The effects of colchicine on hospitalized COVID-19 patients: A randomized, double-blind, placebo-controlled clinical trial

SARS-CoV-2 causes significant morbidity and mortality, often via lung injury and ARDS. While no definitive antiviral exists, remdesivir has shown efficacy. Colchicine, an established anti-inflammatory agent used in gout, FMF, vasculitis, and some cardiovascular conditions, inhibits microtubule polymerization and modulates IL-1β, IL-6, and IL-18, as well as neutrophil chemotaxis. Given the role of cytokine storm (notably IL-6) in severe COVID-19, colchicine may mitigate hyperinflammation. This trial aimed to assess the efficacy and safety of colchicine on clinical improvement in hospitalized COVID-19 patients.

Rationale for colchicine use in COVID-19 stems from its anti-inflammatory and immunomodulatory actions and acceptable safety profile. Prior studies report mixed results: GRECCO-19 (open-label) found improved time to clinical improvement but no biomarker differences, with diarrhea common. A small RCT (Lopes et al.) reported reduced oxygen need and shorter hospital stay but did not examine ICU/mortality and used different regimens/doses. Observational cohorts suggested improved survival, though with heterogeneity and confounding. In outpatients (COLCORONA), colchicine reduced death/hospitalization only in PCR-confirmed subgroup. The large RECOVERY trial showed no benefit on 28-day mortality, hospital stay, or invasive ventilation. Overall, evidence on colchicine’s clinical benefit in COVID-19 has been inconsistent and context-dependent (dose, timing, concomitant therapy).

Design: Prospective, randomized, double-blind, placebo-controlled clinical trial conducted between February and May 2021 at Ibn Sina Hospital, a tertiary referral teaching hospital affiliated with Mazandaran University of Medical Sciences, Sari, Iran. Ethics approval: IR.MAZUMS.REC.1399.914; registered IRCT20190804044429N5; informed consent obtained. Participants: Adults ≥18 years hospitalized with COVID-19 confirmed by RT-PCR or chest CT. Exclusions: Renal impairment (GFR <30 mL/min), hepatic impairment (Child-Pugh C), colchicine hypersensitivity, pregnancy, or breastfeeding. Randomization and blinding: Permuted block randomization (20 blocks of 4), 1:1 to colchicine or placebo for 7 days. Identical-appearing study drugs. Patients, treating physicians, and study pharmacist were blinded. Interventions: Colchicine arm received 2 mg loading dose then 0.5 mg twice daily for 7 days. Control arm received matching placebo. Both groups received standard-of-care per national policy: antivirals (remdesivir 200 mg day 1, then 100 mg daily for 5 days; interferon beta-1b 250 mcg SC every other day for 3 doses), antibiotics if needed, corticosteroids (dexamethasone 4 mg IV three times daily or methylprednisolone 1.5–2 mg/kg IV daily), and prophylactic anticoagulation (UFH 5000 units SC TID or LMWH 40 mg SC daily), plus supportive care. Assessments: Baseline demographics, comorbidities, and medications collected. Daily vitals (SpO2, temperature, respiratory rate, blood pressure). Labs at baseline and every other day: CBC, lymphocytes, ALT, AST, ALP, urea, creatinine, LDH, CRP, ESR, albumin, troponin, electrolytes. Outcomes: Primary—clinical status at days 7 and 14 using 8-point ordinal scale (1: not hospitalized/no limitations to 8: death). Clinical recovery defined as scores 1 or 2. Secondary—ICU admission, ICU days, intubation rate, 28-day mortality, and adverse drug reactions (ADRs). ADRs monitored and classified per WHO; discontinuation if life-threatening. Statistical analysis: Normality by Kolmogorov-Smirnov; quantitative variables as mean (SD) or median (IQR); qualitative as counts (%). Between-group comparisons with t-test or Mann–Whitney U; chi-square or Fisher’s exact for categorical; Kendall’s tau-b for ordinal scores. Recovery rate ratio computed; Kaplan–Meier for time to death; Breslow test for time to recovery. ORs via univariate logistic regression; variables significant in univariate included in multivariable logistic regression; ordinal logistic regression for ordinal outcomes at days 7 and 14. Analyses in R 3.6.1; two-sided p<0.05. Sample size: Minimum 50 per group based on prior study effect on hazard of death (HR 0.151), 80% power, α=0.05.

Participants: 110 enrolled; 106 analyzed. Mean age 54.6±13.9 years; 53.8% female. Baseline characteristics similar between groups except vomiting (p=0.04). Baseline disease severity: Majority had ordinal score 4 (colchicine 58.2%, control 59.6%). Primary outcome—clinical recovery (ordinal 1–2): - Day 7: Control 30/51 (58.8%) vs Colchicine 35/55 (63.6%); p=0.61; OR=1.23 (95% CI 0.560–2.68). - Day 14: Control 42/51 (82.4%) vs Colchicine 48/55 (87.3%); p=0.48; OR=1.47 (95% CI 0.503–4.29). - Overall clinical response rates: Control 88.2% vs Colchicine 89.1%; rate ratio=1.07 (95% CI 0.713–1.60); p=0.746. Time to recovery (survivors): Median 7 days (IQR 6–8) control vs 6 days (IQR 5–8) colchicine; p=0.06. Kaplan–Meier survival: Predicted 72.5% (colchicine) vs 78.5% (control) at day 12; no significant difference (p=0.438). Secondary outcomes: - ICU admission: Control 9/51 (17.6%) vs Colchicine 12/55 (21.8%); p=0.59; OR=1.30 (95% CI 0.497–3.41). - Intubation: Control 6/51 (11.8%) vs Colchicine 6/55 (10.9%); OR=0.918 (95% CI 0.276–3.05); not significant. - Death (28-day): Control 6/51 (11.8%) vs Colchicine 6/55 (10.9%); not significant. Safety: Gastrointestinal ADRs more frequent with colchicine—nausea 0% vs 14.5% (p=0.01), diarrhea 0% vs 25.5% (p<0.01); one discontinuation due to severe diarrhea. Bradycardia 3.9% vs 14.5% (p=0.06). No serious non-GI safety signals observed. Predictors of recovery (multivariable): Higher BMI (OR=0.74; p=0.015) and high-dose corticosteroids (methylprednisolone 1.5–2 mg/kg) (OR=0.48; p=0.033) associated with lower recovery; higher baseline ordinal score reduced recovery (OR=0.14; p=0.057, marginal). Treatment group, age, sex, platelets, and LDH were not significantly associated.

Colchicine, despite its theoretical benefits in modulating cytokine-driven hyperinflammation, did not improve clinical recovery, ICU admission, intubation, or mortality when added to remdesivir- and interferon-based standard care in hospitalized COVID-19 patients. The slight numerical reduction in time to recovery (median 6 vs 7 days) was not statistically significant. Gastrointestinal adverse events were significantly more frequent with colchicine, underscoring safety considerations in this population. These results align with the large RECOVERY trial showing no mortality or hospitalization benefit and are consistent with mixed findings from smaller RCTs and observational studies, which varied in dosing, duration, timing of initiation, and concomitant therapies. The analysis also identified worse recovery among patients with higher BMI and those receiving higher-dose corticosteroids—likely reflecting greater baseline disease severity—highlighting the importance of patient phenotype and disease stage in outcomes. Overall, the findings do not support routine use of short-course colchicine as adjunct therapy in hospitalized COVID-19.

In hospitalized adults with COVID-19 receiving standard-of-care antivirals and supportive therapies, adjunctive colchicine did not confer clinical benefit on recovery, ICU need, intubation, or mortality, and increased gastrointestinal adverse events. Larger multicenter RCTs and studies exploring different dosing regimens, timing of initiation, and patient subgroups may further clarify any potential role, though current evidence does not support its routine inpatient use.

Single-center design; limited resources resulted in some patients lacking RT-PCR confirmation (15/105); inflammatory cytokines were not measured despite the anti-inflammatory hypothesis; relatively short colchicine exposure (7 days) with lower cumulative dosing compared to some positive studies; most patients had baseline ordinal scores 3–4, potentially limiting detection of benefit; potential confounding by indication with higher-dose corticosteroid use in sicker patients; small sample size limits power for mortality differences.