Robert F. Kennedy's book on "The Real Anthony Fauci." Hydroxychloroquine, Part Two

 

 There has been a long interval between my last post and this one. I apologize. I have been very busy.

In this series of critiques of Robert F. Kennedy, Jr.'s book "The Real Anthony Fauci: Bill Gates, Big Pharma, and the Global War on Democracy and Public Health," I left off discussing hydroxychloroquine.

Hydroxychloroquine in Perspective

RFK, Jr. puts forward a dizzying number of treatments for COVID. I counted 28 in his book, but it was a bit difficult to sort out the exact number. He sometimes presents the same treatment by different names as though each were different treatments. Although Kennedy included anti-stomach-acid agents, tea, IV hydrogen peroxide, intramuscular ozone, various vitamins, an antigout agent, an antidepressant, etc., he focused on two particular drugs: ivermectin and hydroxychloroquine. Having discussed ivermectin, a relatively safe compound that is not effective against COVID, I went on to discuss the safety issues and in vitro efficacy of hydroxychloroquine, concluding that hydroxychloroquine is a relatively unsafe drug that does have an in vitro effect against COVID viral replication in concentrations similar to those when it is given for other diseases. 

As I've noted before, Kennedy's book does not have page numbers, so the following citations come from my own hand-numbering. 

In Chapter One, on page 6, Kennedy said something honest, seemingly by accident. He says, "During the centuries that science has fruitlessly sought remedies against coronavirus (aka the common cold), only zinc has repeatedly proved its efficacy in peer reviewed studies." 

Getting beyond the strange notion that Kennedy thinks the coronavirus equals the common cold, he does point out that finding effective antivirals is really, really difficult. 

Contrast the above to the statement to this one, page 21. "Some 200 peer-reviewed studies by government and independent researchers deem HCQ [hydroxychloroquine] safe and effective against Coronavirus [sic], especially when taken prophylactically or when taken in the initial stages of illness along with zinc and Zithromax." Wait a moment, there are lots of peer-reviewed studies that point to a drug works against coronavirus?

He made the above claim at the time of his book being published in October 2001. On page 22, he presents a chart that he says demonstrates that out of 32 studies of hydroxychloroquine, 31 showed benefit. 

As with many presentations in Kennedy's book, there is the problem of having so many things wrong that it is hard to pick and choose what to focus on. Either I'm going to sound like I'm picking on everything, or else that I am implicitly saying, "other than this," it is fine. I've written a lot in my critiques of Kennedy's book so far. I probably could write ten times as much. However, since this table is so full of garbage, I will address it detail.

Chart presented on page 22 of RFK's book

So, he says 31 out of 32 studies showed benefits. That is not what his chart shows. Outcomes are presented with uncertainties (confidence intervals or [CI] as presented in the table). The range of uncertainty is presented as a horizontal line. When a particular line cross the odds ratio mark of one, that means that particular study could not say that the drug worked better than placebo or compared treatment. That degree of uncertainty takes place in 18 of the studies. An extreme is the study by Guérin which has a confidence interval of 0.02 to 9.06 which at the high extreme says treatment could be nine times worse than non-treatment. Only eight of the studies are noted to be randomized controls. Of these, only one has a confidence interval range below one, making the case that hydroxychloroquine doesn't work.

There are a lot more problems with the interpretation of the data. 

1. These are not all the studies of early treatment with hydroxychloroquine. They are not all peer-reviewed.
2. He mixed outcomes. Some refer to hydroxychloroquine in preventing death, some hospitalization, some progression, some viral positive, etc. These are unique and interesting questions that deserve their own analyses. Drugs can work in prophylaxis but not in treatment. You cannot just mash the results together. Often these studies looked at more than one thing. Kennedy only reports one in the table. 
3. In two of the studies there are a total of five controls. You can't get serious results from such a thing. Five more studies have less than 20. 
4. Six of the studies list the hydroxychloroquine dose as n/a, "not available." What kind of study leaves out dosing information from their publications?
5. A huge problem is that many of these studies are combination therapies. Analyzing combination therapies is a very thorny issue. Among other problems: did the hydroxychloroquine plus azithromycin show effectiveness because of the hydroxychloroquine, the azithromycin, or did you need both? Is it possible that the second drug made the combination worse than one drug alone? He says that zinc works against coronavirus. Is the zinc component doing all the work?

Since Kennedy regularly presents conclusions that are contrary to what is presented in the studies he cites, I decided to look at these studies in detail. I hadn't intended to look at them all, but then, once I started and founded study after study to be problematic, I kept on going. Some of the critiques are moderate. Some are jaw-dropping. For example, one study didn't look at hydroxychloroquine. Others weren't ever published in journals. 

Here's Looking at the Studies, Part One.

First one. Gautret. Retracted. 

Second one. Huang. Randomized clinical trial! However, this study doesn't examine hydroxychloroquine. It does have 22 patients divided into chloroquine and a second drug group, lopinavir/ritonavir. It provides several outcomes, including becoming virus results turning negative which interestingly, at 12 days had chloroquine tied with the second drug group with an ~80% success rate. Strangely, the lopinavir/ritonavir group had only a 50% discharge rate in spite of 80% being viral negative.

From Huang, et al. Only the hospital discharge data was dramatically different.

Third one. Esper. Never published, so how it is referred to as "peer-reviewed" is questionable. Maybe the peers did review it and rejected it. The study does not look terrible, but it did make the major mistake of defining those in the control group as those who refused the hydroxychloroquine plus azithromycin treatment. Allowing a control group to self-select is open to bias.

Fourth one. Ashraf. The RFK table says that 10 out 77 treated with hydroxychloroquine died while 2 out of 5 not treated with hydroxychloroquine died. I don't know how they got those figures out of the study. The study had a total of 100 patients with 94 receiving hydroxychloroquine. Further complicating interpretation is that 100% received oseltamivir (an anti-flu antiviral), and 60% received lopinavir/ritonavir (typically used for HIV). There were numerous other drugs. The authors note that azithromycin did not improve outcome. Worrisomely, three patients died during the 14 day follow-up to the study. "Of the 70 discharged patients, 40% had symptom aggravation, 8.6% were readmitted to the hospital, and three patients (4.3%) died." No information on their therapy.

Fifth one. Huang. I could find only one COVID hydroxychloroquine study published in the years 2020-21 with Huang as the first author. It showed a 44% increase in the death rate for those taking hydroxychloroquine-azithromycin to treat COVID (15.0% in the treatment group, 10.4% in the control group.)  Update: While I had searched for Huang studying hydroxychloroquine, Kennedy had referenced a study a Huang study of COVID and chloroquine. Chloroquine is not the same drug. 

Sixth one. Guérin. This study looked at control versus azithromycin versus azithromycin plus hydroxychloroquine. The control group had one death, an eighty-two year old man. Only 58% of the study participants were confirmed to have COVID. Hydroxychloroquine plus azithromycin did not perform better than azithromycin alone (p = 0.26).

Seventh one. Chen. As Kennedy's table says, Chen's study examined 30 patients. RFK identifies viral time as the outcome. The results were not favorable for hydroxychloroquine. "On day 7, COVID-19 nucleic acid of throat swabs was negative in 13 (86.7%) cases in the HCQ group and 14 (93.3%) cases in the control group (P>0.05)." The hydroxychloroquine group had more adverse effects, but the overall study size was small. (Being a common name, there is a study by another Chen of hydroxychloroquine and COVID in 2020, but that involved 61 patients, so it is not the one Kennedy is citing.)

Eighth one. Derwand. A total of 141 patients with confirmed COVID were treated with zinc, low-dose hydroxychloroquine, and azithromycin. Their outcomes were compared to 377 matched controls from the community. One death occurred in the treatment group and thirteen in the control group. All of this sounds impressive and I would have given this study the first thumbs up among the ones I've reviewed so far, except that the senior author is Zelenko, who well before this study made outrageous claims for his triple therapy. The fact that this is a retrospective study and has much fewer numbers than his previous reports (500 patients treated with 100% success rate) which involved treating people without ethical review board permission suggests that he is selecting his data.

Ninth one. Mitjà. His study looked at hydroxychloroquine in patients with mild COVID. As he concluded: "In patients with mild COVID-19, no benefit was observed with HCQ beyond the usual care."

Tenth one. Skipper. He had similar conclusions as did Mitjà: "Conclusion: Hydroxychloroquine did not substantially reduce symptom severity in outpatients with early, mild COVID-19." I don't know how Kennedy got the numbers he reported. They're different from those in the paper.

Eleventh one. Hong. You may have already noticed that in the columns treatment and control, sometimes Kennedy includes both total numbers and how many of those fit the outcomes (for example, for Skipper, 5/231 (treatment) and 8/234 (control)). In six instances, Kennedy just gives single numbers not providing what they mean. For Hong he lists viral positive as the outcome and 42 (treatment) and 48 (control). So what did the study do? They recruited 100 patients: 92 were treated with lopinavir/ritonavir, 90 were treated with hydroxychloroquine, and 22 were treated with dexamethasone. This means the large majority received both lopinavir/ritonavir AND hydroxychloroquine. How do you get 48 controls from that? And what does it mean to be a control? Viral shedding after 30 days was the endpoint of this study. The study did come to the conclusion that hydroxychloroquine was the only variable associated with protection, however Kennedy's presentation of the numbers is questionable.

Twelfth one. Bernabeu-Wittel.  This study looked at a wide range of interventions. "Ten key processes and interventions were established (provision of informatics infrastructure, medical equipment, and human resources, universal testing, separation of "clean" and "contaminated" areas, epidemiological surveys, and unified protocols stratifying for active or palliative care approach, among others)." As you may note in the above hydroxychloroquine treatment was not a main focus. In Kennedy's table it lists deaths as the end point with 189 deaths in the treatment section and 83 in the control. I cannot figure out how they got those figures or related them to those who took hydroxychloroquine in the study. The authors don't even reference hydroxychloroquine in the discussion or conclusions of their study.

Thirteenth one. Yu. According to Kennedy 1/73 died in the hydroxychloroquine treatment group versus 238/2604 in the control group. According to the Yu paper, "We found that fatalities are 18.8% (9/48) in HCQ group, which is significantly lower than 47.4% (238/502) in the NHCQ group (P<0.001)." The discrepancy between the 2604 and 502 can be relegated to the fact that they reported hydroxychloroquine treatment in only those in critical condition and all deaths came among those in critical condition. The reason for the difference between 1/73 and 9/48 is unclear.

Fourteenth one. Ly.  Ly, et al., looked at deaths in a nursing home among those testing positive for COVID with 116 receiving hydroxychloroquine/azithromycin and 110 in the control group. The patients were fairly old: with 43% of those in the treatment group being over 85 along with 56% of those in the control group. There was a dramatic difference in survival rate between women (14.0% death rate) and men (30.7% death rate). Those assigned to the drug treatment group were 60% women, this leaves unresolved whether age and gender were responsible for the better outcome for those treated. From the paper: "Through multivariate analysis, the death rate was positively associated with being male (30.7% vs. 14.0%, OR = 3.95 [1.65–9.44]; P = 0.002), aged > 85 years (26.1% vs. 15.6%, OR = 2.43 [1.04–5.69]; P = 0.041)."

Fifteenth one. Ip.  Ip, et. al., looked at patient charts to find 97 who had been treated with hydroxychloroquine and matched those with 970 who had not and then looked at outcomes. While the matching appeared studious, the hydroxychloroquine-treated patients had more significantly more comorbidities (suggesting any positive results would speak favorably for hydroxychloroquine). While a significant effect on death was not found (p = 0.578), when age was taken into account, they concluded that hydroxychloroquine lowered hospitalization rates (p = 0.045). 

Sixteenth one. Heras.  Heras looked at outcomes in 100 patients in nursing homes (average age, 85). 

Treatment was varied. "As for the treatment, five categories were defined: patients who received hydroxychloroquine and azithromycin, hydroxychloroquine only, hydroxychloroquine plus another antibiotic, beta-lactam, or quinolone antibiotics, and no treatment." Although most of the numbers are unclear in the presentation, 70% of the patients received hydroxychloroquine plus azithromycin and 17% received no treatment (which the paper confusingly says includes beta-lactam or quinolone antibiotics), leaving 13% (13 patients) divided into the other categories. 

Of the whole group, 24 patients died. The paper leaves unclear exactly how many did not receive hydroxychloroquine. It does provide enough information to determine that 8 of 70 who received hydroxychloroquine plus azithromycin died. It also says that 3 of 9 of those receiving hydroxychloroquine died (presumably meaning hydroxychloroquine alone). Their table goes on to say that there were 21 untreated with 13 deaths. (The results are presented like one of those puzzles that say the man with the red hat stood behind the lady with the blue pants.) None of the results in their paper align with Kennedy's numbers in his table. 

Seventeenth study. Kirenga. He looked at 56 patients with COVID, 29 of whom received hydroxychloroquine. Nearly all who received hydroxychloroquine also received an antibiotic. Interestingly, they averaged 34.2 years old, much younger than other studies. I admire the completeness of his study, he looked into or described many cofactors. The study was conducted at two hospitals. One hospital did not prescribe hydroxychloroquine, the other prescribed it to 94% of its patients. The median hospital stay was significantly shorter for the hospital that did not prescribe hydroxychloroquine (p = 0.012). His conclusion? "Outcomes did not differ by HCQ treatment status in line with other concluded studies on the benefit of using HCQ in the treatment of COVID-19." 

It is strange that Kennedy cites Kirenga in favor of hydroxychloroquine. Kirenga conducted a second study "a randomized open label Phase II clinical trial" looking at hydroxychloroquine. His conclusions: "Our results show that HCQ 400 mg twice a day for the first day followed by 200 mg twice daily for the next 4 days was safe but not associated with reduction in viral clearance or symptom resolution among adults with COVID-19 in Uganda."

Eighteenth study. Sulaiman has not published a peer-review article on the efficacy of hydroxychloroquine. He does have this non-peer reviewed article which is referenced by Kennedy. The study compared those who received only supportive care to those who received supportive care plus hydroxychloroquine, zinc, cetirizine (an antihistamine), and acetaminophen. "The primary outcome of interest was hospital admission within 28-days of presentation. The secondary outcome of the study was a composite of ICU admission and/or mortality during the 28-day follow up period." The numbers that Kennedy references in terms of mortality are consistent with those in the document.

Sulaiman does have a peer-reviewed article on the safety of hydroxychloroquine in 2733 patients (data from the above study). Those treated were prescribed hydroxychloroquine plus zinc, acetaminophen and an antihistamine. 6.7% of treated patients experienced drug-related toxicities and 4.1% withdrew from the study due to those toxicities. 

Continued with Entry 14. 

Martin Hill Ortiz is a professor of pharmacology and author of several novels. 

His new novel, The Missing Floor, will be coming out on April 8 from Oliver-Heber books. 

The Missing Floor, available April 8