New ap­proach­es to de-risk­ing ear­ly drug de­vel­op­ment

End­points News talked with bio­phar­ma lead­ers at Mer­ck, Sanofi, Catal­ent, Take­da, and J&J about their ef­forts to im­prove R&D ef­fi­cien­cy

The bio­phar­ma in­dus­try as a whole is spend­ing more mon­ey than ever on drug R&D, but the pay­back on pro­duc­tiv­i­ty keeps shrink­ing. On a long-term ba­sis, that’s un­sus­tain­able for the in­dus­try and has to change. We spoke with sev­er­al of the lead­ing ex­perts in clin­i­cal tri­al de­sign and drug de­vel­op­ment to see what they’re do­ing to pick the right can­di­dates for clin­i­cal de­vel­op­ment and how they’re re­vamp­ing projects to achieve ap­provals as ef­fi­cient­ly as pos­si­ble.

Af­ter all, most of a drug’s de­vel­op­ment costs come late in the process. Run­ning the kind of clin­i­cal tri­als need­ed to sup­port an ap­proval are ex­pen­sive, and if they fail, all cap­i­tal in­vest­ed is lost.

Is it time for an­i­mal stud­ies to be dis­card­ed? What bet­ter tech­nolo­gies are avail­able to im­prove ear­ly-stage de­vel­op­ment? Can op­ti­mal for­mu­la­tion im­prove a mol­e­cule?

We looked at all those ques­tions and more.

Much of the prob­lem is based on an in­com­plete un­der­stand­ing of bi­ol­o­gy and an­i­mal stud­ies of­ten seem to con­fuse the is­sues more than they pro­vide fresh in­sights to re­searchers.

We know that stick­ing with an­i­mal stud­ies for ba­sic pre­clin­i­cal guid­ance will lim­it over­all suc­cess rates in the clin­ic to on­ly 5% to 10%. But the em­pha­sis right now is more on evo­lu­tion than rev­o­lu­tion. Or­gan-on-a-chip tech­nol­o­gy can en­hance pre­clin­i­cal work, just as col­lab­o­ra­tions among CROs and de­vel­op­ers have been re­lied on to iden­ti­fy pre­ferred dosages and de­liv­ery meth­ods.

But you al­so have to keep things sim­ple, to make sure that you don’t over­com­pli­cate your mol­e­cule, mak­ing it im­pos­si­ble to pro­duce at a rea­son­able cost.

And through­out all the changes chron­i­cled in this ar­ti­cle, you will hear of case af­ter case where large com­pa­nies have been strip­ping down in­te­ri­or walls to al­low for more part­ner­ing with in­ves­ti­ga­tors of all stripes, whether in the in­dus­try, gov­ern­ment or acad­e­mia.

In this field, bor­ders are in­creas­ing­ly mean­ing­less ob­sta­cles to change. Sci­en­tists and de­vel­op­ers in North Amer­i­ca, Eu­rope and Asia are look­ing glob­al­ly for the best so­lu­tions to this im­mense prob­lem. Fail­ure is not an op­tion, as NASA might phrase it. These days, when you’re do­ing a moon shot of your own, you need to use every im­por­tant re­source avail­able, wher­ev­er it is, to avoid a fail­ure to launch.

An in­com­plete pic­ture

Robert Plenge, Mer­ck


The vast ma­jor­i­ty of Phase I can­di­dates nev­er make it to mar­ket, putting the true cost of suc­cess­ful drug mak­ing straight in­to the stratos­phere. One re­cent es­ti­mate of the clin­i­cal-ap­proval suc­cess rate was an abysmal 11.8%.

Robert Plenge, Mer­ck vice pres­i­dent and head of its Trans­la­tion­al Med­i­cine de­part­ment, told End­points that one root cause of why bio­phar­ma R&D is so ex­pen­sive and prone to fail­ure is be­cause of our in­com­plete un­der­stand­ing of hu­man bi­ol­o­gy. He said com­pa­nies need to dig fur­ther to find ev­i­dence to sup­port a tar­get be­fore be­gin­ning a drug de­vel­op­ment pro­gram.

De­vel­op­ing a drug based on an­i­mal mod­els just isn’t ef­fi­cient enough any­more, he stressed. That method in­volves mak­ing de­ci­sions about a mol­e­cule where very lit­tle was known in hu­mans, but an in­ter­est­ing ef­fect was seen in an­i­mals. Whether that mod­el trans­lates or not is of­ten un­known, and it’s not un­til the drug gets in­to a Phase I or Phase II tri­al that any­thing is learned about how that drug be­haves in hu­mans.

“We’ve done that ex­per­i­ment for 10 or 20 years, and that mod­el tells us that if we pre­scribe to that mod­el, 5% to 10% of things (de­pend­ing on the dis­ease in­di­ca­tion) will ul­ti­mate­ly lead to a suc­cess­ful drug. That rate of pro­duc­tiv­i­ty is just not suf­fi­cient to sus­tain the in­dus­try.”

Plenge breaks down his for­mu­la for in­creas­ing the ef­fec­tive­ness of R&D in­to four key strate­gies that, when im­ple­ment­ed to­geth­er, could be pow­er­ful change agents for im­prov­ing R&D pro­duc­tiv­i­ty.

Those four com­po­nents are: causal hu­man bi­ol­o­gy, ther­a­peu­tic modal­i­ty, bio­mark­ers of tar­get mod­u­la­tion and clin­i­cal proof-of-con­cept stud­ies.

Im­prov­ing Tar­get Val­i­da­tion

“De­vel­op­ers should know at the be­gin­ning of the de­vel­op­ment pro­gram whether the tar­get, when trig­gered, will achieve the de­sired ef­fect.”

Im­prov­ing tar­get val­i­da­tion process­es ear­li­er in the R&D process could help drug de­vel­op­ers make crit­i­cal de­ci­sions soon­er — in­clud­ing what com­pounds to kill.

Plenge says that safe­ty and ef­fi­ca­cy pro­files are mean­ing­less to guide tar­get iden­ti­fi­ca­tion be­cause drug de­vel­op­ers should know at the be­gin­ning of the drug de­vel­op­ment pro­gram whether the tar­get, when trig­gered, will achieve the de­sired ef­fect. When asked to clar­i­fy whether that means some dis­eases just shouldn’t be stud­ied, he said there is al­ways a bal­ance.

“Even for those things that ‘suc­ceed’ in de­vel­op­ment, too few of those dif­fer­en­ti­ate from stan­dard of care and just don’t de­liv­er val­ue.” For ex­am­ple, even for drugs that work in Phase II or Phase III, they of­ten don’t dif­fer­en­ti­ate much from the stan­dard of care, so the drug dies a slow death.

This “non-dif­fer­en­ti­a­tion” is go­ing to be­come a “big­ger and big­ger deal go­ing for­ward as we think of the things that re­al­ly add val­ue in the re­al world,” Plenge said.

Un­der­stand­ing how drugs are work­ing in pa­tients or how they are not work­ing, why some peo­ple re­spond and why oth­ers don’t, and then think­ing about phar­ma­co­log­ic pur­pos­ing — all of these types of ex­per­i­ments in na­ture can be­gin to pro­vide in­sight.

“Then you need some mech­a­nism to dive in­to the bi­ol­o­gy more deeply in a sys­tem that is hu­man-like and im­per­turbable that of­ten is very dif­fi­cult to do,” he added.

An­i­mal mod­els are use­ful to tease out the un­der­ly­ing bi­ol­o­gy of a known path­way and to un­der­stand gen­er­al phar­ma­col­o­gy and how a drug be­haves in an in­tact or­gan­ism. But they’re not use­ful for pick­ing tar­gets or un­der­stand­ing causal­i­ty in hu­mans, Plenge stressed.

When asked whether adap­tive tri­als could be a so­lu­tion where­by small­er more tar­get­ed stud­ies fo­cused on ear­ly ter­mi­na­tions, Plenge said that would on­ly par­tial­ly solve the prob­lem.

“If you’re still pick­ing the wrong tar­gets, if you haven’t done the ap­pro­pri­ate ther­a­peu­tic re­con­fig­u­ra­tion, you don’t re­al­ly have a ro­bust mod­el you can test in hu­mans. Ei­ther you don’t have the right tar­get dose, you don’t have the right un­der­stand­ing of tar­get en­gage­ment or tar­get mod­u­la­tion — this gets to the bio­mark­er piece — then you end up mak­ing de­ci­sions with in­com­plete in­for­ma­tion.”

“That’s why you have to have that causal hu­man ev­i­dence,” Plenge says. “You have to be con­fi­dent that your mol­e­cule is re­ca­pit­u­lat­ing that bi­ol­o­gy. You have to have some way to mea­sure a read out in hu­mans that al­lows you to make that de­ci­sion. The faster we can do that and the faster we can ac­tu­al­ly in­cor­po­rate and adapt a de­sign, the bet­ter off we will be in the in­dus­try.”

The or­gan-on-a-chip mod­el might help cut through some of that an­i­mal da­ta that doesn’t trans­late to hu­man bi­ol­o­gy. The tech­nol­o­gy en­com­pass­es minia­ture mod­els of hu­man or­gans whose chan­nels are lined with hu­man-de­rived cells. In this way, more can be learned about hu­man sys­tems to make bet­ter de­ci­sions ear­li­er rather than re­ly­ing on in­ac­cu­rate an­i­mal mod­els.

Could or­gan-on-a-chip mod­els re­place an­i­mal mod­els? Mer­ck’s Plenge says that an­i­mal mod­els are still a valu­able tool, but they shouldn’t be used to pick tar­gets.

But or­gan-on-a-chip mod­els can pro­vide in­sight in­to func­tions with­in the hu­man body that could bet­ter in­form re­searchers about tar­get mod­u­la­tion. In some cas­es the or­gan-on-a-chip mod­el could al­so be used for as­says that can be more com­plex, he sug­gest­ed.

Sanofi’s or­gan-on-a-chip plat­form

“At a very ear­ly stage, we are able to screen mol­e­cules vir­tu­al­ly on a com­put­er.”

Philippe De­tilleux, Sanofi

Sanofi has been test­ing a range of tech­nolo­gies that repli­cate the cell dy­nam­ics of ma­jor or­gans in the lab for the last three years, Philippe De­tilleux, glob­al head for Pre­clin­i­cal Safe­ty at Sanofi, told End­points. “These tech­nolo­gies show great promise, not least be­cause they al­low re­searchers to test chem­i­cal com­pounds and their ex­pect­ed tox­i­c­i­ty,” he said, but they’re still at a very ear­ly stage of de­vel­op­ment and are not used out­side of pi­lot pro­grams.

Over the last few years, Sanofi has en­tered in­to nu­mer­ous col­lab­o­ra­tions to ex­plore the or­gan-on-a-chip tech­nol­o­gy. The com­pa­ny has col­lab­o­rat­ed with the Wyss In­sti­tute, Hurel, Mimetas, and CN­Bio as well as the In­no­v­a­tive Med­i­cines Ini­tia­tive Pro­ject MIP-DILI (Mech­a­nism-Based In­te­grat­ed Sys­tems for the Pre­dic­tion of Drug-In­duced Liv­er In­jury).

“One area where we have been quite ac­tive is in de­vel­op­ing in sil­i­co ap­proach­es to an­tic­i­pate po­ten­tial tox­i­c­i­ty li­a­bil­i­ties of small mol­e­cules. At a very ear­ly stage, we are able to screen mol­e­cules vir­tu­al­ly on a com­put­er.”

The com­pa­ny sees po­ten­tial in terms of its pre­clin­i­cal work, where or­gans on a chip can help ac­cel­er­ate tests with chem­i­cal com­pounds for elim­i­na­tion and for tox­i­col­o­gy work, where the re­sis­tance of he­pa­to­cytes in the liv­er mod­el tend to last longer and thus show stronger tox­i­c­i­ty re­sults, De­tilleux said.

While the tech­nol­o­gy will com­ple­ment pre­clin­i­cal pro­grams in the short term and even­tu­al­ly re­duce some an­i­mal work, it’s un­like­ly to be able to re­place it all to­geth­er giv­en the reg­u­la­to­ry frame­work, he added.

The most im­me­di­ate im­pact though may be in pre­dict­ing hu­man me­tab­o­lism and phar­ma­co­ki­net­ics, rather than tox­i­c­i­ty.  “Progress on the DMPK as­pects of liv­er and GI mod­els is be­ing made and these may be im­ple­ment­ed soon. If some of these mod­els are suc­cess­ful in pre­dict­ing hu­man me­tab­o­lism and PK, re­duc­tion or re­fine­ment of hu­man stud­ies could re­duce costs sig­nif­i­cant­ly.”

What about new ge­nom­ic-based tar­gets? “These new modal­i­ties cre­ate new ways to re­ca­pit­u­late the hu­man bi­ol­o­gy,” Mer­ck’s Plenge said. “It might be with small mol­e­cules and bi­o­log­ics we can do these types of ther­a­peu­tic re­ca­pit­u­la­tions. We can in­hib­it en­zymes, we can re­move cir­cu­lat­ing pro­teins from the sys­tem and that’s if small mol­e­cules and bi­o­log­ics do re­spec­tive­ly pret­ty well. There are more com­plex ways in which the hu­man sys­tem caus­es dis­ease and by hav­ing things like CRISPR and mR­NA de­liv­ery in­crease the way in which we can per­turb the tar­gets of in­ter­est. Sud­den­ly things that were in­tractable be­fore maybe be­come tractable to­day.”

For ex­am­ple, there are dis­ease in­di­ca­tions where you can over­lap hu­man ge­net­ics with ap­proved ther­a­pies to un­der­stand how ge­net­ic tar­gets have led to ap­proved ther­a­pies ei­ther di­rect­ly or in­di­rect­ly.

“You can see that in lipid-low­er­ing ther­a­pies, you can see that in os­teo­poro­sis, you can see that in rheuma­toid arthri­tis and in rare dis­eases like cys­tic fi­bro­sis where ge­net­ics have led to ap­proved ther­a­pies or ge­net­ics have ret­ro­spec­tive­ly iden­ti­fied the tar­gets.”

Can op­ti­mal for­mu­la­tion im­prove a mol­e­cule?

Cor­nell Sta­moran

“We kept see­ing prod­ucts in Phase II that we knew we could fix to op­ti­mize the dose or for­mu­la­tion,” said Cor­nell Sta­moran, vice pres­i­dent of Strat­e­gy & Gov­ern­ment Af­fairs for CMO Catal­ent.

“We tell clients not to kill a mol­e­cule un­til they’re sure they have done every­thing they could to im­prove it.”

Once a com­pa­ny iden­ti­fies a lead mol­e­cule, it makes de­ci­sions about which vari­ant of the mol­e­cule to use, and those choic­es can have a di­rect ef­fect on pa­tient ad­her­ence or side ef­fects.

“If you make de­ci­sions right in the de­sign of the mol­e­cule – the route of ad­min­is­tra­tion, de­sign of the dose form, the dos­ing reg­i­men as well as pack­ag­ing– those choic­es can pro­duce a clin­i­cal­ly dif­fer­ent prod­uct.”

Mak­ing in­ten­tion­al de­sign choic­es that are in­formed by pa­tient fo­cus at very ear­ly stages — be­tween late dis­cov­ery and ear­ly clin­i­cal de­vel­op­ment — is one of the places that Catal­ent be­lieves drug de­vel­op­ment can be ac­cel­er­at­ed and the clin­i­cal tri­als out­comes can be im­proved.

He ex­plained that the choic­es com­pa­nies make about for­mu­la­tion and even cap­sule size and shape mat­ter. For ex­am­ple, if you’re try­ing to reach cer­tain first gen­er­a­tion im­mi­grants as part of your tar­get pa­tient pop­u­la­tion, in some cul­tures there are strong as­so­ci­a­tions be­tween col­or and health or sick­ness, so choos­ing the col­or of the pill mat­ters when it comes to drug ad­her­ence in your tar­get pop­u­la­tion.

He said rough­ly 60% to 90% of drugs in de­vel­op­ment face PK/PD and oth­er de­liv­ery chal­lenges, and Catal­ent has built a pre­dic­tive plat­form can iden­ti­fy op­ti­mal for­mu­la­tion ap­proach­es to ad­dress sol­u­bil­i­ty and oth­er is­sues.

The com­pa­ny ac­quired a mol­e­cule op­ti­miza­tion plat­form from GSK, and com­bined it with Catal­ent’s for­mu­la­tion know-how and in­tel­lec­tu­al prop­er­ty to pre­dict which for­mu­la­tion ap­proach is like­ly to give a mol­e­cule the best clin­i­cal re­sults based on avail­able ev­i­dence.

One of the bar­ri­ers that com­pa­nies have had his­tor­i­cal­ly is that be­cause there weren’t plat­forms that pro­vid­ed pre­dictabil­i­ty, they fo­cused on for­mu­la­tion tech­niques they knew in-house, to see if it got them to a “good enough” for­mu­la­tion point. Re­do­ing to im­prove the for­mu­la­tion fur­ther takes time and mon­ey at a stage of the process where time is con­strained.

The prob­lem is, set­tling for some­thing that is “good enough,” can have side ef­fect im­pli­ca­tions or mon­e­tary im­pli­ca­tions down the road.

“We fo­cused on try­ing to build pre­dic­tive tools in­to al­go­rithms and work­flows, so we could do this on an out­sourced ba­sis in a short amount of time – 12 weeks – to make a rec­om­men­da­tion on what plat­form ad­dress­es sol­u­bil­i­ty and avail­abil­i­ty best and to bring a GMP-based prod­uct for an­i­mal stud­ies.

Founder and co-chair of Catal­ent’s Ap­plied Drug De­liv­ery In­sti­tute, Sta­moran says Catal­ent evolved as a drug de­liv­ery com­pa­ny, and it has added many ad­di­tion­al of­fer­ings, from clin­i­cal tri­al sup­ply man­age­ment to bio­man­u­fac­tur­ing and an­ti­body drug con­ju­gate tech­nol­o­gy.

Through it’s Ap­plied Drug De­liv­ery In­sti­tute, Catal­ent is en­gaged in pre­c­om­pet­i­tive col­lab­o­ra­tions with com­pa­nies like Al­ler­gan and Take­da to work on non-in­va­sive de­liv­ery of large mol­e­cules. To­day Catal­ent al­ready pro­duc­ing dose forms for bi­o­log­ics that are de­liv­ered oral­ly, neb­u­lized or de­liv­ered via the eye.

An op­ti­mal Phase I prod­uct, he said, needs to be a for­mu­la­tion that is scal­able – ide­al­ly some­thing that can go to a ra­tio­nal process that is not over­ly com­plex.

“You don’t want to have to do 17 dif­fer­ent things to a mol­e­cule,” he said.

Pre­c­om­pet­i­tive part­ner­ships

“We rec­og­nize that some of the best sci­ence comes from out­side a com­pa­ny’s walls.”

Take­da has been ac­tive in the Struc­tur­al Ge­nomics Con­sor­tium – an open source con­sor­tium that aims to bet­ter un­der­stand ma­jor dis­eases. The group has char­ac­ter­ized 15% of all hu­man pro­tein struc­tures, and the da­ta is open to the pub­lic. Ac­cord­ing to Tufts, the group is gen­er­at­ing char­ac­ter­i­za­tions of three nov­el mol­e­cules each quar­ter, with the goal of val­i­dat­ing tar­get mol­e­cules soon­er and to en­gage sci­en­tists to de­vel­op new ther­a­pies soon­er.

In­dus­try part­ners in the con­sor­tium in­clude Ab­b­vie, Bay­er, Boehringer In­gel­heim, Janssen, Mer­ck, No­var­tis and Pfiz­er.

Daniel Cur­ran

“We rec­og­nize that some of the best sci­ence comes from out­side a com­pa­ny’s walls, and Take­da’s part­ner­ship with the Struc­tur­al Ge­nomics Con­sor­tium has been a valu­able con­duit for Take­da to ac­com­plish this,” Daniel Cur­ran, head of Take­da’s Cen­ter for Ex­ter­nal In­no­va­tion, told End­points.

Be­yond just be­ing a mem­ber of the SGC, Take­da has in­vest­ed time and en­er­gy to un­der­stand what as­sets and tech­nolo­gies al­ready ex­ist with­in the SGC. By con­nect­ing these re­sources to Take­da’s own in­ter­nal re­search, Take­da and SCG have iden­ti­fied mul­ti­ple syn­er­gis­tic op­por­tu­ni­ties to ex­change ma­te­ri­als and col­lab­o­rate.

On the re­ceiv­ing end, Take­da has worked with SGC to gain ac­cess to pa­tient sam­ples for use in iden­ti­fy­ing and val­i­dat­ing nov­el drug dis­cov­ery tar­gets.  Take­da has al­so iden­ti­fied pro­teins and as­says that were al­ready de­vel­oped by SGC mem­ber in­sti­tu­tions that Take­da can read­i­ly use in its ex­ist­ing drug dis­cov­ery pro­grams.

Take­da al­so has pro­vid­ed SGC with com­pounds from its com­pound li­brary to use as tools for dis­cov­er­ing new tar­gets.

Take­da re­cent­ly re­vamped its R&D pro­gram to nar­row in on three key ar­eas: on­col­o­gy, gas­troen­terol­o­gy and the cen­tral ner­vous sys­tem as well as vac­cines.

The fo­cus on those key ther­a­py ar­eas was a strate­gic move to de-risk its R&D pipeline by fo­cus­ing on ar­eas of un­met med­ical need where Take­da had rel­e­vant ex­per­tise and know-how. If the need­ed ca­pa­bil­i­ties to de­vel­op a po­ten­tial med­i­cine are found lack­ing, the com­pa­ny will seek the right part­ner, Cur­ran said.

“To be tru­ly in­no­v­a­tive, Take­da be­lieves you need to be part of a com­mu­ni­ty of ex­perts who lever­age their unique strengths and share a com­mon goal,” he said. To that end, Take­da has been host­ing ‘in­no­va­tor net­work’ events that bring to­geth­er mem­bers of that com­mu­ni­ty to con­nect and share ideas, best prac­tices and learn more about game-chang­ing break­throughs.

“Our ap­proach is more holis­tic,” he said, “mean­ing we start with un­der­stand­ing of bi­ol­o­gy and dis­ease states, and then we con­sid­er the tar­get, di­verse modal­i­ties, tar­get­ed de­liv­ery, trans­la­tion­al sci­ences and ge­nomics. With all of these con­sid­er­a­tions, we be­lieve this should equate to a more risk-bal­anced ap­proach.”

In the last year, Take­da en­tered in­to a num­ber of strate­gic part­ner­ships that build up­on ex­per­tise in di­verse modal­i­ties. A few ex­am­ples in­clude:

• Pros­et­ta Bio­sciences – dis­cov­er­ing and de­vel­op­ing ther­a­pies to treat neu­rode­gen­er­a­tive dis­eases;

• En­terome Bio­science – tar­get­ing GI dis­or­ders us­ing mi­cro­bio­me ther­a­peu­tics;

• Ul­tragenyx Phar­ma­ceu­ti­cal – de­vel­op­ing ther­a­pies to treat rare ge­net­ic dis­eases;

• Fra­zier Health­care – cre­at­ed Out­post Med­i­cine, a new biotech com­pa­ny with a fo­cus on women’s health;

• Mer­sana Ther­a­peu­tics – de­vel­op­ing next-gen­er­a­tion an­ti­body drug con­ju­gates;

• Ther­vance Bio­phar­ma – look­ing at a nov­el agent for gas­troin­testi­nal motil­i­ty dis­or­ders;

• Im­muno­Gen – de­vel­op­ing an­ti-can­cer ther­a­pies us­ing nov­el ADC tech­nol­o­gy; and

• Cour Phar­ma­ceu­ti­cals – de­vel­op­ing GI ther­a­pies in­clud­ing Celi­ac dis­ease.

J&J’s in­vest­ments in ear­ly-stage in­no­va­tion

Robert Ur­ban

J&J is hav­ing some suc­cess get­ting NMEs to mar­ket af­ter it recre­at­ed it­self about four years ago, launch­ing the J&J In­no­va­tion Cen­ters with a mis­sion to in­vest in “high­ly dif­fer­en­ti­at­ed” ear­ly-stage in­no­va­tions. Lo­cat­ed in Boston, Cal­i­for­nia, Lon­don and Shang­hai, the in­no­va­tion cen­ters are re­gion­al hubs cre­at­ed to ac­cess the best sci­ence and tech­nol­o­gy in their re­spec­tive re­gions.

Robert Ur­ban, head of J&J In­no­va­tion, Boston joined J&J about four years ago when John­son & John­son In­no­va­tion was launched.

At the time, there was some re­dun­dan­cy on the phar­ma­ceu­ti­cal side of J&J, so the com­pa­ny sought to trans­form it­self by re­fo­cus­ing on five spe­cif­ic ar­eas: neu­ro­science, on­col­o­gy, im­munol­o­gy, car­dio­vas­cu­lar and me­tab­o­lism, and in­fec­tious dis­eases and vac­cines.

Since that time, the com­pa­ny has had 14 new drugs ap­proved in the last five years, and those were all in un­met need ar­eas, Ur­ban told End­points.

“Each of them are breath­tak­ing in­no­va­tions,” he said, and many of them moved swift­ly through the ap­proval process.

The com­pa­ny ex­pects to have 10 more NMEs filed by 2019, all of which rep­re­sent a sub­stan­tial new type of prod­uct. It an­tic­i­pates an­oth­er 25 or so NMEs will be nom­i­nat­ed be­tween now and 2019.

“The fo­cus has been on build­ing an in­fra­struc­ture of very deep sci­ence on the in­side but an ex­tra­or­di­nary net­work of re­la­tion­ships put in place through out­reach ef­forts all over the world.”

Some of the things that have come from this are first-in-class NMEs, such as J&J’s TB drug Sir­turo, the very first TB drug that’s been launched in 40 years for mul­ti-drug re­sis­tant TB.

There’s al­so been some in­no­v­a­tive kinds of part­ner­ships that show how bold­ly the or­ga­ni­za­tion has em­braced some of these chal­lenges.

The YO­DA Ini­tia­tive — the Yale On­line Da­ta Ac­cess plat­form — is an ex­am­ple where J&J was one of the first com­pa­nies to push out open ac­cess to clin­i­cal da­ta.

“You need to get the da­ta in the hands of many, and that al­lows us to do that. We’ve seen in this last decade a re­al­ly in­ter­est­ing trans­for­ma­tion in both the way we see our­selves, but al­so dou­bling down on how we see our­selves fit­ting to­geth­er in a much more net­worked con­text, with in­ves­ti­ga­tors all over the world.”

The glob­al or­ga­ni­za­tion works 24/7, so as the sun moves across the plan­et, da­ta shifts in­to the hands of oth­er team mem­bers.

J&J spends near­ly $9 bil­lion on R&D per year, and in the last 20 years, the com­pa­ny has in­vest­ed $200 bil­lion in R&D. About $85 bil­lion of that has been in­vest­ed in ideas that came out­side of the or­ga­ni­za­tion.

“Be­ing an ex­pert in the ar­eas that you in­vest in is crit­i­cal,” he said. “As we’ve start­ed to fo­cus on few­er things, we’ve got­ten much, much deep­er in our com­pe­ten­cy, which means we’ve been able to un­der­stand and ap­pre­ci­ate the bi­ol­o­gy and un­der­stand and ap­pre­ci­ate the po­ten­tial of in­ter­est­ing prod­ucts that are emerg­ing in those ar­eas.”

That al­so gives you a much keen­er in­tu­ition about how like­ly some­thing might work when look­ing at in­no­va­tion around the world.

Al­so part of J&J In­no­va­tion is the com­pa­ny’s cor­po­rate ven­ture arm, John­son & John­son De­vel­op­ment Corp., which pro­vides ear­ly-stage re­search fund­ing, seed fund­ing, eq­ui­ty in­vest­ments as well as li­cens­ing op­por­tu­ni­ties and oth­er col­lab­o­ra­tions.

So far, the com­pa­ny has made about 200 in­vest­ments across the world, and the in­no­va­tion cen­ters them­selves make up an ecosys­tem of in­cred­i­bly tal­ent­ed peo­ple across these five busi­ness ar­eas.

“We par­tic­i­pate in help­ing to se­lect the com­pa­nies that are al­lowed to go in and uti­lize these in­cu­ba­tors. It’s quite a rig­or­ous process to get in­to one of those fa­cil­i­ties, and there are no strings at­tached. You don’t have to have a part­ner­ship with us in or­der to get in. You have to sim­ply be work­ing on some­thing quite re­mark­able that’s more like­ly than most to gen­er­ate a growth-ori­ent­ed life sci­ence com­pa­ny.”

Rough­ly 140 com­pa­nies now live some­where with­in one of the JLABs, and the group ex­pects to have as many as 200 com­pa­nies by the end of the year, Ur­ban said.

The cen­ters them­selves are de­signed to be in­spir­ing spaces to fos­ter cre­ativ­i­ty where new start-ups get ac­cess to state-of-the-art fa­cil­i­ties and equip­ment very quick­ly so they can fo­cus on gen­er­at­ing the da­ta they need.

They all pay what is ap­pro­pri­ate rent for the size of space they need with­out hav­ing to sign a mul­ti-year lease, buy new equip­ment or get all the per­mits.

“It’s a trans­for­ma­tive way to start a biotech­nol­o­gy com­pa­ny.”

“The com­pa­nies leave dif­fer­ent than when they came – not just be­cause the sci­ence is al­ways this twisty-turny thing we have to fol­low – but the im­pact that they have on each oth­er by be­ing right there shar­ing ven­tures and shar­ing cof­fee is ex­tra­or­di­nary. We are thrilled by the work.”

Many of J&J’s in­vest­ments have fo­cused on tech­nol­o­gy that en­ables R&D, such as bio­mark­er analy­sis plat­forms. The com­pa­ny is ac­tive­ly ex­plor­ing the mi­cro­bio­me and try­ing to un­der­stand how the mi­cro­bio­me it­self can be ma­nip­u­lat­ed.

The oth­er in­vest­ment that’s quite sub­stan­tial is the hu­man in­vest­ment. Al­though there’s fi­nan­cial par­tic­i­pa­tion, some of the things that turn out to be the most trans­for­ma­tive are ac­cess to in­di­vid­u­als in the or­ga­ni­za­tion.

“At the same time, we do the best we can to an­tic­i­pate what’s go­ing to be around the cor­ner to pre­pare our­selves,” Ur­ban said.