Hydroxychloroquine Use in Patients With COVID-19: A Brief Perspective on Current Clinical Trials

OVERVIEW OF CLINICAL TRIALS

Since the first case of COVID-19 was reported in December 2019, a total of 10 randomized clinical trials (RCTs) have investigated treatment with HCQ (Table).^4-13 These clinical trials included patients at multiple stages of severity, including asymptomatic without hospitalization, symptomatic with mild or moderate illness, and symptomatic with severe illness. From the outset, the external validity of studies investigating HCQ use in patients with COVID-19 has been challenged by heterogeneous methods of patient selection that have ranged from asymptomatic individuals with an identified exposure to hospitalized patients with clinical suspicion to positive reverse transcriptase–polymerase chain reaction (RT-PCR) with chest computed tomography (CT) evidence of pneumonia. In one RCT, methods of diagnostic confirmation of COVID-19 were not even specified.⁴ Drawing conclusions from highly variable clinical and laboratory diagnostic methodologies is fraught with potential error. Irrespective of the ultimate conclusions of the studies, we start first with the appreciation that not all studies are created equal. To add to the variability of the trials, the intervention and treatment dosing fluctuated substantially from center to center. Although the majority of studies elected to compare HCQ to standard of care, the dose and duration of HCQ treatment in the experimental group and what constituted standard of care varied significantly among studies. To add perspective, only 2 trials had an identical intervention regimen of 400 mg HCQ for a cumulative 5 days.^4,5 With these caveats in mind, we turn to the outcomes of these trials.

The largest study as of October 2020 (n=821) observed development of positive molecular assay or COVID-19–related symptoms in previously asymptomatic individuals with exposure to confirmed COVID-19.⁶ No significant difference was observed between the HCQ treatment group (one-time 800-mg dose followed by 600 mg/day for 5 days) and placebo group in the development of COVID-19, with the notable caveat that the majority of participants had limited access to COVID-19 testing.

An open-label RCT among hospitalized patients with COVID-19 used a regimen of HCQ (400 mg twice daily) plus standard of care or HCQ with azithromycin (500 mg daily) plus standard of care. Results indicated no significant differences in the primary outcome of clinical status at day 15 or any secondary outcomes, including use of noninvasive ventilation, in-hospital mortality, or duration of hospital stay.⁷ Two relatively large (n>100) RCTs also failed to demonstrate improvement in viral parameters (viral load) or clinical outcomes (hospitalization, mortality, symptom resolution) in nonhospitalized patients treated with HCQ compared to standard of care or placebo.^12,13

As we turn to 5 small trials from China, we observe a possible suggestion of clinical improvement. An RCT comparing HCQ (400 mg daily) plus standard of care to standard of care alone in 62 COVID-19–positive patients with chest CT confirmation and mild illness (partial pressure arterial oxygen/fraction of inspired oxygen [PaO₂/FiO₂] >300) demonstrated a significantly quicker time to clinical recovery (defined as afebrile body temperature and resolution of cough) and improvement in chest CT imaging in the HCQ group.⁵ L. Chen et al corroborated a trend to shorter recovery with HCQ (200 mg twice daily) in a small study (n=48) of HCQ and standard of care vs CQ (500 to 1,000 mg daily) and standard of care vs standard of care.⁸ Only patients with mild or moderate illness (generally patients with oxygen saturation [SaO₂] >93% and/or PaO₂/FiO₂ >300) were enrolled in these 2 studies. Standard of care was based on clinician judgment and was either not specified or varied widely to include antivirals, antibiotics, immunoglobulins, and/or corticosteroids.^5,8 Thus, the generalizability of such studies to patients with severe COVID-19 who may need more aggressive intervention is questionable.

The 3 other small studies (n ranging from 30 to 150) from China were uniform in identifying a lack of significant difference in proportion or time to negative seroconversion in patients with confirmed COVID-19.^4,9,10 Tang et al also observed no difference in clinical course, inflammatory markers, or mortality when HCQ 800 to 1,200 mg/day was added to standard of care therapy.⁹ Similarly, J. Chen et al and C. P. Chen et al observed no difference in mortality or side effects when HCQ 200 to 400 mg/day was added to standard-of-care therapy.^4,10 These studies also primarily focused on patients with mild or moderate illness.

Turning back to trials outside of China, the potential drawbacks of HCQ and CQ regimens emerge. The Borba et al trial terminated prematurely because of the increased incidence of QTc interval prolongation and lethality in a high-dose (600 mg twice daily) CQ group.¹¹ Boulware et al observed an increased risk of mild adverse events, including nausea (22.9% vs 7.7%) and diarrhea/abdominal discomfort (23.2% vs 4.3%), in the HCQ treatment group compared to the placebo group, a finding corroborated by the Mitjà et al study.^6,12 The unblinded Mitjà et al trial had a high enrollment of health care workers (86.7% of study subjects), and 72.0% of patients taking HCQ reported adverse events vs 8.7% of patients in the control arm.⁶ Elevated aminotransferases were also noted as an adverse effect of HCQ in multiple trials and required discontinuation of the study drug in 1 patient in the J. Chen et al study.^4,7,8 While these adverse events were mild in many cases, decreased adherence to HCQ compared to placebo was noted in 2 studies of HCQ use in outpatient populations; thus, mild adverse events—especially gastrointestinal symptoms including nausea, abdominal discomfort, and diarrhea—may affect the efficacy of HCQ treatment for asymptomatic patients or patients with low-acuity cases of COVID-19.^6,13 Another consideration is that several RCTs specifically excluded patients with preexisting cardiac pathology, underlying QTc interval prolongation, or concomitant use of QTc-prolonging medications, therefore perhaps providing insufficient information about the deleterious cardiac outcomes of HCQ in the population at large.^6,8,10,12

Confusion about what role, if any, HCQ should play in COVID-19 treatment is driven in part by significant study limitations, especially in terms of heterogeneous standard-of-care treatments and limited external validity. Themes that originated in the first studies from China were small sample sizes and skew of the patient populations to mild and moderate disease.^4,5,8-10 More concerning from a methodology perspective were the frequent protocol deviations and the lack of placebo groups, control group treatment specifications, and blinding. We have summarized the limitations and challenges with the study designs for these 10 clinical trials in the Table. Ultimately, robust statistical understanding beyond simple P value dichotomy may be necessary to understand the nuances of and draw reasonable conclusions from underpowered trials.

STUDY CHALLENGES DURING A PANDEMIC

COVID-19 has presented major challenges to the medical-academic community in terms of conducting clinical trials in an epidemiologically valid yet timely manner. From the studies presented here, we have determined that treatment with HCQ in patients with COVID-19 has not been shown to consistently improve clinical outcomes, although the majority of studies had significant design limitations. HCQ may not become part of the standard treatment for patients with COVID-19, but we can still glean lessons that can inform research in future pandemics. Even in the midst of a rapidly evolving pandemic, potential therapeutics should be rigorously tested. Although avenues for timely data dissemination should exist, the peer review process must continue to be held to a high standard and remain uninfluenced by political or personal conflicts of interest. Standard-of-care treatments used as comparisons should be truly standardized and specified in detail, even in preliminary scientific manuscripts. In addition, patient populations included in early studies must be chosen carefully; discussion of the utility of therapeutics that were only investigated in patients with mild or moderate illness must be heavily tempered when considering their use in patients with more serious disease. Further, the safety profile of novel interventions should be rigorously investigated in the general population. COVID-19 has provided fertile soil for the flourishing of clinical research, but both study designers and the reading audience must take great care to determine how the combined body of research ought to affect clinical care.