The need for a standardized approach in COVID-19 clinical research
However, in order to make feasible secondary analysis of individual studies, these should follow guidelines for standard protocols and minimal requirements for outcome measures such as the Standard Protocol Items Recommendations for Interventional Trials (SPIRIT)X and the Core Outcome Measures in Effectiveness Trials (COMET) initiativeXI.
In fact, the major criticism emerging from a review of the ongoing trials is the heterogeneity of protocols. Following the decision of both FDA and EMA, respectively, to allow the use of chloroquine or hydroxichloroquine and remdesivir in COVID-1923,XII, several small trials have started using these drugs on a compassionate basis. The EMA has expressed concern for these small studies and the compassionate use programmes across Europe, as they are unlikely to produce the required level of evidence of efficacy and safety of investigational drugs. On the contrary, the EMA strongly recommends that a more coordinated approach and efforts are put in place to prioritize large multi-country randomized trials and multi-arm clinical trials investigating different agents simultaneouslyXIII.
Unarguably, high quality clinical trials cannot be easily performed in the setting of an outbreak, when investigators are often asked to make patients’ care a priority. Nonetheless, even if adapted to this challenging context, high quality research is still needed24. Accordingly, a more standardized approach to clinical research on COVID-19 should be warranted, including a rigorous but realistic study design, a well characterized study population stratified on the basis of age and severity of the disease, a rationale behind the use of the investigational drug and the choice of comparator, and optimal minimal primary outcome(s).
With regard to the study design, while masking might be difficult in studies of COVID-19, randomization should be mandatory. Multi-centres trials are highly advisable provided that standard operational procedures are set out. Adaptive designs might be considered and the setting should be specified. Rules for informed consent must adapted to the chosen population. Centralized Ethics Committees or IRBs might favour a more rapid start of the trial. Investigators should be encouraged to publish the study protocol to be drafted according to the SPIRIT and COS-STAP statement25.
Age, gender, ethnicity, previous diseases and undergoing treatments have been reported to influence the incidence of COVID3,26-28. Demographic and history data should be obtained, possibly through a standard questionnaire, along with other patients’ characteristics such as social status and environmental exposure.
While waiting for a consensus on criteria for the diagnosis of COVID-19 and a classification of the disease, patients should be stratified on the basis of the study setting, severity, and the predominant pathophysiological abnormality in different phases of the disease (viral, pulmonary, inflammatory; Fig.1)29.
Several research groups are working on a variety of preventive and therapeutic interventions. However, despite the accelerated development pathways adopted by many regulatory bodiesXIV, the marketing authorization pathway for new drugs is a long process, with a high attrition rate. Therefore, there has been considerable interest in repurposing existing drugs – such as antiviral and immunosuppressant agents - for use in COVID-1922. Studies with these drugs should keep in consideration the putative mechanisms of action of the investigational drug in relation to the different phases of COVID-19. Antiviral agents, for example, should be used during the early viral phase of the disease while immunosuppressant may be promising candidate drugs in the severe inflammatory phase (while they might dampen the immune response if used early on14). Combination or sequential treatment might also be considered.
Standard of care seems to be a more realistic comparator than placebo, in view of the emergency nature of the epidemic.
Collaborative trials and multi-arm studies comparing different drugs should also be considered. A commendable example of this kind of approach is represented by the Solidarity clinical trial launched by the WHOXV. The Solidarity trial is a randomized multi-countries open label trial which compares the efficacy of four treatment options (Remdesivir, Lopinavir/Ritonavir, Interferon beta-1a and Chloroquine or Hydroxychloroquine) against standard of care in hospitalized adult patients with COVID-19. Underlying conditions are recorded and severity of illness at entry is determined by a reduced set of end-points that can be recorded even in overwhelmed hospitals. Clinically relevant outcomes undergo an interim analysis by an independent Global Data and Safety monitoring Committee. The simplicity of the trial is balanced by the thousands of patients that are expected to be recruited in more than 70 countries. Preliminary results of this trial are expected by June 2020.
In more rigorous study designs, primary end-points should be chosen in relation to the phase of the disease and the drug under investigation. While SARS CoV-2 RNA clearance and the effects on the progression of the disease may represent significant outcome measures for antivirals in the mild forms of the disease, hard end-points such survival/death are advisable in the most severe forms. However, a core outcome standard set is urgently needed to define a minimal set of outcome measures relevant to patients, investigators and regulators30.