Oxford, the authors of the British clinical trial Recovery attempt to hide deaths by overdose

ANALYSIS: The authors of the recovery clinical trial (Peter Horby and Martin Landray) attempt to cover up a despicable fault in the hydroxychloroquine arm. Several elements are concerning : results that are hiding reality, unforgivable errors in the documents, the author of the appendix of the documents of the Recovery study (British clinical trial) is Dr Hayden known to be historically close to Gilead having taken on several occasions the defence of Remdesivir (drug that has recently been approved by the European Medicines Agency without evidence of therapeutic benefit and  very harmful side effects).  Hence the Recovery study cannot be considered serious.

Retrospective foreword

We note that there is mounting evidence that hydroxychloroquine is active and tolerable against Covid-19.

Before exposing the intellectual fallacy and the real deadly implications of the results of Recovery which conclude that hydroxychloroquine is ineffective, we seek to expose the reader to the comprehensive analytical work carried out by FranceSoir in search for truth. In particular, we have revealed problems of medical ethics (1)and notorious incompetence (2), and even potentially criminal activities (3, 4), related to the Recovery trial, on which we have published a complete document demonstrating the existence of obvious conflicts of interest (5).

This foreword developed here will serve as a reminder and verification that what we affirm in this article is not gratuitous.

France Soir, with the help of scientists and clinical trial experts, has carried out a rigorous and meticulous analysis of the majority of studies published or submitted for pre-publication on the MedRxiv site of Cornell University.  FranceSoir is the only print media to have scientifically dismantled, point by point, the highly questionable prepublications, whose conclusions were biased against hydroxychloroquine or in flagrant contradiction with the data exposed, as for the AP-HP study (6) or Epiphare (7). On the other hand, it is clear from other studies and clinical trials that hydroxychloroquine (HCQ), in combination with azithromycin which gives it considerable synergy, appears to be the only tolerable and active treatment against Covid-19 (8,9,10,11) and also suitable for prophylaxis (9).

We also bring to bear that the current pandemic has allowed members of the public, concerned about the health of their loved ones and themselves, to perceive the extent to which misinformation could circulate in the mainstream media. This led is evidence by the fact that only 7% of the French have confidence in what the messages broadcasted by television media about the pandemic (9).

"With a self-awarded white knight role, the media are no longer content to give a voice to one side or the other by commenting, in a quest for impartiality, but are setting themselves up as true referees of what is or is not scientific and medical truth.

In this respect, hydroxychloroquine (HCQ) has been systematically denigrated as a target. Why has it been systematically denigrated? It was put in their heads that the randomized controlled clinical trial was the universal panacea of the reasoned medical scientific approach. This is particularly false in this case and in opposition to the medical ethics of the Hippocratic Oath (1)," says a medical research specialist.

However, this certainty, instilled by intense brainwashing by public health authorities and Big Pharma-funded television presenters, is crumbling and vacillating in the face of accumulated evidence (8,9,10,11) in favour of hydroxychloroquine and raises questions about how a molecule such as remesivir could have slipped through the cracks of the European Medicines Agency without toxicity testing.

What else can we say about the statement made by the French President on July 14th, in response to a question from a television journalist who asked him, if he would take hydroxychloroquine in the event of Covid-19 contamination?

"From what I understand about science, there is no such thing as a stabilized treatment. France is the country of “Lumière” and I believe in rationality ... If there is no treatment, I'm not going to take it," he said.

He added: "It is not for the President of the Republic or a politician to decide a scientific debate."

But that is exactly what he just did live on the air!

And then speaking about Professor Raoult:

"Nor is it for a man of science, even if he becomes a public figure, to act on scientific beliefs."

Is the President suggesting that Pr Raoult doesn't act as a man of science, but on the basis of scientific belief? Would we have come to the apotheosis of denigrating a man whose entire career speaks for him? Is our President so unsecure, that he cannot help but say, that perhaps in a few months’ time we will learn that hydroxychloroquine is a really active treatment? When we have obviously known this for quite some time. These statements reveal how little our President cares about the possibility that this treatment could have saved lives.

The conditions surrounding Recovery (boundary conditions) give us an indepth

We believe the principal investigators of the Recovery trial are attempting to conceal the results of the very dangerous, even fatal, overdose of the hydroxychloroquine arm.  The authors of the trial came very close to pre-publishing significant results demonstrating the harmful nature of HCQ!

First of all, it should be remembered that this is not a real publication, but a text filed on the MedRxiv site of Cornell University and that it was therefore not submitted to the proofreading and critical questions of other researchers in the clinical field. But never mind. The article still subtly suggests an adverse effect by presenting a survival curve showing a 2% increase in mortality at 28 days (from 25 to 26.8%) in the HCQ arm compared to standard care. This curve is presented with a Y-axis not going to 100% which exaggerates the 2% increase. This difference is not significant according to the value of the calculated statistical power p = 0.18.

We are here in very subtle communication effects which consist in hiding the reality one does not want to talk about. We demonstrated this phenomenon in the phase III study of remdesivir (Veklury®) published in the New England Journal of Medicine (NEJM) where the results of the secondary endpoint of 28-day mortality were deliberately masked.

We state that if HCQ had been used at an acceptable dose level on the first day and on subsequent days a beneficial effect could have been measured. This effect is masked by the premature death or the premature transfer to ICU of hospitalized patients overdosed with HCQ within 48 hours of initiation of treatment. We remind the reader that patients received a cumulative dose of 3.2 g of HCQ in 48 hours, including 2.4 g on the first day, which represents a potentially fatal overdose on patients in this category (4). It should also be remembered that HCQ overdose is characterized by acute respiratory failure (4) which a priori cannot be distinguished from the respiratory symptoms due to Covid-19. On the other hand, only heart failure can be demonstrated (prolongation of the QT interval and twisting of the electrocardiogram peaks).

We also asked in a previous article if, in the Recovery trial, hydroxychloroquine had not killed as many patients as it had saved?

Is it possible that the beneficial effect of a treatment may be masked by a harmful effect such as overdose, comedication or an increased risk factor for certain categories of patients?

We have already recently highlighted such a problem of masking the beneficial effect of HCQ by contradictory effects in our careful reverse engineering analysis of the EPIPHARE study (7). EPIPHARE sought to determine whether HCQ conferred protection from Covid-19-related hospitalization and mortality in patients receiving long-term treatment for chronic inflammatory conditions (lupus and rheumatoid arthritis). The authors concluded that HCQ does not confer protection on these autoimmune patients, who are a priori more likely to develop viral infections than the rest of the population. We showed that the authors of this study were withholding data that they had available and that could have led to the opposite conclusion of a protective effect of HCQ. A Chinese study published in the Lancet on July 3 confirms this.  In the Chinese study, patients with rheumatoid arthritis taking hydroxychloroquine had a 91% reduced risk of infection with the COVID-19 virus (with a statistically significant power p = 0.044) compared to those with the same chronic inflammatory diseases, but not on long-term HCQ treatment.

The parameters of the Recovery results do not deliver the essentials

Mortality data within the first 48 hours after initiation of HCQ treatment and comparison with the control group would provide an indication of whether or not excess mortality occurred within this critical time frame. But this figure is not given! Only one sentence specifies that in the first 2 days of treatment no excess mortality was observed in the HCQ arm as would be expected based on dose-dependent toxicity (sic)!  So why not just give the numbers for both arms?

Part of the reality of the numbers in the HCQ arm is given in an encrypted and anecdotal way by the median treatment time of 6 days and the interquartile range (IQR) of treatment duration between 3 and 10 days. This means that 25% of the patients who received the treatment died after 3 days at the most, and half of them after 6 days at the most (no interruption for toxicity is mentioned, which does not seem very credible in view of the doses administered)! The corresponding figures for the standard care arm are not reported in the pre-published text. However, an idea of the values can be inferred from the results of the study by the Assistance Publique - Hôpitaux de Paris (AP-HP) (themselves deposited on the MedRxiv website) and of which we published a detailed analysis in France Soir.

The samles of these 2 studies are relatively comparable in terms of average age (AP-HP: 69 years old; Recovery: 65 years old) and similar for pulmonary comorbidities, around 15% in the AP-HP study and 23% in the Recovery study. This makes it possible to take the figures provided by AP-HP as an indicative basis for comparison. The median time to death was 7.54 days with an interquartile range of 3.94 to 13 days. In the HCQ arm of the AP-HP study, the median time was 8.66 days (IQR interval: 4.72 to 15.4 days).

We can therefore see that the HCQ arm in the AP-HP study, where the HCQ dosage on the first day was 4 times less (600 mg) and 2 times less on the following days (400 mg), corresponds to a median time of almost 3 extra days and an extended first quartile of 1.7 days compared to the Recovery figures. Furthermore, we see that in the HCQ arm of Recovery the first quartile time is decreased by 1 day (3 instead of 4 days) compared to the standard care arm!

This cross-analysis, despite its limitation, allows us to consider that the data missing in the pre-publication of Recovery would show that in the short period of 2 days following the initiation of HCQ treatment there is indeed an excess mortality compared to standard care.

Another particularly strange issue, but at the very least very revealing of the lack of control of the trial, not to say amateurism, is that 11% of the population in the HCQ arm did not receive hydroxychloroquine! This level of wrongful treatment allocation in the HCQ arm is completely unacceptable and therefore highly questionable. There have necessarily been some very unserious randomization manoeuvres, probably to rebalance the arms between each other? But in any case, this necessarily led to a bias in the results since all patients in the HCQ arm were analyzed for Intent to Treat (ITT). This alone invalidates the results of the trial.  

We therefore continue to state that :

"Almost certainly, the dose levels administered to inpatients in the Recovery trial certainly exacerbated the deteriorated lung condition of a number of patients, which a thorough and systematic review of medical records would reveal. » (4).

Elementary mistakes

• The most laughable elements of the study can be found in the tables, which have not even been checked by their authors.
• Diabetes there are 918 diabetics among the 3199 eliminated, which makes a percentage of 29%.
• Problem on the 4716 retained (HCQ + placebo) there are 1994 diabetics and they indicate that it is 27% but it is 27% of "4716 + 3199" i.e. of the starting group, not of the 4716 of the study.
• Of course these 1994 diabetics were randomized afterwards, therefore statistically neutral on the study, but obviously they used "this artifice" so that the percentages would be about the same between "those kept and those eliminated".
• One could even believe that in percentage they "did not retain" more patients with heart problems when it is exactly the opposite, because in reality in the 4716 of the study there are 1939 patients with previous heart problems.  And yes, this represents 42%. So with 2.4g of hydroxychloroquine on those, the overdose effect may have been immediate.
• Plus 1641 with prior lung problems, 35%.
• As the authors say, "we didn't intensively test the patients (we don't even know if they actually had VIDOC) and the mortality rate is 'all causes'".
• The conclusion is "let's quickly hide what the patients died of and conclude that HCQ is without significant benefit".

Table S1

To continue with the results here are some of what the author states

Results: 1561 patients randomly assigned to receive hydroxychloroquine were compared to 3155 patients simultaneously assigned to usual care. Overall, 418 (26.8%) patients assigned to hydroxychloroquine and 788 (25.0%) patients assigned to usual care died within 28 days (ratio 1.09 95% confidence interval [CI] 0.96 to 1.23 P=0.18). Consistent results were seen in all pre-specified patient subgroups.

Patients assigned to hydroxychloroquine were less likely to be discharged alive from hospital within 28 days (60.3% vs. 62.8% rate ratio 0.92 95% CI 0.85-0.99 p missing) and those not on invasive mechanical ventilation at baseline were more likely to achieve the composite endpoint of invasive mechanical ventilation or death (29.8% vs. 26.5% risk ratio 1.12; 95% CI 1.01-1.25 p missing). There was no excess of new major cardiac arrhythmias.

Several questions arise:

• How with 5000 patients we get such a high p for mortality when we are told; the advantage and necessity of the randomized trial is to have a very small "p".  The advantage and necessity of the randomized trial is to have a very small "p". 500 patients per group would suffice. It is mathematical.
• Why is the "p"not given  for the 2 other tests when on these 2 measures the authors conclude a significant difference?

Our clinical trial expert tells us:

"In fact, the general question is, what went wrong with the data that made such a large trial yield no significant result?"

One gets the impression that the "results are deliberately insignificant" in order to hide a disturbing reality.

To finish off, the icing on the cake: the author of the appendix to Recovery is none other than Frederic Hayden, a doctor historically close to Gilead.

The author of the document is not one of the members of Recovery, but Frederic Hayden of the University of Virginia.  One could almost believe that the Recovery team no longer wants to write the results of the study and is subcontracting it to another university. We had already mentioned this professor in a previous paper that was used primarily to get a valid clinical trial number in the United States. This same professor is a strong advocate of Gilead's recovery being quoted as saying of this drug that "this is the first convincing evidence that an antiviral drug can really benefit Covid-19 patients, especially patients hospitalized with Covid-19".

He participated in the Chinese remdesivir study and is quoted in Fortune.com as having defended the remdesivir study.  He is also known to have been close to Gilead since the HIV.

Finally he was one of the key investigators on Gilead's Tamiflu.

As in a bad movie, Gilead will have pushed its remdesivir, authorized by Europe without the slightest toxicity study, but it will have gone a long way to disqualify its effective, inexpensive, innocuous competitor, hydroxychloroquine.  This battle with unequal weapons does not serve the interest of public health but benefits the mastodons of the pharmaceutical industry, prepared to anything.  The story of recovery is not over and we would not be surprised to see a mixture of dexamethasone and remdesivir point its nose shortly as a potential combination of drugs.  The marketing techniques already used by Gilead are repeating themselves.