The RECOVERY platform trial: No benefit to Hydroxychloroquine in Covid-19. St Emlyn’s

We’ve talked about the RECOVERY trial on the blog and podcast before. This is a UK based platform trial (more on this later) looking at patients admitted to hospital with Covid-19.

It’s an open label RCT that was developed out of the ASAP trial^​1​. ASAP was (still is) a hibernating trial^​2​ that was designed to test the use of steroids in a future flu pandemic. When Covid-19 came along the systems and plans for ASAP were repurposed to test treatments for Covid-19.

I was the Principal Investigator for ASAP and now co-lead the RECOVERY trial at my hospital, one of 175 UK sites led by the chief investigator (Prof. Peter Horby) and team at Oxford University^​3​.

The first results are now out from RECOVERY, although these have not been fully published and have not been peer reviewed. However, the trial design has previously been peer reviewed as part of its inclusion in the NIHR portfolio.

The statement from the RECOVERY team can be found here^​4​, but in brief.

• Patients recruited to RECOVERY – Over 11000 and rising daily.
• Largest RCT in Covid-19 to date
• 1542 patients randomised to Hydroxychloroquine (HCQ)
• 3132 patients randomised to routine care
• End point is 28-day mortality
  • 25.7% Hydroxychloroquine
  • 23.5% usual care
  • hazard ratio 1.11 [95% confidence interval 0.98-1.26]; p=0.10)

The triallists have concluded that it is highly unlikely that there is any meaningful benefit to HCQ in patients with Covid-19 and its use should be discontinued. Interestingly the confidence intervals are close to suggesting that HCQ may be harmful, and as we were still randomising patients to HCQ until yesterday it is likely that the final number of patients in the HCQ arm of the trial will rise and we may get a tighter confidence interval around the hazard ratio (in 27 days time). The investigators estimate that 80% of HCQ outcomes are currently available so an estimated 934 patients are yet to complete follow up and will be added to the trial results later.

Caveats

The main caveat (and it’s a huge one) is that we have only seen the outcome data and the methods. We cannot for example yet know whether the groups were similar at baseline, when the medications were given (in the course of the disease), how many patients completed the treatment nor whether the purported complications have biological plausibility. We must wait until we see the full data set until we can be confident that this finding can stand up to a formal critical appraisal.

What other trials are out there on Hydroxychloroquine?

This is a fascinating question. Early and small open label trials failed to show a benefit^​5,6​ , or showed only a proxy patient benefit^​7​. A recent study in the Lancet suggested that HCQ may be harmful base on an observational database led by Mehra et al^​8​. However, this trial has now been retracted^​9​ in one of the most controversial scientific events of the pandemic^​10​. Despite these earlier and retracted trial results it’s clear that there was no clear high quality evidence to advocate or refute the use of hydroxychloroquine before this, unless of course you live in the White House. Having said that, Trump has been taking HCQ prophylactically and that is not tested in this trial. From a neutral perspective it is still possible that he is right.

What’s a platform trial?

The astute amongst you will have noticed that the numbers in the HCQ + Control groups do not add up to >11000 and that’s because RECOVERY is a platform trial that is testing a range of treatments.

Platform trials, also referred to as multi-arm, multi-stage (MAMS) design trials, are trials that evaluate several interventions against a common control group and can be perpetual. This design has pre-specified adaptation rules to allow dropping of ineffective intervention(s) and flexibility of adding new intervention(s) during the trial. ^​11​ During a pandemic these trials offer a pragmatic and extremely agile way to test old and new therapies against an emergency threat.

The RECOVERY Trial is a large, randomised controlled trial of possible treatments for patients that has a consistent design (in this case an open label RCT) but which permits treatments to enter and exit the trial as the become available. Thus far we have been randomising patients to the following treatments as part of RECOVERY.

 Lopinavir-Ritonavir
 Low-dose Dexamethasone
 Hydroxychloroquine NOW STOPPED
 Azithromycin
 Tocilizumab
 Convalescent plasma

The use of Tociluzamab and convalescent plasma are in subgroups of patients using the following flowchart, which is a little complex, but will allow us to look at individual and combined therapies.

Interim analyses are pre-planned and overseen by a data monitoring committee. The investigators remain blinded to the outcomes unless the DMC permits the release of data.

Further work

The NIHR approach, developed during the last flu pandemic and now actioned in Covid-19 is a model for all nations to plan for the pandemic that will follow Covid-19. We will see more results from RECOVERY and other trials soon.

The NIHR Urgent Public Health strategy seeks to support a series of platform trials designed to evaluate multiple interventions with the same protocol to be ultra efficient. We have RECOVERY for inpatients, REMAP-CAP (an international platform trial) for critical care [pneumonia], PRINCIPLE for outpatients and ACCORD (plus a couple of new additions) for phase 2 evaluations. Even pharma is being pushed down this route – and it means standardised outcomes, inclusion criteria and methodology. It’s a hugely interesting concept and this paper will be the first fruit borne.

To give some perspective my trust has recruited over 2800 patients to 12 different Covid-19 trials thus far. The pace and scale of this is extraordinary effort with more to come as we continue with anti-viral trials, but now also start to include new studies on immuno-modulation for the most seriously affected.

Watch this space.

vb

S