Many planned randomized trials to test the efficacy of hydroxychloroquine or related drugs for preventing COVID-19 infection have, as of the end of April 2020, failed to include ECG assessment to either exclude people at the highest risk for possibly developing a life-threatening cardiac arrhythmia or to flag people who achieve a dangerous QTc interval on treatment, according to an analysis of the posted designs of several dozen studies.
Dr. Arthur Wilde
Hydroxychloroquine, the related agent chloroquine, and azithromycin have all recently received attention as potentially effective but unproven agents for both reducing the severity and duration of established COVID-19 infection as well as possibly preventing or mitigating an incident infection. As of April 30, 155 randomized, control trials listed on a major index for pending and in-progress trials, clinicaltrials.gov, had designs that intended to randomized an overall total of more than 85,000 healthy people to receive hydroxychloroquine or chloroquine, in some cases in combination with azithromycin, to test their efficacy and safety for COVID-19 prophylaxis, Michael H. Gollob, MD, said in an article posted by the Journal of the American College of Cardiology (2020 May 11. doi: 10.1016/j.jacc.2020.05.008).
The problem is that all three agents are documented to potentially produce lengthening of the corrected QT interval (QTc), and if this happens in a person who starts treatment with a QTc on the high end, the incremental prolongation from drug treatment could push their heart rhythm into a range where their risk for a life-threatening arrhythmia becomes substantial, said Dr. Gollob, a cardiac arrhythmia researcher at Toronto General Hospital and the University of Toronto. As a consequence, he recommended excluding from these prophylaxis trials anyone with a resting QTc at baseline assessment of greater than 450 msec, as well as discontinuing treatment from anyone who develops a resting QTc of more than 480 ms while on treatment.
“Though this may seem like a conservative value for subject withdrawal from a study, this is a prudent QTc cut-off, particularly when the severity of the adverse event, sudden death, may be worse than the study endpoint” of reduced incidence of COVID-19 infection, he wrote in his opinion piece.
“We cannot provide an accurate number for elevated risk” faced by people whose QTc climbs above these thresholds, “but we know that events will occur, which is why most trials that involve QT-prolonging drugs typically have an ECG exclusion criterion of QTc greater than 450 msec,” Dr. Gollob said in an interview.
His analysis of the 155 planned randomized prophylaxis trials on clinicaltrials.gov that he examined in detail had enrollment goals that would translate into more than 85,000 uninfected people who would receive hydroxychloroquine or chloroquine plus, in come cases, azithromycin. Only six relatively small studies from among these 155 included a plan for ECG screening and monitoring in its design, he noted. “It is reasonable to estimate that among the 80,000 patients randomized to a QT-prolonging drug [without ECG screening or monitoring] there will certainly be arrhythmic events.” If some of these people were to then die from a drug-induced arrhythmic event that could have been prevented by ECG screening or monitoring, it would be a “tragedy,” Dr. Gollob said.
“It is not only inexplicable, but also inexcusable that clinical investigators would dare to include healthy individuals in a clinical trial involving QT-prolonging medications without bothering to screen their electrocardiogram,” commented Sami Viskin, MD, an electrophysiologist at Tel Aviv Sourasky Medical Center. “The fact that we needed Dr. Gollob to ring this alarm is, itself, shocking,” he said in an interview.
“ECG screening is a good option to minimize the risk. You don’t eliminate the risk, but you can minimize it,” commented Arthur Wilde, MD, a cardiac electrophysiologist and professor of medicine at the Academic Medical Center in Amsterdam. Both Dr. Viskin and Dr. Wilde agreed with the QTc interval thresholds Dr. Gollob recommended using for excluding or discontinuing study participants.
In his commentary, Dr. Gollob estimated that if 85,000 otherwise healthy adults were randomized to received a drug that can increase the QTc interval, as many as about 3,400 people (4%) in the group could statistically be expected to have an especially high vulnerability to QT prolongation because of genetic variants they might carry that collectively have roughly this prevalence. In some people of African heritage, the prevalence of genetic risk for excessive QTc lengthening can be even higher, approaching about 10%, noted Dr. Wilde.
Dr. Gollob hoped the concerns he raised will prompt the organizers of many of these studies to revise their design, and he said he already knew of one study based in Toronto that recently added an ECG-monitoring strategy in response to the concerns he raised. He expressed optimism that more studies will follow.
“It’s a real issue to have these trials designed without ECG exclusions or monitoring. I’m glad that Dr. Gollob sent this warning, because he is right. ECG monitoring during treatment is important so you can stop the treatment in time,” Dr. Wilde said. Dr. Wilde also noted that many, if not most, of the studies listed on clinicaltrials.gov may not actually launch.
In April, representatives from several cardiology societies coauthored a document of considerations when using hydroxychloroquine, chloroquine, or azithromycin to treat patients with a diagnosed COVID-19 infection, and highlighted a QTc interval of 500 msec or greater as flagging patients who should no longer receive these drugs (J Am Coll Cardiol. 2020 Apr 10. doi: 10.1016/j.jacc.2020.04.016). For patients who do not yet have COVID-19 disease and the goal from treatment is prevention the potential efficacy of these drugs is reasonable to explore, but “does not exclude the need to minimize risk to research participants, especially when enrolling healthy subjects,” Dr. Gollob said.
Dr. Gollob, Dr. Viskin, and Dr. Wilde had no relevant financial disclosures.