Phase IIB, Randomized, Double-Blind, Placebo-Controlled Clinical Trial of Intravenous Defibrotide for the Prevention and Treatment of Respiratory Distress and Cytokine Release Syndrome in COVID-19.

Introduction

The coronavirus disease 2019 (COVID‐19) pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) resulted in > 760 million cases of COVID‐19 and 6.9 million deaths worldwide [ 1 , 2 ]. Severe COVID‐19 frequently progressed to acute respiratory distress syndrome (ARDS) [ 3 , 4 , 5 , 6 ]. The pandemic exposed critical vulnerabilities in global healthcare systems, with a scarcity of personal protective equipment (PPE) [ 7 ] and no vaccines or targeted therapies for severe disease [ 8 ] at the onset in 2020. Clinicians instead relied on repurposed drugs with limited efficacy; for example, dexamethasone reduced mortality by 17% in ventilated patients with COVID‐19 [ 9 ], remdesivir modestly accelerated recovery [ 10 ], and tocilizumab showed inconsistent benefits [ 11 , 12 , 13 ]. In Spain, during the first wave (February–May 2020), mortality rates exceeded 40% in mechanically ventilated patients with COVID‐19, and healthcare worker infection rates exceeded 15% due to PPE shortages [ 14 ]. Thus, the crisis demanded evaluation of therapies addressing both patient outcomes and staff safety through simplified administration protocols.

Early in the pandemic, the vascular endothelium was recognized as a critical therapeutic target in severe COVID‐19 [ 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 ]. SARS‐CoV‐2 infection triggers endotheliitis through direct viral invasion via angiotensin‐converting enzyme 2 receptors and subsequent rupture of alveolo‐capillary membranes, cytokine release, and activation of platelets and coagulation. This manifests clinically with vascular leakage and alveolar edema, an increased cytokine storm, generalized inflammation, and immuno‐thrombosis, plus elevated inflammatory and coagulation biomarkers that are strongly associated with mortality. Therapies mitigating these effects and providing endothelial protection may potentially be of benefit in COVID‐19 management.

Defibrotide, a polydisperse oligonucleotide with endothelial‐protective, anti‐inflammatory, and anti‐thrombotic properties, is approved for treating hepatic sinusoidal obstruction syndrome (SOS) at a dose of 6.25 mg/kg every 6 h (Q6hr) in adults, adolescents, and children aged > 1 year [ 27 , 28 ]. Preclinical studies demonstrated defibrotide could significantly reduce levels of endothelial activation markers intercellular adhesion molecule‐1 (ICAM‐1) and vascular cell adhesion molecule‐1 (VCAM‐1) by 62%–75%, suppress interleukin‐6 (IL‐6) and tumor necrosis factor‐α release ( p < 0.01), and improve survival rates in a murine model of hyperinflammatory disease such as graft‐versus‐host disease (87.5% vs. 0% controls) [ 29 ]. Because defibrotide directly targets endothelial activation, microvascular thrombosis, and cytokine‐driven vascular injury—all hallmarks of severe COVID‐19—we hypothesized that defibrotide would accelerate sustained clinical improvement in hypoxemic, hospitalized patients [ 30 , 31 , 32 ]. Furthermore, defibrotide is not incompatible with the COVID‐19 treatments used during the early stages of the pandemic [ 33 ]. Its efficacy in hepatic SOS is also supportive of a potential role in COVID‐19, as both conditions share features of severe thrombotic microangiopathy [ 34 , 35 ].

We therefore conducted a phase IIb, multicenter, randomized, double‐blind, placebo‐controlled trial to evaluate the efficacy and safety of continuous intravenous (IV) defibrotide in hospitalized patients with COVID‐19 and hypoxemic respiratory failure, that is, World Health Organization (WHO) severity scale grades 4–6 [ 36 ]. This study aimed to determine whether defibrotide's multimodal mechanisms of action could improve outcomes in severe COVID‐19 in terms of clinical improvement time and 30−/60‐day mortality, addressing an urgent unmet need. Furthermore, instead of using standard Q6hr dosing, we hypothesized that continuous IV infusion might maintain or improve defibrotide pharmacokinetics, achieving target coverage in the endothelium while avoiding peak concentrations that could be associated with hypotension, an important safety consideration in critically ill COVID‐19 patients, or other toxicities such as bleeding. The rationale for this approach was also to maintain sterility via closed systems, reducing nursing interventions by 80% compared to Q6hr dosing, and thereby mitigating PPE use and staff exposure at a time of acute resource shortages in Spain.

Methods

Study Design and Patients

This trial (DEFACOVID‐HCUVA; EUDRACT: 2020‐001409‐21; ClinicalTrials.gov NCT04348383 ) was conducted at eight tertiary hospitals in Spain (Table S1 ). The trial protocol and amendments were approved by the institutional review board (IRB) of University Clinical Hospital Virgen de la Arrixaca, IRBs at participating centers, and the Spanish Agency of Medicaments and Health Products. The trial was conducted in accordance with the principles of the Declaration of Helsinki and Good Clinical Practice guidelines and was compliant with the European Union's General Data Protection Regulation and Spanish data protection laws. An independent Data Monitoring Committee reviewed safety data every 30 patients (~2‐monthly schedule) during enrollment and at the end of the trial.

The trial population comprised high‐risk adult patients (aged ≥ 18 years) with a reverse transcription polymerase chain reaction (PCR)‐confirmed SARS‐CoV‐2 infection who were hospitalized and had Grade 4, 5, or 6 clinical status per the WHO 7‐category ordinal scale for COVID‐19 severity (Table S2 ) [ 36 , 37 ]. Other eligibility criteria are listed in Supporting Information Methods. All participants provided written informed consent.

Treatment and Randomization

Patients were randomized 2:1 to defibrotide or placebo, plus standard care and conventional COVID‐19 therapies per institutional protocols at time of inclusion. Defibrotide 25 mg/kg/day was administered as a continuous IV infusion in 250 mL saline (rate: 10.4 mL/h) for 15 days. Matching placebo (0.9% saline) was administered using identical infusion protocols. Shorter treatment duration (minimum 7 days) was permitted in patients who improved clinically (no oxygen requirement/return to baseline oxygen requirement) and could be discharged home. Randomization was stratified by disease severity (WHO scale 4/5 vs. 6) using computer‐generated blocks via a centralized electronic system (Biomédica‐IMIB [Biomedical Research Institute of Murcia] platform). Information on blinding is provided in Supporting Information Methods.

Endpoints and Assessments

The primary endpoint was “clinical improvement time,” defined as the cumulative number of days alive between first achievement of improvement (≥ 1 category) on the WHO scale for COVID‐19 severity and Day 30, inclusive. Patients who died before Day 30 were assigned 0 days. This endpoint, adapted from ordinal‐scale response endpoints in ACTT‐1 and related COVID‐19 trials [ 38 , 39 ], was selected to maximize sensitivity for detecting a treatment effect and integrate both onset and durability of improvement, avoiding informative censoring due to early death or relapse. Secondary safety, efficacy, biomarker, and radiographic endpoints are summarized in Supporting Information Methods. Patients were evaluated at baseline (Day 0), daily during treatment (Days 1–14), with secondary endpoint assessment on Day 7, and during follow‐up (Days 15, 30, 60). Assessments and adverse reactions of interest for defibrotide are summarized in Supporting Information Methods.

Statistical Analysis

The sample size of 150 patients (100/50 in the defibrotide/placebo arms) provided 80% power ( β = 0.20) to detect an increase in mean clinical improvement time from 12 days with placebo (based on data from ACTT‐1 [ 38 , 39 ]) to 16 days with defibrotide (two‐sided α = 0.05), assuming a 10% drop‐out rate. For the primary efficacy analysis, statistical comparison between arms was done using the Wilcoxon rank‐sum test. For other statistical comparisons, p ‐values < 0.05 were considered nominally significant. Full statistical methodology is provided in Supporting Information Methods.

Results

Patients

Between April 14, 2020, and June 29, 2021, 150 patients were enrolled at six participating sites (Table S1 ) and randomized, 100 to defibrotide and 50 to placebo (Figure S1 ). Overall, median age was 61.0 years, 78.0% of patients were male, mean body mass index was 29.6 kg/m 2 , and WHO COVID‐19 severity category was 4, 5, and 6 in 24.0%, 49.3%, and 26.7%, respectively, with no differences between arms (Table 1 ). The most common comorbidity was hypertension (45.3%), with lower prevalence in the defibrotide versus placebo arm (38.0% vs. 60.0%, p = 0.01). Other common comorbidities showed similar prevalences between arms (Table 1 ). Baseline d ‐dimer levels were higher in the defibrotide versus placebo arm (Table 1 ).

Clinical Improvement Time

For the primary endpoint, there was no statistically significant difference between the defibrotide and placebo arms (Figure 1A ). Median clinical improvement time was 15.0 days (IQR: 0–24) with defibrotide and 20.0 days (IQR: 9–25) with placebo, with no statistical difference between arms (Wilcoxon rank‐sum test, p = 0.10). Analyses of clinical improvement time by baseline disease severity did not show differences between arms ( p > 0.05) (Figure 1B–D ).

Using a more stringent measure of clinical improvement—sustained reduction to Category 3 (WHO 7‐point severity scale)—median clinical improvement time was 0 days (IQR: 0–17) with defibrotide and 0 days (IQR: 0–19) with placebo (Figure 1A ), with no significant difference between arms ( p = 0.57). The wide IQRs in both arms for this endpoint, and the median of 0 days reflecting the proportion of patients who did not achieve this more stringent threshold, indicate greater inter‐patient variability in clinical improvement time when applying more stringent improvement criteria compared with the primary endpoint.

Safety

Overall, 131/150 patients (87.3%) had any adverse event (AE), including 91 (91%) and 40 (80%) in the defibrotide and placebo arms, respectively ( p = 0.06) (Table 2 ). In total, 820 AEs were recorded, 626 and 194 in the defibrotide and placebo arms, respectively, of which 57 were serious AEs (SAEs; 38 and 19 in the defibrotide and placebo arms, respectively, occurring in 34 and 18 patients). Overall, 702 AEs (85.6%) were classified as unrelated to treatment. Percentages of non‐serious AEs considered unlikely, possibly, or definitely related to treatment were low (≤ 7.6%), with no differences between arms (Table 2 ). Percentages of SAEs considered unlikely, possibly, or definitely related to treatment were ≤ 5.3%.

Severity information was available for 816/820 AEs: 546 (66.9%) were Grade 1, 135 (16.5%) Grade 2, and 135 (16.5%) Grade ≥ 3, including 31 (3.8%) Grade 5 (fatal) AEs (Table 2 ). The distribution of severity grades did not differ between treatment arms, and no association was observed between treatment arm and overall severity distribution ( p = 0.945).

The most frequently reported AEs were elevated liver enzymes (68 events), anemia (53), fever (39), and hypotension (37) (Table 2 ). Most events of elevated liver enzymes and anemia were grade 1/2 (Table 2 ), suggesting limited clinical impact. At the patient level, anemia occurred in 50 patients (33.3%), elevated liver enzymes in 44 (29.3%), hypoalbuminemia in 26 (17.3%), fever in 23 (15.3%), hypotension in 22 (14.7%), pain in 19 (12.7%), and hemorrhage in 17 patients (11.3%); anemia, elevated liver enzymes, and hypoalbuminemia were more common in the defibrotide versus placebo arm (Table 2 ). Although hypotension and hemorrhage were numerically higher with defibrotide, these differences, plus differences in fever and pain, were not significant. The distribution of severity grade for hypotension and hemorrhage did not differ significantly between arms.

Mortality

Overall mortality was 22.7% (34/150) at Day 30 and 25.3% (38/150) at Day 60. Day‐30 mortality was similar in the defibrotide and placebo arms (23/100, 23%, and 11/50, 22%, respectively). By Day 60, mortality was 27.0% (27/100) with defibrotide and 22.0% (11/50) with placebo, with no significant difference between arms ( p = 0.89) (Figure S2 ). The leading causes of death by day 60 ( n = 38) were multiorgan failure (13, 34.2%; defibrotide 9, placebo 4), respiratory failure (13, 34.2%; defibrotide 9, placebo 4), cardiorespiratory arrest (5, 13.2%; defibrotide 4, placebo 1), and severe thrombotic events (3, 7.9%; defibrotide 1, placebo 2). Four deaths were due to metabolic coma, necrotizing pneumonia, refractory shock with oliguria and metabolic acidosis, and distributive septic shock (each 1, 2.6%); all occurred in the defibrotide arm and were not considered treatment‐related.

Mean time from treatment initiation to death was 19.5 days (standard deviation [SD] 12.0; median 16.0 days, IQR: 11.2–23.2) in the defibrotide arm ( n = 26) and 12.1 days (SD 9.0; median 10.0 days, IQR 6.0–18.5) in the placebo arm ( n = 11). The longer time to death in the defibrotide arm was not statistically significant compared with placebo ( p = 0.065), despite the numerical difference and the exploratory biomarker effects described below.

Evaluation of concomitant COVID‐19 medications as a potential confounding factor in the mortality analysis showed no differences between arms (Table 3 ).

Changes in FiO 2 , and Mechanical Ventilation and Hospital/Intensive Care Unit ( ICU ) Requirements

Fraction of inspired oxygen (FiO 2 ) was assessed over time to compare respiratory improvement between arms (Table 3 ). The mean reduction in mean FiO 2 from Day 1 to end of treatment (Day 15) was numerically greater with defibrotide versus placebo (28.6 vs. 26.6), but this difference was not statistically significant. The ICU admission rate was 72.0% in both arms. Extracorporeal membrane oxygenation (ECMO) was required more often in the defibrotide versus placebo arm (12.0% vs. 4.0%), but this difference was not statistically significant. Dialysis requirement was similar between arms. In surviving patients, mean duration of hospitalization was 25.8 versus 21.7 days in the defibrotide versus placebo arm (median 16.5 vs. 14.0 days; p = 0.13) (Table 3 ).

Biomarkers

Biomarker analyses were exploratory and hypothesis‐generating. Seven biomarkers were assessed at baseline and on Days 7, 15, 30, and 60: IL‐6 (inflammation), absolute lymphocyte count (ALC; immune recovery), d ‐dimer (coagulation, fibrinolysis), C‐reactive protein (CRP; inflammation), ferritin (inflammation), lactate dehydrogenase (LDH; tissue injury, inflammation), and creatine phosphokinase (CPK; muscle injury, ischemia). Baseline values were generally similar between arms, except for a higher d ‐dimer level in the defibrotide versus placebo arm ( p < 0.01) (Table 1 ).

Changes from baseline were analyzed as the proportions of evaluable patients achieving either a ≥ 50% reduction or (for ALC) increase or normalization of levels (per established reference ranges, Table 1 ) (Figure 2 ). For IL‐6, the proportion of patients with a ≥ 50% reduction increased from 28.0% (Day 7) to 42.5% (Day 60) in the defibrotide arm but remained relatively stable in the placebo arm (38.0%–42.0%). IL‐6 normalization rates were lower with defibrotide on Days 7 and 15 but increased and were similar between arms on Days 30 and 60 (Figure 2A ).

ALC recovery (≥ 50% increase from baseline) and normalization rates were consistently higher or similar with defibrotide compared with placebo (Figure 2B ). The rate of ≥ 50% increase in ALC on Day 7 was significantly higher with defibrotide (67.0% vs. 50.0%, p = 0.05).

Consistently higher rates of ≥ 50% reductions in d ‐dimer levels were seen with defibrotide versus placebo across time‐points, notably on Day 7 (23.0% vs. 8.0%, p = 0.03) and Day 15 (31.0% vs. 12.0%, p = 0.02) (Figure 2C ) that might reflect a trend toward improving hemostatic recovery, although this requires validation. Conversely, d ‐dimer normalization rates were numerically higher in the placebo group across time‐points, potentially associated with higher baseline d ‐dimer levels in the defibrotide arm offering greater potential for achieving ≥ 50% reductions but making normalization more challenging.

Rates of ≥ 50% reductions in CRP levels were high and comparable between arms at all time‐points (Figure 2D ). CRP normalization rates remained low in both arms. Lower rates of ≥ 50% reductions and normalization of ferritin levels were seen with defibrotide versus placebo across time‐points, notably on Day 7 (1.0% vs. 18.0%, p < 0.01, and 5.0% vs. 18.0%, p = 0.02, respectively), although rates in both arms increased during follow‐up (Figure 2E ).

There were relatively low rates of ≥ 50% reductions and normalization of LDH in both arms across time‐points, with gradual increases over time and no consistent difference between arms (Figure 2F ). For CPK, ≥ 50% reductions were similar between arms at Days 7 and 15 but higher with defibrotide at days 30 (34.0% vs. 18.0%, p = 0.04) and 60 (36.0% vs. 22.0%) (Figure 2G ). CPK normalization rates were consistently high and similar in both arms.

Pulmonary Infiltrates

Severity of pulmonary inflammatory disease on serial chest radiographs was determined using an 8‐point severity score [ 40 ], with continuous quantitative analysis of score over time plus categorical evaluation ( Supporting Information Methods). Continuous quantitative analysis using multiple linear regression models showed that baseline severity score was a predictor of ongoing pulmonary involvement at all analysis time points (Days 7, 15, 30, and 60), with the strongest association on Days 7 and 15 ( p < 0.01) and attenuation of effect size by Day 60 ( p < 0.05) (Figure S3 ).

Defibrotide treatment was not a predictor for pulmonary involvement on Days 7, 15, and 30 but showed a marginal association on Day 60, potentially suggesting slightly less improvement in the defibrotide versus placebo arm at this time‐point. The explanatory power of the regression models was strongest on Day 7 and declined over subsequent time‐points. The constant term, representing expected severity scores for patients in the placebo arm with baseline scores of 0, was a predictor of severity on Day 7 ( p < 0.01) and had a marginal effect on Day 15 ( p < 0.1), indicating measurable pulmonary involvement even in ostensibly “normal” patients at baseline, potentially due to contributions from unmodeled clinical factors.

Figure S3 shows regression lines for the relationship between baseline severity score and scores on Days 7, 15, 30, and 60. The steeper slopes of the defibrotide versus placebo regression lines on Days 30 and 60 suggest that severity of pulmonary involvement at these time‐points was more dependent on baseline severity in the defibrotide arm. These findings reflect more variable improvement patterns, by which defibrotide‐treated patients with higher baseline scores tended to maintain relatively worse severity scores at later time points.

Figure 3 shows categorical changes from baseline in severity of pulmonary involvement, with patients classified as remaining stable, improving, or worsening by at least one severity category compared to baseline. With both defibrotide (Figure 3A ) and placebo (Figure 3B ), percentages of patients with improvement from baseline increased from Days 7 to 60, with no differences between arms ( p ≥ 0.31).

Discussion

To our knowledge, this was the first randomized controlled trial of defibrotide as an endothelial‐protective agent in severe COVID‐19. Although continuous IV infusion of defibrotide did not improve clinical outcomes versus placebo, the study confirmed its safety in critically ill, hypoxemic patients, demonstrated pharmacodynamic signals in coagulation/inflammation biomarkers, and identified a likely subtherapeutic exposure profile with continuous IV infusion versus the approved Q6hr dosing. Therefore, these data redefine how defibrotide should be studied in syndromes driven by endothelial injury and suggest an important biological effect in this setting.

The COVID‐19 pandemic highlighted the urgent need for effective treatments, particularly, in severe cases progressing to ARDS and cytokine storm. Given limited therapeutic options beyond remdesivir and dexamethasone, and emerging evidence of endothelial dysfunction in COVID‐19 pathogenesis, we investigated defibrotide as a potential therapy. To our knowledge, there had been no preclinical studies evaluating defibrotide in animal models of SARS‐CoV‐2 infection. However, in a murine model of graft‐versus‐host disease, defibrotide significantly reduced levels of adhesion molecules and cytokines that are critically elevated in the cytokine release syndrome associated with COVID‐19 and, importantly, improved animal survival [ 29 ]. In addition, the mechanistic rationale for defibrotide was supported by the endothelial dysfunction demonstrated in COVID‐19 [ 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 ] and defibrotide's demonstrated safety and endothelial‐protective properties [ 35 , 41 , 42 ].

This multicenter trial enrolled a population that was balanced between arms and reflected both the geographic spread of participating institutions and the variable impact of COVID‐19 caseloads during different pandemic phases in Spain, enhancing the generalizability of findings while accounting for regional differences in patient populations and hospital protocols. Moreover, causes of patient mortality in this trial reflect the severity of respiratory and vascular involvement in critically ill COVID‐19 patients and align with complications described in the literature, particularly in the context of endothelial dysfunction and proinflammatory states [ 4 , 5 , 15 , 16 , 17 ].

Conducting this trial during the early waves of the pandemic, with limited PPE resources and no available vaccine, was challenging but nevertheless justified by the unmet medical need and substantial clinical and mechanistic rationale. The circumstances presented multiple complexities, requiring consideration of numerous factors including logistics of treatment delivery, safety of healthcare professionals, and the evolving understanding of COVID‐19. Thus, although the trial did not demonstrate the benefit of defibrotide in clinical improvement time, it provided valuable lessons and key findings informing further investigation of defibrotide in this and related settings.

A key hypothesis emerging from this trial, in the context of other evaluations of defibrotide during the COVID‐19 pandemic [ 30 , 31 , 32 ], was the potential importance of intermittent Q6hr dosing rather than continuous IV infusion. At the time of trial design, given the challenges associated with clinical administration of therapy during lockdown, the decision to use continuous IV infusion instead of the approved Q6hr schedule was operationally justified, as it reduced exposure of healthcare staff to patients with severe COVID‐19 and provided a simplified workflow for the medical and nursing teams. However, this strategy may have limited the ability to detect a clinically significant benefit. In contrast, other studies using conventional Q6hr administration of defibrotide have demonstrated promising efficacy in COVID‐19 ARDS [ 30 , 31 , 32 ], including the Italian DEFI‐VID19 study, which showed a significant increase in post‐recovery days and a trend toward improved respiratory failure‐free survival with defibrotide versus a contemporaneous, real‐world control cohort receiving standard‐of‐care therapy only [ 31 ]. These differing results may reflect exposure differences between continuous IV infusion of 25 mg/kg/day and Q6hr dosing of 6.25 mg/kg by 2‐h infusion. Continuous IV infusion (25 mg/kg over 24 h; ~1.04 mg/kg/h) likely produced a substantially lower peak plasma concentration ( C max ), potentially reducing endothelial protective activity, compared to Q6hr dosing (6.25 mg/kg over 2 h; ~3.125 mg/kg/h), for which the reported pharmacokinetic profile shows a C max of 17.3 ± 3.83 μg/mL [ 41 ]. Although clinical data at subtherapeutic exposure are lacking, this hypothesis is supported by dose–exposure pharmacodynamic relationships: lower doses are associated with markedly reduced tissue factor pathway inhibitor release, a key mechanism of action of defibrotide, in preclinical and clinical studies [ 43 , 44 ].

Additionally, DEFI‐VID19 enrolled its defibrotide‐treated patients over an 8‐month period [ 31 ], which was shorter than the 14‐month enrollment window of our trial that spanned four distinct COVID‐19 waves in Spain (Table S3 ). It is therefore possible that a higher proportion of patients in our trial were infected with the original SARS‐CoV‐2 strain, whereas the Italian cohort may have been dominated by infections during the Alpha‐variant wave [ 31 ]. Differences in circulating variants and in vaccination coverage between the two study periods could have influenced the observed differences in defibrotide efficacy.

The rapidly evolving context of the COVID‐19 pandemic also posed major challenges to trial conduct. For instance, the WHO 7‐point ordinal COVID‐19 severity scale [ 36 ] used at the time of study design was subsequently superseded by 8‐point [ 38 ] and later 10‐point [ 45 ] versions, the latter incorporating oxygen saturation levels. Moreover, evolving clinical practice and logistical constraints made it unfeasible to assess several of the originally planned secondary endpoints during the trial.

From a safety standpoint, this trial is important: continuous IV infusion of defibrotide 25 mg/kg/day did not increase serious bleeding, hemodynamic instability, or treatment‐related SAEs versus placebo, even in patients requiring high levels of ventilatory support, ECMO, or renal replacement therapy. This supports the feasibility of administering defibrotide in severely ill COVID‐19/ARDS populations and, more broadly, in syndromes of systemic endotheliitis.

Continued evaluation of defibrotide in patients with severe COVID‐19 was planned based on the present study's findings and the dosing hypothesis described above. Protocol amendments were approved by local IRBs and national regulatory authorities to continue the trial using Q6hr dosing of defibrotide. However, due to very slow accrual associated with the changing clinical course of COVID‐19 following widespread vaccination, the pharmaceutical sponsor recommended closure of the trial, and the study leadership complied. Evaluation of biomarkers of defibrotide activity remains an important ongoing objective. Despite baseline clinical characteristics being generally representative of COVID‐19 cases, imbalances were identified that may have influenced comparative efficacy. Baseline D‐dimer levels were significantly higher and CRP levels trended higher in the defibrotide versus placebo arm, suggesting higher endothelial stress at enrollment. Nevertheless, overall biomarker patterns might reflect a trend toward favorable effects on thrombotic pathway modulation, although these data require validation, and continuous IV infusion may have limited the endothelial protective effects of defibrotide.

In conclusion, continuous IV defibrotide was safe but did not improve clinical outcomes in severe COVID‐19. Further studies exploring intermittent (Q6hr) dosing and/or the potential role of defibrotide in mitigating post‐COVID‐19 vascular sequelae are ongoing. Specifically, additional insights are anticipated from an ongoing phase II study of Q6hr defibrotide for patients with severe COVID‐19 ( NCT04652115 ). The DEFACOVID study group is continuing analysis of biologic samples from this trial to further elucidate the mechanistic actions of defibrotide and the pathophysiology of endothelial dysfunction in COVID‐19, findings that may have broader relevance for other conditions characterized by endotheliitis and ARDS, for example, complications of immunotherapy in hematologic malignancies, including multiple myeloma [ 34 , 46 ].

Author Contributions

Conceptualization and study design: R.J.R., A.J.M.‐M., D.G.‐B., F.I., P.G.R., J.M.M. Treated patients and collected data: R.J.R., A.J.M.‐M., B.G.‐P., P.C., M.D.‐R., A.C.‐A., A.L.‐B., E.B.M., A.R., S.M., F.‐J.R.‐L., E.S.‐M., D.M.‐B., C.L.A.M., E.A.S., A.T.S., S.F.M., J.A.N.V., I.C.L., A.P., A.E.‐M., M.B., A.S.‐S., C.H.G., M.M., A.M.P., A.M.‐D., M.M.G., J.M.M. Analyzed the data: R.J.R., A.J.M.‐M., J.K.‐C., J.G.C., A.M.T.‐C., M.S., M.I., C.C.‐S., L.‐J.W., P.G.R., J.M.M. Performed statistical analyses: J.K.‐C., J.G.C., A.M.T.‐C., L.‐W. Interpreted the data: R.J.R., A.J.M.‐M., M.S., M.I., C.C.‐S., L.‐J.W., P.G.R., J.M.M. Wrote the paper: R.J.R., A.J.M.‐M., P.G.R., J.M.M. Reviewed and edited the paper: All authors. Approved the final version of the paper for submission: All authors.

Funding

This study was sponsored by Jazz Pharmaceuticals (IST‐20‐11291) and the Instituto de Salud Carlos III (ISCIII) through the project COV20/00399 to José M. Moraleda. The study was also sponsored in part by the Paula and Rodger Riney Multiple Myeloma Research Fund to José M. Moraleda and Paul G. Richardson. José G. Clavel thanks Fundación Seneca for its support (22905/EE/25) of the final stage of this study.

Ethics Statement

The trial protocol and amendments were approved by the institutional review board (IRB) of University Clinical Hospital Virgen de la Arrixaca, IRBs at participating centers, and the Spanish Agency of Medicaments and Health Products. The trial was conducted in accordance with the principles of the Declaration of Helsinki and Good Clinical Practice guidelines and was compliant with the European Union's General Data Protection Regulation and Spanish data protection laws.

Consent

All participants provided written informed consent.

Conflicts of Interest

J.K.‐C. has received an honorarium for a lecture on herpes zoster vaccine from GSK. P.C. has received honoraria for scientific collaborations from Pfizer, MSD, Gilead, AbbVie, and Mundipharma and support for attending meetings and/or travel from Menarini and Pfizer, and has participated in advisory boards for Alexion, Janssen, Sanofi, and Gilead. M.D.‐R. has received honoraria for lectures from Jazz Pharmaceuticals and Zacros and equipment, materials, drugs, medical writing, gifts or other services from Zacros. S.F.‐M. has received honoraria for lectures from Gilead and support for attending meetings from Pfizer. P.G.R. has received grants from Oncopeptides and Karyopharm, and consulting fees from Celgene/BMS, GSK, Karyopharm, Oncopeptides, Regeneron, and Sanofi. J.M.M. has received honoraria for lectures from Roche and payments for Data Safety Monitoring Board participation from Rocket Pharmaceuticals Inc. The other authors declare no conflicts of interest.

Supporting information

Table S1: Tertiary hospitals in Spain that participated in the trial, plus site enrollment.

Table S2: The World Health Organization 7‐category ordinal scale for COVID‐19 severity [5].

Table S3: Summary of COVID‐19 waves in Spain, 2020–2021.

Figure S1: CONSORT flow diagram.

Figure S2: Kaplan–Meier analysis of survival/mortality through to Day 60.

Figure S3: Severity of pulmonary involvement on chest radiography at baseline and over time.

Acknowledgments

The authors gratefully acknowledge other members of the Internal Medicine, Pulmonology and Intensive Care Medicine departments of the hospitals that participated in the recruitment and care of the patients, as well as the patients and their families. The authors thank the Biobank of the Instituto Murciano de Investigación Biosanitaria (IMIB, Murcia, Spain; registration number B.0000859, Spanish National Biobank Network, ISCIII) for its collaboration in the storage and management of the biological samples used in this study. The authors acknowledge Mr. Ángel Esteban, from the Information Technology Service of the IMIB, for his valuable collaboration and technical support. The authors thank the Clinical Research Unit (IMIB's Clinical Trials Platform) for its comprehensive support in regulatory affairs, pharmacovigilance, data monitoring, and GCP compliance. The authors also acknowledge its role as a SCReN (Spanish Clinical Research Network) node (PT23/00029) for monitoring all participating centers. The authors gratefully acknowledge Steve Hill, PhD, of Ashfield MedComms, an Inizio company, for medical writing and editing support, funded by Dana‐Farber Cancer Institute, the RJ Corman Multiple Myeloma Research Fund, and the Paula and Rodger Riney Multiple Myeloma Research Fund.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

References

1 B. Wichmann and R. M. Wichmann , “ Big Data Evidence of the Impact of COVID‐19 Hospitalizations on Mortality Rates of Non‐COVID‐19 Critically Ill Patients ,” Scientific Reports 13 ( 2023 ): 13613 , 10.1038/s41598-023-40727-z . 37604881 PMC10442321

2 World Health Organization , “Coronavirus Disease (COVID‐19),” 2023, accessed September 17, 2025, https://www.who.int/news‐room/fact‐sheets/detail/coronavirus‐disease‐(covid‐19) .

3 C. Wu , X. Chen , Y. Cai , et al., “ Risk Factors Associated With Acute Respiratory Distress Syndrome and Death in Patients With Coronavirus Disease 2019 Pneumonia in Wuhan, China ,” JAMA Internal Medicine 180 ( 2020 ): 934 – 943 , 10.1001/jamainternmed.2020.0994 . 32167524 PMC7070509

4 F. Zhou , T. Yu , R. Du , et al., “ Clinical Course and Risk Factors for Mortality of Adult Inpatients With COVID‐19 in Wuhan, China: A Retrospective Cohort Study ,” Lancet 395 ( 2020 ): 1054 – 1062 , 10.1016/S0140-6736(20)30566-3 . 32171076 PMC7270627

5 G. Grasselli , T. Tonetti , A. Protti , et al., “ Pathophysiology of COVID‐19‐Associated Acute Respiratory Distress Syndrome: A Multicentre Prospective Observational Study ,” Lancet Respiratory Medicine 8 ( 2020 ): 1201 – 1208 , 10.1016/S2213-2600(20)30370-2 . 32861276 PMC7834127

6 E. Iglesias‐Julian , M. Lopez‐Veloso , N. de‐la‐Torre‐Ferrera , et al., “ High Dose Subcutaneous Anakinra to Treat Acute Respiratory Distress Syndrome Secondary to Cytokine Storm Syndrome Among Severely Ill COVID‐19 Patients ,” Journal of Autoimmunity 115 ( 2020 ): 102537 , 10.1016/j.jaut.2020.102537 . 32843231 PMC7439828

7 Y. Alhassan , Z. Zaizay , L. Dean , et al., “ Perceived Impacts of COVID‐19 Responses on Routine Health Service Delivery in Liberia and UK: Cross‐Country Lessons for Resilient Health Systems for Equitable Service Delivery During Pandemics ,” BMC Health Services Research 23 ( 2023 ): 304 , 10.1186/s12913-023-09162-8 . 36991477 PMC10057690

8 M. A. Martinez , “ Compounds With Therapeutic Potential Against Novel Respiratory 2019 Coronavirus ,” Antimicrobial Agents and Chemotherapy 64 ( 2020 ): e00399‐00320 , 10.1128/AAC.00399-20 . 32152082 PMC7179632

9 RECOVERY Collaborative Group , P. Horby , W. S. Lim , et al., “ Dexamethasone in Hospitalized Patients With Covid‐19 ,” New England Journal of Medicine 384 ( 2021 ): 693 – 704 , 10.1056/NEJMoa2021436 . 32678530 PMC7383595

10 G. Lapadula , D. P. Bernasconi , G. Bellani , et al., “ Remdesivir Use in Patients Requiring Mechanical Ventilation due to COVID‐19 ,” Open Forum Infectious Diseases 7 ( 2020 ): ofaa481 , 10.1093/ofid/ofaa481 . 33204761 PMC7651598

11 J. Aguareles , C. Forne , A. Garcia‐Casas , et al., “ Influence of Anti‐Spike Protein Antibody Levels on Tocilizumab Efficacy in Hospitalized Patients With Severe COVID‐19 Pneumonia: A Post‐Hoc Analysis of the COVACTA Trial ,” BMC Infectious Diseases 25 ( 2025 ): 676 , 10.1186/s12879-025-11001-6 . 40340657 PMC12063319

12 C. C. Price , F. L. Altice , Y. Shyr , et al., “ Tocilizumab Treatment for Cytokine Release Syndrome in Hospitalized Patients With Coronavirus Disease 2019: Survival and Clinical Outcomes ,” Chest 158 ( 2020 ): 1397 – 1408 , 10.1016/j.chest.2020.06.006 . 32553536 PMC7831876

13 I. O. Rosas , N. Brau , M. Waters , et al., “ Tocilizumab in Hospitalized Patients With Severe Covid‐19 Pneumonia ,” New England Journal of Medicine 384 ( 2021 ): 1503 – 1516 , 10.1056/NEJMoa2028700 . 33631066 PMC7953459

14 Red Nacional de Vigilancia Epidemiológica (RENAVE) , “Informe sobre la situación de COVID‐19 en España: Informe COVID‐19 n° 15, 25 de marzo de 2020,” 2020, accessed August 12, 2025, https://cne.isciii.es/documents/d/cne/informe‐n‐15‐situacion‐de‐covid‐19‐en‐espana‐a‐25‐marzo‐de‐2020‐pdf .

15 M. Ackermann , S. E. Verleden , M. Kuehnel , et al., “ Pulmonary Vascular Endothelialitis, Thrombosis, and Angiogenesis in Covid‐19 ,” New England Journal of Medicine 383 ( 2020 ): 120 – 128 , 10.1056/NEJMoa2015432 . 32437596 PMC7412750

16 A. Bonaventura , A. Vecchie , L. Dagna , et al., “ Endothelial Dysfunction and Immunothrombosis as Key Pathogenic Mechanisms in COVID‐19 ,” Nature Reviews. Immunology 21 ( 2021 ): 319 – 329 , 10.1038/s41577-021-00536-9 . PMC8023349 33824483

17 E. Calabretta , J. M. Moraleda , M. Iacobelli , et al., “ COVID‐19‐Induced Endotheliitis: Emerging Evidence and Possible Therapeutic Strategies ,” British Journal of Haematology 193 ( 2021 ): 43 – 51 , 10.1111/bjh.17240 . 33538335 PMC8014053

18 P. Castro , M. Palomo , A. B. Moreno‐Castano , et al., “ Is the Endothelium the Missing Link in the Pathophysiology and Treatment of COVID‐19 Complications? ,” Cardiovascular Drugs and Therapy 36 ( 2022 ): 547 – 560 , 10.1007/s10557-021-07207-w . 34097193 PMC8181544

19 P. C. Evans , G. E. Rainger , J. C. Mason , et al., “ Endothelial Dysfunction in COVID‐19: A Position Paper of the ESC Working Group for Atherosclerosis and Vascular Biology, and the ESC Council of Basic Cardiovascular Science ,” Cardiovascular Research 116 ( 2020 ): 2177 – 2184 , 10.1093/cvr/cvaa230 . 32750108 PMC7454368

20 G. Goshua , A. B. Pine , M. L. Meizlish , et al., “ Endotheliopathy in COVID‐19‐Associated Coagulopathy: Evidence From a Single‐Centre, Cross‐Sectional Study ,” Lancet Haematology 7 ( 2020 ): e575 – e582 , 10.1016/S2352-3026(20)30216-7 . 32619411 PMC7326446

21 P. Libby and T. Luscher , “ COVID‐19 Is, in the End, an Endothelial Disease ,” European Heart Journal 41 ( 2020 ): 3038 – 3044 , 10.1093/eurheartj/ehaa623 . 32882706 PMC7470753

22 A. B. Moreno‐Castaño , S. Fernandez , M. Palomo , et al., “ Circulating Biomarkers of COVID‐19‐Triggered Endotheliopathy: From Conjecture to Certainty ,” Blood 136 ( 2020 ): 31 – 32 , 10.1182/blood-2020-142311 .

23 L. A. Teuwen , V. Geldhof , A. Pasut , and P. Carmeliet , “ COVID‐19: The Vasculature Unleashed ,” Nature Reviews. Immunology 20 ( 2020 ): 389 – 391 , 10.1038/s41577-020-0343-0 . PMC7240244 32439870

24 Z. Varga , A. J. Flammer , P. Steiger , et al., “ Endothelial Cell Infection and Endotheliitis in COVID‐19 ,” Lancet 395 ( 2020 ): 1417 – 1418 , 10.1016/S0140-6736(20)30937-5 . 32325026 PMC7172722

25 J. Zhang , K. M. Tecson , and P. A. McCullough , “ Endothelial Dysfunction Contributes to COVID‐19‐Associated Vascular Inflammation and Coagulopathy ,” Reviews in Cardiovascular Medicine 21 ( 2020 ): 315 – 319 , 10.31083/j.rcm.2020.03.126 . 33070537

26 H. G. Augustin and G. Y. Koh , “ A Systems View of the Vascular Endothelium in Health and Disease ,” Cell 187 ( 2024 ): 4833 – 4858 , 10.1016/j.cell.2024.07.012 . 39241746

27 Gentium Srl , “Defitelio: EPAR—Product Information,” 2023, May, accessed March 10, 2025, https://www.ema.europa.eu/en/documents/product‐information/defitelio‐epar‐product‐information_en.pdf .

28 Jazz Pharmaceuticals Inc , “Defitelio (Defibrotide Sodium) Injection, for Intravenous Use—Prescribing Information,” 2022, December 2022, accessed October 17, 2024, https://defitelio.com/dosing‐and‐administration/index .

29 D. Garcia‐Bernal , M. Palomo , C. M. Martinez , et al., “ Defibrotide Inhibits Donor Leucocyte‐Endothelial Interactions and Protects Against Acute Graft‐Versus‐Host Disease ,” Journal of Cellular and Molecular Medicine 24 ( 2020 ): 8031 – 8044 , 10.1111/jcmm.15434 . 32519822 PMC7348164

30 D. Frame , G. B. Scappaticci , T. M. Braun , et al., “ Defibrotide Therapy for SARS‐CoV‐2 ARDS ,” Chest 162 ( 2022 ): 346 – 355 , 10.1016/j.chest.2022.03.046 . 35413279 PMC8993696

31 A. Ruggeri , F. Corrado , A. Voza , et al., “ Use of Defibrotide in COVID‐19 Pneumonia: Comparison of a Phase II Study and a Matched Real‐World Cohort Control ,” Haematologica 109 ( 2024 ): 3261 – 3268 , 10.3324/haematol.2024.285345 . 38779740 PMC11443376

32 M. H. Kocoglu , P. G. Richardson , C. C. Mo , A. P. Rapoport , and D. Atanackovic , “ Defibrotide Improves COVID‐19‐Related Acute Respiratory Distress Syndrome in Myeloma Patients After Chimeric Antigen Receptor T‐Cell Treatment Without Compromising Virus‐Specific and Anti‐Myeloma T‐Cell Responses ,” Haematologica 109 ( 2024 ): 2372 – 2377 , 10.3324/haematol.2023.284793 . 38385273 PMC11215353

33 E. Richardson , D. Garcia‐Bernal , E. Calabretta , et al., “ Defibrotide: Potential for Treating Endothelial Dysfunction Related to Viral and Post‐Infectious Syndromes ,” Expert Opinion on Therapeutic Targets 25 ( 2021 ): 423 – 433 , 10.1080/14728222.2021.1944101 . 34167431

34 E. Richardson , C. C. Mo , E. Calabretta , et al., “ Defibrotide for Protecting Against and Managing Endothelial Injury in Hematologic Malignancies and COVID‐19 ,” Biomolecules 15 ( 2025 ): 1004 , 10.3390/biom15071004 . 40723876 PMC12292847

35 P. G. Richardson , E. Carreras , M. Iacobelli , and B. Nejadnik , “ The Use of Defibrotide in Blood and Marrow Transplantation ,” Blood Advances 2 ( 2018 ): 1495 – 1509 , 10.1182/bloodadvances.2017008375 . 29945939 PMC6020812

36 W. H. Self , M. W. Semler , L. M. Leither , et al., “ Effect of Hydroxychloroquine on Clinical Status at 14 Days in Hospitalized Patients With COVID‐19: A Randomized Clinical Trial ,” JAMA 324 ( 2020 ): 2165 – 2176 , 10.1001/jama.2020.22240 . 33165621 PMC7653542

37 B. Cao , Y. Wang , D. Wen , et al., “ A Trial of Lopinavir‐Ritonavir in Adults Hospitalized With Severe Covid‐19 ,” New England Journal of Medicine 382 ( 2020 ): 1787 – 1799 , 10.1056/NEJMoa2001282 . 32187464 PMC7121492

38 J. H. Beigel , K. M. Tomashek , L. E. Dodd , et al., “ Remdesivir for the Treatment of Covid‐19 ‐ Final Report ,” New England Journal of Medicine 383 ( 2020 ): 1813 – 1826 , 10.1056/NEJMoa2007764 . 32445440 PMC7262788

39 Z. R. McCaw , L. Tian , J. L. Vassy , et al., “ How to Quantify and Interpret Treatment Effects in Comparative Clinical Studies of COVID‐19 ,” Annals of Internal Medicine 173 ( 2020 ): 632 – 637 , 10.7326/M20-4044 . 32634024 PMC7350552

40 H. Y. F. Wong , H. Y. S. Lam , A. H. Fong , et al., “ Frequency and Distribution of Chest Radiographic Findings in Patients Positive for COVID‐19 ,” Radiology 296 ( 2020 ): E72 – E78 , 10.1148/radiol.2020201160 . 32216717 PMC7233401

41 P. G. Richardson , S. A. Grupp , A. Pagliuca , A. Krishnan , V. T. Ho , and S. Corbacioglu , “ Defibrotide for the Treatment of Hepatic Veno‐Occlusive Disease/Sinusoidal Obstruction Syndrome With Multiorgan Failure ,” International Journal of Hematologic Oncology 6 ( 2017 ): 75 – 93 , 10.2217/ijh-2017-0015 . 30302228 PMC6171967

42 E. Richardson , C. Carlo‐Stella , R. Jara , et al., “ Response to Maccio et al, “Multifactorial Pathogenesis of COVID‐19‐Related Coagulopathy: Can Defibrotide Have a Role in the Early Phases of Coagulation Disorders?” ,” Journal of Thrombosis and Haemostasis 18 ( 2020 ): 3111 – 3113 , 10.1111/jth.15088 . 32860297 PMC9770769

43 K. Umemura , T. Iwaki , T. Kimura , et al., “ Pharmacokinetics and Safety of Defibrotide in Healthy Japanese Subjects ,” Clinical Pharmacology in Drug Development 5 ( 2016 ): 548 – 551 , 10.1002/cpdd.262 . 27870480 PMC5132046

44 L. Benimetskaya , S. Wu , A. M. Voskresenskiy , et al., “ Angiogenesis Alteration by Defibrotide: Implications for Its Mechanism of Action in Severe Hepatic Veno‐Occlusive Disease ,” Blood 112 ( 2008 ): 4343 – 4352 , 10.1182/blood-2008-04-149682 . 18711003

45 WHO Working Group on the Clinical Characterisation and Management of COVID‐19 Infection , “ A Minimal Common Outcome Measure Set for COVID‐19 Clinical Research ,” Lancet Infectious Diseases 20 ( 2020 ): e192 – e197 , 10.1016/S1473-3099(20)30483-7 . 32539990 PMC7292605

46 C. C. Mo , E. Richardson , E. Calabretta , et al., “ Endothelial Injury and Dysfunction With Emerging Immunotherapies in Multiple Myeloma, the Impact of COVID‐19, and Endothelial Protection With a Focus on the Evolving Role of Defibrotide ,” Blood Reviews 66 ( 2024 ): 101218 , 10.1016/j.blre.2024.101218 . 38852017