key: cord-0857714-0nnru9fx authors: Rojas, Manuel; Anaya, Juan-Manuel title: Why will it never be known if convalescent plasma is effective for COVID-19 date: 2020-11-04 journal: J Transl Autoimmun DOI: 10.1016/j.jtauto.2020.100069 sha: cd87286ce320a65604bd6062d980b45a76562bef doc_id: 857714 cord_uid: 0nnru9fx High expectations have been set around convalescent plasma (CP) for the treatment of COVID-19. However, none of the randomized controlled trials (RCTs) conducted so far have reached their primary endpoints. Herein we report that RCTs of CP disclose a high methodological variability in inclusion criteria, outcomes, appropriate selection of donors, dosage, concentration of neutralizing antibodies and times of transfusion. Therefore, at this time there is insufficient evidence to recommend for or against the use of CP as a treatment for COVID-19. The current pandemic has challenged health systems given the uncontrolled spread and high mortality in critically ill patients with COVID-19. During the last months, several clinical trials have been conducted to discover new treatments that may reduce the burden of the disease. However, none of these studies have reached the expected primary endpoint, and only the RECOVERY study (treatment with Dexamethasone), showed a significant difference of 3% in mortality [1] . However, a reduction of this magnitude in a condition with about 20% of mortality in critically ill patients, is disappointing. Convalescent plasma (CP) emerged as potential treatment for COVID-19 at the beginning of the pandemic. This is a strategy of passive immunization that has been used in prevention and management of infectious diseases since early 20 th century [2] . The CP is obtained using apheresis in survivors with prior infections caused by pathogens of interest and that developed antibodies against the causal agent of disease. The major target is to neutralize the pathogen for its eradication [3] . Given its rapid obtaining, CP has been considered as an emergency intervention in several pandemics, including the Spanish flu, SARS-CoV, West Nile virus, and more recently, Ebola virus [4] [5] [6] [7] [8] [9] . Several studies have shown the potential efficacy of CP in COVID-19. A recent meta-analysis of randomized and matched-control studies, showed a reduction of 57% of mortality in COVID-19 patients treated with CP [10] . However, some concerns arise around this analysis. First, quality of evidence across the studies was not evaluated, and pooled estimation of effect came from the mixture of J o u r n a l P r e -p r o o f randomized and case-control studies. This approach may produce biased estimations of the effects and may not reflect the true efficacy of CP in COVID-19. In addition, high methodological heterogeneity has been observed in most of the studies on CP. Inclusion criteria, outcomes, dosage, and concentration of neutralizing antibodies (NAbs), are some factors that may influence the efficacy of this therapy and hinder the pooling of evidence in this topic. Herein, we analyzed randomized controlled trials (RCTs), their risk of bias, comparability and the potential confounding factors that may disturb conclusions on this treatment. A systematic review of the literature about RCTs in COVID-19 was done following the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines [11] . MEDLINE, EMBASE, and LILACS were systematically searched for published and unpublished studies. Additional manual searches of the references cited in the articles were done. The search included articles up to October 21, 2020 . No restrictions were placed on study period or sample size. Only those articles written in Spanish or English were considered. Other information sources such as personal communications, author's repositories and preprint servers were included. Title and CP text terms in combination with "COVID-19", and "Randomized" were used. Risk of bias analysis included the following: random sequence generation, allocation concealment, blinding participants and personnel, blinding of outcome assessment, selective reporting and incomplete outcome data. Risk of bias was J o u r n a l P r e -p r o o f conducted for mortality and clinical improvement outcomes. Furthermore, the Cochrane GRADE approach was used to assess quality of evidence [12] . Studies were graded from very low quality to high quality in a 4-tiered system. All studies for this review were RCTs and started the evaluation as high quality of evidence. A narrative summary of evidence was performed. Five articles were obtained from the systematic review [13] [14] [15] [16] [17] . General characteristics of studies are shown in Table 1 . Clinical features varied across studies and represented different stages of disease. Three of 5 five studies included critically ill patients, as well as non-severe COVID-19 cases [13, 14, 16 ]. In addition, two studies did not mention severity of disease and just reported a positive RT-PCR SARS-CoV-2 test [15, 17] . This represents the first aspect to discuss in the comparability and reproducibility of RCTs. Although the pathogenesis of COVID-19 is still not completely understood, four overlapping and escalating phases have been proposed to explain the clinical course of the disease [18] . First, there is a viral phase that may well be asymptomatic or mild in the majority, perhaps 80% of patients. In the remaining 20% of cases, the disease may become severe and/or critical. In most patients of this latter group, a hyperresponsiveness of the immune system is characteristic. A third phase corresponds to a state of hypercoagulability. Finally, in the fourth stage organ injury and failure occur [18] . Severity of disease at inclusion is pivotal and may influence outcomes in unknown ways. At this respect, all RCTs have included patients at different stages of disease (Table 1) . For example, 24.3% of patients in the study of Li et al. [13] were on mechanical ventilation (MCV) at the moment of the inclusion. Remaining subjects were on high-flow, low-flow or without requirement of oxygen. Since CP may help patients with non-severe disease [9] , selection of patients by severity of disease is critical. Fusion of all these groups into one may not reflect the real effect of CP. In addition, sample size estimation may require an adjustment by this confounder. Thus, further RCTs should aim to estimate the efficacy of CP in every stage of disease with an appropriate sample size. Little is known about the accurate plasma concentration of NAbs required to produce a significant clinical effect, as well as timing to transfusion. The US food and drug administration (FDA) recommended a minimal titer of NAbs of 1/160 to treat COVID-19 patients [19] . NAbs concentration varied across RCTs, and in some cases, NAbs were below the minimal cut-off proposed by the FDA ( Table 1 ). The first completed RCT conducted in India by Agarwal et al. [17] , CP with low As shown in Table 1 , most studies focus on mortality and early discharge. However, variability in inclusion criteria, and the clinical stage of disease, make studies not equivalent. Other outcomes such as intensive care unit (ICU) admission, and number of patients requiring orotracheal intubation haven been poorly studied. Thus, these outcomes should be considered given the current public health emergency and the high rate of ICU admission in this condition. Risk of bias analysis revealed that none of the RCTs included in this review was on low risk of bias ( Table 1) . Lack of blinding of participants and personnel, and blinding of outcome assessment were the most common bias in these studies. In addition, there were not information about allocation concealment in 3 out of 5. These observations influence the decision to rate all the studies in either high or J o u r n a l P r e -p r o o f unclear risk of bias for mortality and clinical improvement outcomes (Figure 1 ). This is of critical relevance since selection and measurement bias could have influenced the results of these trials. In addition, the current analysis allowed us to downgrade the quality of evidence in at least 2 points indicating that the evidence on CP is of low quality according to GRADE [12] . Double blinded RCTs are required to provide better evidence for the use of CP in COVID-19. About 10% of critically ill patients with COVID-19 present antibodies against type I and II IFN [21] . In addition, other antibodies against IFN-γ, GM-CSF, IL-6, IL-10, IL-12p70, IL-22, IL-17A, IL-17F, and/or TNFβ may also be found [21] . This is of critical relevance in production of CP, since some donors, especially those recovered from critical disease, may have this type of antibodies with unpredictable effects on CP receptors. In addition, it is unknown whether some CP contains pro-inflammatory cytokines that could exacerbate the disease. Thus, besides the measurement of NAbs, standards for cytokine concentration and cytokine autoantibodies are recommended. Recent evidence has emerged on the evolutionary processes associated with adaptation of SARS-CoV-2 to humans. Two major strains of SARS-CoV-2 were described in Wuhan (i.e., strain "L" and "S") [22] . It has been suggested that S strain could be considered more aggressive. In the same line, two studies have shown that the non-synonymous mutation D614G in the spike gene is associated with an increased infectivity of SARS-CoV-2 [23, 24] . Evolutionary models have suggested that the novel coronavirus could change infectivity and mortality over J o u r n a l P r e -p r o o f time influenced by lockdowns and other unpredictable evolutionary factors [25] . Altogether, these data indicate that efficacy of CP could be influenced by the evolutionary change of SARS-CoV-2 over time and argue for the inclusion of mutation analysis in epidemiological surveillance. In the early 1950s, purification and concentration of immunoglobulins from healthy donors or recovered patients (i.e., intravenous immunoglobulins -IVIg), provided an option to treat serious infectious diseases as well as immune conditions including primary immunodeficiencies, allergies, and autoimmune diseases [9] . Recently an observational study on IVIg in COVID-19 showed that this treatment may improve hypoxia, hospital length and reduce progression to mechanical ventilation [26] . Since different concentrations of NAbs were found in this review ( Table 1) . Production of IVIg from recovered COVID-19 donors may provide an option to standardize doses and concentration of NAbs transfused. Not applicable Dexamethasone in Hospitalized Patients with Covid-19 -Preliminary Report Convalescent plasma: new evidence for an old therapeutic tool? Ebola virus convalescent blood products: where we are now and where we may need to go The effectiveness of convalescent plasma and hyperimmune immunoglobulin for the treatment of severe acute respiratory infections of viral etiology: a systematic review and exploratory meta-analysis Ebola virus disease: An emerging and re-emerging viral threat West Nile virus neutralization by US plasmaderived immunoglobulin products The role for intravenous immunoglobulin in the treatment of West Nile virus encephalitis Treatment of West Nile virus encephalitis with intravenous immunoglobulin Convalescent plasma in Covid-19: Possible mechanisms of action Evidence favouring the efficacy of convalescent plasma for COVID-19 therapy. medRxiv Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement How to GRADE the quality of the evidence Effect of Convalescent Plasma Therapy on Time to Clinical Improvement in Patients With Severe and Life-threatening COVID-19: A Randomized Clinical Trial Early Anti-SARS-CoV-2 Convalescent Plasma in Patients Admitted for COVID-19: A Randomized Phase II Clinical Convalescent Plasma for COVID-19: A multicenter, randomized clinical trial. medRxiv Convalescent plasma in the management of moderate COVID-19 in India: An open-label parallel-arm phase II multicentre randomized controlled trial (PLACID Trial) Autoinflammatory and autoimmune conditions at the crossroad of COVID-19 Convalescent plasma therapy in patients with COVID-19 Treatment of Coronavirus Disease 2019 Patients with Convalescent Plasma Reveals a Signal of Significantly Decreased Mortality Auto-antibodies against type I IFNs in patients with life-threatening COVID-19. Science (80-) On the origin and continuing evolution of SARS-CoV-2 Spike mutation pipeline reveals the emergence of a more transmissible form of SARS-CoV-2. bioRxiv The D614G mutation in the SARS-CoV-2 spike protein reduces S1 shedding and increases infectivity On the evolutionary epidemiology of SARS-CoV-2 Intravenous Immunoglobulin (IVIG) Significantly Reduces Respiratory Morbidity in COVID-19 Pneumonia: A Prospective Randomized Trial. medRxiv The authors thank all the members of the CREA for contributions and fruitful discussions. This work was supported by Universidad del Rosario (ABN-011), Bogota, The authors declare that they have no competing interests.