UP­DAT­ED: FDA re­vis­es and ex­pands draft guid­ance on ad­just­ing for co­vari­ates in RCTs

The FDA on Thurs­day re­vised and ex­pand­ed a 2019 draft guid­ance that spells out how to ad­just for co­vari­ates in the sta­tis­ti­cal analy­sis of ran­dom­ized con­trolled tri­als (RCTs).

Build­ing on the ICH’s E9 guide­line on the sta­tis­ti­cal prin­ci­ples for clin­i­cal tri­als, the 3-page draft has been trans­formed in­to an 8-page draft, with more de­tailed rec­om­men­da­tions on lin­ear and non­lin­ear mod­els to an­a­lyze the ef­fi­ca­cy end­points in RCTs.

Base­line co­vari­ates are the de­mo­graph­ic fac­tors, dis­ease char­ac­ter­is­tics, or oth­er in­for­ma­tion col­lect­ed from par­tic­i­pants be­fore they are ran­dom­ized in a tri­al, FDA ex­plains. “Co­vari­ate ad­just­ment refers to the use of base­line co­vari­ate mea­sure­ments for es­ti­mat­ing and test­ing treat­ment ef­fects be­tween ran­dom­ized groups,” the guid­ance notes.

Such ad­just­ments will gen­er­al­ly re­duce the vari­abil­i­ty of es­ti­mat­ing treat­ment ef­fects and lead to nar­row­er con­fi­dence in­ter­vals and more pow­er­ful hy­poth­e­sis test­ing.

While lin­ear mod­els are re­gard­ed by FDA as gen­er­al­ly pro­vid­ing “re­li­able es­ti­ma­tion and in­fer­ence for the av­er­age treat­ment ef­fect,” ad­di­tion­al is­sues should be con­sid­ered be­fore spon­sors use non­lin­ear mod­els, the guid­ance says.

“Spon­sors should dis­cuss with the rel­e­vant re­view di­vi­sions spe­cif­ic pro­pos­als in a pro­to­col or sta­tis­ti­cal analy­sis plan con­tain­ing non­lin­ear re­gres­sion to es­ti­mate con­di­tion­al treat­ment ef­fects for the pri­ma­ry analy­sis,” FDA says. “When es­ti­mat­ing a con­di­tion­al treat­ment ef­fect through non­lin­ear re­gres­sion, the mod­el will gen­er­al­ly not be ex­act­ly cor­rect, and re­sults can be dif­fi­cult to in­ter­pret if the mod­el is mis­spec­i­fied and treat­ment ef­fects sub­stan­tial­ly dif­fer across sub­groups.”

The re­vised draft al­so fea­tures an ex­am­ple of six steps for one sta­tis­ti­cal­ly re­li­able method of co­vari­ate ad­just­ment for an un­con­di­tion­al treat­ment ef­fect with bi­na­ry out­comes that pro­duce a re­sult­ing es­ti­ma­tor.

Frank Har­rell, bio­sta­tis­tics pro­fes­sor at Van­der­bilt’s med­ical school, told End­points News via email, “Over­all it’s re­al­ly ex­cel­lent and there is noth­ing there to pre­vent mod­ern flex­i­ble sta­tis­ti­cal mod­els from be­ing used in this con­text. Sev­er­al mod­ern meth­ods are fos­tered by this guid­ance.

“The on­ly reser­va­tion I have is its em­pha­sis on causal in­fer­ence meth­ods that are not re­al­ly need­ed when the de­sign is al­ready causal (es­pe­cial­ly when in­tent-to-treat is re­spect­ed).  To be fair, those meth­ods may be more ro­bust if you are re­al­ly bad at pre-spec­i­fy­ing the co­vari­ate ad­just­ment.  Causal meth­ods es­ti­mate a quan­ti­ty that de­pends on the in­clu­sion/ex­clu­sion cri­te­ria and the achieved mix of pa­tients, so does not trans­port to pa­tient pop­u­la­tions that don’t have the same co­vari­ate dis­tri­b­u­tion as what’s in the tri­al,” he said.

Why it Works: Man­u­fac­tur­ing a Vac­cine in a Mul­ti-Prod­uct Fa­cil­i­ty.

COVID-19 launched the pharmaceutical industry to the frontline in the battle against the fast-spreading global pandemic. The goal: distribute a safe, effective vaccine as quickly as possible. Major players in the vaccine market needed to partner with contract development and manufacturing organizations (CDMOs) to achieve the goal of mass vaccine quantities under expedited timelines. With CDMOs stepping up to play a critical role in the vaccine manufacturing process, multi-product CDMO facilities took the spotlight. Partnerships quickly formed as the race to save lives and fight a pandemic was on.

Habib Dable, Acceleron CEO

Days of heat­ed ru­mors cul­mi­nate in a re­port that Ac­celeron is in ad­vanced buy­out talks

Days of frothy rumors about possible M&A discussions at Acceleron were capped late Friday with a Bloomberg report asserting that the biotech company is in advanced talks for an $11 billion buyout deal.

Bloomberg was unable to identify any bidders in the deal, but speculation has been running rampant that the surging value of Acceleron stock had to be the result of leaks around the auction of the company. As of early Monday morning, we’re still awaiting the final word.

Endpoints News

Keep reading Endpoints with a free subscription

Unlock this story instantly and join 118,600+ biopharma pros reading Endpoints daily — and it's free.

Safe­ty fears force Pfiz­er to change piv­otal DMD gene ther­a­py tri­al pro­to­col

As one of the biggest players in an increasingly packed gene therapy space, Pfizer has taken an early lead over specialists like Sarepta in taking a Duchenne muscular dystrophy (DMD) candidate into late-stage testing. But new safety fears have led Pfizer to scale back that trial, cutting out patients with certain genetic mutations.

Pfizer has amended its enrollment protocol for a Phase III test for gene therapy fordadistrogene movaparvovec in DMD after investigators flagged severe side effects tied to specific mutations, according to a letter the drugmaker sent to Parent Project Muscular Dystrophy, a patient advocacy group.

Endpoints News

Keep reading Endpoints with a free subscription

Unlock this story instantly and join 118,600+ biopharma pros reading Endpoints daily — and it's free.

Merck CEO Rob Davis

Mer­ck emerges as lead bid­der in po­ten­tial Ac­celeron buy­out with deal pos­si­ble this week — re­port

With rumors swirling about a potential buyout of biotech Acceleron and its lead PAH drug sotatercept, market watchers have been keeping close tabs on industry movers and shakers due up for an expensive bolt-on. According to a new report, it appears Merck may be the one.

Merck is in “advanced talks” on a deal to acquire Cambridge, MA-based Acceleron in what previous reports pegged as a potential $11 billion buyout, the Wall Street Journal reported Monday. A deal could come as early as this week, according to the Journal.

Alexander Lefterov/Endpoints News

The coro­n­avirus vac­cine that the world for­got could still help save it

Back at the beginning of the pandemic — back when we still called the virus “novel” and a single case in Washington state could make headlines — there emerged the story of the coronavirus vaccine that the world forgot.

It was an allegory for our pandemic ill-preparedness. At a time when the world had been caught so flat-footed, there were a pair of scientists who had seen the crisis coming, lab-coated Cassandras with an antidote if only the world had listened sooner.

Endpoints Premium

Premium subscription required

Unlock this article along with other benefits by subscribing to one of our paid plans.

Paul Hudson, Sanofi CEO (Cyril Marcilhacy/Bloomberg via Getty Images)

Sanofi calls it quits on mR­NA Covid-19 shots, scrap­ping vac­cine from $3.2B Trans­late Bio buy­out

Sanofi is throwing in the towel on mRNA-based Covid-19 vaccines.

The French drugmaker will halt development on its unmodified mRNA Covid-19 shot despite what it said were positive Phase I/II results, a spokesperson told Endpoints News on Tuesday morning. Sanofi said the reason it’s stopping the Covid-19 mRNA program, developed in partnership with its new $3.2 billion acquisition Translate Bio, is because the market is too crowded.

Peter Marks (Jim Lo Scalzo/Pool via AP Images)

Tur­moil at CBER: Pe­ter Marks grabs con­trol of FDA's Of­fice of Vac­cines ahead of 2 key ca­reer leader de­par­tures

FDA’s top vaccine official Peter Marks is pulling the plug on a months-long transition for two top career vaccine officials who abruptly called it quits in late August.

Marion Gruber, director of the FDA’s Office of Vaccines Research & Review and 32-year veteran of the agency, her deputy director Phil Krause, announced their departures and then raised concerns with Covid-19 booster shots ahead of and during a recent Covid-19 booster vaccine advisory committee.

Brian Hubbard, Anji Pharmacetuticals CEO

Look­ing to rewrite the rules of drug li­cens­ing, start­up An­ji is on the hunt for 'dy­nam­ic eq­ui­ty' joint ven­tures

Licensing is one of the most common ways big drugmakers leverage biotech innovation to drive gains across their pipelines — and the structure of those deals is pretty well established. But one biotech with home bases in China and the US thinks it may have a better way.

On Tuesday, Cambridge-based biotech Anji Pharma closed a $70 million Series B with two late-stage molecules in the fold and a mission to rewrite the rules of drug licensing through what it calls “dynamic equity” deals and a joint venture-heavy game plan. The round was funded in whole by Chinese hedge fund CR Capital.

Con­tract re­search is hav­ing a mo­ment right now. Will M&A splash­es dri­ve the in­dus­try to even greater heights?

Contract research organizations are a fairly mysterious bunch. They’re typically considered the skilled laborers behind big drug development — the stage crews who run the trials behind some of the most (and least) successful data reveals in biopharma history.

But all that is changing.

This year, a couple of huge, out-of-the-blue M&A deals sounded the alarm on just how much money is flying around in this corner of the industry.

Endpoints Premium

Premium subscription required

Unlock this article along with other benefits by subscribing to one of our paid plans.