The new england journal of medicine

Health Policy Report

The FDA Breakthrough-Drug Designation — Four Years of Experience

Jonathan J. Darrow, S.J.D., J.D., M.B.A., Jerry Avorn, M.D., and Aaron S. Kesselheim, M.D., J.D., M.P.H.

In 2012, Congress created the “breakthrough for faster approval of treatments for rare dis- therapy” designation to expedite Food and Drug eases by authorizing federal research grants and Administration (FDA) testing and approval of a 50% tax credit (reduced to 25% in 2017) for medications that were intended to treat serious clinical testing expenses, providing 7 years of or life-threatening conditions and that prelimi- market exclusivity, and encouraging flexibility in nary evidence suggested may provide a substan- the conduct of clinical trials.7,8 To address HIV– tial improvement over existing treatments with AIDS and other serious illnesses, the FDA estab- regard to one or more clinically significant end lished an “expanded access” program in 1987 to points.1 The creation of this designation was allow the use of experimental therapies before motivated by the concept that advances in preci- FDA approval9 and promulgated “” sion medicine would enable the development of regulations in 1988 to promote quicker evalua- therapies with large treatment effects that were tion and approval after phase 2 trials.10 In 1992, seen early, such that random assignment to receive the agency formalized “accelerated approval” placebo might be unethical and phase 2 trials based on surrogate measures, which can be could provide sufficient evidence for approval.2 evaluated earlier than clinical end points,11 for One of the sponsors of the law, Senator Michael drugs that are intended for the treatment of seri- Bennet (R-CO), explained that the pathway was ous illnesses, as well as “” of ap- intended to speed the approval of drugs that plications for drugs that appear to offer an im- showed “exceptional results for patients.”3 provement over current treatments in safety, Manufacturers reacted vigorously. According efficacy, or convenience. Priority review appli- to Janet Woodcock, the head of the FDA Center cations would be reviewed 6 months earlier than for Drug Evaluation and Research, the agency standard review applications (in 6 months vs. was “inundated” with hundreds of requests,4 12 months, subsequently reduced to 10 months). despite expectations that only about 2 drugs per Each program thus shortened different aspects year would receive the designation.5 From 2014 of and review. Manufacturers to 2016, a total of 26 (24%) of the 108 new were permitted to pursue multiple programs to molecular agents and original biologic agents combine different benefits, thereby lowering that were approved by the FDA received the development costs and enabling earlier revenue breakthrough designation. These 26 drugs, plus generation. These programs did not formally 3 breakthrough-designated drugs that were ap- change the legal standard for approval, which proved in late 2013 and 2 breakthrough-designated continues to require the demonstration of safety vaccines, are summarized in Figure 1. and “substantial evidence of effectiveness.”12 How- ever, the programs created more flexible criteria A Growing Toolbox of Expedited for meeting this standard (the orphan, fast- Programs track, and accelerated programs) and reduced the time available to the FDA to determine The program is the sixth whether the standard had been met (the priority program that has been implemented over the program). The breakthrough program continued past four decades to expedite drug availability.6 this trend, encouraging the use of historical The 1983 Orphan Drug Act provided incentives controls or other alternatives to traditional con-

1444 n engl j med 378;15 nejm.org April 12, 2018 Health Policy Report

Oncology Rare and Other Antiinfective (orphan) (excluding oncology) Agents Blood Cystic fibrosis Meningitis B vaccines Ibrutinib Ivacaftor–lumacaftor Bexsero Oncology Blinatumomab Dabigatran reversal (GlaxoSmithKline) (nonorphan), Idelalisib Idarucizumab Trumenba (Pfizer) 6.5% Parkinson’s delusions Hepatitis C Obinutuzumab Pimavanserin Sofosbuvir Daratumumab Hyperphosphatasia Sofosbuvir–ledipasvir Elotuzumab Asfotase alfa Sofosbuvir–velpatasvir Lung Lysosomal acid lipase Dasabuvir–ombitasvir– Alectinib deficiency paritaprevir– Antiinfective Osimertinib Sebelipase alfa ritonavir agents Oncology Ceritinib Hereditary orotic aciduria Elbasvir–grazoprevir 22.6% (orphan), Skin Uridine triacetate 45.2% Idiopathic pulmonary fibrosis Rare and other Ovary Pirfenidone (excluding Rucaparib Nintedanib oncology), Soft-tissue sarcoma 25.8%

Oncology (nonorphan) Breast Palbociclib Bladder Atezolizumab

Figure 1. Indications of Breakthrough Therapies at Initial Approval. current control groups, as well as smaller trials this program, the FDA has advised sponsors that require less time to complete.6,13 Between about interim analyses, methods for data bridg- 2014 and 2016, a total of 73 (68%) of 108 new ing between studies, study-size reduction, and drugs qualified for one or more expedited pro- custom-designed end points.20 The FDA response grams (excluding expanded access), including 37 timelines are 60 days or less for many break- (34%) that qualified for three or more.14 Of the through-related submissions,18 and discussion of 31 breakthrough-designated therapies, 29 (94%) certain topics, such as proprietary names, manu- benefitted from three or more programs and 20 facturing inspections, and postmarketing studies, (65%) benefitted from four or more (Table 1). can begin earlier in the development process.21 These features have been associated with Benefits of Breakthrough measurable changes. From 2012 to 2016, break- Designation through-designated, non–fast-track drugs were associated with substantially shorter development The 2012 breakthrough-therapy statute offered and review times than fast-track, non–break- benefits to sponsors that were similar to those through-designated drugs or nonexpedited drugs contemplated under the FDA 1988 fast-track (median, 4.8 years vs. 7.1 and 8.0 years, respec- regulations, including early meetings with the tively).22 Breakthrough-designated drugs were ap- FDA, timely communication to promote efficient proved a median of 1.7 months (or 2.8 months for trial design, and involvement of senior FDA of- oncology drugs) before their target review dates. ficials, with the goal of expediting the approval of promising new drugs (Table 2). In carrying “Breakthrough” — A Powerful out its directions from Congress, the FDA devel- Label oped policies that were applicable to break- through-designated therapies: the agency created One controversial feature of the breakthrough- well-defined staff responsibilities, shortened its therapy designation is its name.23,24 Unlike earlier response times, and offered intensive guidance programs, which have titles that relate to the to corporate applicants.17-19 For example, under speed of development (“fast track”), speed of FDA

n engl j med 378;15 nejm.org April 12, 2018 1445 The new england journal of medicine

Table 1. Breakthrough-Designated, Approved Drugs, Excluding New Indications, 2013–2016.

Approval Concurrent Participation Drug Year Indication in Other FDA Programs* Fast Orphan Track Accelerated Obinutuzumab 2013 Cancer + Ibrutinib 2013 Cancer + + + Ceritinib 2014 Cancer + + Idelalisib† 2014 Cancer + + Pembrolizumab 2014 Cancer + + Blinatumomab 2014 Cancer + + Nivolumab 2014 Cancer + + + Palbociclib 2015 Cancer + Osimertinib 2015 Cancer + + + Daratumumab 2015 Cancer + + + Elotuzumab 2015 Cancer + Alectinib 2015 Cancer + + Venetoclax 2016 Cancer + + Atezolizumab 2016 Cancer + Olaratumab 2016 Cancer + + + Rucaparib 2016 Cancer + + Ivacaftor–lumacaftor 2015 Cystic fibrosis + + Idarucizumab 2015 Dabigatran + + + Sofosbuvir–ledipasvir 2014 Hepatitis C genotype 1 + Dasabuvir–ombitasvir–paritaprevir– 2014 Hepatitis C genotype 1 + ritonavir Elbasvir–grazoprevir 2016 Hepatitis C genotypes 1 and 4 Sofosbuvir 2013 Hepatitis C genotypes 1–4 + Sofosbuvir–velpatasvir 2016 Hepatitis C genotypes 1–6 + Uridine triacetate 2015 Hereditary orotic aciduria + Asfotase alfa 2015 Hypophosphatasia + + Sebelipase alfa 2015 Lysosomal acid lipase deficiency + + Meningococcal group B vaccine 2014 Vaccine + + (Trumenba, Pfizer) Meningococcal group B vaccine 2015 Vaccine + + (Bexsero, GlaxoSmithKline) Pimavanserin 2016 Parkinson’s disease psychosis Pirfenidone 2014 Pulmonary fibrosis + + Nintedanib 2014 Pulmonary fibrosis + +

* All drugs that were approved and had a breakthrough designation received priority review. † Idelalisib received accelerated approval on July 23, 2014, for relapsed follicular B-cell non-Hodgkin’s lymphoma and re- lapsed small lymphocytic lymphoma, but its breakthrough designation was granted for relapsed chronic lymphocytic leukemia (also approved on July 23, 2014).

1446 n engl j med 378;15 nejm.org April 12, 2018 Health Policy Report

Table 2. Comparison of Selected Features of the Fast-Track and Breakthrough Designations.

Element Fast Track (1988) Breakthrough (2012) Early meetings, efficient trial “A key component of the [fast track] procedures is early “Intensive Guidance . . . Beginning . . . Early”13; design consultation . . . . Such consultation is intended “taking steps to ensure that the design of the to improve . . . efficiency” by ensuring “proper clinical trials is . . . efficient.”1 ­[trial] design”10 Speed of development and “These procedures are modeled after the . . . devel­ “the Secretary shall . . . expedite the develop- ­review opment, evaluation, and approval of ,” ment and review of such drug”1 which “took only 2 years”10 Senior FDA staff involvement “the [FDA] Commissioner and other agency officials “involving senior managers and experienced will monitor [and facilitate] progress”15 ­review staff . . . in a collaborative, cross-­ disciplinary review”1 Cross-disciplinary lead Not discussed “assigning a cross-disciplinary project lead”1 Rolling review of new drug Included (authorized in 1997)16 Included13 ­application Magnitude of expected benefit “dramatic responses (i.e., great benefit)”10 “substantial improvement”1 review (“priority review”), or speed of overall ap- the task of identifying deserving drugs. The stat- proval (“accelerated approval”), the designation ute required breakthrough-designation requests “breakthrough” appears to signify an exceptional to be supported by preliminary clinical data — magnitude of therapeutic benefit. When drugs unlike requests for fast-track designation, which are designated as “breakthroughs” by the govern- may be based on animal or pharmacologic data ment’s expert medicines agency and then described or on a mechanistic rationale. In 2014, the FDA as “game-changers,” “miracles,” and “cures” in issued guidance explaining that clinical data the popular press,25 the public may reasonably might be available at the time of an investiga- expect them to provide a cure or to dramatically tional application for use in the designation improve how patients feel, function, or sur- process; these data could be based on foreign vive.24,26 Influential legislators, senior FDA offi- clinical trials or the published literature or on cials, and others have contributed to such expec- experience with the same chemical entity if it tations by explaining that the designation is had previously been approved as a different for- intended to be applied to drugs that “knock your mulation or for a different indication. In general, socks off,”27,28 or that are “homerun[s],”28,29 however, the FDA expects that clinical data will “quantum leaps,”30 or even “curative”4,5 and “life- be derived from phase 1 or 2 trials. saving.”5,21,31 Yet the designation does not in fact According to the 2014 guidance from the FDA, require drugs to be curative, or even dramati- the statutory requirement of “substantial im- cally better than existing alternatives. Indeed, provement” depends on both the magnitude and designation requests can be made and designa- the importance of benefit and may be demon- tion assigned at around the time the investiga- strated, for example, by showing that the drug tional-new-drug application is submitted — long produces a response that is “much greater” than before the evidence to substantiate such claims that obtained with existing therapies, that it has is available or the traditional approval threshold a “substantial and clinically meaningful effect” has even been established.13 if there are no existing therapies, or that it treats the underlying cause of a disease or inhibits Qualifying for the Breakthrough disease progression rather than merely treating Designation symptoms.13 The designation may also be based on a safety advantage, even if the drug has effi- Although Congress provided the broad outlines cacy similar to that of existing therapies. Break- of the breakthrough program, it left to the FDA through status may be rescinded if new evidence

n engl j med 378;15 nejm.org April 12, 2018 1447 The new england journal of medicine

shows the drug to be less beneficial than ini- accelerated approval: to meet the criteria for tially believed or if recent drug approvals mean breakthrough status, the FDA must find evidence that the designated drug is no longer likely to suggesting that the drug might provide substan- offer a substantial improvement over existing tial improvement over existing therapies, and at therapies. the time of approval the drug can lawfully be Manufacturers often issue press releases an- described as an approved, FDA-designated break- nouncing breakthrough designations shortly af- through therapy. However, when the FDA consid- ter they occur, but until a new drug or new use ers future breakthrough designations, the previ- of an existing drug is approved, the FDA does ous drug will not be considered as a comparator not publicly disclose the rationale for the break- in the determination of breakthrough status be- through designation or even the fact that the cause the FDA considers the clinical benefit of designation has been given. Instead, the FDA accelerated approval drugs — even those desig- formally recognizes the breakthrough status only nated as breakthroughs — to be uncertain until when the product is approved. postapproval studies are completed,13 which may not occur for many years.37 Of the 108 new Existing Therapies drugs that were approved between 2014 and 2016, a total of 20 (19%) were granted acceler- To ascertain “substantial improvement,” the FDA ated approval, whereas 16 (52%) of the 31 break- is required to compare a new drug with existing through-designated therapies followed the accel- therapies — but Congress did not define the erated approval pathway (Table 1). meaning of “existing therapies.” The FDA there- fore had to determine which available treat- “Substantial Improvement” ments constitute appropriate comparators. A nar- row definition, particularly one that excluded The most important measure of success of the newer or more similar medicines, would lower breakthrough program will be whether the drugs the baseline of comparison and thereby make it that are subject to faster approval through the easier for new drugs to qualify for the break- program turn out to be actual breakthroughs, through designation. as determined on the basis of solid evidence. When the FDA finalized its 2014 guidance, it Among the 31 breakthrough-designated drugs, defined “existing therapies” as excluding a num- 16 (52%) addressed oncology indications, some ber of treatments that were actually available, of which offered improvement over existing including off-label uses of approved products,32 therapies for patients who had no other options drugs approved under the accelerated approval and historically poor prognoses. Some measure pathway for which confirmatory studies had not of overall response rate was reported for all 16 yet been completed, some drugs with restricted drugs, yielding a median response rate of 43% availability under a Risk Evaluation and Miti­ (interquartile range, 32 to 68). Thirteen (81%) of gation Strategy (REMS), dietary supplements, the 16 drugs produced complete responses in no and surgery or other non–FDA-regulated treat- more than 8.5% of patients, including 7 (44%) ments.13,33 Many recently approved, similar, or that produced complete responses in no more identical treatments fell within these exclusions, than 2.8% of patients. thereby lowering the baseline of “existing thera- Even when response rates are encouraging, pies” and expanding the universe of drugs that the incremental benefit over existing treatments could qualify for breakthrough designation. can be unclear because of the limited nature of Accelerated approval can be based on a sur- the evidence, differences in trial design, absence rogate measure — for example, a laboratory test of concurrent controls, and exclusions from the or imaging study — that is “reasonably likely” to definition of existing therapies.38 For example, predict clinical benefit and requires postapproval, alectinib (Alecensa, Hoffmann–La Roche), a confirmatory studies.34-36 The accelerated approv- breakthrough-designated treatment for anaplas- al exclusion therefore required the FDA to hold tic lymphoma kinase (ALK)–positive, metastatic, conflicting positions with respect to break- non–small-cell lung cancer in patients who had through-designated drugs that were also granted disease progression while receiving crizotinib,

1448 n engl j med 378;15 nejm.org April 12, 2018 Health Policy Report was shown to have an overall response rate of vir (Zepatier, Merck) and sofosbuvir–velpatasvir 42% in two non–concurrently controlled trials39; (Epclusa, Gilead) after the approval of sofosbu- however, there was no requirement to compare vir–ledipasvir (Harvoni, Gilead) and dasabuvir– it with ceritinib (Zykadia, Novartis), which had ombitasvir–paritaprevir–ritonavir (Viekira Pak, been granted accelerated approval 19 months AbbVie), which were highly active against hepa- earlier for an identical indication on the basis of titis C virus infection, curing 94% or more of a 44% response rate (also with a breakthrough patients.50,51 But in both cases, new breakthrough designation).40 Overall, of the 31 breakthrough- designations were then granted for these drugs designated therapies, 16 (52%) (including 12 on the basis of their benefits in particular sub- [75%] of 16 oncology drugs) were approved on groups of patients. the basis of phase 1 or phase 2 data, 14 (45%) Three treatments for noninfectious diseases (including 12 [75%] of 16 oncology drugs) were also were shown to have notable efficacy. Idaru- supported by only a single pivotal trial, and 13 cizumab (Praxbind, Boehringer Ingelheim) effec- (42%) (including 10 [63%] of 16 oncology drugs) tively reversed the anticoagulant effects of dabi- were approved on the basis of either non–concur- gatran (Pradaxa, Boehringer Ingelheim) in 88% rently controlled or dose-comparison trials. or more of patients who were undergoing ur- The frequent use of surrogate measures also gent surgery. Asfotase alfa (Strensiq, Alexion), an hinders the evaluation of clinical benefit. Experi- enzyme-replacement therapy for the treatment ence has shown that the correlation of surrogate of hypophosphatasia, was associated with a me- measures with clinical outcomes in oncology is dian age at death of 3.7 years, as compared with often poor, and even a strong correlation does not 9 months in a historical control group. Anoth- necessarily indicate a large magnitude of clinical er enzyme-replacement therapy, sebelipase alfa benefit.41-43 Of the 16 breakthrough-designated (Kanuma, Alexion), was approved to treat lyso- cancer treatments, 13 (81%) received accelerated somal acid lipase deficiency and was associated approval and were therefore approved on the with a median age at death of 12 months, as basis of surrogate measures (Table 1). compared with 3.5 months in a historical con- When early evidence is strongly positive, pa- trol group. tients and researchers may reasonably believe For the remaining five non-oncology drugs that clinical equipoise is not present.44 In these that were approved after breakthrough-therapy exceptional cases, single-group trials that do not designation, efficacy proved less notable. Pirfeni- randomly assign any patients to older, standard- done (Esbriet, ) and nintedanib (Ofev, of-care treatments may be the only ethical choice. Boehringer Ingelheim) were approved in 2015 for Yet historically controlled, single-group trials may the treatment of idiopathic pulmonary fibrosis show larger treatment effects than randomized after receiving breakthrough designations. Among trials,45 and early optimism that new drugs will the three pivotal trials of pirfenidone, one did substantially outperform old ones often fails to not show efficacy and two others showed smaller be borne out when drugs are more rigorously declines in predicted forced vital capacity due to tested.46-49 pulmonary fibrosis — differences of 4.4 per- The 15 non-oncology drugs that were ap- centage points in one trial (an 8.0% decline in proved after a breakthrough designation included forced vital capacity for pirfenidone vs. 12.4% 7 highly effective antiinfective agents. Two men- for placebo) and 2.9 percentage points in the ingitis B vaccines met immunogenicity criteria other (3.7% vs. 6.6%),52 both of which were be- in approximately 80% or more of patients, and low the 10-percentage-point threshold that had 5 hepatitis C treatments cured approximately been discussed at an advisory committee meet- 90% or more of patients. When breakthrough- ing as potentially clinically important. In recom- designated drugs are curative in some patients, mending approval, the FDA noted that “patients further improvement in outcomes may be ob- and physicians are desperate for a non-surgical tained through higher cure rates, fewer side ef- therapy that may have clinical benefits.”52 Both fects, and efficacy in particular subpopulations. medicines satisfied FDA approval standards and For example, breakthrough status was rescinded may reflect important scientific advances, but for the hepatitis C treatments elbasvir–grazopre- their clinical benefits could still fail to meet the

n engl j med 378;15 nejm.org April 12, 2018 1449 The new england journal of medicine

reasonable expectations of patients that break- and the ivacaftor–lumacaftor trials both showed through-designated drugs will offer improve- similar point estimates of efficacy but that ments in lung function and not just a slower the ivacaftor trial was much smaller than the decline.53-55 ivacaftor–lumacaftor trials, which were “powered Pimavanserin (Nuplazid, Acadia Pharmaceuti- to detect even small effects.”61 In determining cals), an atypical antipsychotic indicated for treat- that ivacaftor–lumacaftor met the approval stan- ing the hallucinations and delusions associated dard on the basis of “statistically significant but with Parkinson’s disease psychosis, illustrates the small improvements” over placebo,61 the FDA distinction between approval and breakthrough observed that “there is a desperate need for designation. The drug initially failed to show those homozygous for the F508del defect” and benefit in two double-blind, placebo-controlled “any improvement is welcomed.”62 trials. Assessment questions with responses that Uridine triacetate (Xuriden, Wellstat Therapeu- indicated favorable results with pimavanserin in tics) was approved with a breakthrough designa- the two trials were then compiled into a new tion for the treatment of hereditary orotic acid- custom assessment that showed, in a single uria, a rare and potentially fatal genetic disorder pivotal trial, a 3-point advantage for pimavanserin in which a patient’s ability to produce uridine is over placebo on a 45-point scale indicating the impaired. Uridine triacetate is rapidly converted severity of psychosis.56,57 Despite the modest ben- in the body to uridine, a widely available dietary efit on this scale and the negative results in the supplement that has been used for decades to two trials, the FDA granted the request by Acadia treat hereditary orotic aciduria.63 Because the FDA for the breakthrough-therapy designation, ob- excluded dietary supplements from the baseline serving that there were no other approved drugs of “existing therapies,” the manufacturer was for the treatment of Parkinson’s disease psycho- not required to show substantial improvement sis (the use of clozapine for this indication was over the administration of uridine. Indeed, the supported by a randomized controlled trial, but pivotal trial was intended to show the “same that use was off-label). The primary medical re- clinical benefit” in patients whose treatment viewer from the FDA recommended that pima- was switched from uridine.64 The breakthrough vanserin not be approved, and three additional designation may therefore lead the medical com- FDA clinical reviewers concluded that its benefit munity to incorrectly conclude that a new drug over placebo “does not clearly represent a clini- is superior to another treatment, even if no evi- cally meaningful” change.56 However, 10 of 12 dence supports this conclusion. advisory committee members concluded that the benefits of pimavanserin outweighed its risks, and Policy Recommendations a senior FDA official observed that some patients had an improvement of 10 points or more.58,59 For serious diseases for which no treatments are When ivacaftor (Kalydeco, Vertex) was ap- currently available, patients are likely to welcome proved for the treatment of patients who had new options that may alleviate their symptoms cystic fibrosis with a G551D mutation, the label- or improve their clinical outcomes to any extent, ing indicated that the drug was “not effective in and many investigational drugs will continue to patients with cystic fibrosis who are homozy- be approved by the FDA on this basis.65 However, gous for the F508del mutation.” This exclusion when drugs are designated as breakthroughs was based on a that, according to and expedited through the FDA approval process the FDA, provided “strong evidence” that the with the use of substantial resources, many pa- drug “does not offer any clinical benefit” to tients and physicians will reasonably expect such these patients.60 Three years later, however, a com- drugs to offer major advances.22-25 Yet on the basis bination of ivacaftor and lumacaftor (Orkambi, of experience with the breakthrough designation Vertex) was approved with a breakthrough des- thus far, it seems that new drugs can meet tech- ignation for these patients even though the FDA nical requirements for the designation, and then concluded that lumacaftor may not contribute to be approved, despite having only modest efficacy. the therapeutic effect of the drug combination. To better align expectations with product The FDA explained that the initial ivacaftor study qualities while preserving program benefits,

1450 n engl j med 378;15 nejm.org April 12, 2018 Health Policy Report

Congress could eliminate the potentially mislead- so early in development, the attachment of the ing term “breakthrough” along with the separate designation to approved drugs can also promote designation and incorporate the most salutary the use of drugs that do not live up to expecta- breakthrough features, such as shortened re- tions, even if less expensive or equally effective sponse timelines, into the fast-track program. alternatives are available. With numerous studies Until Congress takes such action, the FDA could showing a higher incidence of safety-related label raise the low baseline of comparison by broad- changes and other risks associated with expedit- ening the definition of “existing therapies” to in- ed premarket testing of new drugs, close post- clude any product that is actually available to market follow-up of breakthrough-designated patients, such as off-label uses of existing prod- drugs should also be undertaken. ucts, accelerated approvals, and dietary supple- The unmet needs of patients with serious and ments. Although new drugs could still be ap- life-threatening illnesses must be promptly ad- proved without such a comparison, the term dressed, but these urgent needs cannot be met “breakthrough” implies superiority over the cur- by laudatory labels that promote the use of new rent standard of care — a comparison that may drugs that frequently offer limited additional not occur under the current requirements. benefits. The gap between the products envi- FDA approval of new products and indications sioned for breakthrough designation and the ac- facilitates insurance coverage and generates im- tual characteristics of many of the approved break- portant new information to guide prescribing, through drugs serves to highlight these unmet but such approval can occur without use of the needs, raising the question of whether this new- potentially misleading term “breakthrough.” est designation can ever live up to the high ex- Other incentives exist to promote the develop- pectations it engenders. ment and approval of off-label uses or new inves- Supported by the Laura and John Arnold Foundation, with tigational therapies, such as the 3-year, 5-year, additional support from the Harvard Program in Therapeutic Science and the Engelberg Foundation (to Dr. Kesselheim). and 7-year statutory exclusivity that is offered Disclosure forms provided by the authors are available with when applicants submit new clinical data in sup- the full text of this article at NEJM.org. port of a new indication, new chemical entity, 66,67 From the Program on Regulation, Therapeutics, and Law (PORTAL), or new orphan indication, respectively. In Division of Pharmacoepidemiology and Pharmacoeconomics, the case of accelerated approval, additional infor- Department of Medicine, Brigham and Women’s Hospital and mation is generated by required confirmatory Harvard Medical School, Boston. studies.68 1. Pub. L. No. 112-144, 126 Stat. 993 (2012). The breakthrough-therapy program has placed 2. Fleming T, Sekeres M, Lieberman G, et al. Issue brief:​ panel 4:​ high resource demands on the FDA,4 and since development paths for new drugs with large treatment effects its creation it has been associated with short- seen early. Presented at the Conference on Clinical Cancer Re- search, Washington, DC, November 10, 2011 (https:/​/​www​.focr​.org/​ ened approval times and some important drugs. sites/default/​ files/​ pdf/​ Panel4FINAL11411​ .pdf).​ In 2017, the FDA Reauthorization Act provided 3. McCrimmon KK. ‘Breakthrough’ drugs speed path to cures more user-fee funding to support the program, and the NBA. Health News Colorado. May 15, 2013. 4. Eastman P. FDA’s Breakthrough Therapy designation: up- which suggests that the FDA and industry plan date from Janet Woodcock. Oncology Times 2014;​36(12):​1,18-19 for it to continue to expand. The percentage of (https://​ journals​ .lww​ .com/​ oncology-times/​ Citation/​ 2014/​ 06250/​ ​ drugs that were approved with breakthrough FDA_s_Breakthrough_Therapy_Designation__Update.1​ .aspx).​ 5. Continuing America’s leadership: the future of medical in- designations was 22% in 2014 and 2015, 32% in novation for patients — a hearing of the U.S. Senate Committee 2016, and 37% in 2017. Ideally, the FDA would on Health, Education, Labor, and Pensions, April 28, 2015. Wash- have sufficient resources to offer the efficiency ington, D.C.: ​Government Publishing Office, 2017 (https:/​/​www​ .gpo​.gov/​fdsys/​pkg/​CHRG-114shrg94476/​pdf/​CHRG-114shrg94476​ benefits of the breakthrough designation, such .pdf). as early consultation, to the sponsors of all ex- 6. Darrow JJ, Avorn J, Kesselheim AS. New FDA breakthrough- perimental drugs. In a resource-constrained en- drug category — implications for patients. N Engl J Med 2014;​ 370:​1252-8. vironment, tethering benefits to only the most 7. H.R.Rep. 97-840(I), 1982 U.S.C.C.A.N. 3577 (1982). promising drugs is sensible, but it also creates 8. Food and Drug Administration. Draft guidance for industry:​ incentives for industry to seek breakthrough rare diseases: ​common issues in drug development. August 2015 (https://​ www​ .fda​ .gov/​ downloads/​ Drugs/​ GuidanceCompliance​ designation even for drugs that do not offer RegulatoryInformation/Guidances/​ UCM458485​ .pdf).​ large benefits. Because the designation is made 9. Darrow JJ, Sarpatwari A, Avorn J, Kesselheim AS. Practical,

n engl j med 378;15 nejm.org April 12, 2018 1451 The new england journal of medicine

legal, and ethical issues in expanded access to investigational events/​breakthrough-therapy-designation-exploring-the-qualifying drugs. N Engl J Med 2015;​372:​279-86. -criteria/​). 10. , antibiotic, and biological drug 29. Taylor L. US FDA announces first Breakthrough Therapy desig- product regulations; procedures for drugs intended to treat life- nations. PharmaTimes. January 9, 2013 (http://www​.pharmatimes​ threatening and severely debilitating illnesses. Fed Regist 1988;​ .com/news/​ us_fda_announces_first_breakthrough_therapy​ 53:41516-24.​ _designations_1004195). 11. New drug, antibiotic, and biological drug product regula- 30. Horning SJ, Haber DA, Selig WK, et al. Developing stan- tions; accelerated approval — FDA: final rule. Fed Regist 1992;​ dards for breakthrough therapy designation in oncology. Clin 57:58942-60.​ Cancer Res 2013;​19:​4297-304. 12. 21 U.S.C. § 355(d) (2017). 31. McClellan M. Breakthrough therapy designation:​ two and a 13. Food and Drug Administration. Guidance for industry:​ ex- half years in [panel discussion, video at 3:​59]. Breakthrough pedited programs for serious conditions — drugs and biologics. therapy designation:​ exploring the qualifying criteria. April 24, May 2014 (https://​ www​ .fda​ .gov/​ downloads/​ Drugs/​ Guidances/​ ​ 2015. Washington, DC: ​Brookings Institution (https:/​/​www​ UCM358301​.pdf ). .brookings​.edu/​events/​breakthrough-therapy-designation-exploring 14. Darrow JJ, Kesselheim AS. Drug development and FDA ap- -the-qualifying-criteria/​). proval, 1938–2013. N Engl J Med 2014;​370(26):​e39. 32. Jenkins J. Breakthrough therapy designation: ​two and a half 15. 21 C.F.R. § 312.87 (1989). years in [panel discussion, video at 20:​45]. Breakthrough ther- 16. FDA Modernization Act, Pub. L. No. 105-115, Sec. 112(a), apy designation:​ exploring the qualifying criteria. April. 24, 2015. 111 Stat. 2296, 2310 (1997). Washington, DC:​ Brookings Institution (https:/​/​www​.brookings​ 17. Food and Drug Administration. Manual of policies and pro- .edu/events/​ breakthrough-therapy-designation-exploring-the​ cedures — good review practice: ​review of marketing applica- -qualifying-criteria/​). tions for breakthrough therapy-designated drugs and biologics 33. Food and Drug Administration. Guidance for industry: ​avail- that are receiving an expedited review. March 9, 2015 (https:/​/​www​ able therapy. July 2004 (https://​ www​ .fda​ .gov/​ OHRMS/​ DOCKETS/​ ​ .fda​.gov/​downloads/​AboutFDA/​CentersOffices/​OfficeofMedical 98fr/​01d-0582-gdl0002​.pdf). ProductsandTobacco/​CDER/​ManualofPoliciesProcedures/​ 34. 21 C.F.R. § 314.510 (2016). UCM437281​.pdf). 35. Avorn J. Approval of a tuberculosis drug based on a para- 18. Food and Drug Administration. Manual of policies and pro- doxical surrogate measure. JAMA 2013;​309:​1349-50. cedures — good review practice: ​management of breakthrough 36. Kesselheim AS, Avorn J. Approving a problematic muscular therapy-designated drugs and biologics. July 29, 2014 (https:/​/​www​ dystrophy drug: implications for FDA policy. JAMA 2016;​316:​ .fda​.gov/​downloads/​AboutFDA/​CentersOffices/​OfficeofMedical 2357-8. ProductsandTobacco/​CDER/​ManualofPoliciesProcedures/​ 37. Woloshin S, Schwartz LM, White B, Moore TJ. The fate of UCM407009​.pdf). FDA postapproval studies. N Engl J Med 2017;​377:​1114-7. 19. Food and Drug Administration. Manual of policies and pro- 38. 21 C.F.R. § 314.126 (2016). cedures — good review practice:​ good review management prin- 39. Food and Drug Administration. Medical review: ​New Drug ciples and practices for effective IND development and review. Application 208434 (alectinib). November 10, 2015. April 29, 2013 (https:/​/​www​.fda​.gov/​downloads/​AboutFDA/​Centers 40. Food and Drug Administration. Medical review: ​New Drug Offices/OfficeofMedicalProductsandTobacco/​ CDER/​ Manualof​ Application 205755 (ceritinib). January 21, 2014. PoliciesProcedures/UCM349907​ .pdf).​ 41. Prasad V, Kim C, Burotto M, Vandross A. The strength of 20. Poirier AF, Murphy WR. The impact of Breakthrough Therapy association between surrogate end points and survival in oncol- designation on development strategies and timelines for nonon- ogy: a systematic review of trial-level meta-analyses. JAMA In- cology drugs and vaccines. Clin Pharmacol Ther 2016;​100:​603-5. tern Med 2015;​175:​1389-98. 21. Kwok M, Foster T, Steinberg M. Expedited programs for se- 42. Booth CM, Eisenhauer EA. Progression-free survival: mean- rious conditions: an update on Breakthrough Therapy designa- ingful or simply measurable? J Clin Oncol 2012;​30:​1030-3. tion. Clin Ther 2015;​37:​2104-20. 43. Cherny NI, Sullivan R, Dafni U, et al. A standardised, ge- 22. Hwang TJ, Darrow JJ, Kesselheim AS. The FDA’s expedited neric, validated approach to stratify the magnitude of clinical programs and clinical development times for novel therapeutics, benefit that can be anticipated from anti-cancer therapies: the 2012-2016. JAMA 2017;​318:​2137-8. European Society for Medical Oncology Magnitude of Clinical 23. Ross JS, Redberg RF. Would a Breakthrough Therapy by any Benefit Scale (ESMO-MCBS). Ann Oncol 2015;​26:​1547-73. other name be as promising? JAMA Intern Med 2015;​175:​1858-9. 44. Hey SP, Weijer C. What questions can a placebo answer? 24. Krishnamurti T, Woloshin S, Schwartz LM, Fischhoff B. A Monash Bioeth Rev 2016;​34:​23-36. randomized trial testing US Food and Drug Administration 45. Sacks H, Chalmers TC, Smith H Jr. Randomized versus his- “breakthrough” language. JAMA Intern Med 2015;​175:​1856-8. torical controls for clinical trials. Am J Med 1982;​72:​233-40. 25. Abola MV, Prasad V. The use of superlatives in cancer re- 46. Hwang TJ, Carpenter D, Lauffenburger JC, Wang B, Franklin search. JAMA Oncol 2016;​2:​139-41. JM, Kesselheim AS. Failure of investigational drugs in late-stage 26. Kesselheim AS, Woloshin S, Eddings W, Franklin JM, Ross clinical development and publication of trial results. JAMA Intern KM, Schwartz LM. Physicians’ knowledge about FDA approval Med 2016;​176:​1826-33. standards and perceptions of the “Breakthrough Therapy” desig- 47. Darrow JJ, Kesselheim AS. Nearly one-third of new drugs are nation. JAMA 2016;​315:​1516-8. no better than older drugs, and some are worse. Health Affairs 27. McClellan M. Summary and discussion of lessons learned Blog. October 6, 2017 ( https://​ www​ .healthaffairs​ .org/​ do/​ 10​ .1377/​ ​ [panel discussion; ​video at 30:​43]. Breakthrough therapy designa- hblog20171021​.268271/​full/​). tion:​ exploring the qualifying criteria. April 24, 2015. Washington, 48. Pease AM, Krumholz HM, Downing NS, Aminawung JA, DC:​ Brookings Institution (https:/​/​www​.brookings​.edu/​events/​ Shah ND, Ross JS. Postapproval studies of drugs initially ap- breakthrough-therapy-designation-exploring-the-qualifying proved by the FDA on the basis of limited evidence: systematic -criteria/​). review. BMJ 2017;​357:​j1680. 28. Sherman R. Summary and discussion of lessons learned 49. Food and Drug Administration. Guidance for industry: ​clin- [panel discussion;​ video at 26:​53]. Breakthrough therapy desig- ical trial endpoints for the approval of non-small cell lung cancer nation:​ exploring the qualifying criteria. April 24, 2015. Wash- drugs and biologics. April 2015 (https:/​/​www​.fda​.gov/​downloads/​ ington, DC:​ Brookings Institution (https:/​/​www​.brookings​.edu/​ drugs/​guidances/​ucm259421​.pdf).

1452 n engl j med 378;15 nejm.org April 12, 2018 Health Policy Report

50. Food and Drug Administration. Medical review:​ New Drug Materials/Drugs/​ PsychopharmacologicDrugsAdvisoryCommittee/​ ​ Application 208261 (elbasvir and grazoprevir). January 15, UCM502050​.pdf). 2016. 59. Food and Drug Administration. Office director memoran- 51. Food and Drug Administration. Medical review: ​New Drug dum:​ 207318 (pimavanserin). April 29, 2016. Application 208341 (sofosbuvir and velpatasvir). June 1, 2016. 60. Food and Drug Administration. Medical review: ​New Drug 52. Food and Drug Administration. Office director memo:​ New Application 203188 (ivacaftor). November 29, 2011. Drug Application 022535 (pirfenidone). October 15, 2014. 61. Food and Drug Administration. Medical review: ​New Drug 53. Raghu G, Selman M. Nintedanib and pirfenidone: New anti- Application 206038 (ivacaftor/lumacaftor). June 2, 2015. fibrotic treatments indicated for idiopathic pulmonary fibrosis 62. Food and Drug Administration. Office director memorandum:​ offer hopes and raises questions. Am J Respir Crit Care Med New Drug Application 206038 (ivacaftor/lumacaftor). July 2, 2015. 2015;​191:​252-4. 63. Becroft DM, Phillips LI, Simmonds A. Hereditary orotic acid- 54. Spagnolo P. Novel treatments for idiopathic pulmonary fibro- uria: long-term therapy with uridine and a trial of uracil. J Pediatr sis. Am J Med 2015;​128:​447-9. 1969;75:​ 885-91.​ 55. Ratner M. Landmark approvals in idiopathic pulmonary fibro- 64. Food and Drug Administration. Medical review: ​New Drug sis. Nat Biotechnol 2014;​32:​1069-70. Application 208169 (uridine triacetate). September 4, 2015. 56. Food and Drug Administration. Medical review: ​New Drug 65. Darrow JJ. Pharmaceutical efficacy: the illusory legal stan- Application 207318 (pimavanserin). February 3, 2016. dard. Wash Lee Law Rev 2013;​70:​2073-136. 57. Food and Drug Administration. Statistical review: ​New Drug 66. 21 C.F.R. § 314.108(b) (2016). Application 207318 (pimavanserin). February 22, 2016. 67. 21 C.F.R. § 316.31(a) (2016). 58. Food and Drug Administration. Meeting of the Psychopharma- 68. 21 C.F.R. § 314.510 (2016). cologic Drugs Advisory Committee. March 29, 2016 (https:/​/​www​ DOI: 10.1056/NEJMhpr1713338 .fda.gov/​ downloads/​ AdvisoryCommittees/​ CommitteesMeeting​ Copyright © 2018 Massachusetts Medical Society.

apply for jobs at the nejm careercenter Physicians registered at the NEJM CareerCenter can apply for jobs electronically. A personal account created when you register allows you to apply for positions, using your own cover letter and CV, and keep track of your job-application history. Visit nejmjobs.org for more information.

n engl j med 378;15 nejm.org April 12, 2018 1453