Adrian Hayday. Crick

Adri­an Hay­day re­futes chal­lenge on in­flu­en­tial au­toan­ti­body pa­per by Karolin­s­ka sci­en­tists

Back in 2016, Adri­an Hay­day and his col­leagues at King’s Col­lege Lon­don — to­geth­er with some col­lab­o­ra­tors in Ger­many and Es­to­nia — made a splash in the im­munol­o­gy world with a pa­per, pub­lished in Cell, sug­gest­ing that pa­tients with a par­tic­u­lar ge­net­ic de­fect pro­duce more au­toan­ti­bod­ies than pre­vi­ous­ly known. These self-gen­er­at­ed an­ti­bod­ies, they pro­posed, of­fered ther­a­peu­tic po­ten­tial for au­toim­mune dis­eases, most promi­nent­ly Type 1 di­a­betes.

The re­sults formed the foun­da­tion of a Ger­man biotech dubbed Im­muno­Qure, which lat­er inked a part­ner­ship with France’s Servi­er to pur­sue an au­toan­ti­body that found to neu­tral­ize in­ter­fer­on-α. Servi­er pledged up to €164 mil­lion to the deal.

Nils Lan­de­gren

It al­so piqued the in­ter­est of Nils Lan­de­gren, a pro­fes­sor at the Karolin­s­ka In­sti­tute who’s spent years de­cod­ing the au­toim­mune reg­u­la­tor gene (AIRE) and its de­fi­cien­cy. While he’s helped iden­ti­fy a num­ber of bona fide au­toanti­gens in au­toim­mune poly­en­docrine syn­drome type 1, the vast num­ber of au­toan­ti­bod­ies iden­ti­fied in the Cell pa­per caught him by sur­prise. And he want­ed to see if there was re­al­ly a link be­tween one of the known au­toanti­gens to Type 1 di­a­betes. So he took a cou­ple of ex­ist­ing datasets and then ran his own analy­sis with a team from Swe­den and the US.

Long sto­ry short: They couldn’t repli­cate the re­sults.

“We could show that us­ing ran­dom da­ta, we got ba­si­cal­ly the same re­sult they did,” Lan­de­gren told Med­Scape. “We have a very strong case their re­sults are false.”

In a re­sponse to the pa­per, Hay­day stood by his re­sults and the meth­ods brought in­to ques­tion, coun­ter­ing that “flaws in de­sign and im­ple­men­ta­tion in­val­i­date this chal­lenge.”

“While we wel­come open, con­struc­tive dis­course about sci­ence, we are dis­ap­point­ed by this dis­pute be­cause we be­lieve it re­flects sim­ple but im­por­tant dif­fer­ences be­tween our ap­proach­es that could have been eas­i­ly re­solved, had Lan­de­gren and co-work­ers ap­proached us di­rect­ly,” his team wrote.

The dis­putes are two-pronged. The first has to do with the num­ber of au­toan­ti­bod­ies APS1 pa­tients har­bor. Hay­day’s group — with Stef­fen Mey­er and Mar­tin Wood­ward — as co-first au­thors found that col­lec­tive­ly hit over more than 3,700 hu­man pro­teins out of the 9,000 screened. But when Lan­de­gren’s team ran their sam­ples against the same as­say, they found “on­ly a very re­strict­ed set of pro­teins (<1% of the pro­tein pan­el) are tar­get­ed by au­toan­ti­bod­ies in mul­ti­ple APS1 pa­tients.”

The sta­tis­ti­cal method de­ployed for the pa­tient group ver­sus the con­trol, Lan­de­gren wrote, skewed the re­sults to show high­er lev­els of au­toan­ti­body sig­nals than in re­al­i­ty.

To bet­ter de­ter­mine how the an­a­lyt­i­cal bias in­flu­enced the over­all re­sults, we ap­plied a re­versed cut­off to our dataset based on the val­ues ob­tained for the 51 APS1 pa­tients in­stead of those for the 21 healthy con­trols. The healthy con­trols now turned out to show greater num­bers of high-lev­el (Z ≥ 5) au­toan­ti­body sig­nals than the APS1 pa­tients, re­veal­ing that the skew­ing ef­fect of the da­ta analy­sis was dom­i­nat­ing the re­sults.

To be sure, the 2016 pa­per had a stat­ed in­ter­est in find­ing pa­tients’ “pri­vate au­toan­ti­body reper­toire” — which might not be shared with oth­ers.

“When you un­der­take a sta­tis­ti­cal treat­ment of da­ta, there is no sil­ver bul­let, you have to use what makes sense,” he said to Med­Scape.

Even if they ap­ply a more con­ser­v­a­tive sta­tis­ti­cal ap­proach, he wrote in his re­sponse, they “iden­ti­fied re­ac­tiv­i­ties over­lap­ping 81% with our orig­i­nal study: again, these com­prised broad­ly shared au­toanti­gens and from ~30 to~100 pri­vate speci­fici­ties that col­lec­tive­ly com­posed a sub­stan­tial frac­tion of the pro­teome.”

Per­haps more point­ed­ly, Lan­de­gren’s team at­tempt­ed to poke holes in the cor­re­la­tion be­tween hav­ing neu­tral­iz­ing au­toan­ti­bod­ies of in­ter­fer­on-α and be­ing free of Type 1 di­a­betes. Hay­day’s the­o­ry was that an­ti-IFNα an­ti­bod­ies ap­peared to pro­tect APS1 pa­tients from ul­ti­mate­ly de­vel­op­ing that dis­ease, sup­port­ing pre­vi­ous claims that type I IFN con­tributes to Type 1 di­a­betes,

Writ­ing in eLife — where they went af­ter get­ting turned down by Cell — Lan­de­gren re­port­ed strong neu­tral­iza­tion of type 1 IFN in all APS1 sam­ples, not just the ones from di­a­bet­ics (a find­ing that Hay­day dis­missed as a re­sult of a “shod­dy ex­per­i­ment” con­duct­ed with less sen­si­tive meth­ods.)

“[A]t high con­cen­tra­tions, the sera of pa­tients with and with­out T1D showed com­pa­ra­ble ac­tiv­i­ties, but at low­er, sub-sat­u­ra­tion con­cen­tra­tions [50-fold di­lu­tions], the co­hort with­out T1D showed sig­nif­i­cant­ly greater ca­pac­i­ty to lim­it IFNα ac­tiv­i­ty,” Hay­day et al wrote in their re­sponse.

They added that the oth­er cen­tral ob­ser­va­tions from their 2016 pa­per re­mained un­chal­lenged: that APS1 pa­tients share nat­u­ral­ly aris­ing au­toan­ti­bod­ies to a small sub­set of pro­teins that are of ex­treme­ly high affin­i­ty.

Im­muno­Qure is still in the CMC stage for its drug can­di­date tar­get­ing IFN-α, but they won’t ex­act­ly be the first. As­traZeneca re­cent­ly con­ced­ed a Phase III de­feat for an­i­frol­um­ab in lu­pus, an an­ti­body that hits the re­cep­tors for IFN-α, IFN-β and IFN-ω. The hope is that by hit­ting the cy­tokines di­rect­ly — a “very dif­fer­ent modal­i­ty” — their nov­el pan an­ti-Type I IFN neu­tral­iz­ing an­ti­body will work “far more ef­fec­tive­ly” than the ex­ist­ing mAbs pro­duced by As­traZeneca or Genen­tech, Hay­day added.

Brian Kaspar. AveXis via Twitter

AveX­is sci­en­tif­ic founder fires back at No­var­tis CEO Vas Narasimhan, 'cat­e­gor­i­cal­ly de­nies any wrong­do­ing'

Brian Kaspar’s head was among the first to roll at Novartis after company execs became aware of the fact that manipulated data had been included in its application for Zolgensma, now the world’s most expensive therapy.

But in his first public response, the scientific founder at AveXis — acquired by Novartis for $8.7 billion — is firing back. And he says that not only was he not involved in any wrongdoing, he’s ready to defend his name as needed.

I reached out to Brian Kaspar after Novartis put out word that he and his brother Allen had been axed in mid-May, two months after the company became aware of the allegations related to manipulated data. His response came back through his attorneys.

Endpoints News

Basic subscription required

Unlock this story instantly and join 57,900+ biopharma pros reading Endpoints daily — and it's free.

UP­DAT­ED: An em­bold­ened As­traZeneca splurges $95M on a pri­or­i­ty re­view vouch­er. Where do they need the FDA to hus­tle up?

AstraZeneca is in a hurry.

We learned this morning that the pharma giant — not known as a big spender, until recently — forked over $95 million to get its hands on a priority review voucher from Sobi, otherwise known as Swedish Orphan Biovitrum.

That marks another step down on price for a PRV, which allows the holder to slash 4 months off of any FDA review time.

Endpoints News

Basic subscription required

Unlock this story instantly and join 57,900+ biopharma pros reading Endpoints daily — and it's free.

Martin Shkreli [via Getty]

Pris­on­er #87850-053 does not get to add drug de­vel­op­er to his list of cred­its

Just days after Retrophin shed its last ties to founder Martin Shkreli, the biotech is reporting that the lead drug he co-invented flopped in a pivotal trial. Fosmetpantotenate flunked both the primary and key secondary endpoints in a placebo-controlled trial for a rare disease called pantothenate kinase-associated neurodegeneration, or PKAN.

Endpoints News

Basic subscription required

Unlock this story instantly and join 57,900+ biopharma pros reading Endpoints daily — and it's free.

We­bi­nar: Re­al World End­points — the brave new world com­ing in build­ing fran­chise ther­a­pies

Several biopharma companies have been working on expanding drug labels through the use of real world endpoints, combing through the data to find evidence of a drug’s efficacy for particular indications. But we’ve just begun. Real World Evidence is becoming an important part of every clinical development plan, in the soup-through-nuts approach used in building franchises.

I’ve recruited a panel of 3 top experts in the field — the first in a series of premium webinars — to look at the practical realities governing what can be done today, and where this is headed over the next few years, at the prodding of the FDA.

ZHEN SU — Merck Serono’s Senior Vice President and Global Head of Oncology
ELLIOTT LEVY — Amgen’s Senior Vice President of Global Development
CHRIS BOSHOFF — Pfizer Oncology’s Chief Development Officer

A premium subscription to Endpoints News is required to attend this webinar. Please upgrade to either an Insider or Enterprise plan for access. Already have Endpoints Premium? Please sign-in below. You can contact our Subscriptions team at with any issues.

Endpoints Premium

Premium subscription required

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

Am­gen, Al­ler­gan biosim­i­lar of Roche's block­buster Rit­ux­an clears an­oth­er US piv­otal study 

Novartis $NVS may have given up, but Amgen $AMGN and Allergan $AGN are plowing ahead with their knockoff of Roche’s blockbuster biologic Rituxan in the United States.

Their copycat, ABP 798, was found to have a clinically equivalent impact as Rituxan — meeting the main goal of the study involving CD20-positive B-cell non-Hodgkin’s lymphoma patients. This is the second trial supporting the profile of the biosimilar. In January, it came through with positive PK results in patients with rheumatoid arthritis.

BeiGene and Mus­tang nail down spe­cial FDA sta­tus for top drugs; Roche bags added cov­er­age for Hem­li­bra

→ BeiGene $BGNE is getting a boost in its drive to field a rival to Imbruvica. The FDA has offered an accelerated review to zanubrutinib, a BTK inhibitor that has posted positive results for mantle cell lymphoma. The PDUFA date lands on February 27, 2020. The drug scored breakthrough status at the beginning of the year.

→ BeiGene isn’t the only biopharma company to gain special regulatory status today. Mustang Bio $MBIO and St. Jude Children’s Research Hospital announced that MB-107, a lentiviral gene therapy for the treatment of X-linked severe combined immunodeficiency, also known as bubble boy disease, has been granted Regenerative Medicine Advanced Therapy status.

Trump ad­min­is­tra­tion re­vives bid to get drug list prices on TV ads

The Trump administration is not giving up just yet. On Wednesday, the HHS filed an appeal against a judge’s decision in July to overturn a ruling obligating drug manufacturers to disclose the list price of their therapies in television adverts — hours before it was stipulated to go into effect.

In May, the HHS published a final ruling requiring drugmakers to divulge the wholesale acquisition cost— of a 30-day supply of the drug — in tv ads in a bid to enhance price transparency in the United States. The pharmaceutical industry has vehemently opposed the rule, asserting that list prices are not what a typical patient in the United States pays for treatment — that number is typically determined by the type of (or lack thereof) insurance coverage, deductibles and out-of-pocket costs. Although there is truth to that claim, the move was considered symbolic in the Trump administration’s healthcare agenda to hold drugmakers accountable in a climate where skyrocketing drug prices have incensed Americans on both sides of the aisle.

Bob Smith, Pfizer

Pfiz­er is mak­ing a $500M state­ment to­day: Here’s how you be­come a lead play­er in the boom­ing gene ther­a­py sec­tor

Three years ago, Pfizer anted up $150 million in cash to buy Bamboo Therapeutics in Chapel Hill, NC as it cautiously stuck a toe in the small gene therapy pool of research and development.

Company execs followed up a year later with a $100 million expansion of the manufacturing operations they picked up in that deal for the UNC spinout, which came with $495 million in milestones.

And now they’re really going for it.

Endpoints News

Basic subscription required

Unlock this story instantly and join 57,900+ biopharma pros reading Endpoints daily — and it's free.

Video: Putting the AI in R&D — with Badhri Srini­vasan, Tony Wood, Rosana Kapeller, Hugo Ceule­mans, Saurabh Sa­ha and Shoibal Dat­ta

During BIO this year, I had a chance to moderate a panel among some of the top tech experts in biopharma on their real-world use of artificial intelligence in R&D. There’s been a lot said about the potential of AI, but I wanted to explore more about what some of the larger players are actually doing with this technology today, and how they see it advancing in the future. It was a fascinating exchange, which you can see here. The transcript has been edited for brevity and clarity. — John Carroll