A Trial of Lopinavir-Ritonavir in Adults Hospitalized with Severe Covid-19

[NNadir]

The paper I'll discuss in this brief post is this one: A Trial of Lopinavir–Ritonavir in Adults Hospitalized with Severe Covid-19 (in Cao, M.D., Yeming Wang, M.D., Danning Wen, M.D., Wen Liu, M.S., Jingli Wang, M.D., Guohui Fan, M.S., Lianguo Ruan, M.D., Bin Song, M.D., Yanping Cai, M.D., Ming Wei, M.D., Xingwang Li, M.D., Jiaan Xia, M.D., et al. New England Journal of Medicine, March 18, 2020.)

Lopinavir and Ritonavir are proteases developed to treat HIV in the 1990s. Although I was personally not involved in the commercialization of the latter, I was involved - fairly peripherally - in issues connected with its scale up through phase 2. I was not involved in any way in the successor compound, Lopinavir, although I was aware of its development.

This paper is open sourced, as are all publications connected with Covid-19 in the primary scientific literature. The point of bringing it up is to show how clinical trials are conducted. Please note the summary conclusions:

CONCLUSIONS
In hospitalized adult patients with severe Covid-19, no benefit was observed with lopinavir–ritonavir treatment beyond standard care. Future trials in patients with severe illness may help to confirm or exclude the possibility of a treatment benefit. (Funded by Major Projects of National Science and Technology on New Drug Creation and Development and others; Chinese Clinical Trial Register number, ChiCTR2000029308. opens in new tab.)

Note also this statement in the body of the text:

Patients in the lopinavir–ritonavir group had a shorter stay in the intensive care unit (ICU) than those in the standard-care group (median, 6 days vs. 11 days; difference, ?5 days; 95% CI, ?9 to 0), and the duration from randomization to hospital discharge was numerically shorter (median, 12 days vs. 14 days; difference, 1 day; 95% CI, 0 to 3). In addition, the percentage of patients with clinical improvement at day 14 was higher in the lopinavir–ritonavir group than in the standard-care group (45.5% vs. 30.0%; difference, 15.5 percentage points; 95% CI, 2.2 to 28.8) (Fig. S5).

The first sentence in the latter excerpt, which I have bolded, would seem to suggest a positive result to someone who knows nothing at all about clinical trials and their supporting statistics. However, upon statistical analysis, the authors, all of whom are highly trained physicians, conclude that there is no significant difference.

This would have no bearing of course, on "Dr./Professor" Snake Oil Trump and his even more stupid son in law "Dr./Professor" Slumlord Jared, but it means something to scientists, even if Snake Oil Trump and Slumlord Jared having never been accomplished enough to have opened a science book in their pathetic criminal lives and thus might advise people to take these drugs based on a stupid reading of the first sentence in the last excerpt.

As I used to say on another website where I used to write, "Ignorance kills."

Stay healthy. Be safe.

[Igel]

1. I struggle with stats.

But at least I know there's something there to struggle with.

Maybe that's what I should do while staying at home. Beef up my stats skill.

Then again, if you look at a spectrogram when somebody says "stats kill" and "stats skill" in fluent speech there's absolutely no difference.

[EarnestPutz]

2. Mr. NNadir - I'm pleased to have your input on this current medical issue and ....

political hot-potato. I think that Dr. Fauci does his best at the daily Trumpian circus of mis-information, but it's hard given the level of stupid that he has to overcome. Help me with one though, it's been a long time since I studied statistical analysis. Am I right in thinking that a difference of -5 days has a 95% chance of being accurate only when considered as being within an interval that ranges from -9 to 0, and that this is where the reported statistics lose any significance?

[NNadir]

3. I think the issue is that the study was under powered for the lopinavir ritinovir group.

To be clear, clinical trials are always supported by a professional statistical team, and in fact, in the clinical trial industry, there are companies which do nothing but statistics. The area of clinical trials in which I work is more connected with analysis and relatively simple analytical statistics. (Note that there is reference to SAS software, standard in many clinical trials.)

Note that 13 patients in the lopinavir/ritinovir group dropped out because of adverse events.

The paper also contains the following text:

The planned enrollment of 160 patients in the trial occurred quickly, and the assessment at that point was that the trial was underpowered; thus, a decision was made to continue enrollment by investigators. Subsequently, when another agent (remdesivir) became available for clinical trials, we decided to suspend enrollment in this trial.

This means that the study did not have a large enough sample to show that the difference in ICU time was not random.

Note also that this was a secondary outcome.

The authors are not claiming that they have established that lopinavir/ritinovir does not have a clinical effect. What they have said is that their study does not validly show a clinical effect. Any evidence of an effect is too weak to suggest an effect.

The point of the OP was to show that an under powered study has no meaning. From what I've read on the topic, the Trump assholes probably googled their way to some under powered chlorohydroxyquinine study, didn't understand what they were reading, maybe found a friend, a colleague or even themselves who could make money off this snake oil claim, and thus placed people's lives in danger.

I hope this helps.

[Red Pest]

4. You raise some good points

Statistics in studies on treatments really do matter. Any population of patients will have a wide range of recovery times (or death) and trying to show that a particular treatment shortened the time to recovery in a statistically valid way requires a large treatment group with a matched control group. This can be relatively easy to do in an experimental animal system, but in people the medical staff are trying to prolong the lives of the diseased patients. Further, the control and experimental groups are not identical - only matched as well as possible. Human studies are really tough unless you have a really big effect with the treatment.

A couple of other thoughts: 1) not surprised that lopinavir did not have a significant effect, as the processing proteases in HIV and SARS-CoV-2 appear to be somewhat different (I think different cut sites). 2) Remdesivir (an inhibitor of RNA dependent RNA polymerase) may have a better chance of working, but we shall see.

Hydroxychloroquine: very poor data in the original French paper (no matched groups, poor follow-up of patients, senior author with a history of data manipulation/falsification, etc.). The only chance the this stuff works is that the mechanism is through modulation of the immune response to decrease the cytokine storm in very ill patients or perhaps prevent the cytokine storm.

A soluble form of human ACE2 has a better chance of working that OH-chloroquine - see this new paper accepted for publication in Cell (link: https://www.cell.com/pb-assets/products/coronavirus/CELL_CELL-D-20-00739.pdf)

Stay safe & healthy everyone!