Among 240 non-oncology BTD requests, 93 requests (39%) were granted and 147 requests (61%) were denied. Table 1 shows a summary of BTD decisions in therapeutic areas, including pulmonary, allergy, and rheumatology (14%); dermatology and dental (13%); psychiatry (13%); neurology (12%); gastroenterology and inborn errors (10%); and cardiology and renal (9%). Of note, four of these therapeutic areas were among the top five non-oncology therapeutic areas identified in a previous analysis conducted by CDER [3]. This analysis did not consider the number of BTDs granted relative to the number of active INDs in each area; accordingly, any differences observed may be related to the number of active development programs.

Additional analyses of the BTD requests examined whether an FDA-approved therapy was available for the proposed indication in the BTD request at the time of submission; see Table 2. Among 107 (45%) of the 240 BTD requests for which FDA-approved therapy was available at the time of BTD submission, 30 (28%) were granted and 77 (72%) were denied. Among the 133 (55%) of the 240 requests that were submitted for indications for which no approved therapies were available, 63 (47%) were granted and 70 (53%) were denied.

The orphan-designation status of the BTD requested drugs is summarized in Table 3. Of note, this assessment did not discern whether the request pertained to an orphan subset of persons among those with the disease or condition involved. Among 240 BTD requests, 90 (42%) were for orphan drugs, of which 47 (52%) were granted and 43 (48%) were denied. In contrast, among 150 (58%) requests that were not for an orphan drug, 46 (31%) were granted and 104 (69%) were denied.

Table 4 shows the combined pattern of orphan status of candidate drugs and availability of therapy for the respective indications for granted and denied BTD requests. The results show that 26% (n = 63) of the BTD requests were for drugs designated as orphan and with no available therapy; the rate of granting BTD requests is highest (56%) for this subset of requests. In contrast, among the 33% (n = 80) of BTD requests for drugs that are not designated as orphan and where therapy is available, the rate of granting the request was lowest (23%).

The 93 granted BTD requests were further evaluated to obtain information on (1) clinical characteristics of the supporting trials; (2) preliminary clinical evidence of substantial improvement over available therapy; (3) rationale for withdrawal of BTD requests or rescission of BTDs granted where applicable; and (4) marketing approval status.

Table 5 shows that the latest trial phase supporting a majority of the BTD requests was Phase 2 (n = 50) or Phase 3 (n = 15). Data from patients receiving candidate drugs under expanded access (EA) were included in four requests. In two of these requests, the EA data were used to supplement the data collected from clinical trials. In one request, for a very rare disease, EA data from one patient and limited information from another patient were used to support granting the request. In another case, also for a very rare disease, EA data from 13 patients were included to support the request for breakthrough therapy.

The studies supporting granted BTD requests included at least one randomized trial in 71 (76%) of requests, the majority of which (n = 58; 82%) were blinded. Although data from blinded trials supporting BTD were primarily from double-blind studies (n = 53), single-blind (n = 4), and blinded-to-dose-strength (n = 1) trials were also noted. In 71 (76%) BTD requests, a placebo, vehicle, active control, or no treatment was used in a concurrent control group. In 22 (24%) of granted BTD requests without a control arm, study characteristics included use of historical or baseline controls (e.g., extrapolation from trials of approved indications or contemporaneous external controls).

As mentioned previously, among 63 (68%) of 93 granted BTD requests, no therapy was available for the indication for which the BTD was requested. For these requests, CDER’s decision to grant BTD was based most often on requests (n = 62) with specific data showing a clinically meaningful treatment effect on important outcomes, including six instances in which the effect size was reported as relatively small but was still considered meaningful because of, in part, the absence of available therapy. For one request related to medical countermeasures, the decision to grant BTD was based on extrapolation of data from a clinical trial of a related approved indication and data from animal studies, given that conducting a human clinical trial for the BTD requested indication was determined not to be appropriate due to ethical considerations.

For 30 (32%) of 93 granted BTD requests, therapy was already available for the indication for which BTD was requested. For these requests, CDER’s decision to grant BTD included consideration of (1) preliminary evidence that the BTD drug had the potential to be a substantial improvement over available therapy for the specified indication (n = 20); (2) efficacy of the BTD drug being comparable to available therapy that was known to be discontinued by the manufacturer (n = 2); (3) BTD drugs having an improved safety profile (n = 4); and (4) other (n = 4), including the BTD drug being intended for patients who do not respond adequately to the available therapy.

BTD requests were received for a wide variety of indications (e.g., cytomegalovirus infection, neurotrophic keratitis, sickle cell disease, weight management), and diverse clinical endpoints were used to provide preliminary clinical evidence of efficacy. Accordingly, the type of point estimate used to evaluate efficacy varied and included hazard ratios, relative or mean changes compared to control, change in least-square mean values, etc. This variability in diseases and clinical endpoints limited our ability to directly compare the preliminary clinical evidence and decisions made to grant or deny BTD across various therapeutic areas. Nonetheless, in 60 (65%) of 93 granted BTD requests, information about the statistical significance of the data supporting the requests was included in the BTD request review, and in 90%, the p value was less than 0.05 (including 57% < 0.01). In the remaining 10% of granted BTD requests, nominal statistical significance was noted for a key secondary endpoint (and the p value for the primary endpoint was typically less than 0.10).

BTD requests for which a p value was not reported or assessed (n = 33) most often involved orphan diseases. Among these requests, various attributes were observed, including a small number of patients receiving the drug (< 30, and often < 10), with such requests often having a summary-level external control.

Several other factors are noted in our review of the requests which may have influenced decision making by review divisions. For example, for nine granted BTD requests, the reviewers noted that the mechanism of action was either novel or complementary to existing treatment, and therefore, the potential for additional efficacy using combination therapy was recognized. In addition, preliminary evidence suggesting a sustained response to the drug was noted as supportive information for nine requests. Preliminary clinical evidence suggesting not only efficacy but also the potential for improved safety compared to available therapy, including less abuse potential, was mentioned for nine requests. In three cases, an apparent improvement in the safety profile was emphasized as a reason for granting the BTD.

Eight granted BTDs were subsequently rescinded by FDA for reasons including (1) approval of new therapies for the respective indications such that the BTD therapy was no longer found likely to show a substantial improvement over available therapy; (2) sponsor’s plan to discontinue the drug development program; and (3) additional data from clinical trials subsequent to BTD indicate that the drug is unlikely to demonstrate substantial improvement over available therapies.

When a BTD request is denied, the sponsor has the option to submit a new request to provide additional data. Among the 93 BTD requests that were granted, 8 (9%) were new submissions that were previously denied by FDA but were resubmitted with a revised indication or with additional data, to support the request.

Three BTD requests were withdrawn by the sponsor after being granted BTD. For one such request, the review division raised concerns about subsequent safety data that became available after BTD was granted, and the sponsor requested the withdrawal. For two other requests, the sponsor requested withdrawal of BTD per the advice of the review divisions. In one case, a collaborative decision was made to contemporaneously grant a new BTD for a broader indication for the same drug development program that encompassed the original BTD; in another case, the sponsor was asked to submit a new request under a new IND for the specified drug development program, and subsequently the original request was withdrawn. Overall, and excluding the granted BTD request that was withdrawn due to granting BTD for a broader indication, ten (11%) of 93 drugs did not retain the BTD designation.

As of December 31, 2022, 28 drugs involving 29 BTD requests that were granted BTD between calendar years 2017 and 2019 received marketing approval for the indication for which BTD was granted; see Table 1, Supplementary Materials. Of note, for 11 drugs the NDA/BLA was received within a year after the corresponding BTDs were granted, indicating that some requests were granted relatively late in drug development. Twenty one of these 28 drugs (75%) are first in class, which indicates that the drugs have mechanisms of action different from those of existing therapies.

The BTD requests denied without review by the MPPRC (n = 61; 42%) were denied due to clearly not aligning with BTD criteria, as discussed in the “Methods” section. Examples include instances where the indication was not a serious or life-threatening disease, the participants enrolled in the trial were not representative of the patient population for which BTD was being sought, or demonstration of potential improvement over available therapy was lacking. The current analysis focused on the 86 BTD requests denied by the review division after taking specific feedback from the MPPRC into consideration. In 26 (30%) of 86 denied requests, a single reason was cited for denial. In the remaining 60 (70%) denied requests, the decision to deny was based on multiple reasons. Reasons for denial are shown in Table 6, with insufficient treatment effect identified as the most frequent reason for denial (n = 50; 58%). The specific examples included small, or no, treatment effect observed compared to an appropriate comparator; lack of statistical significance for the effectiveness endpoint, when applicable; pattern of dose response that raised uncertainty about efficacy; and loss of efficacy over time.

As shown in Table 6, aspects of trial design were the second most common (n = 46; 53%) reason for denial, including problems related to short study duration making it difficult to assess preliminary evidence of substantial benefit, lack of an appropriate control arm, concern regarding multiplicity when secondary endpoints were being relied on, and an external control group matched poorly to patients in a single-arm trial. In 33 (38%) of denied BTD requests, data-related issues were cited as reasons for denial. Corresponding examples, whether as a single reason or multiple reasons, included lack of trial participants’ pre-treatment history, timing of improvement incongruent with anticipated mechanism of drug, or inadequate data regarding the comparator arm.

As another reason (n = 28; 33%) for denial, the review division determined that the endpoint was not clinically relevant for patients with the disease. Examples of endpoint issues identified include a lack of supportive evidence that the endpoint used by the sponsor predicted a clinically meaningful benefit, or the endpoint did not represent a serious aspect of the disease.

Safety was cited as a reason for denial in 14 (16%) requests. An example in this category includes an excess of adverse events in the investigational arm versus the control arm. Flawed (e.g., unblinded) post hoc analyses were noted as a reason for denial in five requests (6%). An in-depth description of the various characteristics (e.g., phase, blinding) of the trials for denied requests was not conducted.