect on ischemic symptoms and mortality.

Cardiac healing by regeneration rather than scarring could tilt the IHD with its complications towards an increasingly manageable, even curable, disease. While the hearts of some vertebrates heal by regeneration throughout their lifespan, in rodents this capacity is limited to the first week of life. Very early in life, also the human heart seems to possess capacity to regenerate after ischemia.

It has proved complex to activate cardiac regenerative repair in adult human heart. Many stem, progenitor and differentiated cells have been tested in this regard. While these investigations have provided promising results, the therapies remain complex and costly, highlighting the need for more clinically straightforward approaches. Cells derived from atrial appendages have been shown to be capable of stimulating regenerative cardiac healing in the context of ischemic cardiac damage. As positioned by the European Society of Cardiology, many tissue-engineered approaches, including epicardial extracellular matrix (ECM) patch transplantation, are highlighted as promising future therapies for ischemic HF. These approaches could improve the local persistence and viability of the co-transplanted cells-a major obstacle identified in previous studies.

GENERAL CONCEPT

In this trial, the patient's own heart tissue from the right atrial appendage is used for therapy. Neither the left nor the right atrial appendage (LAA and RAA, respectively) directly contribute to the heart's pumping function. A piece of the RAA can be safely harvested upon insertion of the right atrial cannula during the set-up of the heart and lung machine at the beginning of CABG. In the AAMS2 trial, a piece of the RAA tissue is used as epicardially transplanted, patch-encased, and mechanically expanded atrial appendage micrografts (AAMs). This therapy can be administered during single CABG surgery.

PREVIOUS RESULTS

In a preclinical mouse model of MI and HF, the effects of epicardial AAMs-patches were compared to acellular ECM patches. The results demonstrated myocardial tissue protection, attenuated scarring, and retained cardiac function. Further, mass-spectrometry-based quantitative proteomics demonstrated widespread regenerative and cardioprotective effects in the myocardium, including decreased oxidative stress and AAMs-mediated induction of myocardial glycolytic metabolism, a process associated with an increased regenerative capacity of myocardium. The AAMs-patch therapy has proceeded to clinical use. Following the first-in-man application of AAMs, the safety and feasibility of the epicardial AAMs transplantation during CABG was recently confirmed. Moreover, as this study suggested increased thickness of the viable myocardium in the scar zone, it provided the first indication of therapeutic benefit.

OBJECTIVES AND OVERVIEW

This AAMS2 randomized double-blinded and controlled trial evaluates the effect of epicardially transplanted AAMs as an adjuvant therapy to CABG surgery. The trial's primary endpoints are changes in cardiac function and structure as evaluated using late gadolinium enhancement cardiac magnetic resonance imaging (LGE-CMRI) at 6-month follow-up after surgery as compared to preoperative LGE-CMRI. The trial enrolls 50 patients in a 1:1 group allocation ratio to the two study groups: 1.) collagen-based patch + AAMs + CABG (treatment arm) and 2.) collagen-based patch + CABG (control arm). Autologous RAA tissue is harvested from the RAA during CABG from all participants and based on randomization, the piece of RAA tissue is either processed to AAMs perioperatively or cryostored for biochemical analyses. The AAMs, embedded in fibrin matrix gel, are placed on a collaged-based patch, which is then epicardially sutured in place. To pinpoint the ischemic scar area as the epicardial transplantation site, LGE-CMRI is done preoperatively. Study blood samples are collected preoperatively as well as at 3- and 6-month follow-up after surgery. Transthoracic echocardiography (TTE), LGE-CMRI, symptom-scaling measures, and 6-minute walking test (6MWT) are performed preoperatively and at the 6-month follow-up.

METHODS

• Patient selection, enrolment, ethics, and timeline-The patients meeting the both eligibility, inclusion as well as exclusion criteria, as evaluated by either a cardiologist or a cardiac surgeon, are selected from the hospital's list of elective cardiac surgeries. The patients' medication is optimised according to the current guidelines by the treating cardiologist. The usual waiting time on the list ranges between 2 and 8 weeks. This time allows medication changes to take effect before surgery. Later, the recruited patients are called for a clinical control visit (denoted as the 3-month follow-up) and a dedicated trial visit (at 6-8 months postoperatively, denoted as the 6-month follow-up).

All patients are provided with information describing the trial. Before a subject undergoes any study procedure, an informed consent discussion will be conducted and written informed consent to participate is required. The trial will be conducted following the Declaration of Helsinki on Ethical Principles for Medical Research Involving Human Subjects. The study has been approved by the Ethics Committee of Hospital District of Helsinki and Uusimaa (HUS; Dnr. HUS/12322/2022), and the Finnish Medicines Agency Fimea (FIMEA; Dnr. FIMEA/2023/004090). The estimated start of the patient recruitment is March 2024 with an estimated full study completion date on January 2027. The participant is excluded from the trial (screening failure), if, after optimisation of medication, a visible scar cannot be identified or left ventricular ejection fraction (LVEF) is ≥50% in the preoperative LGE-CMRI. This applies also if the LGE-CMRI has not been performed prior to CABG.

• Endpoints-The trial endpoints are listed in a separate section. The primary endpoints focus on changes in cardiac fibrosis as evaluated by LGE-CMRI and circulating levels of N-terminal prohormone of brain natriuretic peptide (NT-proBNP). The secondary endpoints center on other efficacy parameters as well as both safety and feasibility of the therapy.
• Randomization and blinding-Patients are randomized into CABG or AAMs groups using sex-stratified block randomization via the 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. 50 participants that are treated during CABG either with Hemopatch® or the AAMs-patch are randomized. The research nurse texts the Sealed Envelope service phone number 'AAMS2' (the trial abbreviation) with and command 'randomise', and the participant pseudonym. The research nurse receives then the allocation by a text message. This is done on a previous day of patient's CABG to allow adequate time to setup practicalities required for the perioperative AAMs-patch assembly. The study nurse oversees the allocation in a double-blinded manner, where the patient and the evaluating cardiologist and radiologists remain blinded to the study group allocation. Given the nature of the treatment (transplant vs. no transplant) it is impossible to blind the operating surgeon or the study nurse to the allocations intraoperatively. However, the cardiac surgeon is blinded when planning the anastomoses prior CABG. All LGE-CMRI and TTE measurements as well as laboratory analyses are done by researchers blinded to the group allocations. In this trial, if considered acutely mandatory 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. The RAA tissue is weighed and mechanically processed into micrografts as previously described by using the Rigeneracon blade (Rigenera-system, HBW s.r.l., Turin, Italy). Dedicated CE-marked instrumentation kits to support tissue processing in the operating room are obtained from EpiHeart Oy (Helsinki, Finland). After RAA grinding, the cold cardioplegia supernatant is removed and the AAMs pellet is collected with 0,4mL of the Tisseel (Baxter AG, Vienna, Austria) fibrinogen solution diluted with 0.9% NaCl (1:1 relation). Then, the AAMs-fibrinogen mixture is spread onto a cooled sterile metallic dish to be mixed in situ with 0,2mL of thrombin solution diluted in 1:30 relation with 0.9% NaCl. The AAMs-fibrinogen-thrombin mixture then undergoes spontaneous gelling for at least 10-15 minutes. Then, the gelled AAMs-fibrin gel is maintained cooled (+6 - +8oC), covered, and sterile waiting for transplantation. When all the anastomoses are ready, the AAMs-fibrin gel is lifted onto a collagen-based matrix (HEMOPATCH Sealing Hemostat [45 mm × 45 mm; catalog ref. 1506256; Baxter International Inc., Illinois, USA). Next, the edges of the Hemopatch® are moistened with sodium bicarbonate (4.2-8.4%), which activates the polyethylene glycol-coating of the patch. Immediately after, 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 for 2 minutes. The transplantation site is assessed prior CABG with LGE-CMRI to the ischemia-induced scar. In this trial, the AAMs patch shall wait, at maximum, 6 hours prior transplantation.
• General data protection regulation-The data collected during the trial will fulfil EU regulations for personal health data protection, including General Data Protection Regulation (GDPR).
• Adverse events-As a part of overall safety evaluation of the trial method, adverse events are subdivided to (1) major adverse cardiac and cerebrovascular events (MACCE), (2) anticipated SADE (serious adverse device effects), (3) other SAE (serious adverse events), (4) unanticipated SADE. All of these events are continuously monitored and reported to regulatory bodies. Here, the MACCE comprise i.) death (all-cause), ii.) MI, iii.) any acute coronary revascularization, or iv.) stroke. MI is defined in line to the fourth universal definition on MI as either a.) perioperative myocardial injury (i.e., type 5 MI ≤48 hours post-CABG, creatinine kinase muscle-brain isoenzymes [CK-MB] ≥ 10 times the upper reference) or b.) 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 with lesion(s) concordant to the clinical presentation. New revascularizations span any postoperative coronary intervention.

Anticipated SADE are events identified by the trial investigators with, at a general level, possible causal relationship to the AAMs-patch therapy. These include: i.) mediastinitis, ii.) postoperative pericardial effusion requiring subxiphoidal drainage or resternotomy, iii.) major bleeding (BARC classes 4-5) from the RAA biopsy site, or iv.) major postoperative arrhythmia (ventricular fibrillation, or ventricular tachycardia over 30 seconds). Mediastinitis is diagnosed according to Centers for Disease Control and Prevention guidelines. The published data in mice, pigs, and human on the method does not indicate increased risk for any of 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: myocarditis, pericardial effusion, HF exacerbation, resternotomy, atrial tachycardia or fibrillation, atrial flutter, transient ischemic attack, major bleeding (BARC 3-5), acute kidney injury, or other hospitalization due to ischemic cause.

• Blood samples-A blood sample for NT-pro-BNP measurement, a gold standard biomarker for HF evaluation, is collected preoperatively and at both 3-month and 6-month follow-ups. In addition, the AAMS2 trial assesses blood, plasma, and RAA tissue samples for their contained RNA transcripts with a novel focus on their post-transcriptional modifications, as previously described. However, to ensure adequate yield of native RNA, instead of collecting 3 mL x 5 of TEMPUS(TM) RNA-stabilized blood per visit, 3 mL x 8 is collected. These modifications comprise a biologic frontier in genetics that is unveiling a key contributor and regulator of many cellular functions and pathophysiologic conditions, also IHD and ischemic HF. As the information regarding blood epitranscriptomics in human IHD and HF remains scarce, the AAMS2 trial aims to provide insight into the AAMs-treatment-induced alterations in the blood epitranscriptomes. Especially, the focus is in the two most common modifications, N6-methyladenosine (m6A) and adenosine-to-inosine (A-to-I) editing.
• RAA tissue samples-The removed piece of RAA tissue from the control group is collected as a sample for biochemical analyses. The piece of RAA tissue will be divided in two and stored (RNAlater or formaldehyde-ethanol) for subsequent epitranscriptomics-targeted evaluations as previously described.
• Echocardiography-All participants are assessed with electrocardiogram and TTE preoperatively and postoperatively at hospital discharge and at 3-month follow-up. The TTE recordings are performed with designated cardiologists. These recordings include both anatomical and functional assessments of ventricles, atria and valves. An especial focus is granted during preoperative TTE, prior to the above-described RAA sample collection for AAMs, to the atrial appendages and the atria proper for anatomic characterization. Moreover, the presence or absence of pericardial effusion, thrombus and aneurysm is recorded. Perfusion anesthesiologist will perform transesophageal echocardiography (TEE) in the operating room during anesthesia to evaluate both left and right atrial appendages for blood flow velocities, possible sludge and thrombus before CABG.
• Late gadolinium enhancement cardiac magnetic resonance imaging (LGE-CMRI)-A whole body 1.5-T MRI scanner (Siemens Sola or Avanto-fit, Siemens AG, Erlangen, Germany) is used for LGE-CMRI image acquisition. Cardiac structure and function are evaluated with a standardized LGE-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. LGE is assessed to measure infarction volume and mass using 5-SD semiautomatic gain estimate, as previously suggested for semiautomatic thresholding for infarction detection. 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 measured using the RAND36 (SF36) short form questionnaire. The questionnaire is standardized with specified mean and standard deviation values for eight dimensions that range from physical functioning and subjective feeling of vitality and health to bodily pain. The obtained scores are compared to a Finnish cohort with any chronic disease. Also, a subject's symptom-evaluation is performed for the two cardinal symptoms of IHD and HF, angina pectoris and exertional dyspnea, with standardized classification systems developed by the Canadian Cardiovascular Society (CCS) and the New York Heart Association (NYHA), respectively. Finally, a six-minute walking test (6MWT) is assessed to gain objective morbidity measures. All these parameters are assessed both preoperatively and at the 6-month follow-up visit postoperatively. NYHA and CCS classes are also recorded at 3-month clinical follow-up.
• Statistical analyses-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. Power analysis sample data was derived from the previous open-label AAMs trial. 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 of 0.05. 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 means 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. The data can be analyzed and published in phases during the trial.
• Data collection preoperatively and postoperatively during follow-up-Clinical, laboratory and drug treatment data are collected to the hospital electronic health records. After discharge and during follow-up patients can visit the primary healthcare. No concomitant care or interventions are prohibited. 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 that patient. The participant adherence is monitored by updating an anonymized trial progression file. In case of deviation from the trial protocol, the participant is followed and examined according to the trial protocol as accurately as possible after the deviation. In case of participant discontinuation, the study data accumulated until the discontinuation is included in the trial datasets and reported separately.
• Data storage, management and sharing-The produced data are stored in the network hard drives of HUS and UH during analyses as well as on the servers of the CSC - IT CENTER FOR SCIENCE LTD. (Finland) specifically designed for sensitive data storage, all with automated backups. Directly identifying patient data with corresponding pseudonymization keys and randomization codes are stored in a key registry located within the safe hospital systems with automated backup and access control. The accession to the registries is controlled via role-based accession rights, and only those researchers specified in the registry description approved by the HUS Ethics Committee can access the data therein. PI-researcher docent Antti Nykänen is the responsible party for the key registry and its contents.

The TTE data of participants is accessed via IntelliSpace software (Philips, Netherlands) that is ultimately stored on the Microsoft® Azure Cloud, which fulfills the HUS data security guidelines. The LGE-CMRI data and reports are stored to digital HUS picture archiving and communication system (PACS). When needed, the LGE-CMRI data are transferred internally between HUS Medical Imaging Center's servers to allow image analysis with appropriate CMR software. Case report formats are both physically stored in the HUS premises with an access control and electronically in the research registry.The accessions to the research registry are controlled via role-based accession rights and only those research team members singly specified in the registry description document, approved by the HUS Ethics Committee, can access the data therein. All workstations, network drives and servers are password protected.

Prior any sharing of pseudonymized data with the academic study collaborators inside or outside European Union take place, whether performed via CSC - IT CENTER FOR SCIENCE LTD. (Finland) servers or with strong-password-protected hard drives transported by either official courier of the University of Helsinki or Helsinki University Hospital, the responsible collaborating scientist or the representative of the affiliated institution will sign Material Transfer Agreement (MTA). These MTAs will use the EU commission's Standard Contractual Clauses (SCCs) to protect the access, i.e. transfer, of the pseudonymized data. Also, all personnel handling pseudonymized data will be required to sign an official HUS secrecy and data security commitment. Moreover, the CSC - IT CENTER FOR SCIENCE LTD. requires its own data secrecy handling agreement for each collaborator to sign prior accessing the pseudonymized data.

Principally, after the active phase of the trial, the produced data with pseudonyms will be stored on the servers of the Finnish IT Center for Science CSC (SD-Apply) for 15 years. After this, the data is anonymized via erasing all the pseudonyms and curated indefinitely. A distinct Data Access Committee will monitor the re-use of the stored data. Also, the sequencing datasets with group-level anonymized metadata can be made available upon publication via uploading into repositories such as the European Nucleotide Archive (ENA) of the European Molecular Biology Laboratory European Bioinformatics Institute (EMBL-EBI, Cambridge, UK) or the Gene Expression Omnibus (GEO) functional genomics database repository (National Center for Biotechnology Information NCBI, Bethesda, MD, USA).

• Trial monitoring-The AAMS2 trial is monitored by the Clinical Research Institute, Helsinki University Central Hospital (HUCH) to ensure trial subjects' rights, safety, and well-being. A detailed monitoring plan is effective. The monitor audits the trial data registries for appropriateness and completeness regularly. Also, 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.

Inclusion criteria

• Informed consent obtained
• Left ventricular ejection fraction (LVEF) between ≥ 15% and ≤ 40% at recruitment (transthoracic echocardiography)
• New York Heart Association (NYHA) Class II-IV heart failure symptoms

Exclusion criteria

• Heart failure due to left ventricular outflow tract obstruction
• Acute myocardial infarction (AMI) within last 30 days
• History of life-threatening and possibly repeating ventricular arrhythmias or resuscitation, or an implantable cardioverter-defibrillator
• Stroke or other disabling condition within 3 months before screening
• Severe valve disease or scheduled valve surgery
• Renal dysfunction (GFR <45 ml/min/1.73m2)
• Other disease limiting life expectancy
• Contraindications for coronary angiogram or LGE-CMRI
• Participation in some other clinical trial

Screening Failure:

• After optimization of medications, no visible scar or LVEF ≥ 50% in preoperative LGE-CMRI
• Preoperative LGE-CMRI has not been performed prior scheduled CABG