John Jenk­ins, Di­rec­tor, Of­fice of New Drugs

Now that the FDA has capped Sarep­ta’s long and con­tro­ver­sial cam­paign to win an ap­proval for its Duchenne mus­cu­lar dy­s­tro­phy drug with an ac­cel­er­at­ed ap­proval, FDA of­fi­cials would like to send a very clear mes­sage to any oth­er biotech look­ing to slug their way for­ward us­ing the same tac­tics: For­get about it.

In a pre­sen­ta­tion at a sum­mit meet­ing of the Na­tion­al Or­ga­ni­za­tion of Rare Dis­eases on Tues­day, John Jenk­ins, di­rec­tor of the of­fice of new drugs at the Cen­ter for Drug Eval­u­a­tion and Re­search, sin­gled out the eteplirsen case as a new text­book ex­am­ple of what oth­er com­pa­nies should avoid.

“Path tak­en by Sarep­ta NOT a good mod­el for oth­er de­vel­op­ment pro­grams,” de­clared one of Jenk­ins’ talk­ing points un­der a fi­nal sec­tion marked “lessons learned” from eteplirsen and oth­er cas­es.

What fol­lowed was a litany of sins that FDA in­sid­ers have laid at the door of Sarep­ta, which gained the FDA’s stamp of ap­proval to sell Ex­ondys 51 for $300,000 a year, de­spite the heat­ed ob­jec­tions raised by se­nior of­fi­cials over the lack of safe­ty and ef­fi­ca­cy da­ta laced with barb ob­jec­tions over what was pre­sent­ed. In the end, Sarep­ta was able to gain an ap­proval af­ter pa­tient ad­vo­cates and in­flu­en­tial play­ers in Wash­ing­ton DC de­mand­ed it, with lit­tle more than da­ta from a tiny study that was re­peat­ed­ly chal­lenged by re­view­ers but en­dorsed by Janet Wood­cock and per­mit­ted by Com­mis­sion­er Robert Califf.

Jenk­ins had quite a few points to make, and they re­flect­ed the laun­dry list of is­sues that the agency had spot­light­ed to crit­i­cize Sarep­ta re­peat­ed­ly over the course of its re­view and a de­fense of its po­si­tion in light of the biotech’s chal­lenge, plus a few new ones. The FDA, he said, want­ed to make clear that bad­ly planned pro­grams could on­ly de­lay a re­view and OK; as­says, par­tic­u­lar­ly ones that re­ly on in­va­sive pro­ce­dures with chil­dren, need to be well val­i­dat­ed; stud­ies need to be care­ful­ly blind­ed and ran­dom­ized; ev­i­dence of an im­pact on bio­mark­ers has to be care­ful­ly eval­u­at­ed, ex­clud­ing ‘any’ ef­fect that might reg­is­ter, re­view­ers should not be sub­ject­ed to at­tacks, and so on.

Since Sarep­ta won an ap­proval for eteplirsen, there have been some se­ri­ous con­cerns that the FDA has low­ered its stan­dards to the point that more com­pa­nies were like­ly to fol­low suit, wag­ing a lob­by­ing cam­paign to back in­ad­e­quate tri­al re­sults. NYU bioethi­cist Art Ca­plan told End­points re­cent­ly that more de­vel­op­ers are bound to use Sarep­ta as a role mod­el, not­ing that the in­dus­try and reg­u­la­tors need to come up with a new set of rules that could make some sense when it comes to more such pitch­es on the rare dis­ease front.

Jenk­ins, though, wants to hold the line with a more tra­di­tion­al ap­proach, leav­ing Sarep­ta as the ex­cep­tion that could help prove the rule.

Here’s the full man­i­festo:

• A poor­ly planned and ex­e­cut­ed de­vel­op­ment pro­gram for a rare dis­ease mis­us­es valu­able pa­tient re­sources and serves to de­lay ob­tain­ing the knowl­edge re­quired to un­der­stand the ben­e­fits and risks of a drug to sup­port reg­u­la­to­ry re­view and ap­proval

• FDA pro­vides valu­able ad­vice and guid­ance to spon­sors, we can­not re­quire spon­sors to fol­low our ad­vice

• Path tak­en by Sarep­ta NOT a good mod­el for oth­er de­vel­op­ment pro­grams

• As­says for bio­mark­ers should be well val­i­dat­ed be­fore use to avoid ob­tain­ing mis­lead­ing in­for­ma­tion and wast­ing clin­i­cal spec­i­mens

• Par­tic­u­lar­ly true when in­va­sive pro­ce­dure re­quired to col­lect tis­sue in chil­dren

• Rig­or­ous blind­ing and con­trol pro­ce­dures should be in place to min­i­mize bias in as­say in­ter­pre­ta­tion

• Pro­to­col should spec­i­fy blind­ing pro­ce­dures, ad­ju­di­ca­tion meth­ods, in­de­pen­dence of read­ers, etc.

• In many cas­es, ran­dom­ized con­trolled clin­i­cal tri­als rep­re­sent the fastest way to de­ter­mine if a drug is ef­fec­tive

• Ran­dom­ize as ear­ly as pos­si­ble in de­vel­op­ment to avoid po­ten­tial­ly mis­lead­ing and un­in­ter­pretable find­ings from open-la­bel tri­als

• Em­ploy meth­ods to lim­it time on place­bo (e.g., dose-re­sponse, de­layed start, ran­dom­ized with­draw­al, in­ter­im analy­sis)

• Re­port ear­ly tri­al re­sults ac­cu­rate­ly, post hoc analy­ses of failed tri­als are gen­er­al­ly hy­poth­e­sis gen­er­at­ing for next tri­al, not ev­i­dence to sup­port ap­proval

• Knowl­edge of nat­ur­al his­to­ry of dis­ease crit­i­cal to in­tel­li­gent de­sign of clin­i­cal tri­als

• Con­duct nat­ur­al his­to­ry tri­als be­fore clin­i­cal tri­als be­gin

• If a nat­ur­al his­to­ry ex­ter­nal con­trol group is pro­posed, it should be iden­ti­fied prospec­tive­ly to en­sure com­pa­ra­bil­i­ty to treat­ment group

• Nat­ur­al his­to­ry ex­ter­nal con­trol group cre­at­ed post hoc is very dif­fi­cult to in­ter­pret, un­less ef­fect of test drug is very large, due to known and un­known con­found­ing

• Use of ac­cel­er­at­ed ap­proval path­way should be prospec­tive­ly planned, NOT as a “res­cue” for a failed pro­gram

• Spon­sor and FDA should agree on the sur­ro­gate and drug ef­fect con­sid­ered “rea­son­ably like­ly” to pre­dict clin­i­cal ben­e­fit be­fore un­blind­ing da­ta

• “Any” ef­fect of a drug on a bio­mark­er is not a ba­sis for AA

• Ide­al­ly, the con­fir­ma­to­ry tri­al to fur­ther de­fine clin­i­cal ben­e­fit should be start­ed be­fore AA is grant­ed to en­sure the tri­al will be com­plet­ed in a time­ly man­ner

• FDA wel­comes the en­gage­ment of pa­tients and care­givers in help­ing to de­sign de­vel­op­ment pro­grams that will re­sult in drugs that pro­vide mean­ing­ful clin­i­cal ben­e­fit to those with dis­ease

• Ap­proval de­ci­sions must be based on da­ta from ad­e­quate and well-con­trolled clin­i­cal tri­als, which may in­clude PROs and oth­er pa­tient-de­rived mea­sures

• Ex­pe­ri­ence of pa­tients en­rolled in tri­als can be very help­ful; dis­cor­dant re­sults be­tween tri­al da­ta and pa­tient anec­dotes are very hard to rec­on­cile

• FDA re­view­ers are com­mit­ted to fa­cil­i­tat­ing de­vel­op­ment of ef­fec­tive and safe drugs for rare dis­eases

• Up­hold­ing statu­to­ry stan­dards for ap­proval in face of hopes and de­sires of pa­tients, fam­i­lies, spon­sors, and in­vestors is a very dif­fi­cult job

• Per­son­al at­tacks on FDA re­view­ers cre­ates an at­mos­phere of dis­trust and iso­la­tion rather than col­lab­o­ra­tion

• Re­cruit­ment and re­ten­tion of qual­i­fied re­view staff is very chal­leng­ing in such an en­vi­ron­ment

Fas­ci­nat­ing pre­sen­ta­tion by FDA. Slides 21 & 27 w/ es­pe­cial­ly firm lan­guage: paint Sarep­ta as a one-off, con­cern w/ at­tacks on FDA re­view­ers https://t.co/XW­b­vCT­MXNq

— Rachel Sachs (@RE­Sachs) Oc­to­ber 19, 2016