Antiporphyrics, antithrombotics and erythropoiesis stimulating agents to treat Covid-19: A living systematic review and meta-analysis

Abstract

Aims: Covid-19 hematology is hallmarked by porphyrin dysmetabolism, altered hemostasis and erythropoietic distress. We hypothesize that antiporphyrics, antithrombotics and erythropoiesis stimulating agents (ESAs) may prove beneficial in the treatment of severe Covid-19. The objective of the current study was to systematically review and meta-analyse the mortality risk profile of hospitalized Covid-19 patients, when treated with each of the three interventions of interest - versus placebo, standard of care or non-intervention.Methods: Eligible studies were identified from: biorxiv and medrxiv, clinical trials govR4, covid19 trials tracker netR5, DISCOVERR2 (Ebsco), ICTRPR7, PubMedR1 and Google ScholarR3 during search period 27 May-07 June 2022. Risk of bias was assessed with SIGN, NIH and JBI checklists and results quantitatively synthesised for antiporphyrics plus antithrombotics studies. Quantitative synthesis was performed in RevMan 5.4R9 with random effects model meta-analysis. Publication bias was assessed in R by funnel plot, Egger’s regression and trimfill analysis (metafor).Results: Number and type of studies included was 6 RCTs for antiporphyrics, 16 cohort for antithrombotics, and 3 case report/series for ESAs. Both unadjusted and adjusted analyses suggested that hydroxychloroquine treatment was not significantly associated with mortality risk in hospitalised Covid-19 patients (RR 0.95 [0.71, 1.29], P=0.76; aHR 0.78 [0.51, 1.22], P=0.28). Interestingly, the direction of effect shifted further towards favoring intervention when only high quality studies were included in the analysis (unadjusted RR 0.80 [0.42, 1.55]) compared with low quality studies (unadjusted RR 1.59 [0.71, 3.53]) although neither were statistically significant (P=0.51 and P=0.26, respectively). Exposure to antithrombotics was associated with reduced in-hospital mortality risk in adjusted analysis (aOR 0.54 [0.33, 0.86], P=0.01; aHR 0.65 [0.52, 0.82], P=0.0003). Subgroup analysis of adjusted estimates showed that exposure to aspirin (aHR 0.62 [0.50, 0.77], P<0.00001) and heparin (aOR 0.35 [0.14, 0.90], P=0.03) individually were associated with reduced mortality risk. The finding was robust to type of regimen, with pre-admission exposure to antithrombotics (aHR 0.79 [0.67, 0.93], P=0.005) and aspirin (aHR 0.66 [0.52, 0.84], P=0.0009] both showing reduced mortality risk. Additionally, prophylactic dose antithrombotics were associated with reduced mortality (aHR 0.44 [0.32, 0.60], P<0.00001). Publication bias was negative by Egger’s regression test (P=0.95). The direction of effect remained significantly in favor of intervention after trim-and-fill analysis (P=0.0034). Discussion: We make an interim recommendation in favor of antithrombotics, with very low certainty of evidence. We make an interim recommendation in favor of either intervention or non-intervention for both antiporphyrics and ESAs, with low and very low certainty of evidence.