30 Ridiculously Low Number To Look At Efficacy Of A Drug: Medicine Specialist On Biocon Trial

Like everything else, the medical world too has been roiled by Covid-19 as the search for life-saving treatments proceeds at a frenetic pace. This week, the debate veered to Biocon Ltd’s Itolizumab—the first Indian drug to get approval for restricted emergency use from the Drugs Controller General of India (DCGI)—and whether its clinical trial sample of 30 patients was too small to make an assessment. On the other side is the context of its usage – on compassionate grounds. There aren’t easy answers to these questions. But the DCGI can clear the air by being more forthcoming with its information, says Prof O C Abraham, specialist in Internal Medicine & Infectious Diseases at Christian Medical College (CMC), Vellore. In an interview with Outlook’s Ajay Sukumaran, Abraham says large trials can be done quickly even during a pandemic. Edited excerpts:

What do you make of the discussion about the small sample size of randomised control trial for Itolizumab based on which it was cleared for restricted emergency use?

First of all, there is no data published or made available on the DCGI website. All we have is a press release from PIB and an announcement by the manufacturer (Biocon). By itself, 30 is a ridiculously low number to look at the efficacy and safety of a drug. To give a comparison, the RECOVERY Trial in the UK, which said that Dexamethasone, a steroid, saves lives, compared a little over 2,100 patients who received the drug to about 4,300 patients on usual care alone. They had to study that number of patients to find out if there was a true difference.

The common thing you are hearing during this pandemic is “people are dying, we have to save them, we don't have time to wait for studies.” But the RECOVERY Trial produced fantastic information in the middle of a pandemic—we know that Hydroxychloroquine does not work in hospitalised Covid-19 patients, we know that Dexamethasone saves lives. Similarly, a large randomised US trial called the ACTT1 done in the middle of a roaring pandemic said Remdesivir is not a magical cure, but it does improve symptoms.

Several institutions in India, including CMC Vellore, are participating in the SOLIDARITY Trial sponsored by the WHO in which close to 5,000 patients across the world have been enrolled. That is giving us very valuable information about what works, what does not work, what saves lives and what reduces the need for ICU admission. We went through all these things with Ebola and with AIDS epidemics as well.

To me, the way DCGI has approved Favipiravir and Itolizumab for Covid-19 is like a “black box”. Is there any information at all on its website? Let's compare and contrast. When the US Food and Drug Administration put out the emergency use authorisation for Hydroxychloroquine (HCQ) and Remdesivir, the full information was there on the website. Anybody can access it; you don't have to be a doctor. The Dexamethasone trial was done across the UK. The only thing is it is not yet published in a peer-reviewed publication... The study results are only available as a pre-print and a press release by one of the chief investigators. That is the only drawback. However, the protocol is available for everybody on the RECOVERY website to download. Everything is transparent.

What is a reasonable sample size for a clinical trial?

That depends on the research question and the magnitude of difference being anticipated between the new treatment and the standard treatment. You have some rough idea (from previous studies, personal experience etc.), based on which you do a calculation of how many patients you need to study. If the expected difference is small, you will have to study a large number of patients. If the difference is huge, you study fewer patients. So there could be 500 or 700 or 1,000 patients.

The RECOVERY Trial of Dexamethasone looked at people dying due to Covid-19. In the ACTT1 trial for Remdesivir, the question was whether the symptoms improved. So they had a score based on whether the fever subsided and oxygen requirements became less and so on. SOLIDARITY and RECOVERY are looking at mortality.

During this pandemic, however, we have seen some experts question the efficacy of randomised trials as well. Can you put this into perspective?

Whenever you say that a drug works, that is, it reduces death or reduces admission to ICU or reduces the need for putting a patient on a mechanical ventilator, what are you comparing it with? The best way of comparison is a randomised clinical trial (RCT). The beauty of an RCT is that everything else is similar between two patient groups—for example, a patient who gets the new drug and one who gets standard treatment. Everything else is comparable, so if there is a difference, you can say the drug has made the difference. With low sample sizes, there could be what are called chance associations...anything in the world could not be the effect of the drug, right? Why is the National Health Service in the UK or the World Health Organization spending millions doing large randomised clinical trials. That is where you find whether it truly works or not. So RCTs are the best way.

But that is not feasible in several situations. Or even needed. Take insulin for young diabetics, for example. Because the response to insulin in children dying from Type1 diabetes and ketoacidosis is so dramatic, there is no need to do a randomised control trial. Or the original study of penicillin in pneumonia-caused by pneumococcus. There is no need for an RCT. But if you look at HIV, diabetes, breast cancer, heart attack—all these treatments are based on good, large randomised control trials. It's not like you swear by them. There are many examples whether it is neither needed nor possible. We should not take a stand that it is not needed at all or it is only RCTs that work -- you have to be pragmatic.

In the medical world are these discussions getting more heated up in some sense, because of Covid-19?

It was always so. You may have heard about a movement called Evidence based Medicine in the late 1980s. These are ongoing debates. There are enough people supporting and arguing vehemently for randomised control trials in academic circles, published journals and in the lay press. Also, the other way.

Can you tell us a bit about the SOLIDARITY Trial currently ongoing at CMC Vellore?

The Solidarity Trial started with five arms—among them were standard of care, HCQ, Lopanavir/Ritonavir, Remdesivir and a drug called Interferon. It’s an adaptive design. Across the globe you are recruiting patients and statisticians are looking at the data very frequently. If a drug is useless or causing harm, it is taken off. It mimics real life practice, very suited for a condition like a pandemic when we need information fast.

So, they have taken off Lopanavir and they have taken off HCQ. Now, it is Remdesivir, Interferon and a combination of Remdesivir/Interferon. There will soon be an interim analysis on Remdesivir and we will know whether to continue it or not. There is also talk about a protocol amendment to include new therapies, we don't know the details yet.