Beneficent Human Research Clinical Trial Tribulations During a Pandemic Michael Para, MD May 14, 2020
The Pandemic The Perfect Storm new infectious agent transmitted and acquired by just breathing nosocomial transmission exponential spread - an infected person would infect 2-3 others significant morbidity explosive spread put heavy demands on health care resources societal disruption and the worse - social media went crazy Viral Infectious Agent no effective therapeutics innumerable agents show activity in tissue culture
Pandemic Do Something Now! Rushed Science Pandemic started in China Suddenly overwhelming medical care required Unknown disease, unknown pathophysiology, unclear spread Rushed Publications Majority of early publications came from China Poor quality science in midst of crisis Poor quality publications Rapid manuscript review if any Usually released in electronic form often labeled “peer review pending” Lay media reporting results before manuscript review – OMG!
Do Something Now! Clinical Trials The Good, the Bad and the Ugly Rushed Treatment Communications As disease subsided in China, reports flooded out Reports having 2 or 3 patients flourished a lead JAMA article had 5, a PNAS paper had 10 patients Often treated with multiple agents simultaneously European publications minimally better Social Media Political commentary with little scientific input Discuss 3 therapeutics showing the good, the bad and the ugly of trials
Trial 1 – the Ugly Hydroxychloroquine and Azithromycin as Treatment of COVID-19: Open-label non-randomized clinical trial Tweeted to be a “Game changer”
Trial 1 Hydroxychloroquine and Azithromycin as Treatment of COVID-19: Open-label non-randomized clinical trial Rationale – Hydroxychloroquine active in culture, reportedly effective in China Subjects – treated subject had COVID-19 with spectrum of illness from Marseille, controls were either study ineligible patients or from 4 other cities Protocol – Daily NP swabs for viral PCR among treated and controls Endpoint – PCR for virus from NP on day 6 Sample size – with 50% efficacy at 85% power there should be 24 patients each arm “ Enrolled 36 out of 42 patients meeting inclusion” 26 treated, 16 controls (?10) Results Clinical status at entry of treated and untreated was not significantly different 6 treated exited early – 3 to ICU, 1 death, 1 toxicity, 1 LTFU; excluded in analysis 20 treated were in analysis 14 hydroxychloroquine alone + 6 with azithro to prevent bacterial infection
Trial 1 Hydroxychloroquine and Azithromycin as Treatment of COVID-19: Open-label non-randomized clinical trial Results – Reportedly 14/20 treated were undetectable at day 6 but only 2/16 of controls Analysis excluded 6 treated patients dropping out counted untreated pts as infected although they lacked the endpoint of a day 6 PCR Referenced manuscript for their PCR Assay used in this trial gave limited technical details This assay reported finding no COVID infection among 273 persons with fever and cough who arrived in Marseille from China or Italy during their epidemics – very suspect assay Confirmatory study by same authors of 80 patients on HCQ + azithro had no controls In contrast to 1 st report of 100% undetectable, < 50% undetectable at day 5 Good News > 25 placebo controlled trials of HCQ/AZ for COVID listed on clinicaltrials.gov a 3000 person placebo controlled trial for prevention of COVID in HCW and first responders Dr. Fauci asked the ACTG to conduct 2 large (1000+) placebo controlled trials one trial for ambulatory COVID+ and one trial for hospitalized COVID patients
Convalescent Plasma for COVID-19 – the bad Rationale – Plasma from recovered COVID-19 patients should have viral antibodies An infusion should neutralize virus and lead to improvement. Background – Serotherapy dates back to 1918 Flu epidemic Plasma has show some beneficial activity in hep B, polio, measles, flu, ? Ebola H1NI 2009-11, Clin Inf Dis, severe dz in 93 pts, mortality 20% plasma vs 55% none Uncontrolled trials of plasma for MERS and SARS showed faster viral clearance There was some reported but unpublished success of convalescent plasma in China Do Something Trials JAMA online published 5 COVID patients on ventilators received 2 units of plasma Plasma had SARS-CoV-2 ELISA Ab titer >1:1000 and neutralizing ab titer >1:40 All 5 patients treated with steroids and some antivirals of unknown efficacy. A PNAS paper reported on 10 patients who were treated similarly.
Convalescent Plasma for COVID-19 – the bad Rationale – Plasma from recovered COVID-19 patients should have viral antibodies An infusion should neutralize virus and lead to improvement. Do Something Trials JAMA online published 5 COVID patients on ventilators received 2 units of plasma SARS-CoV-2 ELISA antibody titer >1:1000 and neutralizing ab titer >1:40 Patients had been symptomatic for over 2 weeks, and in hospital > 10 days All treated with steroids and multiple other antivirals of unknown efficacy JAMA Results - After 2 wks, 3 left hospital and two were improved on ventilator WHO CAN TELL IF IT DID ANYTHING ? (other than give authors a JAMA paper) Editorial in the issue of JAMA stated if the results of a rigorously conducted investigation, such as a large-scale randomized clinical trial, demonstrated efficacy, and included data supporting biomarkers of success, this therapy could help change the course of this pandemic. Such biomarkers would accelerate our vaccine efforts.
Convalescent Plasma for COVID-19 – the bad Do Something March 24 - 3 days before any data (even pre-print) reported FDA approved infusion of “investigational convalescent plasma” if O2 sat<93% “since no known cure exists and a vaccine is more than 1 year away“ Any licensed physician could request a single patient Emergency IND (eIND) COVID-19 convalescent plasma was to be provided by a blood bank. March 29 – FDA arranges with Mayo Clinic to establish an Expanded Access Protocol Protocol Patients in acute care facilities infected with SARS-CoV-2 Severe or life-threatening COVID-19, having O2 sat<93% on room air Consent signed by the patient or legal representative Any site can apply and any licensed physician can participate Minimal data collected – duration of hospitalization, ventilator, ICU stay, death
Convalescent Plasma for COVID-19 – the bad Do Something Mayo Clinic Expanded Access Protocol as of May 11 2200 sites approved to give plasma 5400 physicians approved to order convalescent plasma 14,000 patients prescribed convalescent plasma 8,800 plasma units infused 25 units infused since EAP in place (7 days) at OSU AND WE STILL HAVE NO IDEA IF IT DOES ANY GOOD OR HARM !!!! The Not Quite So Bad About 60 trials are listed on clinicaltrials.gov using convalescent plasma Sadly under 10 of these are actually controlled trials only 3 are controlled using non-convalescent plasma others involve not so sick COVID patients in ED given plasma or not others involve infusions of 0, or 1 or 2 units Will this FDA promoted EAP subvert controlled trials so we never learn if it works?
Remdesivir for COVID-19 – the good “Do Something Helpful Now” Remdesivir Nucleotide analogue that inhibits viral RNA polymerases, daily parenteral agent Made by Gilead who developed blockbuster nucleotides against HIV and HCV Primate models showed effective against both MERS and SARS and Ebola Studies in Africa did not show efficacy vs Ebola but it was shown to be safe. COVID-19 studies Gilead initiated 2 trials in early Feb in China 450 patient placebo controlled trial of SOC vs remdesivir for severe COVID 310 patient placebo controlled trial of SOC vs remdesivir for moderate COVID Both closes in April with less than 50% accrual and inconclusive results NIAID initiated Adaptive COVID-19 Treatment Trial (ACTT) on Feb 23 At that time there were 15 cases and no deaths from COVID in the US. It was a placebo controlled trial enrolling 800 COVID PCR+ with O2 sat ≤94%
Remdesivir for COVID-19 – the good “Do Something Helpful Now” COVID-19 studies Gilead initiated 2 trials on March 3 at multi-nation sites – “Simple Trials” 400 patient REM for 5 day vs 10 day for severe COVID but not on ventilator Endpoint was improvement on day 11 600 patient REM 5 d vs REM 10 d for moderate COVID O2 sat >94% similar design Limited Compassionate Use Protocol multi-nation study initiated mid-Feb Reported in NEJM on April 10 Number of authors exceeded the number of eligible subjects - 55 68% of subject improved but study conclusion was - need a controlled trial March 27-REM compassionate expanded access for ventilated patients in US
Remdesivir for COVID-19 – the good “Do Something Helpful Now” COVID-19 studies ACTT April 29 - NIAID press release based on DSMB report Median time to recovery from severe COVID 11 days in REM arm vs 15 days inSOC arm (p<0.001) Survival for REM arm 92% vs 88.4% for SOC arm (p<0.059) No other reporting yet Simple Trial for severe disease April 29 press release from Gilead 5 days of treatment was as effective as 10 days, Those treated with Sx<10 days did better than treated later Based on this, FDA authorized emergency use on May 1 for O2 sat ≤94% Good science even during a pandemic wins out!
Questions?
Recommend
More recommend
