Portal de Periódicos da CAPES

Pragmatic randomized clinical trials: a proposal to enhance evaluation of new cancer therapies with early signs of exceptional activity

2016; Elsevier BV; Volume: 27; Issue: 7 Linguagem: Inglês

10.1093/annonc/mdw143

1569-8041

María Koehler, Erling T. Donnelly, D. Kalanovic, Ramzi Dagher, Mace L. Rothenberg,

Protein Degradation and Inhibitors

The number of regulatory approvals for oncology new molecular entities has risen rapidly over the past 5 years. This is likely due to our ability to design targeted agents, a consequence of our improved insight into the molecular basis of cancer. As a result, several new drugs have demonstrated exceptional clinical activity in small phase I or phase II trials. This has led to the increased use of regulatory mechanisms that allow initial approval on the basis of these early data and, more recently, the introduction of the breakthrough therapy designation (BTD) in the United States in 2012 and the anticipated piloting of Sakigake in Japan and PRIME in Europe in 2016 [1.Sherman R.E. Li J. Shapley S. et al.Expediting drug development—the FDA's new ‘breakthrough therapy’ designation.N Engl J Med. 2013; 369: 1877-1880Crossref PubMed Scopus (108) Google Scholar, 2.Reflection paper on a proposal to enhance early dialogue to facilitate accelerated assessment of priority medicines (PRIME).Eur Med Agency. 2015; (http://www.ema.europa.eu/ema (12 April 2016, date last accessed))Google Scholar]. An analysis of oncology regulatory approvals since 2011 reveals that ‘exceptional activity’ in phase I or phase II trials strongly predicts sustained efficacy in subsequent phase III studies (Table 1). For our proposal, ‘exceptional activity’ is defined as (i) an objective response rate (ORR) of ≥50% to a single agent (with relevant response durability and some complete responses) or (ii) a hazard ratio (HR) of ≤0.5 in an early randomized study (Table 1). Examples of drugs that showed ‘exceptional activity’ in phase I or phase II trials include ceritinib, ibrutinib, and palbociclib, which received BTD and subsequent accelerated approval in the United States, and crizotinib, which was granted accelerated approval in the pre-BTD era. However, in spite of the ‘exceptional activity’ leading to accelerated approval, full regulatory approval was contingent upon subsequent verification of clinical benefit in confirmatory phase III randomized controlled trials (RCTs).Table 1New chemical entity therapeutics, approved since January 2011, which demonstrated exceptional activityaExceptional activity defined as: (1) ORR ≥50% observed in a single-arm trial for a therapeutic used as a single agent; (2) HR ≤0.50 in a randomized phase II study. in phase I or phase II and their phase III confirmatory resultsTherapeuticIndicationPhase II resultsPhase III resultsBTDUS approval in indication testedEU approval in indication testedORR % (95% CI)bPhase II results are objective response rate (ORR) except where specified.PFS HR (95% CI)OS HR (95% CI)ORR ≥50% in single-arm phase I or phase II TAGRISSO (osimertinib)2L EGFRm NSCLC (1L T790M+)57 (50–64) for study 160 (54–68) for study 2Not yet availableYesAcceleratedNo ODOMZO (sonidegib)Basal cell carcinoma58 (45–70)cData are shown from the 200mg arm; however, the study was designed as a randomized trial of 200mg ODOMZO versus 800mg ODOMZO.Not applicableNoFullNormal ZYKADIA (ceritinib)2L ALK+ NSCLC54.6 (47–62)Not yet availableYesAcceleratedConditionalIMBRUVICA (ibrutinib)2L CLL58.3 (43–72)0.22 (0.15–0.32)0.43 (0.24–0.79)YesAccelerated; converted to fullNormal IMBRUVICA (ibrutinib)MCL65.8 (56–75)Not yet availableYesAcceleratedNormalGILOTRIF (afatinib)1L EGFRm+610.58 (0.43–0.78)–NoFullNormalTAFINLAR (dabrafenib)1L BRAF+ melanoma59 (48–70)0.33 (0.20–0.54)–NoFullNormal ICLUSIG (ponatinib)Resistant/intolerant CMLMCyR: 54 (48–60)Not yet availablePre-BTDAcceleratedNormalXALKORI (crizotinib)ALK+ NSCLC50 (42–59)61 (52–70)0.45 (0.35–0.60) for 1L0.49 (0.37–0.64) for 2L–Pre-BTDAccelerated; converted to fullConditional ADCETRIS (brentuximab vedotin)ALCL86 (77–95)Not yet availablePre-BTDAcceleratedConditionalADCETRIS (brentuximab vedotin)Hodgkin's lymphoma73 (65–83)0.57 (0.40–0.81)–Pre-BTDAccelerated; converted to fullConditionalZELBORAF (vemurafenib)BRAFm+ melanoma52 (43–61) for 2L0.26 (0.20–0.33) for 1L0.47 (0.35–0.62) for 1LPre-BTDFullNormalPFS HR ≤0.50 in randomized phase I or IIIBRANCE (palbociclib)1L ER+HER2- MBCPFS HR: 0.49 (0.32–0.75)0.42 (0.32–0.56)–YesAcceleratedNoTherapeutics shown in bold have confirmatory phase III results verifying exceptional activity observed in phase 1/phase II.1L, first-line therapy; 2L, second-line therapy; ALCL, anaplastic large cell lymphoma; ALK, anaplastic lymphoma kinase; BTD, breakthrough therapy designation; CLL, chronic lymphocytic leukemia; CML, chronic myeloid leukemia; EGFR: epidermal growth factor receptor; ER, estrogen receptor; HER, human epidermal growth factor receptor; MCyR, major cytogenetic response; MCL, mantle cell lymphoma; MBC, metastatic breast cancer; NSCLC, non-small-cell lung cancer; OS, overall survival; PFS, progression-free survival.a Exceptional activity defined as: (1) ORR ≥50% observed in a single-arm trial for a therapeutic used as a single agent; (2) HR ≤0.50 in a randomized phase II study.b Phase II results are objective response rate (ORR) except where specified.c Data are shown from the 200 mg arm; however, the study was designed as a randomized trial of 200 mg ODOMZO versus 800 mg ODOMZO. Open table in a new tab Therapeutics shown in bold have confirmatory phase III results verifying exceptional activity observed in phase 1/phase II. 1L, first-line therapy; 2L, second-line therapy; ALCL, anaplastic large cell lymphoma; ALK, anaplastic lymphoma kinase; BTD, breakthrough therapy designation; CLL, chronic lymphocytic leukemia; CML, chronic myeloid leukemia; EGFR: epidermal growth factor receptor; ER, estrogen receptor; HER, human epidermal growth factor receptor; MCyR, major cytogenetic response; MCL, mantle cell lymphoma; MBC, metastatic breast cancer; NSCLC, non-small-cell lung cancer; OS, overall survival; PFS, progression-free survival. Recently, there has been a growing interest in the systematic collection of real-world data to gain further insight into the impact of new drugs as they are introduced into clinical practice. This is exemplified by the new technology platforms CancerLinQ™ (by the American Society of Clinical Oncology) and Flatiron (by Flatiron Health, a health care information technology company). Furthermore, payers, physicians, and patients increasingly require more data on alternative comparators, specific subgroups, patient-reported outcomes, and long-term toxicities—data that may be more appropriately collected in real-world studies than in separate confirmatory trials. This reality raises a few questions:(i)When a drug demonstrates ‘exceptional activity’ and receives accelerated/conditional approval and/or BTD, are confirmatory phase III RCTs an optimal way to provide additional evidence to patients, physicians, regulators, and payers?(ii)For drugs with exceptional activity, are phase III RCTs necessary to generate additional indications in the same tumor if a different combination partner or a different line of therapy is studied?Intelligent flexibility is required for the approval of novel targeted drugs. A proposal for ‘Limited Approval’ of such drugs on the basis of limited initial clinical investigation was recently described in this very journal [3.Dhingra K. Oncology 2020: a drug development and approval paradigm.Ann Oncol. 2015; 26: 2347-2350Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar]. We recognize that our proposal described in this article is a substantial departure both from the current expectations of clinicians and from the requirements of regulators. Why change the current paradigm of confirmatory phase III RCTs?(i)Many patients prefer not to enroll in RCTs because of the requirement for additional testing and strict adherence to the trial protocol.(ii)Only 3%–7% of patients with cancer meet trial inclusion criteria [4.Transforming Clinical Research in the United States: Challenges and Opportunities: Workshop Summary. Institute of Medicine (US) Forum on Drug Discovery, Development, and Translation. Washington, DC: National Academies Press, 2010.Google Scholar]. Patients with co-morbidities are generally excluded from RCTs; thus trial populations are typically homogeneous and not representative of patients in the real world. Surveys of cancer RCTs show that participating patients are typically younger than 65 years (71%), female (84%), not African American (93%), and lack socioeconomic diversity [5.Murthy V.H. Krumholz H.M. Gross C.P. Participation in cancer clinical trials: race-, sex-, and age-based disparities.JAMA. 2004; 291: 2720-2726Crossref PubMed Scopus (1492) Google Scholar, 6.Javid S.H. Unger J.M. Gralow J.R. et al.A prospective analysis of the influence of older age on physician and patient decision-making when considering enrollment in breast cancer clinical trials (SWOG S0316).Oncologist. 2012; 17: 1180-1190Crossref PubMed Scopus (79) Google Scholar, 7.Unger J.M. Gralow J.R. Albain K.S. et al.Patient Income Level and Cancer Clinical Trial Participation: a prospective survey study.JAMA Oncol. 2016; 2: 137-139Crossref PubMed Scopus (79) Google Scholar]. Thus, extrapolation from RCTs to the real world is difficult and uncertainty remains about the risks/benefits of interventions in the most commonly treated populations.(iii)Many practices have limited financial and time resources to meet the stringent requirements of RCTs.(iv)Phase III RCT data are frequently supplemented with guidelines to help practitioners with treatment decisions. Guidelines are more comprehensive than just phase III RCT data, are based on different datasets that may not have been collected in a systematic or validated fashion, and are critical to determine treatment pathways and reimbursement.We propose that when novel oncology drugs meet certain criteria, specifically (i) demonstrate ‘exceptional activity’ in phase I or phase II studies and (ii) have received accelerated/conditional approval and/or BTD, a prospectively designed pragmatic randomized clinical trial (pRCT), pre-agreed with regulatory authorities, could provide sufficient evidence for verification of clinical benefit leading to full regulatory approval (Figure 1). Although pRCTs may be suitable for the assessment of any drug including those with limited activity, our proposal is that pRCTs be used instead of industry-sponsored pivotal trials specifically for drugs with early signs of exceptional activity and a strongly favorable benefit:risk ratio. We also propose that a pRCT be accepted for subsequent full approval of a different line of therapy or different combination in the same disease if a phase III RCT leading to the original approval in that disease was stopped at interim analysis because of exceptional efficacy (Table 2). A historical example is bortezomib, which demonstrated overwhelming efficacy compared with standard therapy, resulting in early stopping of trials at interim analysis.Table 2Registrational Phase III studies, since January 2011, which were stopped at interim analysis because of exceptional efficacyTherapeuticIndicationPhase I or phase II resultsPhase III results at interim analysisBTDUS approval in indication testedEU approval in indication testedORR % (95% CI)aPhase II results are ORR% except where specified.PFS HR (95% CI)OS HR (95% CI)OPDIVO (nivolumab)2L RCC20–0.73 (0.60–0.89)YesFullNot yetIBRANCE (palbociclib) + FASLODEX (fulvestrant)2L Hormone receptor-positive, HER2- MBC–0.422 (0.32–0.56)–YesNot yetNoOPDIVO (nivolumab)2L Non-squamous NSCLC18–0.73 (0.60–0.89)YesFullNot yetIMBRUVICA (ibrutinib) + TREANDA (bendamustine) + RITUXAN (rituximab)Refractory CLL930.20 (0.15–0.28)–NoNot yetNot yetKEYTRUDA (pembrolizumab)1L Melanoma34 (28–40)0.58 (0.47–0.72)0.69 (0.52–0.90)YesNot yetNormalGAZYVA (obinutuzumab)Refractory NHL430.55 (0.40–0.74)NoNot yetNot yetOPDIVO (nivolumab)2L Squamous NSCLC17.1 (11.0–24.7)–0.59 (0.44–0.79)YesFullNormalTAFINLAR (dabrafenib) + MEKINIST (trametinib)BRAFm+ melanoma76 (62–86)0.56 (0.46–0.69)0.69 (0.53–0.89)NoFullNormalIMBRUVICA (ibrutinib)2L CLL58.3 (43–72)0.22 (0.15–0.32)0.43 (0.24–0.79)YesFullNormalXTANDI (enzalutamide)1L Prostate cancer36–0.71 (0.60–0.84)NoFullNormalXTANDI (enzalutamide)2L Prostate cancer12–0.63 (0.53–0.75)Pre-BTDFullNormalZYDELIG (idelalisib) + RITUXAN (rituximab)Relapsed CLL810.18 (0.10–0.32)–YesNot yetNormalSTIVARGA (regorafenib)2L Colorectal cancerDCR: 74–0.77 (0.64–0.94)Pre-BTDFullNormalZYTIGA (abiraterone)1L Prostate cancer380.53 (0.45–0.62)0.75 (0.61–0.93)Pre-BTDFullNormalAFINITOR (everolimus)2L ER+, HER2- MBC–0.43 (0.35–0.54)–Pre-BTDFullNormal1L, first-line therapy; 2L, second-line therapy; BTD, breakthrough therapy designation; CLL, chronic lymphocytic leukemia; DCR, disease control rate; ER, estrogen receptor; HR, hazard ratio; HER, human epidermal growth factor receptor; MBC, metastatic breast cancer; NHL, non-Hodgkin lymphoma; NSCLC, non-small-cell lung cancer; ORR, objective response rate; OS, overall survival; PFS, progression-free survival; RCC, renal cell carcinoma.a Phase II results are ORR% except where specified. Open table in a new tab 1L, first-line therapy; 2L, second-line therapy; BTD, breakthrough therapy designation; CLL, chronic lymphocytic leukemia; DCR, disease control rate; ER, estrogen receptor; HR, hazard ratio; HER, human epidermal growth factor receptor; MBC, metastatic breast cancer; NHL, non-Hodgkin lymphoma; NSCLC, non-small-cell lung cancer; ORR, objective response rate; OS, overall survival; PFS, progression-free survival; RCC, renal cell carcinoma. pRCTs would be prospective clinical studies in which patients are randomized to two or more interventions and then treated and followed up according to the investigators' usual practice [8.Methodology Committee of the Patient-Centered Outcomes Research Institute (PCORI) Methodological standards and patient-centeredness in comparative effectiveness research: the PCORI perspective.JAMA. 2012; 307: 1636-1640Crossref PubMed Scopus (201) Google Scholar, 9.Jones P.B. Barnes T.R. Davies L. et al.Randomized controlled trial of the effect on quality of life of second- vs first-generation antipsychotic drugs in schizophrenia: Cost Utility of the Latest Antipsychotic Drugs in Schizophrenia Study (CUtLASS 1).Arch Gen Psychiatry. 2006; 63: 1079-1087Crossref PubMed Scopus (952) Google Scholar]. The Salford Lung Study was the first pRCT of a pre-licensed medication [10.Bakerly N.D. Woodcock A. New J.P. et al.The Salford Lung Study protocol: a pragmatic, randomised phase III real-world effectiveness trial in chronic obstructive pulmonary disease.Respir Res. 2015; 16: 101Crossref PubMed Scopus (32) Google Scholar, 11.Woodcock A. Bakerly N.D. New J.P. et al.The Salford Lung Study protocol: a pragmatic, randomised phase III real-world effectiveness trial in asthma.BMC Pulm Med. 2015; 15: 160Crossref PubMed Scopus (38) Google Scholar]. Table 3 shows the contrasting features between RCTs and pRCTs. pRCT studies, agreed with regulatory agencies at the time of accelerated/conditional approval, would be large enough to address safety questions and provide evidence of clinical benefit/value in the real-world population. With larger patient numbers, pre-agreed safety reporting, and enhanced reporting of postapproval safety events (compared with spontaneous reporting), pRCTs would provide a reasonable ability to detect rare adverse events that are unlikely to be captured in typical phase III RCTs. Safety data could be collected via electronic health records and follow-up calls. Formal study visits would only be required at baseline (consent and randomization) and at study end, with interim study visits as per standard practice and phone follow-up as needed. The financial and economic implications of pRCTs as used for regulatory purposes will have to be assessed on a case-by-case basis.Table 3Comparison of conventional RCT versus pRCTRandomized controlled trialPragmatic randomized clinical trialGeneral/strategic Purpose of studyHypothesis testing in well-defined study population;Regulatory approvalHealth Technology Assessment/policy decision making;Quality assurance;Reimbursement support;Comparative assessment researchProposed: Regulatory approval Target stakeholdersRegulatory authoritiesPayers, physicians, patients, guideline committeesProposed: Regulatory authorities Primary objectiveEfficacy and safety (versus specific comparator) in well-defined study population, aim to equalize non-specific effectsEffectiveness (benefit/risk) in real-world setting versus standard practice;Determine relative value of a novel treatment in a generalizable population, aim to optimize non-specific effects.Proposed: Efficacy and safety in confirmatory settings Impact on clinical practiceProvides initial and subsequent guidance in selected indication; frequently high impact on change in practiceHigh impact on physicians and patients due to a priori real-world study settingMethodological considerations Outcome measures, end pointsObjective efficacy parameters/surrogates (ORR, PFS, OS);Safety according to NCI CTCAE criteriaPractical impact of therapy on daily life, cost-effectivenessProposed: Objective efficacy parameters/surrogates (ORR, PFS, OS);Further patient-relevant outcomes (e.g. time-to-subsequent-therapy);Co-primary end points of PRO and efficacy;Safety according to NCI CTCAE criteria Blinding and placebo-controlDesirable; focus is on internal validityNot practical in real-world setting; focus is on external validity RandomizationIndividual patients are randomizedCluster/Center Level randomization preferred to allow for real-world practice within study centers;Individual patient's randomization possible PatientsHomogeneous; careful selection of patients to ensure well-defined study populationHeterogeneous; broad inclusion criteria including comorbidities;Physicians' judgment as in real-world situations Physicians/Study centers participatingLimited number of specialized/well-trained study centers, often organized in study networksLarger number of less specialized study centers, patient referral as in daily routineProposed: Same centers as for RCT as well as above (training in GCP necessary) Sample sizeDepends on hypothesis and end pointLarge sample size to account for heterogeneous population, less precise timing of follow-up, and individual or cluster randomization.Proposed: Size of the study agreed with respect to expected effect on primary end point;Primary end point may be hybrid of efficacy and patient-related outcomes Follow-upShort with clearly defined assessment intervalsStudy setting may offer more flexible and longer follow-up AdherenceNon-adherence can be a challenge in study conduct and a serious confounderFlexible treatment/diagnostic schedules lower adherence burden; Non-adherence and their causes may be part of the outcomes measured;Measurement of adherence may be part of a trialPractical considerations Data acquisitionComplex data acquisition, heavy follow-up, queries Validity and relevance to practiceHigh internal validity, lower external validity, frequently low relevance/impact on practice.High external validity, lower internal;High relevance/impact on practice. Drug utilizationDosing precisely defined by protocol criteria and adherence rulesRealistic assessment of drug utilization and use of health care resources (costs) Sponsor involvement in trial conductHigh, with standardized monitoring and study conduct in each phase of trialLow, sponsor mainly responsible for enabling real-world study setting;Commercial drug supplyProposed: sponsor provides funding for data collection and trial conduct;Sponsor and regulatory authorities agree a priori any specific safety monitoring criteria beyond GCP, possible added efficacy mandatory monitoring in selected studies CostsTo be evaluated case by case:Costs are driven by sample size, interventions, diagnostic and follow-up measures, recruitment dynamic, training effort of study centers etc.pRCT may have higher sample size, longer follow-up, higher recruitment rate, fewer number of trial specific procedures, most costs covered by routine medical care.pRCT potentially more efficient in answering reimbursement-relevant questions, i.e. better return-on-investment than (additional) conventional RCTCTCAE, common terminology criteria for adverse events; GCP, good clinical practice; NCI, National Cancer Institute; ORR, objective response rate; OS, overall survival; PFS, progression-free survival; RCT, randomized controlled trial; pRCT, pragmatic randomized clinical trials; PRO, patient-reported outcome. Open table in a new tab CTCAE, common terminology criteria for adverse events; GCP, good clinical practice; NCI, National Cancer Institute; ORR, objective response rate; OS, overall survival; PFS, progression-free survival; RCT, randomized controlled trial; pRCT, pragmatic randomized clinical trials; PRO, patient-reported outcome. How would pRCTs address the limitations of phase III RCTs?(i)Even though patients will be randomized in a pRCT, strict adherence to a protocol including ‘trial-prescribed’ visits to the clinic would not be necessary. These flexibilities would allow the assessment of the treatment during patients' routine care.(ii)pRCTs would be representative of real-world patients, regardless of comorbidities, age, race, or income level, and so allow the assessment of effectiveness in largely unselected populations.(iii)pRCTs would provide real-world data necessary for full regulatory approval and would also address questions of the new drug's value for reimbursement purposes. Data from electronic health records would provide long-term outcomes including health economics and would be free from interviewer or recall bias; such data could be supplemented by limited additional data gathering. Despite their benefits, pRCTs are not without limitations: (i) specific training of healthcare professionals would be required before recruitment; (ii) trial end points may be limited by routine clinical care and must be carefully selected for expected precision; (iii) bias is possible because most pRCTs would be open-label; (iv) independent review, data monitoring, and statistical analysis may be challenging; and (v) treatment switching/sequencing may dilute efficacy end points. The key objectives of pRCTs would remain enrollment of a sufficiently large population to support the conversion of initial approval to full regulatory approval, collection of data with minimal disruption to patient care, and appropriate safety monitoring to meet ethical and regulatory requirements. The sponsor would be expected to consult experts, the FDA/EMA/Japan PMDA, patient advocacy groups, and payers to select trial end points and other design features. An important hurdle remains the end point of the pRCT. Overall survival (OS) would be the simplest measurable outcome, especially in later-stage disease when subsequent therapies would unlikely affect OS. Duration of subsequent progression-free survival [12.Engelsberg A. Cross-over-it's a feature, not a bug.Ann Oncol. 2015; 26: 2000-2002Abstract Full Text Full Text PDF PubMed Scopus (1) Google Scholar] or time-to-subsequent treatment could also be end points that could be measured with sufficient precision. Regulatory agencies, payers, and biopharmaceutical companies could work in partnership to prospectively agree to the degree of preliminary efficacy and safety as well as to the type and size of an acceptable pRCT. If pRCTs were an alternative to Phase 3 RCTs for drugs with exceptional activity, let us look at an example of how pRCTs could provide sufficient confirmatory evidence for full or subsequent approval in the same tumor type. Bortezomib represents a pre-BTD era compound that revolutionized the treatment of multiple myeloma (MM). The initial open-label non-randomized phase II trial (n = 202) in relapsed/refractory MM showed an ORR of 35% with a median duration of 12 months and a median OS of 16 months [13.Richardson P.G. Barlogie B. Berenson J. et al.A phase 2 study of bortezomib in relapsed, refractory myeloma.N Engl J Med. 2003; 348: 2609-2617Crossref PubMed Scopus (2423) Google Scholar]. At the time of the initial approval in the United States, two trials were ongoing: a phase III study (n = 669) comparing monotherapy bortezomib with dexamethasone in recurrent disease [14.Richardson P.G. Sonneveld P. Schuster M.W. et al.Bortezomib or high-dose dexamethasone for relapsed multiple myeloma.N Engl J Med. 2005; 352: 2487-2498Crossref PubMed Scopus (2209) Google Scholar] and another trial (n = 682) comparing bortezomib plus melphalan and prednisone versus melphalan and prednisone in previously untreated patients with MM [15.San Miguel J.F. Schlag R. Khuageva N.K. et al.Bortezomib plus melphalan and prednisone for initial treatment of multiple myeloma.N Engl J Med. 2008; 359: 906-917Crossref PubMed Scopus (1632) Google Scholar]. Both studies were stopped at the time of pre-specified interim analyses due to exceptional efficacy and led to rapid full approval for the first-line and refractory conditions. However, data from these studies did not answer the value question in the real-world setting. In retrospect, a pRCT to support the first-line MM indication could have provided important data on a novel therapeutic in a real-world setting. The example of bortezomib suggests that a meaningful result in one randomized setting (bortezomib versus dexamethasone), in the context of impressive single-agent activity with compelling biological rationale, is likely to be repeated in a different setting (i.e. first-line therapy) and could be tested in a pRCT. In conclusion, we propose that for oncology drugs that demonstrate ‘exceptional activity’ in phase I or phase II trials and receive accelerated/conditional approval and/or BTD, and for certain expanded indications, regulatory authorities should consider accepting data from prospectively agreed pRCTs to grant full regulatory approval. These real-world pRCT trials would focus more on physicians' and patients' experience and outcomes, a dataset that is difficult and impractical to obtain in burdensome RCTs. The ‘exceptional activity’ would be defined as an ORR of ≥50% for a single agent or an HR of ≤0.5 in a randomized phase II study. We believe that pRCTs would offer sufficient precision to assess the efficacy and safety of new treatments and required evidence in situations where the preliminary data are adequately robust, the biological evidence and mechanism of action are strong, and the observed early benefit is convincing. Such an approach for drugs with exceptional efficacy could reshape the future of clinical trials, meet the demands for confirmatory studies, accelerate the availability of therapy to patients, and satisfy the demand by patients, physicians, and payers for value-based medical evidence. The authors thank Robert Glassman (Credit Suisse) for contribution to an early version of the manuscript, Prakash Naik (Pfizer) and Mary Skousen (Pfizer) for their help in analyzing prior approvals and registrational studies supporting the proposal, and Sashi Nadanaciva (Pfizer) for writing support. There is no grant or funding related to this work.