Upcoming results from hydroxychloroquine and chloroquine Covid-19 clinical trials undercut by study design concerns

GlobalData Healthcare 20th April 2020 (Last Updated April 23rd, 2020 17:44)

Despite the randomised and controlled nature of numerous ongoing Covid-19 trials of hydroxychloroquine and chloroquine, several interviewed experts noted they were skeptical of the relevance of upcoming results and overall trial successes.

Upcoming results from hydroxychloroquine and chloroquine Covid-19 clinical trials undercut by study design concerns

by Jennifer C. Smith in London, with additional reporting from Reynald Castaneda and Bernarda Tundzhay.

Despite the randomised and controlled nature of numerous ongoing Covid-19 trials of hydroxychloroquine and chloroquine, several interviewed experts noted they were skeptical of the relevance of upcoming results and overall trial successes. Others, though, said it is too early to judge, and earlier clinical trial data are too limited to draw conclusions.

Hydroxychloroquine has consistently been in the news for the past several weeks after US President Donald Trump’s pronouncement of the drug as a ‘game-changer’ in the race for a Covid-19 treatment. GlobalData cites 117 planned and ongoing clinical studies for hydroxychloroquine and chloroquine. Both drugs are now generic, with hydroxychloroquine approved for malaria, lupus and rheumatoid arthritis (RA).

Despite the hype surrounding the use and evaluation of hydroxychloroquine and chloroquine, a key hedge for interviewed experts is that ongoing trials are comparing these drugs to the so-called standard of care (SOC), but in fact there is no one standard and this would affect the generalizability of trial results. In discussing the Phase III trial being funded by the World Health Organization (WHO), some experts noted the adaptive design could help with finding an early signal but another noted such protocols can also make data interpretation tougher.

Prior trial data also did not impress experts overall, except to warrant further study. Whilst there is some mechanistic rationale to conducting clinical trials, it may have weak antiviral activity. And while hydroxychloroquine is thought to have a better side-effect profile than chloroquine, both have the potential for cardiac events.

Screening patients for cardiac arrhythmias and also concomitant use of certain central nervous system (CNS) drugs is important, but experts downplayed concerns around hydroxychloroquine’s possible retinopathy side effect. Retinopathy issues are observed with higher doses than what is being evaluated in COVID-19, but the optimal dosing for SARS-CoV-2 is a key unanswered question, with trials evaluating a wide range of regimens.

SOC variations set extrapolation limits

Amongst the numerous studies testing hydroxychloroquine and chloroquine, the WHO-sponsored SOLIDARITY study (ISRCTN83971151) in hospitalized patients has an interesting pragmatic design that could make it possible to see an early efficacy signal, said Derek Lowe, a drug discovery scientist. Janet Wittes, PhD, founder, Washington, DC-based biomedical statistics firm Statistics Collaborative, also praised the adaptive design, which allows treatments to get removed from the trial and others added. Yet, a second US-based statistician noted that it is also dicey to interpret data based on patients randomised at different periods during the trial.

Besides hydroxychloroquine or chloroquine in one arm, the study tests interferon-beta, Gilead Sciences’ remdesivir and AbbVie’s Kaletra (lopinavir plus ritonavir), each in separate arms. The randomised, controlled study, which is expected to enroll several thousand patients, tests each of these therapies plus local SOC versus SOC alone.

Despite their interest in the pragmatic design, Lowe and Wittes noted even slight variations in SOC could affect results. SOCs range from oxygen and ventilation to antibiotics, depending on the symptoms, said Dr Chris Butler, principal investigator for a UK-based study (ISRCTN86534580) of hydroxychloroquine in older patients.

Furthermore, SOLIDARITY excludes the US and UK where SOC could be more similar, Wittes noted. The WHO trial recruits from Argentina, Brazil, Canada, Germany, Indonesia, Iran, Norway, Peru, Qatar, South Africa, Spain, Switzerland and Thailand. All experts noted the potential issues expected in SOLIDARITY related to variable SOC use could be applicable to other randomised, controlled trials.

Although Dr Lewis Nelson, chair, Department of Emergency Medicine, Rutgers New Jersey Medical School, agreed SOLIDARITY has merit, he said the study’s one-year timeline is too long considering an answer is imminently needed. Even a randomised, controlled, single-centre study of about 500 patients testing hydroxychloroquine and chloroquine could provide an answer, he noted. However, both statisticians said there remains the concern that physicians will not be able to generalize the results to a broad patient population.

MOA rational although early evidence faulty

Hydroxychloroquine is thought to work against viruses by increasing pH, which creates a challenging environment, explained Dr Olav Dalgard, professor of infectious diseases, Akershus University Hospital, Norway. However, hydroxychloroquine does not work in influenza so it could be a weak antiviral, added Dalgard, who is also leading a hydroxychloroquine trial (NCT04316377). Vanya Loroch, PhD, professor of Biology in Health and Disease, Business School Lausanne, Switzerland, agreed, noting it is likely that both viral replication and the excessive immune response will have to be targeted. It is possible that for influenza, earlier treatment would have worked better, so hydroxychloroquine may be ineffective in hospitalized Covid-19 patients, Dalgard noted.

Yet, Lowe said, influenza and Covid-19 are triggered by such different families of viruses that a drug would have to have a broad-spectrum mode of action to be effective in both. Dalgard also agreed it is unclear if the influenza experience will translate to Covid-19. It could be that hydroxychloroquine works as a longer-acting treatment in Covid-19, he added. SOLIDARITY’s primary endpoint is all-cause mortality, subdivided by disease severity at randomization. Secondary endpoints are hospital stay duration and time to first ventilation.

Further casting skepticism on hydroxychloroquine’s potential are several Chinese studies and a recent French study. Despite the randomised, controlled nature of the former, the trial sizes are small, with less than 100 patients, thus making it difficult to parse out a true efficacy signal, both statisticians noted. In addition, the studies are not peer-reviewed, which makes the data somewhat suspect, Wittes said. For the highly maligned French study by Didier Raoult and colleagues, all experts agreed the lack of a control group made the study—an investigation of hydroxychloroquine and the antibiotic azithromycin—dubious.

The clinical course is so varied among Covid-19 patients that without a control group, it is difficult to know whether results indicate a true clinical benefit or random variation, Nelson said. Wittes also criticized the lack of examination of patients who died early and were thus removed from the analysis. Since the earlier studies, more data has come out on hydroxychloroquine, including from a nonrandomised French study and a randomised, open-label study in China.

Whilst a letter published in the electronic platform J-stage (2020, Volume 14 Issue 1, pp. 72-73) noted chloroquine phosphate “has shown apparent efficacy” in Covid-19-associated pneumonia in multicenter clinical studies in China, experts said they were skeptical until they viewed the data and study designs.

In terms of safety, both hydroxychloroquine and chloroquine could have concerns about QT prolongation, experts said. To minimize such events, studies could screen patients for cardiac arrythmias, said Lowe and a US-based infectious disease doctor. Azithromycin has also been associated with cardiac events, Lowe noted. Additionally, certain therapies for CNS disorders are contraindicated with hydroxychloroquine and chloroquine, Lowe said, noting patients taking these medications should be screened out of studies. On the flip side, although retinopathy can be an issue with very high doses and with hydroxychloroquine and chloroquine use over an extended period of time, this is less of a concern in Covid-19, said Lowe and Dr Armando Gabrielli, professor, Department of Clinical and Molecular Sciences, Marche Polytechnic University Faculty of Medicine, Ancona, Italy. In Covid-19, the treatment is only used for 2–4 weeks, Gabrielli added.

Dosing variety precludes forecasts

There is no firm evidence on the optimal dosing for hydroxychloroquine or chloroquine, hence the range of regimens across trials, said Butler and Lowe. Butler, who is also a professor of Primary Care, Nuffield Department of Primary Care Health Sciences, Oxford, said the dosage choice in his study of 200mg twice daily for seven days was based on pre-existing patient conditions and concomitant medications, among other factors.

Other trials are basing the dose regimen on hydroxychloroquine’s approved doses. Dalgard said his trial’s dose of 400mg twice daily for seven days is double the dose approved for lupus. The dose was also decided based on Chinese studies and an in vitro study of lung tissue, he added. SOLIDARITY’s evaluation of hydroxychloroquine includes two 800mg loading doses followed by 400mg twice daily for 10 days.

Whilst an analyst suggests a 1,200mg loading dose could address questions of hydroxychloroquine’s efficacy, Lowe said this is quite high, and although it may be used in RA, the drug is mechanistically different in Covid-19. However, countered Barry Robson, PhD, distinguished scientist, University of Wisconsin-Stout, such a high dose may be appropriate for a short period of time to create more endosomes and in turn block virus entry.

Jennifer C. Smith is Senior Editor and Reynald Castaneda and Bernarda Tundzhay are Reporters for Clinical Trials Arena parent company GlobalData’s investigative journalism team. A version of this article originally appeared on the Insights module of GlobalData’s Pharmaceutical Intelligence Center. To access more articles like this, visit GlobalData.