How to Ac­cel­er­ate De­vel­op­ment in Nov­el & Ad­vanced On­col­o­gy Ther­a­pies - From the Start­ing Line

Ear­ly-phase on­col­o­gy clin­i­cal tri­al de­sign is evolv­ing rapid­ly of­fer­ing biotechs new op­por­tu­ni­ties to ac­cel­er­ate their clin­i­cal de­vel­op­ment. To sup­port biotechs nav­i­gate this chang­ing land­scape, Novotech and End­points News re­cent­ly as­sem­bled an ex­pert pan­el to share their ear­ly phase on­col­o­gy tri­al de­sign in­sights.

“If you don’t get it right in the be­gin­ning if you’re a small biotech, it’s hard to go back and start over,” said Dr Pa­tri­cia Muc­ci LoRus­so, the Di­rec­tor of the Ear­ly Phase Clin­i­cal Tri­als Pro­gram and As­so­ciate Cen­ter Di­rec­tor, Ex­per­i­men­tal Ther­a­peu­tics of the Yale Can­cer Cen­ter in New Haven, Con­necti­cut.

She said tri­als have be­come much more com­plex, “and the de­mands and the da­ta points have be­come sig­nif­i­cant­ly greater”.

“The num­ber of ex­plorato­ry end­points has re­al­ly ex­plod­ed in many of these tri­als rel­a­tive to what we used to see sev­er­al years ago.”

New on­col­o­gy study de­signs have al­so been re­cent­ly rec­om­mend­ed by the FDA. The On­col­o­gy Cen­ter of Ex­cel­lence (OCE) Project Op­ti­mus is an ini­tia­tive to re­form the dose op­ti­miza­tion and dose se­lec­tion par­a­digm in on­col­o­gy drug de­vel­op­ment.

“Too of­ten, the cur­rent par­a­digm for dose se­lec­tion—based on cy­to­tox­ic chemother­a­peu­tics—leads to dos­es and sched­ules of mol­e­c­u­lar­ly tar­get­ed ther­a­pies that are in­ad­e­quate­ly char­ac­ter­ized be­fore ini­ti­at­ing reg­is­tra­tion tri­als,” ac­cord­ing to the FDA.

Dr. LoRus­so said: “Project Op­ti­mus is try­ing to iden­ti­fy what the right dose is in­stead of giv­ing too much of a dose be­cause it’s the max­i­mum tol­er­at­ed dose. Per­haps less of a dose that can al­low the pa­tient to main­tain on that agent more chron­i­cal­ly with­out dose in­ter­rup­tions or dose re­duc­tions is one of the re­cent dri­ving forces with­in the FDA guide­lines. Try­ing to max­i­mize ef­fi­ca­cy but min­i­mize tox­i­c­i­ty.”

“There’s been a big push for that, not on­ly by the FDA, but ad­vo­ca­cy groups have pushed for it in hopes that we’re no longer push­ing a drug so high that you have to have so many in­ter­rup­tions that it may ac­tu­al­ly, in the long run, be less ef­fi­ca­cious.”

“Al­so, there’s a lot more in­put in­to more nov­el bio­sta­tis­ti­cal de­signs.”

“Bayesian op­ti­mal in­ter­val (BOIN) de­signs are be­com­ing more com­mon­place in ear­ly phase tri­als.”

“I think the way for­ward is go­ing to be away from 3+3 and more to­wards these de­signs so that we can max­i­mize the da­ta that we’re ob­tain­ing from each in­di­vid­ual pa­tient so we can get it right as ear­ly as pos­si­ble with­in the con­text of a Phase 1 tri­al.”

“It’s a well-known fact, es­pe­cial­ly with tar­get­ed drugs, that we would get a rec­om­mend­ed Phase 2 dose, and 45 to 50% of the time con­ser­v­a­tive­ly it’s been shown that that dose is not the right dose in Phase 3 and that’s a very ex­pen­sive er­ror.”

“9% of Phase 3 tri­als had his­tor­i­cal­ly been abort­ed be­cause the wrong dose was cho­sen to ad­vance for­ward.”

Dr. Vishal Na­vani, the Staff Med­ical On­col­o­gist and Pro­fes­sor of the Tom Bak­er Can­cer Cen­tre, Uni­ver­si­ty of Cal­gary in Cal­gary said there are so many end­points need­ed and da­ta re­quired from these ear­ly phase tri­als that “us­ing a Bayesian ap­proach or a BOIN ap­proach gets the most out of each pa­tient’s kind vol­un­tary time that they spend with us on a tri­al”.

“I like mod­el-as­sist­ed de­signs like BOIN be­cause they’re flex­i­ble. Let’s say, for ex­am­ple, you have a bis­pe­cif­ic tri­al and you’re wor­ried about a spe­cif­ic tox­i­c­i­ty like cy­tokine re­lease. Well, you can set a dose-lim­it­ing tox­i­c­i­ty rate of 10%, let’s say, for ar­gu­ment’s sake with BOIN. Where­as us­ing a rule-based de­sign like 3+3, you can’t re­al­ly do that,” said Dr Na­vani.

“If you use a time-to-event point, it can cap­ture tox­i­c­i­ties through­out the en­tire time the pa­tient’s on ther­a­py, rather than just wait­ing for one ex­pe­ri­ence of dose do­main tox­i­c­i­ty.“

Michele Ger­ber, the Chief Med­ical Of­fi­cer at Myeloid Ther­a­peu­tics in Cam­bridge, USA said Myeloid Ther­a­peu­tics has in­cor­po­rat­ed Bayesian de­signs in their drug de­vel­op­ment pro­grams.

“The 3+3 de­sign was re­al­ly de­signed for chemother­a­py and for de­ter­min­ing the max­i­mum tol­er­at­ed dose. With gene and im­mune tar­get­ed drugs, we re­al­ly need to find out the op­ti­mal dos­es which may not be at the max­i­mum tol­er­at­ed dose,” she said.

“Bayesian de­signs were the best way for­ward for our project. We be­lieve that they im­prove the pre­ci­sion of the safe­ty pro­file and al­low ear­li­er in­for­ma­tion on the ther­a­peu­tic in­dex.”

“There’s re­al­ly two main Bayesian de­signs, one on mod­el-based and the oth­er mod­el-as­sist­ed de­sign. Which one you choose is de­pen­dent on op­er­a­tional fea­si­bil­i­ty, in­clud­ing the ex­per­tise and in­fra­struc­ture with the mod­el-based de­signs be­ing much more dif­fi­cult to im­ple­ment. The mod­el-as­sist­ed de­signs in­clude the BOIN de­sign. That’s what we’ve cho­sen to use in our in vi­vo pro­gram.”

In ad­di­tion to tri­al de­sign op­tions, ge­o­graph­ic ad­van­tages can al­so ac­cel­er­ate ear­ly phase on­col­o­gy tri­als.

Prof Jayesh De­sai, Med­ical On­col­o­gist and Clin­i­cal Re­search Head, Ear­ly Drug De­vel­op­ment at the Pe­ter Mac­Cal­lum Can­cer Cen­tre in Mel­bourne, Aus­tralia said Aus­tralia of­fers rapid ac­ti­va­tion time­lines.

He said in Aus­tralia, we have ex­cel­lent align­ment be­tween in­ves­ti­ga­tors’ sites and our reg­u­la­to­ry agen­cies and ethics com­mit­tees. “It’s been an en­vi­ron­ment where we’ve man­aged to get things go­ing very quick­ly. Our ac­ti­va­tion time­lines are very good and I think that’s been at­trac­tive for the in­dus­try.”

“I’d like to hope that in­dus­try comes to Aus­tralia be­cause of the qual­i­ty of the in­ves­ti­ga­tors and the sites, not just be­cause we do things quick­ly.”

“The strat­e­gy of be­gin­ning a first time in hu­man tri­al in Aus­tralia and then, for ex­am­ple, mov­ing that to the US, as you move through dose es­ca­la­tion or even to­wards the lat­ter parts of dose es­ca­la­tion. That’s some­thing that we’ve done many, many times.”

Kedan Lin, the Se­nior Vice Pres­i­dent at Har­bour Bio­Med in South San Fran­cis­co, USA said her com­pa­ny had al­so opened sites in Aus­tralia.

Kedan Lin said Aus­tralia of­fers “a very quick start and the da­ta qual­i­ty is ex­cel­lent”.

“We can ac­tu­al­ly lever­age the da­ta we ob­tain from Aus­tralia to open up the site tri­als in the US or in Chi­na or some­where else.”

She said Aus­tralia’s pop­u­la­tion di­ver­si­ty can al­so help meet FDA tri­al di­ver­si­ty re­quire­ments.

She said CROs like Novotech have a very strong pres­ence in Aus­tralia. “They help the spon­sor get to the key de­ci­sion point very quick­ly.”

“We have had a won­der­ful ex­pe­ri­ence in Aus­tralia.”

Michele Ger­ber said her com­pa­ny was ini­ti­at­ing the tri­al for their in­vi­vo pro­gram in Aus­tralia. “We have de­cid­ed to go there first and fore­most be­cause of the ex­cel­lent med­ical sys­tem that ex­ists there and the very strong Phase 1 units. Al­so, we are us­ing a lot of aca­d­e­mics to sup­port our trans­la­tion­al pro­gram. In ad­di­tion, the reg­u­la­to­ry path­way is very well de­fined. It’s easy to nav­i­gate and very ef­fi­cient.”

“Last­ly, very im­por­tant­ly for a small biotech com­pa­ny, the tri­als are less ex­pen­sive than in the US, and that sup­ports our needs to re­al­ly con­serve funds.”

Dr. LoRus­so said: “I can hon­est­ly tell you that I’ve re­al­ly en­joyed work­ing with Aus­tralian sites and out­stand­ing in­ves­ti­ga­tors.”

Dr. Na­vani ad­dressed the is­sue of pro­to­cols and en­roll­ment bar­ri­ers –“There’s been some re­cent FDA guid­ance on im­prov­ing el­i­gi­bil­i­ty and mi­nor­i­ty rep­re­sen­ta­tion in tri­als and ex­pand­ing ac­cess to tri­als so that they al­ready re­flect the pop­u­la­tions that the end drug may serve.”

“I think that spon­sors need to work hard in help­ing a broad­er pan­el of po­ten­tial pa­tients re­cruit on­to tri­als and deal with the lan­guage, com­mu­ni­ca­tion, health lit­er­a­cy bar­ri­ers that may pre­vent the pa­tient from en­rolling on­to an ear­ly phase tri­al.”

“I al­so think we need to think as a field more broad­ly about ex­pand­ing in­clu­sion cri­te­ria, es­pe­cial­ly in the ear­ly phase set­ting.”

Con­tact Novotech for more in­for­ma­tion about op­ti­mal tri­al de­sign for clin­i­cal de­vel­op­ment ac­cel­er­a­tion here.  

View the ex­pert pan­el video here.