WHO Leak Sheds Light on Covid-19 Drug Tri­als

Data posted ac­ci­den­tally to WHO web­site shows remde­sivir didn’t im­prove pa­tients’ con­di­tion in rig­or­ous Chi­nese tri­als, which have since been dis­con­tin­ued

Caixin China Economics & Finance - - SPECIAL REPORT -

By Flynn Mur­phy, Di Ning and He Jing­wei

Shares in Gilead Sci­ences Inc. tum­bled af­ter dis­heart­en­ing data from a closely-watched Wuhan trial of its ex­per­i­men­tal coron­avirus drug was posted on­line late April by the World Health Or­ga­ni­za­tion (WHO), ap­par­ently by ac­ci­dent.

The since-re­moved data was from a study of 237 pa­tients with se­vere Covid-19, 158 of whom were given the an­tivi­ral remde­sivir and 79 a placebo drug. The re­sults show remde­sivir did not im­prove pa­tients’ con­di­tion or re­duce the amount of virus they had in their blood­stream, ac­cord­ing to the Fi­nan­cial Times, which was among the first to re­port the leak.

The study was one of two rig­or­ous Chi­nese tri­als that were both qui­etly dis­con­tin­ued in April af­ter strug­gling to re­cruit pa­tients, rais­ing ques­tions about the com­plete­ness and ac­cu­racy of promis­ing early ev­i­dence for remde­sivir’s ef­fec­tive­ness and safety.

Gilead said the in­for­ma­tion was re­leased with­out per­mis­sion and was not peer re­viewed, and said the post “in­cluded in­ap­pro­pri­ate char­ac­ter­i­za­tions of the study,” in a state­ment.

The study was aborted due to dif­fi­culty find­ing test sub­jects and was “un­der­pow­ered to en­able sta­tis­ti­cally mean­ing­ful con­clu­sions,” the com­pany said. It said peer-re­viewed re­sults were ex­pected soon.

The news has am­pli­fied the voices of skep­tics who said more ro­bust ev­i­dence was needed to en­sure coron­avirus pa­tients treated with ex­per­i­men­tal drugs were not im­prov­ing for other rea­sons. While the dis­ease can be deadly and un­pre­dictable, and has the full spec­trum of sever­ity, most peo­ple who con­tract Covid-19 re­cover with­out treat­ment.

Gilead is forg­ing ahead with two Phase 3 clin­i­cal tri­als, one which hopes to re­cruit 1,600 pa­tients with moder­ate Covid-19, and an­other which tar­gets 6,000 pa­tients with se­vere cases of the dis­ease, ac­cord­ing to trial records posted on Clin­i­cal­tri­als.gov, a U.S. govern­ment-backed clin­i­cal tri­als database. Place­bos will not be used in ei­ther trial.

Allen Cheng, a lead­ing in­fec­tious dis­eases physi­cian who is ad­vis­ing the Aus­tralian govern­ment on its coron­avirus re­sponse, said it was un­usual that the Chi­nese tri­als were not moved over­seas to other Covid-19 hotspots in or­der to col­lect more data.

“What you’d nor­mally do in that sit­u­a­tion is just re­cruit in an­other coun­try. It wouldn’t be un­usual for big pharma com­pa­nies to do a trial in 10 coun­tries,” Cheng, said, adding dif­fer­ent stan­dards of med­i­cal care could be fac­tored into the anal­y­sis.

He said the fact Gilead was seek­ing to en­roll such large num­bers of pa­tients in its own on­go­ing tri­als could be a bad sign. “If you need that many pa­tients to show a dif­fer­ence, they’re prob­a­bly look­ing for a pretty small ef­fect size … which means it may not have a dra­matic ef­fect.”

“If it re­duced (deaths) by half, you’d only need a few hun­dred par­tic­i­pants to know that.”

On the other hand, there are still hopes the drug may be ef­fec­tive if given ear­lier in the course of in­fec­tion, while the pa­tients’ vi­ral load is lower, Cheng said. Gilead did not re­spond to de­tailed ques­tions.

In a state­ment, the WHO said: “In re­sponse to WHO ask­ing for in­for­ma­tion and stud­ies to be shared early, a draft doc­u­ment was pro­vided by the authors to WHO and in­ad­ver­tently posted on the web­site and taken down as soon as the mis­take was no­ticed.”

High hopes

Quadru­ple-blinded con­trolled tri­als — where the pa­tient, care provider, in­ves­ti­ga­tor and asses­sor are all un­aware who has an ac­tive drug or a placebo — are tricky to pull off even in ideal con­di­tions.

And the con­di­tions in China’s twin remde­sivir tri­als were far from ideal, tak­ing place in 10 hos­pi­tals over­run with peo­ple sick from a pre­vi­ously unseen and poorly un­der­stood dis­ease, at the epi­demic ground zero of Wuhan as a health­care cri­sis un­folded around them.

But they were led by Cao Bin, head of pul­monary and crit­i­cal care medicine at Bei­jing’s China-ja­pan Friend­ship Hos­pi­tal, a well-re­garded and fre­quently pub­lished fig­ure China’s Covid-19 fight. And they were backed at the high­est level by Chi­nese au­thor­i­ties and reg­u­la­tors. Gilead Sci­ences sup­plied remde­sivir for free, and had what it called “in­put on study de­sign and con­duct.”

They had the am­bi­tious goal of re­cruit­ing around 750 pa­tients hos­pi­tal­ized with Covid-19 — 308 with mild and moder­ate symp­toms, and 453 with se­vere cases.

On Feb. 3, U.S. time, as air­lines were al­ready ground­ing flights amid fears about pos­si­ble pan­demic spread of Covid-19, a batch of remde­sivir was dis­patched on the fi­nal United Air­lines flight of the day to Bei­jing. SF Ex­press de­liv­ered the drug to Wuhan’s

Jinyan­tan Hos­pi­tal, one of the first fa­cil­i­ties to treat

Covid-19 pa­tients, on Feb. 6. Light­ning fast ap­proval from China’s drug reg­u­la­tor and even faster trial ap­proval from the Min­istry of Science and Tech­nol­ogy had al­ready been granted, and the first Chi­nese

Covid-19 pa­tient was in­jected with remde­sivir that same day.

The trial in pa­tients with mild or moder­ate symp­toms started Feb. 12 and was to con­clude April 27, ac­cord­ing to records posted on Clin­i­cal­tri­als.gov. The trial in­volv­ing se­vere pa­tients was ex­pected to con­clude on May 1.

But in late Fe­bru­ary, Cao re­vealed the dif­fi­cul­ties the team was hav­ing in re­cruit­ing enough pa­tients amid com­pe­ti­tion with other drug tri­als. The tri­als’ strict re­quire­ment that pa­tients had not taken any other treat­ment in the prior 30 days also made it dif­fi­cult to find el­i­gi­ble par­tic­i­pants.

Fur­ther bad news was buried in an April 10 note from Gilead Chairman and CEO Daniel O’day. The open let­ter led with hope­ful find­ings from a Gilead­funded in­ves­ti­ga­tional study of a small num­ber of pa­tients: a New Eng­land Jour­nal of Medicine pa­per said most of the 53 pa­tients in the U.S., Europe and Canada who were given remde­sivir un­der com­pas­sion­ate use pro­vi­sions showed im­prove­ment.

But that was no clin­i­cal trial, and the con­clu­sions were un­re­li­able. Fur­ther down in O’day’s let­ter was the bombshell: one of Cao Bin’s tri­als had al­ready been aborted. “The pub­li­ca­tion of data from the China remde­sivir tri­als rests with the Chi­nese in­ves­ti­ga­tors, but we have been in­formed that the study in pa­tients with se­vere symp­toms was stopped due to stalled en­roll­ment,” the let­ter said.

By April 23, up­dated trial records said the se­cond trial, in­volv­ing mild or moder­ate cases, had also been sus­pended be­cause Covid-19 had been “con­trolled well in China, (and) no el­i­gi­ble pa­tients can be en­rolled at present.”

In­ves­ti­ga­tor ini­ti­ated tri­als

What set the Cao-led tri­als apart was that they were “in­ves­ti­ga­tor-ini­ti­ated tri­als,” which are gen­er­ally bet­ter re­garded than tri­als led by phar­ma­ceu­ti­cal com­pa­nies on their own drugs, ex­perts told Caixin. Though in this case, as is com­mon, the drug­maker did pro­vide test drugs, place­bos and ad­vice.

How the data is now used will de­pend on the agree­ment be­tween the drug com­pany and the in­ves­ti­ga­tor. The de­tails of the deal be­tween Cao and Gilead have not been made pub­lic.

Ran­dom­ized, blinded con­trolled tri­als are con­sid­ered the gold stan­dard for ver­i­fy­ing the safety and ef­fec­tive­ness of drugs, and pro­vide the strong­est ev­i­dence that they work or don’t. But be­cause they re­quire some peo­ple to go with­out a po­ten­tially life­sav­ing medicine, and need the re­searchers

The study was aborted due to dif­fi­culty find­ing test sub­jects and was “un­der­pow­ered to en­able sta­tis­ti­cally mean­ing­ful con­clu­sions,” the com­pany said. Photo: Xinhua

Newspapers in English

Newspapers from China

© PressReader. All rights reserved.