https://orcid.org/0000-0001-5066-9018
Abstract
Great expectations have been set around the clinical potential of regenerative and reparative medicine in the treatment of cardiovascular diseases [i.e. in particular, heart failure (HF)]. Initial excitement, spurred by encouraging preclinical data, resulted in a rapid translation into clinical research. The sobering outcome of the resulting clinical trials suggests that preclinical testing may have been insufficient to predict clinical outcome. A number of barriers for clinical translation include the inherent variability of the biological products and difficulties to develop potency and quality assays, insufficient rigour of the preclinical research and reproducibility of the results, manufacturing challenges, and scientific irregularities reported in the last years. The failure to achieve clinical success led to an increased scrutiny and scepticism as to the clinical readiness of stem cells and gene therapy products among clinicians, industry stakeholders, and funding bodies. The present impasse has attracted the attention of some of the most active research groups in the field, which were then summoned to analyse the position of the field and tasked to develop a strategy, to re-visit the undoubtedly promising future of cardiovascular regenerative and reparative medicine, based on lessons learned over the past two decades. During the scientific retreat of the ESC Working Group on Cardiovascular Regenerative and Reparative Medicine (CARE) in November 2018, the most relevant and timely research aspects in regenerative and/or reparative medicine were presented and critically discussed, with the aim to lay out a strategy for the future development of the field. We report herein the main ideas and conclusions of that meeting.
1. Background
Great expectations have been set around the clinical potential of regenerative and reparative medicine in the treatment of cardiovascular diseases [i.e. in particular, heart failure (HF)]. Initial excitement, spurred by encouraging preclinical data, resulted in a rapid translation into clinical research. The sobering outcome of the resulting clinical trials suggests that preclinical testing may have been insufficient to predict clinical outcome. A Number of barriers for clinical translation include the inherent variability of the biological products and difficulties to develop potency and quality assays, insufficient rigour of the preclinical research and reproducibility of the results, manufacturing challenges, and scientific irregularities reported in the last years.1 The failure to achieve clinical success led to an increased scrutiny and scepticism as to the clinical readiness of stem cells and gene therapy products among clinicians, industry stakeholders, and funding bodies. The present impasse has attracted the attention of some of the most active research groups in the field, which were then summoned to analyse the position of the field and tasked to develop a strategy, to re-visit the undoubtedly promising future of cardiovascular regenerative and reparative medicine, based on lessons learned over the past two decades.
During the scientific retreat of the ESC Working Group on Cardiovascular Regenerative and Reparative Medicine (CARE) in November 2018, the most relevant and timely research aspects in regenerative and/or reparative medicine were presented and critically discussed, with the aim to lay out a strategy for the future development of the field (Figure 1). We report herein the main ideas and conclusions of that meeting.
Priority concepts in cardiac regenerative and reparative medicine according to the field of research.
2. Insights into therapeutic products
Multiple strategies have been used in the past two decades in an attempt to achieve cardiac regeneration. Some have been abandoned because of safety issues,2,3 but most are still under investigation and share some common or more specific types of bioactivity.4,5 However, the number of studies comparing different products is very limited, making it difficult to determine which product may be the most promising in terms of therapeutic efficacy. The question of whether, based on the available scientific evidence, future research and investments should be focused on any particular type of product(s), was discussed during the meeting. The specific disease phenotype, associated comorbidities, concomitant treatments, as well as the genetic background of each individual may be important in deciding which cell or cell-free product should be used to achieve the desired effect and stratification of patients in clinical trials taking into account these conditions is recommended. Moreover, the pathophysiological stage of the disease, acute or chronic, and the predominantly inflammatory or fibrotic underlying background, are also important factors to consider when opting for a particular product. While in early stages of the disease, products with more immunomodulatory, anti-inflammatory and cardioprotective (anti-apoptotic) mechanisms of action may be preferred, proangiogenic, anti-fibrotic, and direct re-muscularization strategies may be more beneficial in advanced stages of the disease. For an advanced therapy medicinal product (ATMP) to be approved by regulatory authorities such as national regulatory bodies and the European Medicines Agency (EMA), a link between potency, mode of action, and the pathophysiology of the disease should be demonstrated.6,7
Another important consideration is the standardization of the production and the characterization of the different ATMPs, given their biological complexity and heterogeneity. Comparison between different strategies or even different studies is possible only if standardized conditions as to ATMP manufacturing, preclinical models, and target patient population profiles are used by different researchers. At present, allogeneic cell-free products seem also promising as future therapeutics.8 However, there is evidence that a multipronged approach, using a combination of different products in association with specific delivery systems and/or tissue engineering, may be required to achieve optimal regeneration/repair of the heart. The use of biomaterials containing different components of the cellular secretome (i.e. extracellular vesicles) is believed to increase their retention in the tissue and improve long-term effects, something that may also be achieved by repeated dosing.9–11 In addition, recent results have shown that appropriate preconditioning of allogeneic stem cells drives regulation of miRNA to release proangiogenic secretome12 and that engineering of therapeutically eleterious fibroblasts may turn them into factories for sustained release of cell-protective factors.13 Further research in these areas is much needed.
Cells and at some extent cell-derived products are considered ATMPs by regulatory authorities such as EMA and FDA. Thus, they must be extensively characterized according to a well-defined quality control strategy, which includes standardized testing in line with the European or US Pharmacopoeia.
The identification of a potency test—preferably in vitro—to reproducibly measure the ability of the ATMP to produce a given result in line with the proposed mode of action in a given disease state is also considered a priority.6 Here it is important to realize that many regenerative products elicit their effects via complex (multi-modal) mechanisms, and it is not uncommon that additional unknown mechanisms may be involved (which is, in fact, also true for many classical pharmacological compounds and commonly referred to as pleiotropic effects). To develop an adequate potency assay and successfully quantify the potential efficacy of an ATMP, it is essential to identify the most relevant mechanism of action of the therapeutic product and at least one of the disease-related significant pathophysiological pathways expected to be counteracted by the former. For example, tube formation assay (in vitro) to measure the angiogenic capacity of a given cell-product for the treatment of refractory angina due to microvascular disease. Although complex, all attempts should be made to develop robust and reproducible potency assays that reflect the product’s relevant biological properties.
Finally, taking into consideration the economics of health systems, new therapies which could be realistically adopted by Pharma and embedded as standard-of-care in hospitals’ systems are to be considered and prioritized.
3. Insights into preclinical research
In the preclinical arena, an increase in the general rigour of studies was considered the main priority in order to reduce ultimate translational failure.14,15 Blinding is the single most powerful measure that would greatly impact the quality of the studies. Randomization, proper power analysis, and definition of the inclusion/exclusion criteria before study initiation—were also identified as conditions necessary to improve the quality of preclinical research. In this regard, reporting the protocol design and rationale, parameters, and readouts (outcomes) is now a requirement in several scientific journals before submitting a paper for publication. Finally, the reproducibility of the findings should ideally be independently corroborated by different groups and laboratories encouraged to share protocols and knowledge. For this particular purpose, co-ordinated proof-of-concept studies which involve several centres are of a special interest.14 Although we recognize that the practical implementation of multicentre preclinical studies may be difficult, such studies should be considered when large-animal models are used to validate a therapy, which require highly specialized infrastructure, staff, and knowledge to provide compelling data for the support of clinical translation. Setting up such core-labs is a main challenge even at the industry level and may require a concerted action by stakeholders to enhance the quality of translational research in general. Given the highly individualized requirements for ATMP testing, validated assays for testing of quality and potency should also be developed for large-animal models and funding provided for setting up such assays. Although at a higher cost, reporting should be ideally done according to Good Laboratory Practice (GLP) standards and therefore subject to inspection by regulatory authorities. Hypothesis-generating and initial testing phase I ‘proof-of-concept’ experiments can be performed effectively in a single laboratory. But the confirmation of the efficacy in robust and intensively monitored phase II and III experiments, with transparent analysis and reporting may be easier in multicentre studies with pre-existing platforms and with independent centralized core-labs. These platforms can be open to academia, industry, and other parties for preclinical testing of any kind of therapeutic product and not necessarily limited to reparative ones. The main goals of these studies should be similar to the ones pursued in clinical trials: defining feasibility, dosing, timing, and safety in phase I/II and efficacy in phase III experiments. Pre-registration with information on the study protocol (blinding, randomization, and sample size calculation), similar to that performed for clinical trials, is highly recommended14 to increase awareness for a proper trial design, reduce the chance of bias, and increase the accessibility to study-related data repositories. Of note, adaptive clinical trial designs are increasingly being used in preclinical research. Collectively, these measures will contribute to increased transparency as well as reproducibility and will thus provide important insights to inform the design of clinical trials. Finally, particularly in large-animal studies (similar to the first-in-human studies) adaptive clinical trial designs are increasingly used to expedite trials according to the ongoing gain in knowledge and be in accordance with 3R principles.16,17
The availability and utility of adequate preclinical disease models are another key concern. While models of ischaemic heart dysfunction were considered well-developed (although not well standardized), models of non-ischaemic heart disease seemed inadequate. For example, in light of the increasing awareness of the burden and unmet medical need for treatments in patients with HF with preserved ejection fraction (HFpEF), the development of corresponding animal models would be highly desired.18–20 Animals with naturally occurring ageing diseases such as dilated cardiomyopathy or valvular diseases in some breeds of dogs and cats are an interesting source also for academic preclinical research. Other models of special interest nowadays are chemotherapy-induced systolic dysfunction, ageing-related diastolic dysfunction, metabolic derangements (diabetes, etc.), induced coronary microvascular dysfunction for studies of HF with reduced and preserved ejection fraction, and capillary rarefaction for refractory angina. Other factors such as the age of the experimental animals (which should be post-pubertal) and the relevance of gender balance should be emphasized. Additionally, there was debate on the importance of animal vs. human origin of the product used in these studies. It was agreed that since immunity is one of the drivers of the regenerative/reparative response, the results of any animal study using human products together with immunosuppression should be interpreted with the necessary caution. Moreover, testing immune-suppression regimens and outcomes, especially in studies using allogeneic cells in repeated dosing, would be advisable. Homologous studies of immunogenicity (allo- and autograft instead of xenograft) should be carefully considered for translatability and 3R reasons.
Correct selection and definition of study endpoints are crucial to adequately evaluate product efficacy/safety. For studies evaluating systolic function, left ventricular ejection fraction (LVEF) and left ventricular (LV) volumes are the most appropriate endpoints. The clinical relevance of isolated scar reduction evaluated by magnetic resonance imaging (MRI) is debatable if not paralleled by an increase in LVEF. For studies focused on diastolic function, intra-LV cavity pressure-volume loops were acknowledged as the most reliable parameter. For studies evaluating myocardial perfusion, stress gated-SPECT or MRI should be considered. Evaluation of Major Adverse Cardiac Events (i.e. mortality) in naturally occurring animal diseases with recognized high mortality rates (i.e. in dogs with dilated cardiomyopathy) may be feasible. Furthermore, the development of more robust and reliable biomarkers to inform accurately on treatment outcomes is a significant future challenge. On this regard, the group suggested that serum levels of brain natriuretic peptides (BNP) may be useful in monitoring systolic and diastolic dysfunction. Myocardial fibrosis and cardiomyocyte hypertrophy were considered the best histologic readouts. Finally, whether regenerative therapy-induced cardiomyocyte proliferation—if any—is just an epiphenomenon or can be responsible for the improvement of cardiac function still has to be clarified. In part, this will depend on the number of new cardiomyocytes formed, which must be carefully quantified and reported on in experimental studies; simply showing examples of new myocytes is not adequate. Regarding imaging techniques, echocardiography was recommended for small animal studies (preferably using long- and short-axis images rather than M-mode for higher accuracy), while large animals should be investigated by echocardiography and preferably MRI (i.e. pigs).
4. Insights into clinical trials
One of the most important unanswered questions is which clinical conditions (diseases) should be prioritized in research to increase the probability of therapeutic success. Acute myocardial infarction was no longer considered to be a primary target, at least with first- and second-generation stem cells, since several trials showed no benefit in terms of clinical events or changes in left ventricular function.21–23 On the other hand, three conditions were identified as clinical priorities: (i) HF with reduced ejection fraction (HFrEF), preferably of non-ischaemic origin, (ii) HFpEF, and (iii) refractory angina.
For clinical trials to be approved by national regulatory authorities in EU (in alignment with the ATMP regulatory) or by FDA, an unmet clinical need has to be identified.6,7 In fact, patients not responding to optimal medical treatment (OMT) or if a proper OMT is not defined, like in HFpEF, seem to be ideal candidates to satisfy EMA’s requirements. The following selection criteria for a hypothetical HFrEF randomized clinical trial (RCT) were suggested: OMT compliance, New York Heart Association (NYHA) functional class II or III, LVEF <40% and elevated BNP levels. For an RCT in HFpEF, given the heterogeneity of the syndrome and the failure of all previous trials, criteria should be very restrictive and specific. So, together with an NYHA class II–III (in the absence of overt non-cardiac causes) and LVEF ≥50%, the recently described Echocardiographic and Natriuretic Peptide HFpEF diagnostic score (≥5 points) was pointed out as a possibility if restricted to specific underlying aetiologies (i.e. senile mild-type transthyretin deposition, microvascular disease, immune/inflammatory disorders).24 For an RCT in refractory angina, Canadian Cardiovascular Society functional class III or IV along with reversible ischaemia on imaging tests were considered the main criteria.
So, which is the most appropriate design for an ATMP RCT? Which should be the endpoints? For both HF and refractory angina trials, the 6-min walking and cardiopulmonary tests, and quality of life (QoL) questionnaires were considered the best options. Moreover, QoL tests should be reported separately by physicians and patients and supported by structural and functional surrogates, which serve to underpin plausible mechanisms of action. Endpoints should be ideally maintained throughout the whole process of clinical testing. Ideally, primary endpoints should be used to estimate the sample size for phase III trials and the secondary ones for phase II trials. Whenever possible, sham/placebo groups should be used in a double-blind fashion. Otherwise, endpoints should be blindly assessed by an independent review and monitoring board.
Regarding manufacturing, the ‘quality by design’ methodology, which consists of implementing Good Manufacturing Practice (GMP)-compliant manufacturing procedures as early as possible, even in preclinical studies before starting phase I trials, is advisable. For any type of product, autologous or allogeneic, identity, potency, sterility, quality, and release criteria need to be clarified and approved by the regulators. Furthermore, the same product that was used in animal models should be the one applied in RCT (if the mode of action is presumed to be the same), trying to keep also the same manufacturing processes. Whenever feasible or possible, issues pertaining to intellectual property and freedom of operation should be clarified at an early stage to guarantee a smooth transition from clinical trial testing to clinical use.
From the regulatory point of view, it is crucial that investigators seek scientific and regulatory advice early on in the ATMP development process.6,7 In 2020, the new Clinical Trial Portal of EMA will facilitate harmonized trial authorizations, enabling quicker timelines and obviating the need for submissions to different national competent authorities.
Finally, an important issue is how to rekindle the interest of stakeholders into supporting the preclinical and early clinical developmental process of ATMPs. Commercial interest is essential to move the field from proof-of-concept academic studies to larger clinical trials and, eventually, to product approval and commercialization. ‘Stakeholders’ is broadly defined not only by the involvement of industry partners but also by participation of other interested partners (public healthcare institutions, private equity funds, or other funding bodies), patients, and healthcare professionals. The group agreed that only by performing high-quality, innovative science, with a credible pipeline for product development, with patient-targeted research and high-quality preclinical and clinical evidences with hints of efficacy and cost-effectiveness, will we be able to move regenerative therapies into clinical practice.
5. Recent advances in the field
Some advances have been made in the field since this ESC Working Group on Regenerative and Reparative Cardiovascular Medicine meeting. Briefly, and not trying to be exhaustive, some breakthroughs in the preclinical and clinical settings warrant to be mentioned.
In vitro potency assays, based on anti-apoptotic and pro-angiogenic properties, and capable of predicting the in vivo activity in a repeatable fashion by using more quantitative detection methods, have been reported in a study using cardiac progenitor cells and derived exosomes.25 These optimized tests resulted in more reproducible and suitable assays for formal GMP validation. As for multicentre preclinical studies, the first randomized, control-group, open-label, and endpoint-blinded experimental trial will be carried out within the CIBER-CLAP large-animal platform. This trial has been designed with the aim of testing reproducibility in cardiovascular interventions.26 Finally, in the clinical field, the 3-year follow-up favourable results of the treatment of patients with refractory angina with adipose-derived stromal cells (MyStromalCell Trial) have been recently published.27
6. Conclusions
As in most drug development endeavours, there are phases of premature excitement followed by realistic expectations as to the clinical outcome. Cardiovascular regenerative and reparative medicine has passed the phase of overly optimistic expectations and is building its foundations on scientific evidence more than eminent opinions. The current challenge to critically understand and decipher the accumulated knowledge and to translate it into realistic therapies remains. Standardization, cooperation, and transparency will be key to move the whole field of cardiovascular regenerative and reparative medicine forward.
Conflict of interest: none declared.
