Abstract
Background
Many new molecular entities enter clinical development to evaluate potential therapeutic benefits for oncology patients. We characterized adult and pediatric development of the set of new molecular entities that started clinical testing in 2010-2015 worldwide.
Methods
We extracted data from AdisInsight, an extensive database of global pharmaceutical development, and the FDA.gov website. We followed the cohort of new molecular entities initiating first-in-human phase I clinical trials in 2010-2015 to the end of 2020. For each new molecular entity, we determined whether it was granted US Food and Drug Administration (FDA) approval, studied in a trial open to pediatric enrollment, or stalled during development. We characterized the cumulative incidence of these endpoints using statistical methods for censored data.
Results
The 572 new molecular entities starting first-in-human studies in 2010-2015 were studied in 6142 trials by the end of 2020. Most new molecular entities were small molecules (n = 316, 55.2%), antibodies (n = 148, 25.9%), or antibody-drug conjugates (n = 44, 7.7%). After a mean follow-up of 8.0 years, 173 new molecular entities did not advance beyond first-in-human trials, and 39 were approved by the FDA. New molecular entities had a 10.4% estimated probability (95% confidence interval = 6.6% to 14.1%) of being approved by the FDA within 10 years of first-in-human trials. After a median of 4.6 years since start of first-in-human trials, 67 (11.7%) new molecular entities were tested in trials open to pediatric patients, and 5 (0.9%) were approved for pediatric indications.
Conclusions
More efficient clinical development strategies are needed to evaluate new cancer therapies, especially for children, and incorporate approaches to ensure knowledge gain from investigational products that stall in development.
Over the past 2 decades, an improved understanding of cancer biology and innovations in tumor profiling techniques have promoted the development of many new anticancer products in clinical trials. Moving away from traditional chemotherapies, clinical trials have begun to evaluate investigational drugs and biologics with diverse mechanisms of action, such as kinase inhibitors, antibodies, and other molecularly targeted agents (1-3).
To maximize the benefits of these scientific advances, concomitant development of data-driven approaches is needed to improve efficiencies and success rates in clinical development pipelines. Investigational products frequently stall during clinical development, from first-in-human trials to application for regulatory approval, with common causes including emerging safety or efficacy concerns and difficulties in patient accrual (4,5). Systematically characterizing development programs and their outcomes would inform strategies to increase development efficiencies and identify processes in need of innovation.
In assessing development programs, special consideration is warranted for the evaluation of therapeutic agents in children and adolescents. Historically, most initial or pivotal clinical trials of novel agents excluded pediatric cancer patients, potentially contributing to delays in development and a paucity of approved anticancer therapies for this population (6). When adult drugs are studied in children, initiation of pediatric clinical trials is typically postponed until well after completion of 1 or more in adults (7), promoting off-label use in pediatric patients (8). To accelerate the development of new therapies, US Congress has promulgated several reforms granting the Food and Drug Administration (FDA) authority to request pediatric studies, but data on the impact of these programs in pediatric oncology remain sparse (9,10).
We conducted a longitudinal analysis of a comprehensive set of new therapeutic products starting development for an oncology indication from 2010 to 2015, with follow-up through 2020. Our objective was to characterize the clinical development pipeline of oncology products for adult and pediatric cancers and determine the rate and timing of regulatory deliberations.
Methods
Data sources
We obtained data on novel therapeutic products from AdisInsight, a commercial database of global pharmaceutical research and development (11) that was previously used in other studies of clinical development pipelines (12-16). AdisInsight provides data on individual therapeutic products, including their type (Supplementary Methods, available online), and their clinical trials, including trial start date, phase, and age-related inclusion criteria. For each investigational product, we used the FDA’s public data to obtain information on its approval status for adult and pediatric patients and on the issuance and status of pediatric study requests (17-19).
Selection of new molecular entities
We queried AdisInsight on May 25, 2021, for all oncology products studied for anticancer indication in any phase I clinical trial (including hybrid phase I-II studies) started between January 1, 2010, and December 31, 2015. We included only new molecular entities, which are defined as drugs or biologics containing moieties not previously approved by the FDA (as single drug ingredients or as part of combination products) (20). Conversely, we excluded products not studied for anticancer indications (eg, contrast dyes or anti-emetics), products studied in other clinical trials (of any phase and for any indication) before 2010, and products for which the first identified trial during 2010-2015 was not a phase I study. We identified the first trial started during 2010-2015 that studied an new molecular entity as its first-in-human study and defined the corresponding start date as the start date of the new molecular entities clinical development.
Longitudinal follow-up of individual new molecular entities
We followed each new molecular entity from start of clinical development to December 31, 2020, when data were censored. We identified all clinical trials started during follow-up that studied the product, alone or in combination with other interventions, and for any indication (not distinguishing between company and investigator-sponsored trials). We considered phase I/II and II/III studies as equivalent to phase I and III studies, respectively. We defined trials as pediatric eligible if patients aged younger than 18 years were eligible for participation. Trials with missing age eligibility information (n = 95) were excluded from analyses examining pediatric trials. Similar to prior analyses (21), we considered development of an new molecular entity as stalled if the new molecular entity was neither approved nor studied in new trials of any phase for at least 6 years after start of the last trial, assuming an average trial duration of 4 years and a gap of 2 years between the end of a study and the start of the next. As these assumptions might be conservative for new molecular entities studied in phase I and II studies (which are commonly based on short-term toxicity or tumor response endpoints), in sensitivity analyses we also considered 4- and 5-year periods without new trials.
We identified FDA actions for each new molecular entity through the end of 2020, including drug approval, date of approval, and approval for use in pediatric populations. We also determined if new molecular entities were subject to FDA pediatric study requests under the Best Pharmaceuticals for Children Act (BPCA) or the Pediatric Research Equity Act (PREA) and whether these were fulfilled.
Statistical analyses
We described the distribution of categorical variables using counts and percentages and of quantitative variables using means, medians, and interquartile ranges (IQRs). To account for censored follow-up times, we summarized the distribution of time to FDA approval and time to initiation of pediatric-eligible trials using Kaplan–Meier (KM) cumulative probability estimates (with 95% confidence intervals [CIs]). Because by definition new molecular entities were not at risk of stalling after being approved (competing risk), we summarized the distribution of time until stalled development using Kalbfleisch–Prentice cumulative probability estimates (22). We performed all analyses in R (v4.1.2) (23).
Results
Identification and classification of new molecular entities
We identified 1181 therapeutic products potentially eligible to enter our analysis cohort. After applying our exclusion criteria (Supplementary Figure 1, Supplementary Table 1, available online), we included 572 new molecular entities in our analysis (Supplementary Table 2, available online). On average, 95 new molecular entities started clinical development per year, from a minimum of 76 in 2013 to a maximum of 112 in 2011 (Figure 1). The 3 most studied types of new molecular entities were small molecules (55.2%; Figure 2), antibodies (25.9%), and antibody-drug conjugates (7.7%).
Figure 1.
New molecular entities starting clinical development, 2010-2015. Blue, total number of new molecular entities; orange, new molecular entities that were ultimately studied in at least 1 pediatric-eligible trial before the end of 2020.
Figure 2.
Distribution of new molecular entity types that started clinical development, 2010-2015. Upper bars, total number of NMEs that started clinical development in 2010-2015; Lower bars, new molecular entities that were studied in at least 1 pediatric-eligible trial before the end of 2020. Small molecules included kinase inhibitors (n = 141) or other targeted agents (n = 175). Nonantibody biologics included nonantibody proteins (n = 23) and nucleic acids (n = 12). Other or unspecified new molecular entities included 3 products: a radiopharmaceutical agent, a photosensitizing drug, and an new molecular entity for which we could not identify the type or mechanism of action.
Follow-up of new molecular entities and identified clinical trials
We followed each new molecular entity for a mean of 8.0 (median = 8.1, IQR = 6.4-9.5) years. During follow-up, the 572 new molecular entities were studied in 6141 clinical trials initiated during 2010 to 2020, with a mean of 12 (median = 3, IQR = 1-7) trials per product. Of all trials, 2971 (48.4%) were phase I, 2413 (39.3%) phase II, 490 (8.0%) phase III, and 258 (4.2%) phase IV (phase information was not available for 9 trials). The highest development phase reached by the end of 2020 was first-in-human phase I trials for 173 (30.2%) new molecular entities, non-first-in-human phase I trials for 152 (26.6%), phase II trials for 162 (28.3%), phase III trials for 55 (9.6%), and phase IV trials for 30 (5.2%).
Development and regulatory outcomes
There were 39 (6.8%) new molecular entities approved by the FDA during follow-up, with a median time to approval of 6.1 (IQR = 4.6-8.1) years. For these new molecular entities, a mean of 26 trials (of any phase) were initiated prior to approval (median = 20, IQR = 9-35). Based on KM estimates, at 5 and 10 years from start of clinical development, each new molecular entity had a 2.6% (95% CI = 1.3% to 3.9%) and 10.4% (95% CI = 6.6% to 14.1%) probability, respectively, of reaching FDA approval (Figure 3).
Figure 3.
Cumulative probability estimates for each time-to-event endpoint. Terminal curve heights correspond to the estimated probabilities of achieving the corresponding endpoint during follow-up. FDA = US Food and Drug Administration.
Development of 184 of the 572 (32.2%) new molecular entities stalled during follow-up. These 184 new molecular entities were studied in 275 trials before stalling, with a mean of 1.5 (median = 1.0, IQR = 1.0-2.0) trials per stalled new molecular entity. Most of these new molecular entities stalled after their first-in-human phase I trial (n = 133, 72.3%) or a non-first-in-human phase I trial (n = 45, 24.5%), and 6 (<0.1%) stalled after a phase II trial. Each new molecular entity had a 49.2% (95% CI = 43.4% to 55.7%) probability of stalling within 10 years of development start (Supplementary Figure 2, available online). At the level of drug types, small molecules and antibodies had a 10-year probability of stalling of 43.4% (95% CI = 36.2% to 52.0%) and 54.8% (95% CI = 44.1% to 65.2%), respectively, and for new molecular entities of remaining types of this probability was 58.6% (95% CI = 44.9% to 76.5%; Figure 4).
Figure 4.
Estimates of the cumulative probability of stalled development (based on 4 years without new trials) for new molecular entities of different types. Terminal curve heights correspond to the estimated probabilities of stalled development (based on 4 years without new trials) during follow-up.
In sensitivity analyses using 4 and 5 years of trial inactivity to define stalled development, 341 (59.6%) and 275 (48.1%) new molecular entities stalled during follow-up, respectively. In either case, more than 90% of stalled new molecular entities ended development during phase I trials, with more than 60% never leaving first-in-human phase I trials (Supplementary Table 3, available online). New molecular entities had a 67.2% (95% CI = 62.5% to 72.2%) and 57.6% (95% CI = 52.5% to 63.2%) 10-year probability of stalling, based on 4 and 5 years of trial inactivity, respectively (Supplementary Figure 2, available online).
Pediatric development of new molecular entities
Of the 572 new molecular entities, 67 (11.7%)—mostly small molecules and antibodies (Figure 2)—were tested in pediatric-eligible trials during follow-up, with a median time from start of clinical development to first pediatric-eligible trial of 4.6 (IQR = 2.8-6.4) years. Five new molecular entities began their clinical development with a pediatric-eligible trial (Supplementary Table 4, available online). For the remaining 62 new molecular entities, a mean of 14.9 (median = 8.5, IQR = 3.0-19.8) trials (not open to children) were started before the start of pediatric-eligible trials. The highest clinical trial phase reached by these 62 new molecular entities before the start of pediatric-eligible trials was first-in-human phase I for 5 (8.1%), non-first-in-human phase I for 11 (17.7%), phase II for 18 (29.0%), phase III for 21 (33.9%), and phase IV for 7 (11.3%). Based on KM estimates, the 67 new molecular entities had a 15.4% 10-year probability (95% CI = 11.2% to 19.4%) of being studied in pediatric-eligible trials (Figure 3).
Of the 39 approved new molecular entities, 5 (13% of approved new molecular entities, 0.9% of all new molecular entities) were approved by the FDA for use in pediatric patients by the end of 2020 (Table 1). The 10-year probability of obtaining pediatric approval was 1.4% (95% CI = 0.0% to 2.7%; Figure 3). Of the 39 approved products, 8 received BPCA requests, which resulted in the initiation of pediatric studies for 2 new molecular entities. No approved new molecular entity was subject to PREA study requirements.
Table 1.
New molecular entities initiating clinical development from 2010 to 2015 and approved by the FDA for pediatric oncology indications by the end of 2020
| New molecular entity name (start year of clinical development)a | First FDA approval, year | First pediatric approval, year | Approved pediatric indication | Pediatric ages included in initial pediatric approval | Clinical data supporting the initial pediatric approval | No. of pediatric-eligible trialsb |
|---|---|---|---|---|---|---|
|
2014 | 2017 | Refractory or relapsed classical Hodgkin lymphoma | Any age (0-17 years) | Safety evaluated in a pediatric study (40 patients aged 2-18 years). Efficacy extrapolated from trial results in adults. | 7 |
|
2020 | 2020 | Relapsed or refractory bone or bone marrow neuroblastoma | 1 year or older | Safety and efficacy data from pediatric patients enrolled in 2 clinical trials. | 4 |
|
2017 | 2017 | Metastatic Merkel cell carcinoma | 12 years or older | Safety and efficacy extrapolated from results of adult trials. | 1 |
|
2019 | 2019 | Solid tumors with NTRK gene fusion | 12 years or older | Safety and efficacy extrapolated from results of 3 adult trials and a pediatric study (30 patients aged 2-18 years). | 2 |
|
2018 | 2018 | Solid tumors with NTRK gene fusion | Any age (0-17 years) | Safety and efficacy data from pediatric patients enrolled in 2 clinical trials. | 5 |
Naxitamab, entrectinib, and larotrectinib were subject to BPCA requests. Studies were conducted for entrectinib and contributed to pediatric approval. BPCA = Best Pharmaceuticals for Children Act; FDA = US Food and Drug Administration.
Number of pediatric-eligible trials that started before pediatric approval.
Discussion
In this study, we characterized the clinical development of a comprehensive cohort of 572 new anticancer products that started clinical testing in 2010-2015 worldwide, following them from their first phase I trial to 2020. We found that, over a mean follow-up of 8 years, individual products were studied in 12 trials on average, but less than half progressed beyond phase I. The 10-year probability of stalled development was 67.2% and, of reaching FDA approval, 10.4%. Few products were studied in trials open to pediatric patients, and the 10-year probability of receiving a pediatric approval was 1.4%. Overall, our results highlight the need for improved strategies to reduce unnecessary clinical testing, promote the study of new anticancer agents in pediatric populations, and ensure knowledge gain from investigational products that stall in development.
Our results update and expand previous analyses of clinical development programs (7,21,24-28). Compared with some of these, a strength of our study is that it was based on a database that captures information by continuously monitoring multiple origins, including, among others, different clinical trial registries, press releases from pharmaceutical companies, and publications in medical journals. This database is thus more likely to provide a timely characterization of the clinical development of an individual product beginning their first clinical trial than any single information source considered alone. Another key feature is our focus on adult and pediatric development outcomes, including start of pediatric-eligible trials or pediatric FDA approvals. Although other studies evaluated similar endpoints (7,28), ours is the first to characterize their incidence within a comprehensive cohort of new anticancer products.
In our cohort, development programs that stalled were more likely to have done so after phase I studies, especially first-in-human ones. Although no data on the reason for stalling were available from our data sources, we expect that, in addition to commercial considerations by the sponsor, development was likely paused or terminated during this phase because toxicity rates were higher or tumor response rates lower than predicted by preclinical studies (29). Indeed, meta-analyses of preclinical studies indicated that preclinical models are poor predictors of drug effects in human patients, and their findings often have poor reproducibility (30-33). Improving the robustness and predictive value of preclinical results could thus help reduce the risk of translational failures and waste of resources in unwarranted clinical trials (34).
Although products in our cohort that stalled during follow-up were less likely to have done so after a phase II or III trial, such studies are typically longer, larger, and more costly than phase I trials (35). This highlights the need to identify, develop, and evaluate strategies aimed at reducing potential waste in such studies, including use of external controls, novel surrogate endpoints, and more efficient trial designs.
The increasing availability of patient-level data from completed clinical trials, real-world observational studies, and electronic health records has created new opportunities for the use of trial designs based on external control arms (36). In phase II studies, these designs can help increase statistical power to assess experimental therapies and more quickly predict which ones are unlikely to succeed in phase III studies (37-39). However, careful consideration is required to account for confounding because of differences in prognostic factors across internal and external study groups or other distortions (40,41). Novel statistical methods can leverage multiple external control datasets to predict the impact of such biases and inform decision making during clinical development processes (37,38). Additionally, new Bayesian statistical methods can tailor the weight given to external controls during analyses based on their level of homogeneity with concurrent controls, minimizing the potential impact of confounding bias (42).
Surrogate endpoints have long been used to shorten studies’ duration and provide more statistical power to detect treatment effects. However, the value and predictive strength of common surrogates have often been questioned, such as use of progression-free survival and tumor response rates as surrogates for overall survival (43-45). Identification of novel surrogates with strong predictive power might provide more robust evidence helpful to reduce the clinical and economic burden of conducting clinical trials—either by using these as primary endpoints in study designs in place of harder-to-assess clinical outcomes, as standard, or by incorporating them in novel adaptive designs in which surrogates are only considered in interim analyses for futility (46). Emergent liquid biopsy techniques may allow the development of new blood-based surrogates defined by monitoring circulating tumor cells or DNA to detect residual disease or predict treatment response (43,47,48). Such endpoints will need to be collected in clinical trials alongside outcomes such as overall survival to enable their prospective validation.
Innovative trial designs may also support increased efficiency in the development of new cancer therapies. For example, adaptive multistage designs can enable early termination of studies for investigational therapies lacking efficacy—a common cause of failure in phase II and III trials (29,49)—and support redirection of resources to other products (50,51). Novel platform trial designs that allow new arms to open after study start may help avoid the cost of planning new trials for studying emergent therapies (52-54). In trials of investigational targeted therapies, enrolling only patients predicted to receive its benefits may help reduce the overall study size. However, this approach is only appropriate when validated biomarkers that accurately predict treatment effects are available (55,56).
Given the high rate of products that stall in development, strategies are needed to ensure information generated in clinical trials of unsuccessful therapeutics is made available to inform future research activities (57). Sponsors can discontinue or pause development pipelines without public disclosure of the underlying reasons and without making clinical trial data available (5,58-60). Development stalled for more than half of new products in our cohort, with about 2 clinical trials conducted on average for each stalled product. Access to data from such trials can guide development decisions for other investigational agents, help identify and evaluate resource-saving strategies for early stopping of unpromising development programs, and help reduce exposure of trial participants to futile or unsafe products. Data from trials of stalled development programs can also help reposition a product for alternative indications and contribute information for meta-analyses of toxicity or efficacy endpoints (5,58,60-62).
Most clinical trials in our study excluded pediatric patients aged younger than 18 years. Although not all their studied new molecular entities might be relevant for pediatric indications, many targeted therapies (the majority in our sample) might be used to treat adult and pediatric cancer types if these express a common molecular target (10). Additionally, even in the small sample of drugs studied in pediatric-eligible studies, we found that pediatric studies were not initiated until a median of 4.6 years after study start in adults. This finding supports the observation that investigational drugs relevant for pediatric patients are typically not evaluated in children or adolescents until several years after study start in adults, if at all. Although this practice allows a preliminary evaluation of safety and efficacy in adults, in some cases it may be scientifically unnecessary and might promote off-label use of new drugs in young patients (6,7,63,64). When scientifically appropriate, lowering the minimum age required for study participation—for example, allowing the inclusion of adolescents alongside adults (65-67)—might accelerate the pediatric evaluation of new anticancer products. However, this strategy should be evaluated to ensure acceptable safety risks for children and adolescents and to assess its potential impact on testing in adults (6,67,68).
The pediatric programs implemented by the FDA under the BPCA and PREA legislations (the first originally enacted in 2002, the second in 2003) aimed to increase the number of clinical trials that test therapies developed for adults in young patients. Our results suggest that their effect has been minimal for cancer therapies. This finding is consistent with a recent review of the outcomes of BPCA study requests issued for oncology products (63), which found that of 40 requests issued during 2001-2019, only 23 resulted in a pediatric study. For PREA, prior studies have also found that requirements were waived for all new products developed for adult cancers, either because the cancers did not occur in children (such as small cell lung cancer) or because the original indication was granted orphan designation (10,63).
Our study has several potential limitations. In particular, we cannot exclude that some anticancer products or clinical trials potentially eligible to enter our analyses were not captured by AdisInsight. Still, this database is assembled using information regularly collected from multiple sources and is designed to track individual products from start of clinical development, including through capture of all product names (Supplementary Table 2, available online) (12-16). Thus, any possible missed trial is most likely to be a phase I study, as higher phase trials are subject to more stringent registration requirements. For the same reason, if any therapeutic product was not captured by AdisInsight, it is likely that this was only studied in phase I trials.
More generally, our results depend on the completeness and quality of information collated by AdisInsight from multiple sources. Validation studies of such databases are impaired by the lack of an appropriate gold standard data source. Future investigations like ours might rely on data extracted from multiple databases that report information on individual clinical development programs (69,70) to allow comparisons of results obtained across different data sources.
Lastly, for pediatric-eligible trials, we could only determine if the reported minimum age for trial participation was younger than 18 years, but we did not have information on the actual age distribution of patients enrolled in each study. Thus, it is possible that the number of trials enrolling pediatric patients was lower than the number of pediatric-eligible trials, and we may have underestimated the time between start of clinical trials and testing in pediatric patients.
To characterize clinical development programs in oncology, we focused on endpoints relevant for US regulatory policies and decisions. Alternative endpoint definitions may also be considered, including adult or pediatric approvals from regulatory agencies outside the United States. Metrics of resource utilization or of new molecular entities potential clinical value (like the American Society of Clinical Oncology Clinical Value Framework) could also be investigated (71,72). To study these, data sources such as ours should be integrated with additional ones, such as public registries of international regulatory agencies (73).
In conclusion, our findings call for a renewed focus on improving preclinical research practices, evaluating and promoting more efficient clinical trial design strategies for the study of new anticancer products in adult and pediatric patients, and ensuring the availability of data from unsuccessful development programs. Analyses such as this should be conducted regularly to support data-driven approaches to optimize clinical development processes and assess the impact of new regulatory initiatives and development strategies.
Supplementary Material
Acknowledgements
Funds from the Harvard-MIT Center for Regulatory Science were used to purchase a subscription to AdisInsight, which was used to access and extract the data used in this study. While working on this study, AA was supported by a postdoctoral fellowship from the Harvard-MIT Center for Regulatory Science.
Partial results from this study were the subject of a poster presentation at the 2022 annual meeting of the American Society of Clinical Oncology (meeting abstract: J Clin Oncol. 2022;40(16 suppl): 1563-1563. doi: 10.1200/JCO.2022.40.16_suppl.1563).
Contributor Information
Andrea Arfè, Department of Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York, NY, USA.
Claire Narang, Pediatric Therapeutics and Regulatory Science Initiative, Computational Health Informatics Program (CHIP), Boston Children’s Hospital, Boston, MA, USA.
Steven G DuBois, Dana-Farber/Boston Children’s Cancer and Blood Disorders Center, Boston, MA, USA; Department of Pediatrics, Harvard Medical School, Boston, MA, USA.
Gregory Reaman, Center for Drug Evaluation and Research, US Food and Drug Administration, Silver Spring, MD, USA.
Florence T Bourgeois, Pediatric Therapeutics and Regulatory Science Initiative, Computational Health Informatics Program (CHIP), Boston Children’s Hospital, Boston, MA, USA; Department of Pediatrics, Harvard Medical School, Boston, MA, USA.
Data availability
Data for this analysis were made available to the authors through a third-party license from AdisInsight, a commercial database that is part of Springer Nature. In accordance with the corresponding data use agreement, the authors cannot deposit the data extracted for use in this publication in a publicly accessible database. Investigators may access the data by purchasing a license through AdisInsight. Interested individuals may refer to https://adisinsight.springer.com/ for more information on accessing this database.
Author contributions
Andrea Arfe, PhD (Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Project administration; Software; Validation; Visualization; Writing – original draft; Writing – review & editing), Claire Narang, BSc (Data curation; Investigation; Validation; Writing – original draft; Writing – review & editing), Steven G. DuBois, MD (Investigation; Writing – original draft; Writing – review & editing), Gregory Reaman, MD (Investigation; Writing – original draft; Writing – review & editing), Florence T. Bourgeois, MD, MPH (Conceptualization; Funding acquisition; Investigation; Methodology; Project administration; Resources; Supervision; Writing – original draft; Writing – review & editing)
Funding
This study was supported by funds from the Harvard-MIT Center for Regulatory Science, Harvard Medical School, Boston, MA (USA) and an Innovation in Regulatory Science Award from the Burroughs Wellcome Fund.
Conflicts of interest
AA, CN, GR, and FTB have no conflict of interests to disclose. SGD was a consultant for Amgen and Bayer and received research funding from Bayer, BMS, Curis, Eisai, Lilly, Loxo, Merck, Pfizer, and Turning Point Therapeutics. SGD also received research funding and travel/expenses support from Roche/Genentech Salarius Pharmaceuticals.
References
- 1. Attwood MM, Fabbro D, Sokolov AV, Knapp S, Schioth HB. Trends in kinase drug discovery: targets, indications and inhibitor design. Nat Rev Drug Discov. 2021;20(11):839-861. doi: 10.1038/s41573-021-00252-y. [DOI] [PubMed] [Google Scholar]
- 2. Vogelstein B, Papadopoulos N, Velculescu VE, Zhou S, Diaz LA Jr, Kinzler KW. Cancer genome landscapes. Science. 2013;339(6127):1546-1558. doi: 10.1126/science.1235122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Xin Yu J, Hodge JP, Oliva C, Neftelinov ST, Hubbard-Lucey VM, Tang J. Trends in clinical development for PD-1/PD-L1 inhibitors. Nat Rev Drug Discov. 2020;19(3):163-164. doi: 10.1038/d41573-019-00182-w. [DOI] [PubMed] [Google Scholar]
- 4. Gan HK, You B, Pond GR, Chen EX. Assumptions of expected benefits in randomized phase III trials evaluating systemic treatments for cancer. J Natl Cancer Inst. 2012;104(8):590-598. doi: 10.1093/jnci/djs141. [DOI] [PubMed] [Google Scholar]
- 5. Hwang TJ, Carpenter D, Lauffenburger JC, Wang B, Franklin JM, Kesselheim AS. Failure of investigational drugs in late-stage clinical development and publication of trial results. JAMA Intern Med. 2016;176(12):1826-1833. doi: 10.1001/jamainternmed.2016.6008. [DOI] [PubMed] [Google Scholar]
- 6. Gore L, Ivy SP, Balis FM, et al. Modernizing clinical trial eligibility: recommendations of the American Society of Clinical Oncology-Friends of Cancer Research Minimum Age Working Group. J Clin Oncol 2017;35(33):3781-3787. doi: 10.1200/Jco.2017.74.4144. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Neel DV, Shulman DS, DuBois SG. Timing of first-in-child trials of FDA-approved oncology drugs. Eur J Cancer. 2019;112:49-56. doi: 10.1016/j.ejca.2019.02.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Lim M, Shulman DS, Roberts H, et al. Off-label prescribing of targeted anticancer therapy at a large pediatric cancer center. Cancer Med. 2020;9(18):6658-6666. doi: 10.1002/cam4.3349. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Bourgeois FT, Kesselheim AS. Promoting pediatric drug research and labeling–outcomes of legislation. New Engl J Med. 2019;381(9):875-881. doi: 10.1056/NEJMhle1901265. [DOI] [PubMed] [Google Scholar]
- 10. Hwang TJ, Orenstein L, DuBois SG, Janeway KA, Bourgeois FT. Pediatric trials for cancer therapies with targets potentially relevant to pediatric cancers. J Natl Cancer Inst. 2020;112(3):224-228. doi: 10.1093/jnci/djz207. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Springer. AdisInsight. https://adisinsight.springer.com/. Accessed May 25, 2021.
- 12. Hwang TJ, Bourgeois FT, Franklin JM, Kesselheim AS. Impact of the priority review voucher program on drug development for rare pediatric diseases. Health Affair. 2019;38(2):313-319. doi: 10.1377/hlthaff.2018.05330. [DOI] [PubMed] [Google Scholar]
- 13. Collignon O, Koenig F, Koch A, et al. Adaptive designs in clinical trials: from scientific advice to marketing authorisation to the European Medicine Agency. Trials. 2018;19(1):642–655. doi: 10.1186/s13063-018-3012-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Laba AE, Ziolkowski P. Trends in glioblastoma treatment research: an analysis of clinical trials and literature. Neurol Neurochir Pol. 2021;55(3):269-280. doi: 10.5603/PJNNS.a2021.0024. [DOI] [PubMed] [Google Scholar]
- 15. Freire-Moran L, Aronsson B, Manz C, et al. ; for the ECDC-EMA Working Group. Critical shortage of new antibiotics in development against multidrug-resistant bacteria-time to react is now. Drug Resist Update. 2011;14(2):118-124. doi: 10.1016/j.drup.2011.02.003. [DOI] [PubMed] [Google Scholar]
- 16. World Health Organization. Health products in the pipeline from discovery to market launch for all diseases. 2023. https://www.who.int/observatories/global-observatory-on-health-research-and-development/monitoring/health-products-in-the-pipeline-from-discovery-to-market-launch-for-all-diseases. Accessed May 25, 2021.
- 17. US Food and Drug Administration. Drugs@FDA: FDA-approved drugs. https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm. Accessed May 25, 2021.
- 18. US Food and Drug Administration. Approved active moieties to which FDA has issued a written request for pediatric studies under section 505A of the Federal Food, Drug, and Cosmetic Act. https://www.fda.gov/drugs/development-resources/written-requests-issued. Accessed May 25, 2021.
- 19. US Food and Drug Administration. Reviews of pediatric studies conducted under BPCA and PREA from 2012 – present. https://www.fda.gov/drugs/development-resources/reviews-pediatric-studies-conducted-under-bpca-and-prea-2012-present. Accessed May 25, 2021.
- 20.US Food and Drug Administration. New drugs at FDA: CDER’s new molecular entities and new therapeutic biological products. https://www.fda.gov/drugs/development-approval-process-drugs/new-drugs-fda-cders-new-molecular-entities-and-new-therapeutic-biological-products. Accessed May 25, 2021.
- 21. Wong CH, Siah KW, Lo AW. Estimation of clinical trial success rates and related parameters. Biostatistics. 2019;20(2):273-286. doi: 10.1093/biostatistics/kxx069. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Austin PC, Lee DS, Fine JP. Introduction to the analysis of survival data in the presence of competing risks. Circulation. 2016;133(6):601-609. doi: 10.1161/CIRCULATIONAHA.115.017719. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing. https://www.R-project.org/.
- 24. Kola I, Landis J. Can the pharmaceutical industry reduce attrition rates? Nature Reviews Drug Discov. 2004;3(8):711-715. doi: 10.1038/nrd1470. [DOI] [PubMed] [Google Scholar]
- 25. DiMasi JA, Reichert JM, Feldman L, Malins A. Clinical approval success rates for investigational cancer drugs. Clin Pharmacol Ther. 2013;94(3):329-335. doi: 10.1038/clpt.2013.117. [DOI] [PubMed] [Google Scholar]
- 26. Hay M, Thomas DW, Craighead JL, Economides C, Rosenthal J. Clinical development success rates for investigational drugs. Nat Biotechnol. 2014;32(1):40-51. doi: 10.1038/nbt.2786. [DOI] [PubMed] [Google Scholar]
- 27. Haslam A, Olivier T, Powell K, Tuia J, Prasad V. Eventual success rate and predictors of success for oncology drugs tested in phase I trials. Int J Cancer. 2023;152(2):276-282. doi: 10.1002/ijc.34181. [DOI] [PubMed] [Google Scholar]
- 28. Oliviero E, Kourkopoulos G, Kimmelman J. Success rates for US and Canadian anticancer drug development efforts in pediatric oncology. Pediatr Blood Cancer. 2022;69(8):e29534. doi: 10.1002/pbc.29534. [DOI] [PubMed] [Google Scholar]
- 29. Fogel DB. Factors associated with clinical trials that fail and opportunities for improving the likelihood of success: a review. Contemp Clin Trials Commun. 2018;11:156-164. doi: 10.1016/j.conctc.2018.08.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Pan E, Bogumil D, Cortessis V, Yu S, Nieva J. A systematic review of the efficacy of preclinical models of lung cancer drugs. Front Oncol. 2020;10. doi: 10.3389/fonc.2020.00591. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Begley CG, Ellis LM. Drug development: raise standards for preclinical cancer research. Nature. 2012;483(7391):531-533. doi: 10.1038/483531a. [DOI] [PubMed] [Google Scholar]
- 32. Errington TM, Mathur M, Soderberg CK, et al. Investigating the replicability of preclinical cancer biology. Elife. 2021;10. doi: 10.7554/eLife.71601. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Atkins JT, George GC, Hess K, et al. Pre-clinical animal models are poor predictors of human toxicities in phase 1 oncology clinical trials. Br J Cancer. 2020;123(10):1496-1501. doi: 10.1038/s41416-020-01033-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Prasad V. Translation failure and medical reversal: two sides to the same coin. Eur J Cancer. 2016;52:197-200. doi: 10.1016/j.ejca.2015.08.024. [DOI] [PubMed] [Google Scholar]
- 35. Sertkaya A, Wong HH, Jessup A, Beleche T. Key cost drivers of pharmaceutical clinical trials in the United States. Clin Trials. 2016;13(2):117-126. doi: 10.1177/1740774515625964. [DOI] [PubMed] [Google Scholar]
- 36. Mishra-Kalyani PS, Amiri Kordestani L, Rivera DR, et al. External control arms in oncology: current use and future directions. Ann Oncol. 2022;33(4):376-383. doi: 10.1016/j.annonc.2021.12.015. [DOI] [PubMed] [Google Scholar]
- 37. Ventz S, Khozin S, Louv B, et al. The design and evaluation of hybrid controlled trials that leverage external data and randomization. Nat Commun. 2022;13(1):5783. doi: 10.1038/s41467-022-33192-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Ventz S, Lai A, Cloughesy TF, Wen PY, Trippa L, Alexander BM. Design and evaluation of an external control arm using prior clinical trials and real-world data. Clin Cancer Res. 2019;25(16):4993-5001. doi: 10.1158/1078-0432.Ccr-19-0820. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Psioda MA, Soukup M, Ibrahim JG. A practical Bayesian adaptive design incorporating data from historical controls. Stat Med. 2018;37(27):4054-4070. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Suissa S. Single-arm trials with historical controls: study designs to avoid time-related biases. Epidemiology. 2021;32(1):94-100. [DOI] [PubMed] [Google Scholar]
- 41. Seeger JD, Davis KJ, Iannacone MR, et al. Methods for external control groups for single arm trials or long-term uncontrolled extensions to randomized clinical trials. Pharmacoepidemiol Drug Saf. 2020;29(11):1382-1392. doi: 10.1002/pds.5141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Jiang L, Nie L, Yuan Y. Elastic priors to dynamically borrow information from historical data in clinical trials. Biometrics. 2023;79(1):49-60. doi: 10.1111/biom.13551. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Conforti F, Pala L, Bagnardi V, et al. Surrogacy of pathologic complete response in trials of neoadjuvant therapy for early breast cancer: critical analysis of strengths, weaknesses, and misinterpretations. JAMA Oncol. 2022;8(11):1668-1675. doi: 10.1001/jamaoncol.2022.3755. [DOI] [PubMed] [Google Scholar]
- 44. Prasad V, Kim C, Burotto M, Vandross A. The strength of association between surrogate end points and survival in oncology: a systematic review of trial-level meta-analyses. JAMA Intern Med. 2015;175(8):1389-1398. doi: 10.1001/jamainternmed.2015.2829. [DOI] [PubMed] [Google Scholar]
- 45. Jena AB, Zhang J, Lakdawalla DN. The trade-off between speed and safety in drug approvals. JAMA Oncology. 2017;3(11):1465-1466. doi: 10.1001/jamaoncol.2016.3337. [DOI] [PubMed] [Google Scholar]
- 46. Niewczas J, Kunz CU, König F. Interim analysis incorporating short- and long-term binary endpoints. Biom J. 2019;61(3):665-687. doi: 10.1002/bimj.201700281. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Heller G, McCormack R, Kheoh T, et al. Circulating tumor cell number as a response measure of prolonged survival for metastatic castration-resistant prostate cancer: a comparison with prostate-specific antigen across five randomized phase III clinical trials. J Clin Oncol. 2018;36(6):572-580. doi: 10.1200/jco.2017.75.2998. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Dasari A, Grothey A, Kopetz S. Circulating tumor DNA-defined minimal residual disease in solid tumors: opportunities to accelerate the development of adjuvant therapies. J Clin Oncol. 2018;36(35):JCO2018789032. doi: 10.1200/jco.2018.78.9032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49. Van Norman GA. Phase II trials in drug development and adaptive trial design. JACC Basic Transl Sci. 2019;4(3):428-437. doi: 10.1016/j.jacbts.2019.02.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Zhou H, Lee JJ, Yuan Y. BOP2: Bayesian optimal design for phase II clinical trials with simple and complex endpoints. Stat Med. 2017;36(21):3302-3314. doi: 10.1002/sim.7338. [DOI] [PubMed] [Google Scholar]
- 51. Zhang Q, Freidlin B, Korn EL, Halabi S, Mandrekar S, Dignam JJ. Comparison of futility monitoring guidelines using completed phase III oncology trials. Clin Trials. 2017;14(1):48-58. doi: 10.1177/1740774516666502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52. Angus DC, Alexander BM, Berry S, et al. Adaptive platform trials: definition, design, conduct and reporting considerations. Nat Rev Drug Discov. 2019;18(10):797-807. doi: 10.1038/s41573-019-0034-3. [DOI] [PubMed] [Google Scholar]
- 53. Saville BR, Berry SM. Efficiencies of platform clinical trials: a vision of the future. Clin Trials. 2016;13(3):358-366. doi: 10.1177/1740774515626362. [DOI] [PubMed] [Google Scholar]
- 54. Arfè A, Silverman LB, Bourgeois F. Master protocols and adaptive trial designs to develop tumor-agnostic drugs for children: essential tools in the era of the research to accelerate cures and equity act. JAMA Oncol. 2021;7(9):1281-1282. doi: 10.1001/jamaoncol.2021.1508. [DOI] [PubMed] [Google Scholar]
- 55. Simon R, Maitournam A. Evaluating the efficiency of targeted designs for randomized clinical trials. Clin Cancer Res. 2004;10(20):6759-6763. doi: 10.1158/1078-0432.CCR-04-0496. [DOI] [PubMed] [Google Scholar]
- 56. Freidlin B, McShane LM, Korn EL. Randomized clinical trials with biomarkers: design issues. J Natl Cancer Inst. 2010;102(3):152-160. doi: 10.1093/jnci/djp477. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57. London AJ, Kimmelman J. Why clinical translation cannot succeed without failure. Elife. 2015;4:e12844. doi: 10.7554/eLife.12844. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58. Chan AW, Song F, Vickers A, et al. Increasing value and reducing waste: addressing inaccessible research. Lancet. 2014;383(9913):257-266. doi: 10.1016/S0140-6736(13)62296-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59. Rogawski MA, Federoff HJ. Disclosure of clinical trial results when product development is abandoned. Sci Transl Med. 2011;3(102):102cm29. doi: 10.1126/scitranslmed.3002939. [DOI] [PubMed] [Google Scholar]
- 60. Hakala A, Kimmelman J, Carlisle B, Freeman G, Fergusson D. Accessibility of trial reports for drugs stalling in development: a systematic assessment of registered trials. BMJ. 2015;350:h1116. doi: 10.1136/bmj.h1116. [DOI] [PubMed] [Google Scholar]
- 61. Brown AS, Patel CJ. A standard database for drug repositioning. Sci Data. 2017;4:170029. doi: 10.1038/sdata.2017.29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 62. Rahman R, Ventz S, McDunn J, et al. Leveraging external data in the design and analysis of clinical trials in neuro-oncology. Lancet Oncol. 2021;22(10):e456-e465. doi: 10.1016/S1470-2045(21)00488-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 63. Akalu AY, Meng X, Reaman GH, Ma L, Yuan W, Ye J. A review of the experience with pediatric written requests issued for oncology drug products. Pediatr Blood Cancer. 2021;68(2):e28828. doi: 10.1002/pbc.28828. [DOI] [PubMed] [Google Scholar]
- 64. de Rojas T, Neven A, Terada M, et al. Access to clinical trials for adolescents and young adults with cancer: a meta-research analysis. JNCI Cancer Spectr. 2019;3(4):pkz057. doi: 10.1093/jncics/pkz057. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 65. US Food and Drug Administration. Considerations for the Inclusion of Adolescent Patients in Adult Oncology Clinical Trials Guidance for Industry. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/considerations-inclusion-adolescent-patients-adult-oncology-clinical-trials. Accessed May 25, 2021.
- 66. Noel GJ, Nelson RM, Bucci-Rechtweg C, et al. Inclusion of adolescents in adult clinical trials: report of the institute for advanced clinical trials for children’s pediatric innovation research forum. Ther Innov Regul Sci. 2021;55(4):773-778. doi: 10.1007/s43441-021-00283-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 67. Msaouel P, Hong AL, Mullen EA, et al. Updated recommendations on the diagnosis, management, and clinical trial eligibility criteria for patients with renal medullary carcinoma. Clin Genitourin Cancer. 2019;17(1):1-6. doi: 10.1016/j.clgc.2018.09.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 68. US Food and Drug Administration. Cancer Clinical Trial Eligibility Criteria: Minimum Age Considerations for Inclusion of Pediatric Patients - Guidance for Industry and IRBs. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/cancer-clinical-trial-eligibility-criteria-minimum-age-considerations-inclusion-pediatric-patients. Accessed May 25, 2021.
- 69. Tufts University Center for the Study of Drug Development. Tufts CSDD Database. https://csdd.tufts.edu/databases. Accessed May 25, 2021.
- 70. Citeline. Trialtrove. https://pharmaintelligence.informa.com/products-and-services/clinical-planning/trialtrove. Accessed May 25, 2021.
- 71. Hutchinson N, Carlisle B, Doussau A, et al. Patient participation in clinical trials of oncology drugs and biologics preceding approval by the US Food and Drug Administration. JAMA Netw Open. 2021;4(5):e2110456. doi: 10.1001/jamanetworkopen.2021.10456. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 72. Vokinger KN, Hwang TJ, Grischott T, et al. Prices and clinical benefit of cancer drugs in the USA and Europe: a cost-benefit analysis. Lancet Oncol. 2020;21(5):664-670. doi: 10.1016/S1470-2045(20)30139-X. [DOI] [PubMed] [Google Scholar]
- 73.European Medicines Agency. Medicines. https://www.ema.europa.eu/en/medicines. Accessed May 25, 2021.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data for this analysis were made available to the authors through a third-party license from AdisInsight, a commercial database that is part of Springer Nature. In accordance with the corresponding data use agreement, the authors cannot deposit the data extracted for use in this publication in a publicly accessible database. Investigators may access the data by purchasing a license through AdisInsight. Interested individuals may refer to https://adisinsight.springer.com/ for more information on accessing this database.