Cell‐based therapy to reduce mortality from COVID‐19: Systematic review and meta‐analysis of human studies on acute respiratory distress syndrome

Abstract Severe cases of COVID‐19 infection, often leading to death, have been associated with variants of acute respiratory distress syndrome (ARDS). Cell therapy with mesenchymal stromal cells (MSCs) is a potential treatment for COVID‐19 ARDS based on preclinical and clinical studies supporting the concept that MSCs modulate the inflammatory and remodeling processes and restore alveolo‐capillary barriers. The authors performed a systematic literature review and random‐effects meta‐analysis to determine the potential value of MSC therapy for treating COVID‐19‐infected patients with ARDS. Publications in all languages from 1990 to March 31, 2020 were reviewed, yielding 2691 studies, of which nine were included. MSCs were intravenously or intratracheally administered in 117 participants, who were followed for 14 days to 5 years. All MSCs were allogeneic from bone marrow, umbilical cord, menstrual blood, adipose tissue, or unreported sources. Combined mortality showed a favorable trend but did not reach statistical significance. No related serious adverse events were reported and mild adverse events resolved spontaneously. A trend was found of improved radiographic findings, pulmonary function (lung compliance, tidal volumes, PaO2/FiO2 ratio, alveolo‐capillary injury), and inflammatory biomarker levels. No comparisons were made between MSCs of different sources.


| INTRODUCTION
As a common enveloped RNA virus that crosses species, 1 COVID-19 has been associated with an intensive care unit (ICU) admission rate of 5% of proven infections 3 and an overall mortality rate in the range of 0.5% to 7%. 4 Among patients who require hospitalization, mortality may be approximately 5% to 15%. 3,5 Current treatment for patients with acute lung injuries is supportive, but with a high case fatality rate of 22% to as high as 88% for ICU patients. 3,5,6 Importantly, age represents a major risk factor for mortality. 6 New treatment modalities are needed to save lives by addressing the underlying pathophysiological processes that prevent oxygen exchange and destruction of the alveoli.
Similar to prior findings in SARS and MERS, COVID-19/SARS-CoV-2 in severe cases leads to fatal acute respiratory distress syndrome (ARDS), associated with monocyte and macrophage infiltration, diffuse alveolar damage, and cellular fibromyxoid exudates 7,8 with mortality reported as high as 52.4%. 4 The respiratory distress peaks at 7 to 10 days with manifestations of immune dysregulation, including cytokine release syndrome with elevation of cytokine levels (IL-6, IL-8, IL-1, IL2R, IL-10, and TNF-α), lymphopenia (in CD4+ and CD8+ T cells), and decreases in IFN-γ expression in CD4+ T cells. 8,9 The inverse correlation between cytokine storm with lower CD4+ and CD8+ counts suggests that the cytokine response may dampen adaptive immunity against COVID-19 infection, 10 which is associated with atrophy of the secondary lymphoid tissues. 7 Anti-inflammatory treatment has been proposed but challenged with the dilemma of balancing the risk of secondary infection. 11 Mesenchymal stromal cells (MSCs, also known as mesenchymal stem  12 MSCs have been studied as a promising candidate to treat certain inflammatory conditions and immunologic diseases based upon their well-characterized immunomodulatory effects, especially in the treatment of graft-vs-host disease, where MSC therapy was found to substantially improve complete response and overall survival. 13,14 The immunomodulatory activities are thought to include (a) inhibition of the proliferation and function of T cells, B cells, dendritic cells, and natural killer cells; (b) monocyte polarization to anti-inflammatory M2 macrophages; and (c) production of IL-10 and decreased production of TNF-α and IL-12. [15][16][17] In addition, MSCs have powerful antifibrotic effects and may alleviate lung fibrosis. 18,19 An urgent question is whether large-scale clinical trials and compassionate use of MSC therapy should be instituted to treat COVID-19-induced ARDS. To address this question, we systematically reviewed the available literature on safety, efficacy, and cytokine responses to MSC therapies in patients with ARDS.

| METHODS
The study protocol was finalized at the beginning of the project, which defined objectives, search strategy, inclusion/exclusion criteria, data extraction, outcomes of interest, and analytical approaches.

Significance statement
The potential benefits of mesenchymal stromal cell (MSC)

| Eligibility criteria
We included randomized controlled trials (RCTs), observational studies, case reports, and case series that evaluated safety and/or efficacy of stem cells administered to adult patients with a diagnosis of COVID-19 pneumonia or ARDS from any cause. MSCs that were culture-expanded or minimally manipulated were included. We included studies regardless of language of publication. Studies were excluded if they did not report original data (eg, clinical reviews, editorials, letters, or erratum).

| Outcome measures
The primary outcome was safety based on the frequency of serious adverse events (SAEs), adverse events (AEs), and if they were related to the treatment with MSCs.
Radiographic and computed tomographic (CT) findings were also included. were resolved by a third independent reviewer (W. Q.) ( Figure 1).

| Data collection
A standardized data extraction form was developed and piloted.
We extracted publication characteristics, study population, intervention and comparison details, and outcomes measures.

| Risk of bias/certainty in evidence
For RCTs, we used the Cochrane risk of bias tool to evaluate bias from sequence generation, allocation concealment, blinding of participants, personnel, and outcome assessor, incomplete outcome data, selective outcome reporting, and other sources of bias. For observational studies, case series, and case reports, we used the Newcastle-Ottawa Scale to evaluate representativeness of study population, ascertainment of exposure, comparability between groups (if applicable), outcome data source, and blinding of outcome assessment (Appendix 1, Tables 1 and 2). We used the GRADE approach (Grading of recommendations, development assessment and Evaluation) to rate certainty in the effect of MSC on mortality. 29

| Statistical analysis
We conducted meta-analysis to quantitatively summarize study findings based on the similarities of PICOTS (patient, intervention, control, outcome, timing, setting) presented by the studies. Mortality was the only outcome deemed to be appropriate for meta-analysis. included for the data extraction. Full text exclusion reasons are presented in Table 3.

| Study characteristics
Of the nine studies, there were four phase I clinical trials, three phase I/II or phase II clinical trials, one case report, and one case series. Five were comparative studies with control groups, including three RCTs.
All nine studies reported SAEs, AEs, mortality, and pulmonary function outcomes (n = 200). Six studies evaluated inflammatory markers (n = 98). A total of 117 patients received MSC therapy, while 83 participated as controls ( Table 1). The risk of bias of included randomized trials was high due to unclear random sequence generation and allocation concealment procedures.

| Patient characteristics
The studies were from six countries and regions, including China  (Table 1). MSCs in one study, 25 and umbilical cord blood MSCs in one study. 27 One study did not report the tissue origin of the MSCs and described their cell-product as commercially obtained 20 (

| Adverse events
Mild AEs were reported. One patient experienced a grade I allergic reaction. 23 In another study, patients experienced diarrhea in both treatment groups. 22  tainty in evidence about an effect on mortality was low, due to imprecision and methodological limitations of the included trials, as presented in Figure 2 and Table 5.

| Pulmonary function changes
All studies reported pulmonary function tests as outcome mea-

| Systemic changes and symptoms
Systemic changes were reported as any changes to overall patient status, including mental status, SOFA score, quality of life (SF-36), laboratory markers of organ function, symptoms, and relative improvements from baseline (Table 5). 20-22,24-28

| Inflammatory cytokines
Inflammatory markers and cytokines were evaluated in six studies.
These showed clear decreases in pro-inflammatory cytokines, including IL-1, IL-6, TNF-α, and CRP, and increased lymphocytes and IL-10 within 5 days of MSC therapy ( Findings from this study also suggest the impact of MSC therapy on important immunologic and inflammatory processes that lead to organ injury in COVID-19-infected patients. Cytokine release syndrome is a major underlying pathophysiological process in ARDS and was found to decrease with MSC therapy in seven of the studies analyzed. [20][21][22][23][24][25]28 In addition, MSCs have been hypothesized to neutralize free virus particles through the production of antibiotic proteins like LL37, which bind to virus and lung cell binding sites. 42 As recently reported, COVID-19 not only affects the lung, but also the heart and kidney with reported cardiomyopathy and kidney injury. 43,44 Two of the analyzed studies reported improved resolution of multiple organ failure or increased organ failure-free days with MSC treatment, which further supports their consideration for clinical use. Future clinical trials should include multiorgan failure parameters.
Senior age is associated with higher rate of mortality and has been reported to be associated with higher risk of death in patients with COVID-19. 45,46 In comparative studies included in this review, the age ranges were similar in both the MSC and the control groups at baseline and appeared to have minimal impact on outcomes.

| Limitations
The limitations of this systematic review include the lack of large-scale RCTs, the heterogeneity of outcome measures on pulmonary function, and the lack of studies collecting extra-pulmonary pathology. In addition, variability of MSC products prevents the assessment of which MSC type or preparation may be effective. Moreover, there may exist selection bias in published results and case reports that favors positive findings for MSC therapy for ARDS vs results that were equivocal or negative and not published.