The outbreak of the novel severe respiratory distress syndrome coronavirus 2 (SARS-CoV2) and its adjunct symptomatic, Coronavirus 2019 induced disease (COVID-19), is one of the most significant world health events in recorded history (1). Early reports during the initial outbreak in Wuhan, China, found that up to 14% of patients presented with the severe form of COVID-19 and that mortality was as high as 3% (2–4). Subsequently, the virus became a global pandemic and new variants emerged (5–9). Major variants responsible for surges of virus infections include “Beta” (South Africa, May 2020), “Delta” (India, October 2020), and “Gamma” (Brazil, November 2020) ( Figure 1A ) (7). During the summer and fall seasons of 2022, the predominant circulating variants were sub-types of “Omicron”, first documented in November 2021 across multiple countries (10); E.g. the “Omicron” sublineage BQ.1 was designated as a Variant of Interest (VOI) by the European Center for Disease Prevention and Control (ECDC) as of 20^th of October 2022 and it was expected that by mid-November to beginning of December 2022 more than 50% of SARS-CoV-2 infections were due to BQ.1/BQ.1.1 (11). This demonstrates the rapid dynamics in virus changes (9). By the end of December 2022, the worldwide death toll attributed directly to COVID-19 had surpassed 6.6 million individuals (5–7).

The COVID-19 pandemic created an ideal situation for the convergence of two research quests that had been progressing independently for decades. The first quest came from pulmonology, where researchers have sought to improve mortality from acute respiratory distress syndrome (ARDS) for decades, with mortality levels of 44% in clinical trials since the 1980’s (12). The second quest, dating back from the 1990s, was the scientific effort to demonstrate clinical efficacy for cell therapy products containing multipotent mesenchymal stromal/stem cells (MSCs) (2, 13–18). Preliminary evidence suggested that MSCs might be beneficial for pulmonary disorders (2, 18). This is supported by biodistribution studies which demonstrated that MSCs given intravenously (IV) rapidly localize to the lungs, where they may exert their beneficial properties (2, 15, 16, 18). It is well established that the immunomodulatory and regenerative properties of MSCs entail a plethora of distinct synergistic mechanisms of action (MoAs) that might help ameliorate pulmonary conditions (13, 18).

Between 2011 to 2019, the database CellTrials.org identified 16 clinical trials of MSCs for ARDS, and by April 2020, seven of these trials were completed, and five were published (19). Unfortunately, none of these publications demonstrated clinical efficacy of MSCs against ARDS. Similarly, a literature search based on published studies of MSCs for ARDS between 1990 to 2020 found nine such studies and confirmed that the improvement in mortality was not significant (20). In the past, poor outcomes considering ARDS mortality were often attributed to the high complexity of ARDS etiology and pathology (e.g. many small/difficult to target subgroups) and rapid disease progression (e.g. short time window for interventional treatment). Thus, any studies aiming to prove efficacy in ARDS typically require stringent inclusion/exclusion criteria and large trial cohorts to control for confounders. This has been challenging to achieve, given that patient enrollment only reaches sufficient numbers in larger specialized clinical centers, thus often requiring multi-center studies (12).

This led to the convergence of the quests mentioned above during the COVID-19 pandemic. For the first-time large cohorts were broadly available to effectively study novel therapeutic interventions. Pulmonologists noted very early that COVID-19 differs from classic presentations of ARDS (21), thus calling for an in-depth analysis of clinical trial outcomes. However, current knowledge on the outcome of cell therapy studies for both ARDS and COVID-19 is still fragmented ( Figure 1B ) (2, 18). Covering the most relevant 2-year interval (Jan 2020 to Dec 2021), we here report the outcomes of the available published clinical trials focusing on cell therapy of COVID-19 today. This also covers specific intricacies of product manufacturing and delivery to patients (2). Our analysis found that three quarters of cell therapies deployed against COVID-19 relied on MSCs. During 2020 alone more than 100 clinical trials were registered worldwide that employed MSCs to treat COVID-19 pneumonia and acute respiratory distress. Indeed, published outcomes from those trials that focus on MSC therapy for COVID-19 now appear to be sufficient to warrant a first comprehensive examination of the safety and efficacy profiles of MSCs for treating severe COVID-19.

This study presents three types of data regarding cell-based therapies for COVID-19 ( Figure 1B ): (1) We have collected two years (from Jan 2020 to Dec 2021) of registered clinical trial activity extracted from worldwide registries; (2) We have gathered the published clinical outcomes and extracted any available information on manufacturing and clinical delivery of MSC products from the published studies to study the potential impact of cell product manufacturing and mode of delivery on clinical efficacy; (3) We have performed a relative risk ratio (RR) and odds ratio (OR) analysis for all-cause COVID-19 mortality for studies employing intravenous use of MSCs (most frequent).

To the best of our knowledge, this effort is the first report to date, that comprehensively links clinical trials of advanced cell therapy for COVID-19 with the published outcomes of those trials. This type of linkage requires that the starting database of clinical trials is as complete as possible but avoids/omits any redundancies. Thus, in the process of building the COVID-19 trials database at CellTrials.org, we have incorporated several crucial quality steps, e.g. inclusion of trials from all national registries, exclusion of false positives on keywords, and exclusion of double counting of the same trial. So far, none of the other existing trial compilations that we examined in Table 2 stated that they have employed such steps. We also must point out, that most academic studies of COVID-19 cell therapy trials ran their entire search at a single point in time and selected only for COVID-19 trials. In contrast, CellTrials.org has collected all advanced cell therapy trials monthly and then extracted the COVID-19 trials at the end of each month. In a typical month, CellTrials.org sorts through about 600 clinical trials that hit on keywords and finds that 10% qualify as advanced cell therapy trials.

The advanced cell therapy trials for COVID-19 have been conducted in 30 countries led by the US, China, Iran, and Spain, yet most resulting publications have come from China, Iran, and 11 other countries so far. The initial surge in clinical trials registered to apply cell-based therapy for treating COVID-19 peaked in April 2020 and subsided into an ongoing effort of a few new trials per month. It must be noted that many healthcare policies at both national and local levels influence the ability to launch trials and recruit COVID-19 patients for cell-based therapies. Despite ongoing outbreaks in the US, we have noted that multiple trials have been suspended because they cannot recruit enough patients at a single hospital. Thus, large research consortia with multi-institutional and multi-national collaboration are needed to tackle this shortcoming and more rapidly develop new treatment approaches for COVID-19. In China, the “Zero COVID” policy was so efficient at suppressing outbreaks for two years, that clinical trials had stopped because they could not accrue patients (54). Since this policy was changed in late 2022, new infections and probably also associated severe cases and deaths due COVID-19 are likely to have surged dramatically (101), with a need for effective treatments.

In many Western countries (e.g. Europe being subject to both national and EMA regulation, with considerable variability in regulation between different European nations) (2, 24), strict regulations on human cell therapy mean there were only a limited number of cell therapy products with established safety profiles that could be trialed. While we found that 56% of the trials were below phase 2, we found that 69% of the published outcomes were studies with a control arm. The fraction of the trials employing MSC was 72%, but 92% of the published outcomes are studies that relied on MSCs. The mechanisms of action by which ARDS and COVID-19 patients may benefit from MSC therapy have been exhaustively reviewed by the papers listed in Tables 2 , 3 (37–53, 63, 64, 66–86, 97–100), such as relying on multiple synergistic effector mechanisms, and promoting/triggering beneficial immunomodulatory and regenerative pathways, as well as angiogenesis and antiapoptosis (18). Hence, we will not repeat that discussion here, but only refer to the most crucial key observations in the discussion further below. Most importantly, in this study, we have also endeavored to quantitatively calculate the safety and efficacy profile of MSC infusions as a novel treatment for COVID-19. However, efforts to treat these topics systematically appear to be fraught with difficulties.

On the issue of MSC safety, we found that all of the 24 published trials claimed that they had no severe adverse events related to the MSC infusion (37–60). Many of the studies gave anti-coagulant therapy as a prophylaxis (38, 42, 44, 56, 57, 60). Already in the first reports from the COVID-19 epicenter in Wuhan, and swiftly following global reports, severe coagulopathy was identified as one of the most evident complications arising from SARS-CoV2 infection and critical/severe COVID-19 (102–112). The increased incidence of thrombotic complications in these patients was verified in large population studies in Sweden (110, 111). Indeed, MSC-IV therapeutics carry a risk of thrombotic complications, due to their expression of the highly prothrombotic tissue factor (TF/CD142) (2, 16, 17). Hence, since the first months of the pandemic, we have identified prophylaxis against coagulation as a crucial factor for the safe application of IV MSC therapeutics in a COVID-19 setting (2, 14, 15). We were poised to highlight the safety aspect of IV MSC therapy for COVID-19, but we discovered that none of the published trials had conducted monitoring that could address this issue. For example, in one trial that treated patients suffering from severe COVID-19 in a hospital ICU, a patient that received IV MSC therapy died 13 days later following an arterial thrombosis (56). This event was considered unrelated to MSC therapy because it occurred outside their four-hour window of post-infusion monitoring.

This was common in the 24 publish trials: possible complications arising from MSC infusion were only monitored during or shortly after infusion. Yet, case studies of adverse events associated with MSC infusions found that elevated clotting markers (e.g. TAT and D-Dimer) typically peaked 9-12 hours post-MSC infusion (17), while pulmonary (but not arterial) embolism post MSC therapy could be detected days, weeks, and even months after treatment (17, 113, 114). This should caution us that the impact of the instant blood-mediated inflammatory reaction (IBMIR) post IV MSC therapy for COVID-19 may only become evident at later time points (2, 16, 17). Given that thromboembolism is a well-known side effect of either COVID-19 infection or MSC infusions, it is currently not possible to retrospectively assign that complication to one or the other group (17), when these studies were not constructed to discriminate the potential causes. Thus, we did not attempt to compute the rate of adverse events in the published trials because the few observed/reported events are largely anecdotal in nature and cannot be quantified at the current stage. Hopefully, future clinical trials will incorporate longer monitoring periods and will have large enough patient groups to statistically identify if any adverse events were increased in response to the experimental (MSC) treatment but not in response to the standard of care.

The primary theme of our analysis was to calculate the efficacy of MSC infusions as a therapy for COVID-19 based on the risk of mortality. The mortality data represent a quantitative set of facts that were extracted from each of the published clinical trials and are listed in Table 5 . However, we must caution that not all these trials were designed with survival as an endpoint. For example, one study recruited convalescent patients to determine if MSC infusions would speed up the resolution of their lung lesions and consequently none of the patients died on either arm (49). The previous meta-analyses have also noted that published studies with intravenous use of MSC for COVID-19 have very heterogeneous patient populations, in terms of the severity of their COVID-19 illness as well as the forms of concomitant therapies that they received (97–100). Our statistical model has included more studies and more patients than previous meta-analyses of MSC efficacy for COVID-19 (97–100). We corroborate the conclusions of the previous meta-analyses, that intravenous MSC therapy appears to provide a benefit for the treatment of COVID-19 (97–100). Recently, earlier timing of MSC infusion has emerged as a new factor that may be associated with improved survival for patients receiving MSC therapy for COVID-19 (115). This was established in a single center study that gave an IV MSC dose 3 x 3 million cells/kg. We note that their cumulative dose is about three times higher than the average in our 24 studies of IV infused MSCs. We do find that most published studies (17 out of 24) spread out clinical MSC delivery over two to four doses.

Perinatal sources of MSCs (e.g. tissue of UC and/or placenta, collectively called PT-MSCs) (16, 17), were employed in 18 of 24 published trials ( Table 3 ). Despite a great deal of preclinical and clinical research, there are still conflicting opinions on the biological characteristics of MSCs isolated from perinatal versus adult sources. Multiple earlier studies suggested that PT-MSCs may have superior immunomodulatory properties (116–123), higher proliferation (117, 124), and richer secretome (120, 125–127), compared to adult sources. In turn, some studies also reported that adult sources of MSCs have superior or similar immunomodulatory ability (126, 128, 129), and that adult MSCs have higher production rates of several vital molecular mediators [e.g. VEGF (126, 127), PLGF (126), IL-10 (130), and TGF-β1 (130)] when compared to perinatal cells. Intriguingly, when we performed meta-analysis for subgroups of our data, we found that the RR of the six studies with non-perinatal MSCs was more than factor two better than the RR of the 18 perinatal studies ( Table 6 ). All six of the non-perinatal studies, which employed five different MSC products, reported good outcomes. In contrast, among the 18 perinatal studies there were five studies where mortality on the MSC treatment arm was higher than the control baseline (40, 47, 48, 55, 56). However, this RR comparison was not statistically significant, leading us to argue that more data is needed to confirm or refute this result.

Another focus of our study was to evaluate the diversity of manufacturing methods used to prepare MSC products and to highlight the importance of reporting manufacturing information to enable study comparability ( Figure 3 ; Tables 3 , 4 ). Indeed, diversity in cell product manufacturing parameters, cell dosing, and cell characterization for therapeutic use, but also the completeness of study descriptors, have all been identified as a potential confounder to interpretation of safety and efficacy outcomes in MSC studies and should be monitored/reported more thoroughly in future studies (2, 16, 17, 33, 131–133). Recent reviews (131, 132) have highlighted the frequent lack in reporting of MSC manufacturing and study descriptors as a considerable shortcoming to clinical trial reporting and subsequent study interpretation. This aspect is of such importance, that it has been taken up into: “A modified Delphi Study Protocol” for “Establishment of a Consensus Definition for Mesenchymal Stromal Cells (MSC) and Reporting Guidelines for Clinical Trials of MSC Therapy” (133). For example, one multi-center randomized control trial (RCT) that employed UC-MSCs to treat COVID-19 found that MSC therapy had no efficacy (47). However, the MSCs in that study were manufactured in a laboratory that alternated between cell isolation with explants vs. enzymatic digestion (which may impact cell yield and immunophenotype) (134, 135). Still, the research consortium did not publish any records of which patient received which MSC product (47). Consequently, the outcome of the entire multi-center RCT may be cast into doubt because it is unknown how manufacturing variability may have compromised the respective results. Another manufacturing issue, which has been largely overlooked in the discussion of cell therapy trials for COVID-19 so far, is the ability to scale up the production of successful therapies (31, 136–139). Most trials published so far have relied on 2D monolayer cell expansion methods in flasks, with only one study employing a hollow-fiber bioreactor (42). The monolayer methods are time-consuming, labor intensive, and have limited scalability (135, 140, 141). Given the potential combinatorial detrimental impact of both freeze-thawing and in vitro aging on cell potency (33, 87–91, 135), it is of importance to accurately evaluate the number of population doublings a cell product has accumulated during ex vivo culture expansion for COVID-19 therapy and other indications (2, 14, 16). Of course, it is also relevant to record if the cell product was given either fresh or as a freeze-thawed product derived from cryostorage, since this may impact substantially on the product performance (14, 30, 31, 33, 88–91, 142–144).

The International Society for Cell and Gene Therapy (ISCT) recently published an editorial calling for a global registry of clinical trials that employ MSCs for COVID-19 to harmonize the data on the limited number of patients and “To collect information on critical process parameters used to manufacture the MSCs” (145). We support that call to action. Our review of the manufacturing parameters in clinical trials giving cell-based therapy for COVID-19 has revealed a partial disconnect between clinical centers that treat patients versus laboratories that manufacture cell therapy products. Frequently, the clinicians running the trials have acquired cells and delivered them to patients without keeping any records about the cell production. This disconnect could be closed, if the clinical trials participated in a global registry that required completing standardized categories of information. In this study we have compiled two years of worldwide clinical trials testing cell-based therapies for COVID-19 and linked those trials to their published outcomes. This “end-to-end” survey of the research field has enabled us to learn new insights not published earlier. First, we found that global registrations of advanced cell-based therapies for COVID-19 were more numerous than previously reported, but that they experienced only one single early surge in trial registrations during a time frame, when global COVID-19 infections went through multiple surges. Our analysis also includes the contribution from registered clinical trials that are not listed on the national registries of the United States and China, with 53 and 43 trials, respectively. Hereby, we have learned that Iran (19 trials) is among the three leading nations running advanced cell therapy trials for COVID-19 and the 2^nd in publishing trial outcomes. In turn, Israel, Spain, Iran, Australia, and Sweden are leading in relative contributions to COVID-19 cell therapy trials normalized to population size (N=0.641, 0.232, 0,223, 0.194, and 0.192 trials per million inhabitants).

Although 72% of the COVID-19 cell therapy trials employed tissue-derived MSCs, a significant fraction of clinical trials conducted immunotherapy with blood-derived cells. So far, most of the published trials describe infusions of MSCs, and 75% of those employed MSCs derived from perinatal tissue sources. Throughout these studies there is a strong theme of heterogeneity. The patient groups in the clinical trials are heterogeneous, as are the manufacturing methods used to prepare the MSCs. Most importantly, our statistical analysis shows that infusions of MSCs show a clinical benefit for COVID-19 patients. The risk ratio for all-cause mortality is RR=0.62 [95% CI 0.44 to 0.87] for the 17 MSC studies with control arms, and when we compensated for the missing controls and incorporated all 24 MSC studies the result is RR=0.63 [95% CI 0.46 to 0.85]. Early during the COVID-19 pandemic, it emerged that the respiratory distress caused by COVID-19 is a substantially different clinical entity compared to classic ARDS (21). Hence, we cannot assume that the efficacy results obtained here will similarly influence the long-standing quest to improve ARDS mortality. For this answer, we must await the outcomes of large placebo-controlled randomized trials of MSCs for ARDS, such as the REALIST trial in the UK, and the STAT trial in the US, respectively (146, 147). We close with the concern that there may never be enough data to fully explore the efficacy of cell-based therapy against ARDS from COVID-19. Owing to the evolution of the virus to less lethal variants and the rollout of vaccination, it has recently become difficult to accrue patients for clinical trials that treat severe symptoms of COVID-19 infection. The relative impact of MSC product source, MSC dosing, and the timing and type of MSC delivery, etc., may never be fully explored or known within the current setting of COVID-19.

From a public health perspective, the highest goal is to prevent the development of severe or critical COVID-19 through combined effective pandemic countermeasures ( Figure 1A ) (148). In this regard, vaccination is the most valuable tool available. Also, the standard-of-care for the treatment of severe and critical COVID-19 is continuously improving. Although we here found that infusions of MSCs confer a reduction in the risk for all-cause-mortality from COVID-19 in the studies published to date, more research is needed to clarify this point. Nonetheless, there will always be high-risk patients who develop severe or critical COVID-19, and for them the existence of adjunct treatment with advanced cell therapy may be beneficial. The target groups for whom this therapy may provide benefit include the elderly, immunocompromised individuals, cancer patients, and transplant patients (both stem cell transplants and solid organ transplants) as well as patients with kidney failure on dialysis (17, 149, 150). More research on the efficacy of advanced cell therapy for COVID-19 will reveal the degree to which these groups may benefit. Eventually, the cost and access to advanced cell therapy must also be anticipated, typically requiring advanced medical infrastructure. It is our hope that the testing of advanced therapies will be pursued in parallel to the improvement of standard care.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding authors.

PC and GM and FV conceived and designed this study. PC and AB and FV conducted primary data collection. PC and NA-A and IF and DF and RC and OC-M and GM and FV conducted the primary data analysis. PC and GM and FV wrote the first draft of the article. IH and RAC and OC-M. complemented and revised the manuscript and provided resources to support the study. All authors contributed to the article and approved the submitted version.