Current status of mesenchymal stem cell therapy for immune/inflammatory lung disorders: Gleaning insights for possible use in COVID‐19

Abstract The broad immunomodulatory properties of human mesenchymal stem cells (MSCs) have allowed for wide application in regenerative medicine as well as immune/inflammatory diseases, including unmatched allogeneic use. The novel coronavirus disease COVID‐19 has unleashed a pandemic in record time accompanied by an alarming mortality rate mainly due to pulmonary injury and acute respiratory distress syndrome. Because there are no effective preventive or curative therapies currently, MSC therapy (MSCT) has emerged as a possible candidate despite the lack of preclinical data of MSCs for COVID‐19. Interestingly, MSCT preclinical data specifically on immune/inflammatory disorders of the lungs were among the earliest to be reported in 2003, with the first clinical use of MSCT for graft‐vs‐host disease reported in 2004. Since these first reports, preclinical data showing beneficial effects of MSC immunomodulation have accumulated substantially, and as a consequence, over a third of MSCT clinical trials now target immune/inflammatory diseases. There is much preclinical evidence for MSCT in noninfectious—including chronic obstructive pulmonary disease, asthma, and idiopathic pulmonary fibrosis—as well as infectious bacterial immune/inflammatory lung disorders, with data generally demonstrating therapeutic effects; however, for infectious viral pulmonary conditions, the preclinical evidence is more scarce with some inconsistent outcomes. In this article, we review the mechanistic evidence for clinical use of MSCs in pulmonary immune/inflammatory disorders, and survey the ongoing clinical trials—including for COVID‐19—of MSCT for these diseases, with some perspectives and comment on MSCT for COVID‐19.


| INTRODUCTION
Human mesenchymal stem/stromal cells (MSCs) are multilineage somatic progenitors with broad immunomodulatory properties. Since initial isolation from the bone marrow (BM), MSCs have been found in numerous adult and fetal-derived organs/tissues such as adipose tissue, dental pulp, umbilical cord, and placenta. 1 In addition to trilineage paraxial mesodermal differentiation capacity toward bone, cartilage, and fat, the immunomodulatory properties of MSCs not only allow for expansion of therapeutic use from regenerative medicine to immuneand inflammation-related diseases, but also for third party allogeneic use. 2 The first published full report on clinical use of MSCs for immune/inflammatory disease was in 2004, in which allogeneic haploidentical bone marrow mesenchymal stem cell (BMMSC) infusions were given for a pediatric patient with acute refractory graft-vs-host-disease (GVHD). 3 Of note, median survival at that institution was a mere 2 months for the other 24 patients with similarly severe GVHD, while this patient remained well 1 year after MSC treatment. Surprisingly, prior to this clinical case report, there were only a handful of studies demonstrating MSC immunomodulation, with only one study showing in vivo data of prolonged skin engraftment. [4][5][6] Since then, MSC immunomodulation has shown to be broad-based, best detailed for CD4 lymphocytes but also for dendritic cells and natural killer cells. 7,8 The immunomodulatory properties are clinically relevant, as evidenced by the increasing proportions of MSC trials focusing on immune/inflammatory diseases which in recent years has accounted for approximately one-third of the trials. 9 One of the earliest reports demonstrating MSC immunomodulation was in reduction of bleomycin-induced pulmonary inflammation in mice. 10 It comes as somewhat of a surprise that clinical use of MSCs for lung diseases has been relatively slow to start, with most trials initiated in 2015. Moreover, it had been known for over a decade that intravenous delivery of MSCs-the most typical method of intervention for any cell therapy-results in the overwhelming majority of cells (80%90%) lodging in the lungs which is further increased with inflammation. 10,11 Hence, there is discussion that MSC therapy (MSCT) may be particularly useful in immune/inflammatory pulmonary conditions. 12 However, clinical trials for these diseases were still relatively few until this year: as of 17 May 2020, out of 68 MSC trials for lung immune/inflammatory diseases, 31 trials are specifically for COVID-19 as registered on the NIH Clinical Trial Database (https://ClinicalTrials.gov/) ( Figure 1). Due to the rapid global spread of COVID-19, the high mortality rate of those with severe disease, and no proven effective therapies as of yet, a desperate search for possible treatments is ongoing. 13 MSCT is clearly one such attempt, with new trials being added almost daily despite the lack of COVID-19-related preclinical data. In this review, we will examine the mechanistic evidence for clinical use of MSCs in pulmonary immune/inflammatory disorders, and survey the ongoing clinical trials-including for COVID-19-of MSCT for these diseases, with some perspectives and comments on MSCT for COVID-19.

| PRECLINICAL DATA OF MSCT FOR NONINFECTIOUS PULMONARY IMMUNE/ INFLAMMATORY DISORDERS
The lungs are in direct exposure to the external environment, requiring constant immune surveillance by the native epithelial cells and resident alveolar macrophages for homeostasis and health. 14 Immune dysregulation and inflammation, therefore, are common components in both infectious and many noninfectious pulmonary diseases, such as obstructive diseases including chronic obstructive pulmonary diseases (COPDs), in which injury is mainly mediated by cytotoxic T cells and neutrophils, and asthma, where type 2 helper T (Th2) lymphocytes and eosinophils are more predominant. 15 In restrictive diseases such as idiopathic pulmonary fibrosis (IPF), resident alveolar macrophages appear critical in mediating the fibrosis. 16 17,22,23 and keratinocyte growth factor (KGF) 24 to be involved ( Figure 2, left box). Rodent studies of asthma demonstrate that MSC induction of CD4 regulatory T cells (Tregs), which are immunomodulatory CD4 cells, is critical in decreasing Th2 responses 25,26 and Th2 cytokines interleukin-4 (IL-4), IL-5, and IL-13 as well as immunoglobulin E levels to ameliorate disease severity 27,28 ; one recent study implicated MSC transfer of mitochondria in this process. 29 Surprisingly, in these studies on asthma, no specific MSC paracrine factor was identified, but more recent reports implicate that MSC-expressed micro-RNA and exosomes can improve disease outcome. 30,31 Despite transforming growth factor-β being a prominent paracrine factor of MSCs 7,32 and also known for strongly inducing fibrosis, MSCT appears to be efficacious even for fibrotic pulmonary conditions, as evidenced by the early study of MSC efficacy for bleomycin-induced lung fibrosis, a preclinical disease model of IPF. 10 MSC-secreted IL-1 receptor antagonist (IL-1RA) was subsequently shown by the same group to be the paracrine factor involved. 33 Moreover, MSCs may enhance resident lung bronchioalveolar stem cells to repair and regenerate healthy lung parenchyma. 34 There has also been a growing number of reports using MSCs other than BMMSCs including umbilical cord MSCs (UCMSCs), 35 The immunomodulatory effects of MSCs may lead one reasonably to avoid using these cells in infectious diseases, especially bacterial infectious since a strong effector response is required for clearance of these rapidly growing microorganisms. But surprisingly, the preclinical data have been rather consistent on MSCs actually enhancing antibacterial processes and decreasing overexuberant immune responses leading to pulmonary injury and acute respiratory distress syndrome (ARDS), a complication which is still associated with high morbidity and mortality. 43 In rodent models of lung injury using either and IL-6, reverse pulmonary tissue damage, and improve survival through numerous paracrine factors including TNF-stimulated gene 6 F I G U R E 2 Mechanisms involved in MSC therapy for immune/inflammatory pulmonary disorders. Mechanisms reported in in vivo preclinical studies of MSC therapy for immune/inflammatory lung diseases of non-infectious etiology-including asthma, IPF, and COPDs-and infectious etiology-including bacterial and/or LPS and viral infection and related ARDS. Detailed descriptions can be found in the text. Ang-1, angiopoietin-1; ARDS, acute respiratory distress syndrome; COPD, chronic obstructive lung disease; EGF, epidermal growth factor; EV, extracellular vesicles; HGF, hepatocyte growth factor; IL-1RA, interleukin-1 receptor antagonist; IPF, idiopathic pulmonary fibrosis; KGF, keratinocyte growth factor; LPS, lipopolysaccharide; MΦ, macrophage; miRs, microRNAs; mitoch, mitochondria; MSC, mesenchymal stem cell; OCR, oxygen consumption rate; PMNs, polymorphonuclear leukocytes/neutrophils; Th2, T helper type 2 lymphocytes; TNF-α, tumor necrosis factor-α; Treg, regulatory T lymphocytes; TSF-6, TNF-stimulated gene 6 protein; VEGF, vascular endothelial growth factor; WBCs, white blood cells protein (TSG-6), 44 angiopoietin-1, 45,46 LL-37, 47 lipocalin-2, 48 KGF, 46,48,49 and microRNAs ( Figure 2, middle box). 50 Beneficial effects of MSCT in ex vivo human lung injury/bacterial infection models were seen as well. 51 Similar to a report for asthma, mitochondrial transfer from MSCT-either directly or through exosomesdecreased pulmonary injury and improved macrophage energetics and antibacterial functions. 52,53 Other studies have also found that MSCs modulate macrophages from an M1 inflammatory phenotype to a more immunomodulatory M2 phenotype, [54][55][56] as has been shown in nonpulmonary in vivo models. 57,58 It is surprising, however, that no in-depth investigation of MSCs with neutrophils, the first-line and critical leukocyte involved in bacterial clearance, was carried out any of these animal studies, since in vitro reports and one in vivo sepsis model have shown that MSCs preserve neutrophil viability and antibacterial functions. [59][60][61] But overall, these preclinical studies of bacterial-related lung injury/pneumonia consistently demonstrate that MSCT improves bacterial clearance and pulmonary tissue repair to impact survival.

| PRECLINICAL DATA OF MSCT FOR VIRAL PNEUMONIA AND COMPLICATIONS
Overall, reports on MSCT for viral infections are relatively scarce.

Most of the in vitro studies have been on the H1N1 influenza virus
using MSCs from many different organisms, finding that MSCs can be infected with resultant cell lysis and death. 62,63 For in vivo studies, there are currently only six reports which have examined intravenous MSCT for viral pneumonitis/pneumonia, all focusing on influenza. In the first two studies on the subject, the outcome was negative. Both studies evaluated syngeneic murine as well as allogeneic human BMMSC treatment in mice infected with pulmonary mouse-adapted H1N1 and/or swine H1N1, with no improvement in pulmonary inflammation or survival seen. 64,65 In the four other more recent reports, however, pulmonary inflammation was improved overall, with survival seen to improve in two out of the three studies which evaluated this endpoint; no specific factor was shown to be responsible ( Figure 2, right box). Interestingly, all three reports which evaluated survival used non-H1N1 subtypes. The only one beneficial report using H1N1 was a porcine study in which in vivo infection with swine H1N1 in 8-week-old pigs improved lung inflammation after intratracheal administration of syngeneic BMMSC extracellular vesicles; survival was not evaluated. 66 A report using H9N2 found syngeneic BMMSC treatment suppressed infection and improved survival in infected mice, 67 whereas another study using H5N1 found that conditioned medium and exosomes from human UCMSCs but not BMMSCs improved lung injury in infected mice partly due to two paracrine factors, angiopoietin-1 and HGF, but survival was only minimally improved. 68 A more elaborate murine study found that human BMMSCs reduce H5N1-induced lung injury and survival but only in aged mice, partially through the paracrine factors of angiopoietin-1 and KGF; the improved response in aged mice (8-12 months old) but not young mice (6-8 weeks old) to MSCT was in part attributed to more severe disease in these aged hosts, which may allow for exogenous MSCT to exert a more obvious benefit. 69 This report also further discussed that the tissue reparative properties of MSCs may only be apparent with the severe damage caused by highly pathogenic influenza subtypes including H5N1, which is not seen with the less pathogenic H1N1 subtype. The collective results of these in vivo studies, while few, would seem to support this viewpoint. Such differences in the infecting viral subtype and host conditions are unfortunately rarely tackled in preclinical studies, but in clinical practice, differences in patient profiles, including age and sex, are known to highly influence disease progression and outcome-as is strikingly evident with COVID-19, with higher positivity rates and worse outcome in men, and significantly higher mortality in the elderly and those with underlying chronic diseases. 70,71 Preclinical studies clearly should pay attention to such parameters for improved clinical use and outcome.

| CURRENT CLINICAL TRIALS OF MSCT FOR PULMONARY IMMUNE/ INFLAMMATORY DISEASES
To date, 68 clinical studies using MSCs for pulmonary immune/inflammatory disorders have been registered (Figure 1 and Table S1 for detailed information on each trial). The most commonly targeted disease is COVID-19 with 31 trials. MSCT for ARDS other than for COVID-19 and COPD are the next two most commonly targeted diseases, with 10 trials each. There are six trials for IPF, and two trials for asthma; the rest of the nine trials include trials for cystic fibrosis, lung transplantation, pneumoconiosis, radiation-caused injury, and unspecified lung injury.
MSC sources used are broad (Table 1 and Figure 3), with the two most common types being BMMSCs (22 trials) and UCMSCs (20 trials), then AdMSCs with 12 trials. A few trials use either dental pulp MSCs (two trials), placenta-MSCs (one trial), and olfactory mucosa MSCs (one trial); eight trials did not specify MSC source. As testament to the strong evidence for MSC immunomodulation, the majority of trials use allogeneic MSCs (49 trials, includes two trials using conditioned medium and exosomes), with all UCMSC trials being allogeneic; 10 trials use autologous sources which are either BMMSCs or AdMSCs, and nine trials were unspecified. There are only two trials using MSCderived products such as conditioned medium or exosomes rather than the cells themselves, and these products are derived from allogeneic AdMSCs or UCMSCs. Trials tend to be at early phases, with 30 being phase 1 trials, 17 being combined phase 1/2 trials, 14 trials in phase 2, 2 trials being a combined phase 2/3 trials, and 1 trial in phase 3; 3 trials are unspecified. As expected, the overwhelming majority of trials deliver MSCs intravenously (60 trials) but surprisingly, except for two trials which did not specified delivery method, the remaining six trials deliver MSCs to the lungs more directly, either through intratracheal/endobronchial delivery (three trials), intranasal delivery (one trial using UCMSC-conditioned medium), aerosolized inhalation (one trial using AdMSC-derived exosomes), or bronchial lavage (one trial) ( Figure 4).  88 There has also been much discussion on whether the use of fresh vs cryopreserved MSCs would have therapeutic implications. 89 In the six published clinical reports which all used allogeneic sources, two studies used freshly cultured MSCs 84,85 whereas the other four studies used previously cryopreserved MSCs [81][82][83] ; no clear difference in efficacy could be easily discerned. In addition, the delivery method and dose of cells given, and whether multiple doses should be given, as well as cell numbers used are also critical parameters that likely impact efficacy, but all are difficult to test in human studies. Further accumulation of preclinical data investigating these parameters is urgently needed for better tailoring of specific tissuesource MSCs, MSC preparation, as well as dosing regimens in clinical use to improve outcome.

| MSCT TRIALS FOR COVID-19: THE DEVIL MAY BE IN THE DETAILS
One startling discovery is that the majority of MSC clinical trials for immune/inflammatory lung disorders currently are for COVID-19, a disease that did not exist 6 months ago. 90 As of this writing, there are  (Table S2 with brief details for these trials).   Table S152 for trial details). While it is unclear whether the rapid increase in MSCT trials for COVID-19 is a factor in accelerating the registration of these trials, there is strong interest in using MSCT for severe pulmonary inflammation and complications given the continued lack of curative treatments for ARDS from any cause.   94 The preclinical data on MSCT decreasing TNF-α and IL-6 levels, two pro-inflammatory cytokines highly expressed during cytokine storm, 95

CONFLICT OF INTEREST
The authors declared no potential conflicts of interest.

DATA AVAILABILITY STATEMENT
Data sharing is not applicable to this article as no new data were created or analyzed in this study.