The systematic search resulted in 4278 records, of which 1489 remained after duplicate removal. After screening, a total of 343 clinical trial registrations were included. Figure 1 shows the search flowchart. Results from all included registries are listed in Appendix 2.

Trial characteristics are summarized in Table 1. Among the 343 included registered clinical trials, 284 (82.8%) investigated cell therapy, 40 (11.7%) gene therapy, and 19 (5.5%) tissue engineering. The most targeted disease was heart failure (HF) (177 trials, 51.6%), followed by myocardial infarction (90, 26.2%), coronary artery disease (56, 16.3%), congenital heart disease (19, 5.5%), and arrythmia (1, 0.3%). The most reported etiology of HF was ischemic (100 trials, 56.5%), followed by genetic (38, 21.5%), other forms of HF (7, 4.0%), and HF with preserved ejection fraction (3, 1.7%), while 29 trials did not specify HF type (16.4%). Other forms of HF included heart failure caused by anthracyclines, Chagas cardiomyopathy, and amyloidosis-related cardiomyopathy. Most trials studied adults (18 years or older), whilst 37 studies included children (younger than 18 years). Of these 37 studies, 24 cell therapy trials, one gene therapy and one tissue engineering trial focused exclusively on children. Regarding the inclusion criteria for participant sex, 336 trials were not limited to a specific sex. Two trials included only male participants, while one trials included only females. In four trials age-range was not provided and in four trials the participant sex was not specified. Most registered trials were of North American (36.4%) or European (31.2%) origin, followed by Asia (24.2%).

The earliest registered clinical trial that was retrieved was focused on gene therapy for coronary artery disease and was started in 2001. From 2003 onwards, the number of registered clinical trials increased and reached its peak in 2011, during which most clinical trials were registered (27 trials) (supplementary figure 1). The mean number of yearly registered trials was 15 ± 6 trials. For cell therapy, the mean number of yearly registered trials was 13 ± 6 trials, for gene therapy 2 ± 1, and for tissue engineering 1 ± 1 trial. The first trial using engineered tissue was registered in 2009, the next in 2011, and from 2014 onwards trials investigating engineered tissue were registered yearly.

In 2001 and 2002 coronary artery disease was the most researched condition, surpassed by myocardial infarction in 2003. From 2005 onwards, there is a steep increase in heart failure trials, becoming the most targeted condition from 2009 onwards, surpassing myocardial infarction.

Of all trials, 158 (46.1%) were registered as completed and 56 (16.3%) were registered as ongoing (recruiting, not yet recruiting or active). A total of 64 (18.6%) trials were registered as terminated (43, 12.5%), withdrawn (17, 5.0%), or suspended (4, 1.2%), while the status of 65 (19.0%) trials was unknown. When comparing cell therapy, gene therapy and tissue engineering, there were no differences in percentage of completed trials (p = 0.163) or in the number of trials terminated, withdrawn, or suspended (p = 0.099) (Table 2).

For the terminated, withdrawn, and suspended trials, 53/64 (82.8%) provided an explanation for their status (Table 3). The most frequently reported reasons were recruitment difficulties (32.8%) and funding constraints (21.9%). In other trials the decision for terminating, withdrawing, or suspension were based on study data, including method inefficacy or necessary protocol alternation (3 trials), committee recommendation (2 trials), and safety issues (1 trial). Incontinuity of the research team resulted in termination of 3 trials, another 3 trials were never started. For 2 trials, termination was a sponsor decision; both trials were cell therapy trials. Other reasons included unspecified administrative reasons, the COVID-pandemic, political pressure, and laboratory contamination. In 11 (17.2%) trials, no explicit reason for their non-completion was reported. These trials were all cell therapy trials.

Results were available for 111 trials (32.4%) (Table 1), which included results attached to the trial registration and journal article links with an according NCT-number. When excluding currently ongoing (active, recruiting, not yet recruiting) trials, 181/287 (63.1%) not-ongoing trials had no results available. There was no difference in the availability of results between the three regenerative medicine categories (p = 0.199) (Table 2).

Amongst all trials, 88.3% were registered as phase one or two, with 10 registrations indicating the trial phase as not applicable without further explanation. For 97 trials, a combined trial phase (phase 1 & 2 or phase 2 & 3) was reported (Table 1). For cell therapy 129/284 (45.4%) trials were in phase one and 121 (42.6%) trials in phase two. For gene therapy this was 23/40 (57.5%) and 14 (35.0%), respectively. In the field of tissue engineering most trials were in phase one (14/19 trials, 73.7%), and only two trials in phase two (2 trials, 10.5%).

A total of 28 trials were registered as phase three and only two trials were in study phase four, both in the field of cell therapy as illustrated in supplementary figure 2.

In 135 (39.4%) trials, the trial was (partially) funded by an industrial partner. Industry funding and collaboration was more prevalent in the field of gene therapy (29/40 trials, 72.5%) compared to cell therapy (97/284 trials, 34.2%, p < 0.001). There were no significant differences in the number of industry-affiliated trials between gene therapy and tissue engineering (9/19 trials, 47.4%), and between tissue engineering and cell therapy. When comparing industry-affiliated and non-industry-affiliated trials, no differences were observed in the percentage of completed trials (p = 0.186), terminated, withdrawn, or suspended trials (p = 0.309), and trials reporting results (p = 0.691) (Table 2).

Primary outcome measures and means of outcome assessment are summarized in Table 4. Of all trials, only three trials (0.9%) did not report a primary outcome, while 239 (69.7%) reported a single primary outcome and 101 (29.4%) reported multiple primary outcomes. In 135 trials, the primary outcome was a composite endpoint, with different variables as part of the composite. Outcome and outcome assessment was highly variable among identified trials. When categorizing the primary endpoints, 186 trials (54.2%) assessed efficacy and 121 trials assessed safety (35.3%). A total of 35 trials (10.2%) assessed both.

Safety was most often measured through the incidence of adverse events (114 trials). Of these, a total 70 trials (63%) specified which adverse events were assessed, which yielded 50 different adverse event definitions. In 44 trials the adverse events were not specified. In 23 trials it was not specified which outcome parameter was assessed for safety.

The most frequently used primary outcome for determining therapeutic efficacy, was left ventricular ejection fraction (LVEF; 109/186, 58.6%), followed by myocardial perfusion (26 trials, 14.0%), exercise tolerance (24 trials, 12.9%), left ventricular dimensions (23 trials, 6.7%), and mortality (20 trials, 11.0%). A total of 58 other outcome measures were reported, which are listed in supplementary table 1a.

For outcome assessment, most trials used echocardiography (67 trials), followed by magnetic resonance imaging (58 trials) and computed tomography (33 trials). A total of 32 trials used exercise testing to assess the primary outcome. In 47 trials, it was not specified how outcome assessment was conducted. The median follow-up until endpoint measurement was six months (Q1-Q3: 4 months – 1 year). Nonetheless, there was variability among length of follow-up, ranging from one day up to eight years, as illustrated in Fig. 2. Time until primary outcome assessment was specified in 331 (96.5%) registered clinical trials, the remainder did not specify at which interval the outcome was assessed.

Time upon endpoint measurement in weeks for the 7 most prevalent outcome measures

In cell therapy trials, different cell types were investigated, which originated from various regions of the human body (supplementary table 2). Most investigated cell types were non-cardiac specific bone-marrow derived cells (130 trials, 45.8%), while 29 trials (10.2%) investigated cardiac-associated cells (supplementary table 2). The other trials investigated cells which were less differentiated and not specifically derived from heart tissue or prepared to become a cardiomyocyte, such as umbilical cord- (24, 8.5%), adipose tissue- (17, 6.0%), and skeletal muscle-derived (5, 1.8%) cells. Within the category of umbilical cord-derived cells, 8 out of 24 trials specified that the cells were Whartons’ Jelly-derived. In 77 (27.1%) clinical trials, cell type was not specified.

Most of the utilized cells were of autologous origin (67.3%). Whether the cells were harvested from the treated patient or from a donor was not specified in 26 (9.2%) trials (supplementary table 2).

A heterogeneity in cell therapy administration route nomenclature was observed, with 29 different identified terms (supplementary table 3). However, upon categorization, the majority could be grouped into three main routes of administration, being intracoronary, intramyocardial, and systemic injection (supplementary figure 3). Intramyocardial administration routes can be subdivided into catheter-based (endovascular) and syringe-based (by direct exposure of the heart during cardiac surgery).

A total of 40 trials investigated gene therapy for heart disease (supplementary table 4). Vectors for delivery could be divided into viral and non-viral vectors. In total, 27 (67.5%) trials used viral vectors, of which 13 used the adenovirus and 14 used an adeno-associated virus. Non-viral vectors were used in 10 (25.0%) trials, of which 7 used plasmid DNA, 2 hybrid DNA (recombinant DNA molecules), and 1 lipid nanoparticles. The applied vector was not specified in 3 trials.

Genes delivered included genes encoding for vascular endothelial growth factors (13 trials, 32.5%), fibroblast growth factors (4, 10.1 = 0%), hepatocyte growth factors (4, 10.0%), SERCA2a (7, 17.5%), fraxatin (2, 5.0%), adenylyl cyclase type 6 (2, 5.0%), stromal cell-derived factor 1 (2, 5.0%), protein phosphatase inhibitor (2. 5.0%), lysosome-associated membrane protein (1), hERG potassium channel (1), human plakophilin-2a, myosine binding protein C3, and 1 trial describing a CRISPR/Cas9 gene editing system which results in transthyretin reduction.

Administration routes of gene therapy included intramyocardial injection (16 trials, 40%), intracoronary infusion (16 trials, 40%), systemic administration (6 trials, 15%), retrograde infusion in the coronary sinus (1 trial), and one trial with an unknown administration method. Gene therapies aimed at growth factors were mainly administered intramyocardially (71.4%) instead of intracoronary/systemically (28.6%). Trials in which the gene therapy was administered through intracoronary infusion or systemically mainly used a viral vector (21/22 trials, 95.5%), while trials which used intramyocardial injection as administration route less often used viral vectors (5/16, 31.3%).

Constructed bioartificial tissue with cells or growth factor, in the form of a micrograft, patch or sheet, were investigated in 11 trials (57.9%). The other trials investigated hydrogel-based therapies (5 trials, 26.3%), engineered heart muscle (1 trial, 5.2%), a tissue engineered vascular graft (1 trial, 5.2%), and cells in sprayed form (1 trial, 5.2%) (supplementary table 5).

Regarding administration route, direct placement on the heart was the most prevalent (13 trials, 68.4%), followed by intramyocardial injection (5 trials, 26.3%) and a cell spray (1 trial, 5.2%).