写在前面：今年3月份FDA出台了一份《支持加速批准肿瘤治疗药物的临床试验考虑因素行业指南》。考虑到癌症的严重性、亟需治疗的迫切性，加速审批途径通常用于肿瘤药物的审批。使用单臂试验来支持加速审批有其局限性，基于单臂试验的加速批准可能是不合理的。如果设计和执行得当，随机对照试验可以解决单臂试验的局限性。
故FDA建议申请人可以考虑进行进行随机对照试验，以支持加速批准（ A randomized controlled trial(“one-trial” approach) to support accelerated approval）。在适当的情况下，“单一/一次试验”方法（“one-trial” approach）的长期随访可以满足上市后验证临床益处的要求。这种“单一/一次试验”方法（“one-trial” approach）保持了药物开发的效率，并可以通过加速批准途径提供药物的早期准入，同时确保充分积累并顺利进行上市后试验，以及时验证长期效益。
本指南的目的是向抗癌药物或生物制品的申请人提供建议，说明设计试验以支持加速审批的考虑因素。
一般来说，FDA的指导文件并不具有法律上可执行的责任。相反，指导文件描述了该局目前对某一主题的想法，并应仅被视为建议，除非引用了具体的监管或法定要求。在机构指导文件中使用 "应该 "一词，意味着建议或推荐某事，但不是必须。
加速审批途径通常用于肿瘤药物的审批，部分原因是由于癌症的严重性和对生命的威胁性，以及由于现有的替代终点或中间临床终点被认为有可能预测临床益处。虽然各种试验设计和终点历来被用来支持加速批准，但单臂试验设计和反应终点（反应持续时间为支持性）在肿瘤学中最常被使用。反应率是药物活性的一个标志，因为恶性肿瘤通常不会自行消退，而且这个终点可以在单臂试验中解释为单药治疗的肿瘤药物方案。然而，使用单臂试验来支持加速审批有其局限性，包括但不限于以下几点：
● 安全性数据库通常很小，可能无法识别罕见的、潜在的严重不良事件。对于已确定的严重不良事件，对于已确定的严重不良事件，在缺乏对照组的情况下，将不良事件归因于所研究药物可能受到限制。
● 肿瘤学中常见的时间事件终点:包括（无进展生存期(progression-free survival,PFS)、无病生存期(disease-free surviva)通常是无法解释的，因为在将结果与外部对照比较时，未能考虑到已知和未知的混杂因素。FDA认为这种终点是探索性的，不足以作为旨在支持批准的单臂试验的疗效措施。
● 一般来说，低幅度的反应率可能无法合理地预测临床效益（如免疫疗法）。
● 对于联合治疗方案，个别成分对所声称的效果的贡献一般来说可能难以确定。
● 依靠与历史试验的交叉比较来评估所观察到的治疗效果是否比现有的治疗方法有所改善，是具有挑战性的。各项试验之间可能存在差异（例如，在设计、实施、反应评估间隔、研究人群等方面），这些差异可能很容易辨别，也可能不容易辨别，这可能导致对调查组和历史对照组之间反应估计的观察差异得出错误的结论（例如，错误地将反应率的差异归因于研究药物）。
单臂试验的这些和其他局限性会给药物的安全性和/或有效性评估增加不确定性，因此在特定的临床环境下，基于单臂试验的加速批准可能是不合理的。
如果设计和执行得当，随机对照试验可以解决单臂试验的局限性，包括但不限于以下几个方面：
随机对照试验提供了更有力的疗效和安全性评估，并允许与同时进行的对照组进行直接比较。
● 如果历史上的试验没有专门评估生物标志物选择的相关人群对标准护理治疗的反应率（即现有疗法被批准用于所有人群），在同一试验中评估新药与现有疗法的比较，可以更准确地反映生物标志物定义的患者群中标准护理的疗效和安全性。
● 在现有疗法的试验完成后，治疗情况可能发生了变化，随机对照试验可以对可比较的研究人群进行研究。
● 虽然支持加速批准的试验通常是在难治性疾病患者中进行的，但随机对照试验可以在较早的治疗环境中对新药进行评估，从而在更多患者可能受益的病程早期获得新药的机会。
● 当临床试验地点跨越几个地理区域时，如在国际上招募参与者的试验，随机对照试验可以评估可能来自多种因素的潜在区域差异。
进行随机对照试验以支持加速批准的另一个潜在优势是，在适当的情况下，同一试验的长期随访可以满足上市后验证临床益处的要求。这种“单一/一次试验”方法（“one-trial” approach）保持了药物开发的效率，并可以通过加速批准途径提供药物的早期准入，同时确保充分积累并顺利进行上市后试验，以及时验证长期效益。
鉴于单臂试验的局限性，随机对照试验是支持加速批准申请的首选方法。申请人可以酌情选择进行单一的随机对照试验，以支持加速批准并验证临床效益（即采用 "单一试验 "的方法），或者，可以进行单独的试验--一项支持加速批准，另一项是确证性试验，以验证临床效益。
虽然随机对照试验是首选的方法，但在开发加速批准的药物时，可能会出现适合进行单臂试验的情况，例如，当对随机对照试验的可行性存在重大关切时。在决定是否进行单臂试验时，应仔细考虑。
虽然随机对照试验是首选的方法，但在某些情况下，单臂试验在药物开发中是合适的，以加速批准，例如当对随机对照试验的可行性存在重大担忧时。在确定单臂试验是否适用于特定的临床和监管环境时，应仔细考虑。无论正在考虑的方法是什么，FDA建议在启动试验之前和适当的试验期间与FDA进行早期讨论。
申请人可以进行单独的随机对照试验--一项试验具有早期终点（如应答率），以支持药物的加速审批，另一项试验以长期临床终点（如无进展生存期（PFS）或总生存期（OS））为动力，验证临床获益。或者，申请人可以设计一个单一的随机对照试验来支持加速批准，该试验也为长期临床终点提供动力，并在同一试验中随访以验证临床效益（即 "单次试验 "方法）。以下是关于两个独立的随机对照临床试验的设计、实施和数据分析的建议，或者使用 "单次试验 "（one-trial）的方法来加速批准和验证临床益处。
1.两项随机对照临床试验的考虑
由于药物在临床实践中的可用性，在获得加速批准之前等待启动随机对照验证试验可能会在招募参与者方面带来挑战。因此，为了帮助确保旨在验证临床益处的试验的可行性和及时完成，FDA强烈建议在加速批准之前，如果没有完全注册，该试验应该顺利进行。
为了促进确证性临床的完成，在同一癌症类型但在另一个治疗线中评估该药物可能是可以接受的。例如，对于为难治性癌症的适应症批准的加速审批，确认性试验可以在较早的疾病环境中进行。这种方法有可能为患有早期疾病的患者提供有效的药物，受益可能更大，并且当一种药物已经获得后期适应症的加速批准时，它有助于患者的累积。
鉴于在加速批准时，药物的临床效益存在固有的和残留的不确定性，及时完成旨在验证临床效益的试验至关重要。在提交上市申请时，应进行确证性临床。
2. 支持加速审批和验证临床效益的单一随机对照试验的考虑因素
如果计划采用 "单次试验 "的方法，利用同一试验来支持加速审批，并进行长期随访以验证临床效益，那么申请人在启动试验前应仔细评估现有的初步临床数据。FDA建议为加速批准选择一个适当和可行的终点，以便在疾病的早期和进行试验的早期进行评估。在选择加速批准的终点时，申请人还应该考虑疾病的自然史（例如，惰性癌）、研究性药物的作用机制、可靠地描述可测量的疾病以评估反应的能力，以及其他特定情况的因素。
在评估 "单次试验 "方法的可行性和适当性时，维护试验的完整性至关重要，因为对数据的评估和随后对加速审批申请的监管行动可能无意中引入偏见。在评估偏倚的可能性时，申请人应考虑的因素包括：交叉的预期影响（如允许）；关于药物效果的初步数据，包括毒性情况、治疗情况以及对照组使用的治疗方法等。
在启动试验之前，申请人应考虑并与FDA讨论，根据现有的初步临床数据，对反应率或其他早期终点的预期影响是否有足够的规模，以合理地预测临床效益。根据病程、预期人群和FDA的指导，也可以用 "单次试验 "的方式对反应率以外的其他终点的使用进行评估，同时对临床获益终点进行后续评估。
如果药物开发计划旨在评估一个组合方案，申请人应明确说明证明每个组成部分的贡献的方法。应提供证据支持各成分对所声称的效果的单独贡献，这些证据通常来自多臂试验，并对无效性进行中期分析，或使用其他适应性试验设计元素。
申请人应仔细考虑试验结果是否足以支持提交申请。加速审批的一个要求是，药物必须证明对替代性终点或中间临床终点的影响，有理由预测临床效益，并提供比现有疗法有意义的优势。在评估是否满足这些要求时，FDA考虑的因素包括：在终点上显示的治疗效果的统计学意义和临床意义，其他支持观察到的效果可能预测临床益处的特定背景证据，以及对照组是否代表适当的可用疗法。
如果自试验开始以来，治疗情况发生了变化（例如，对照组的治疗不再反映最佳的可用疗法），应与FDA讨论关于提交加速批准申请与推迟提交申请的决定，直到有了支持传统批准的结果。归根结底，什么是可用的治疗方法是在监管做决定的时候，而不是在开始试验的时候确定的。
试验的设计、执行和分析应确保对疗效终点的有力评估。试验方案应规定一个计划，以有力地控制支持加速批准的终点和支持验证临床效益的终点的总体假阳性率（I型错误）。
试验样本量的选择应使其具有足够的能力来检测加速批准的终点（如应答率）和临床获益验证的终点（如PFS或OS）是否有临床意义和统计学意义上的改善。试验设计可以包含适应性设计元素（例如，样本量的重新估计）。对于适应性设计，申请人应根据臂间（between-arm comparisons）比较的背景考虑I型错误的控制，解决这种方法可能引起的操作问题，并在设计试验时将及时完成试验作为首要考虑。欲了解更多信息，请参考《药物和生物制品临床试验的行业适应性设计指南》（2019年12月）。
对于基于反应的终点，支持加速批准的分析可以基于预先指定的初始随机患者的数量，而对于时间到事件的终点，预先指定事件的数量是合适的；在每种情况下，申请人应确保在较早的分析时间点上进行稳健评估和可靠估计。支持加速审批的疗效分析应避免在试验接近或完全注册之前进行，以减轻加速审批时可能出现的挑战。确定总反应率（ORR）数据是否足以支持加速批准的一般考虑因素在下文B节中描述。
应采取措施，防止出现可能危及试验结果或试验完整性的情况。例如，对支持验证临床获益的终点数据应保持盲法，直到达到终点方案规定的分析时间点，以确保对该终点的有力评估。
在审查加速批准的申请时，FDA的安全评估可能包括评估现有数据是否表明研究组治疗的潜在危害（例如，对临床终点(如OS)的有害影响）。FDA可以要求生存数据分析的总结结果，以支持作为申请提交的一部分的评估，并可以在申请审查过程中要求更新的生存结果。申请人应指定一个计划，说明为这种分析保持研究盲法的措施。
如上所述，在特定的临床和监管背景下，单臂试验是否适合支持加速批准，应与FDA讨论。本节概述了设计、进行和分析旨在支持加速审批的单臂试验数据的考虑因素，以及确定数据是否足以实现这一目的的考虑因素。
1. 研究功效的考虑
终点：在肿瘤学中，当批准是基于单臂试验的数据时，反应率是支持加速批准最经常使用的终点。应使用适当的反应率评估标准（例如，基于实体瘤反应评估标准[RECIST]的ORR）。在某些疾病的情况下，除ORR外的其他反应措施可能更适合于描述疗效（如完全缓解率、主要分子反应等）。使用新的反应评估标准或修改已有的标准应该有强有力的基本理由支持，并应在试验设计阶段与FDA讨论。在可能的情况下，试验中使用的反应评估方法应与产品标签使用的方法相同。
现有疗法：加速批准是为那些有望提供比现有疗法更有意义的优势（包括疗效优势）的药物设置的。为了便于证明其优于现有疗法，申请人应预先指定作为比较基础的历史试验，以及选择试验的理由。试验的时间框架、试验规模、试验人群的临床和人口统计学特征，以及评估反应的任何潜在偏见，是评估历史试验的适用性时需要考虑的一些因素。FDA认识到，确定历史试验可能具有挑战性，特别是对于在分子定义的患者群体中开发的药物;在这种情况下，提供数据来证明分子定义亚组的治疗效果的量级优于历史试验可能是合适的。
样本量：单臂试验的规模应允许在点估计周围有足够的精度，提供反应持续时间的可靠性，并充分描述药物的不良事件概况。
2. 试验分析的考虑
当疗效终点为反应率时，支持加速批准的结果的充分性应基于缓解的程度和持续时间。申请人应考虑必要的随访时间，以充分表征特定疾病环境下的反应率和反应的持久性（例如，快速进展的疾病vs.惰性疾病）。在单臂试验中，没有必要用统计推断程序来评估这些端点。在大多数情况下，需要对大多数应答者在应答后至少随访6个月，以确定应答的持久性特征。然而，在某些情况下，可能需要在应答后进行更长时间的最低随访，以充分说明临床效益的特点。在某些情况下，FDA在审查申请时可能会要求提供有关反应持久性的额外数据。
应预先规定反应的试验样本量和分析人群。鉴于大多数单臂试验的规模较小，分析人群一般预计为整个试验人群。然后，已经接受了至少一剂研究药物的患者将被纳入分析人群，无论他们是否因随访时间短而有机会回应。在没有预先规定的计划的情况下，多次增加研究样本量，反复查看数据，可能会在疗效评估中引入偏差，应避免。
为了减少引入偏见的可能性，并减少评估反应的差异，应该对反应评估进行盲法独立中央审查（BICR）。
包括裁决程序的BICR章程应作为上市申请的一部分提供给FDA。
一般来说，在适当的临床背景下，FDA将反应率定义为部分反应和完全反应之和。当以这种方式定义时，反应是对药物抗肿瘤活性的直接衡量，可以在单臂研究中进行评估。稳定的疾病不应该是反应率的一个组成部分。同样，也不应该使用临床获益率（例如，反应率+稳定的疾病>6个月）等措施。这种措施在很大程度上可以反映疾病的自然史，而肿瘤的缩小则代表了直接的治疗效果。
对于获得加速批准的肿瘤学药物，上市后的确证性临床试验被要求验证和描述预期的临床效益。这种试验有助于解决替代或中间终点与最终临床效益之间关系的剩余不确定性。为了尽量减少这种不确定性的持续时间，FDA可能会酌情要求在批准适用产品之前，或在批准日期之后的特定时间段内，进行旨在验证临床益处的研究。上市后试验必须尽职尽责地进行，并符合FDA规定的上市后试验条件，其中可能包括招募目标、研究方案和里程碑，包括研究完成的目标日期。“单一试验 "（“one-trial” ）方法的优点是可能不需要单独的确认性试验。然而，当单臂试验支持加速批准时，FDA要求进行上市后试验以评估PFS或OS，可能需要进行单独的随机对照试验。建议尽早与FDA讨论旨在支持加速批准的试验和上市后试验的设计和启动，以便迅速提供临床效益的证据。
Clinical Trial Considerations to Support Accelerated Approval of Oncology Therapeutics Guidance for Industry
The purpose of this guidance is to provide recommendations to sponsors of anti-cancer drugs or biological products on considerations for designing trials intended to support accelerated approval.
In general, FDA’s guidance documents do not establish legally enforceable responsibilities. Instead, guidances describe the Agency’s current thinking on a topic and should be viewed only as recommendations, unless specific regulatory or statutory requirements are cited. The use of the word should in Agency guidances means that something is suggested or recommended, but not required.
The accelerated approval pathway is commonly used for approval of oncology drugs in part due to the serious and life-threatening nature of cancer and because of available surrogate or intermediate clinical endpoints considered reasonably likely to predict clinical benefit. While a variety of trial designs and endpoints have historically been used to support accelerated approval, single-arm trial designs and response endpoints (with duration of response as supportive) have most commonly been used in oncology. Response rate is a marker of drug activity because malignant tumors do not typically regress on their own, and because this endpoint can be interpreted in single-arm trials for monotherapy oncology drug regimens. However, there are limitations to the use of single-arm trials in support of accelerated approval, including but not limited to the following:
• Safety databases are typically small and may not allow for the identification of rare, potentially serious adverse events. For identified serious adverse events, attribution of adverse events to the drug under study can be limited in the absence of a comparator arm.
• Common time-to-event efficacy endpoints in oncology (e.g., tumor progression, survival) are generally uninterpretable due to failure to account for known and unknown confounding factors when comparing the results to an external control. FDA considers such endpoints exploratory and not adequate to be used as measures of efficacy in single arm trials intended to support approval.
• Low magnitude response rates generally may not be reasonably likely to predict clinical benefit (e.g., immunotherapy).
• For combination regimens, the contribution of the individual components to the claimed effect(s) generally may be challenging to establish.
• Reliance on cross-trial comparisons to historical trials to assess whether the observed treatment effect represents an improvement over available therapy is challenging. There can be differences across trials (e.g., in design, conduct, response assessment intervals, study population, etc.) which may or may not be easily discernible and which could lead to erroneous conclusions regarding observed differences in the response estimate between the investigational arm and a historical control (e.g., erroneously attributing differences in response rate to the investigational drug).
These and other limitations of single-arm trials can add uncertainty to the assessment of the safety and/or effectiveness of a drug such that accelerated approval based on a single-arm trial may not be justified in a given clinical setting.
When properly designed and executed, a randomized controlled trial can address the limitations of single-arm trials, including but not limited to, the following ways:
A randomized controlled trial provides a more robust efficacy and safety assessment and allows for direct comparison to a concurrent control arm.
In cases wherein historical trials did not specifically evaluate the response rate for the standard of care treatment in a biomarker-selected population of interest (i.e., available therapy is approved for an all-comer population), assessing the new drug compared to the available therapy in the same trial provides a more accurate representation of the efficacy and safety of standard of care in the biomarker-defined cohort of patients.
In settings wherein the treatment landscape may have changed since completion of the trial(s) for available therapy, a randomized controlled trial enables comparable study populations to be studied.
While trials that support accelerated approval have typically been conducted in patients with refractory disease, a randomized controlled trial may allow for the evaluation of a new drug in an earlier treatment setting, thereby enabling access to a new drug earlier in the course of the disease when more patients are likely to benefit.
When clinical trial sites span several geographic regions as would be the case for trials that enroll participants internationally, a randomized controlled trial allows for an assessment of potential regional differences that may stem from multiple factors.
Another potential advantage to conducting a randomized controlled trial to support accelerated approval is that, in appropriate cases, longer term follow-up in the same trial could fulfill a postmarketing requirement to verify clinical benefit. This “one-trial” approach maintains efficiency in drug development and can provide early access to a drug using the accelerated approval pathway, while ensuring that a postmarketing trial is fully accrued and well underway to verify longer term benefit in a timely fashion.
Given the limitations of single-arm trials, a randomized controlled trial is the preferred approach to support an application for accelerated approval. Sponsors can, as appropriate, elect to conduct a single randomized controlled trial to support an accelerated approval and to verify clinical benefit (i.e., follow a “one-trial” approach) or, they can conduct separate trials – one to support the accelerated approval and another, a confirmatory trial, to verify clinical benefit.
Although a randomized controlled trial is the preferred approach, there can be circumstances wherein a single-arm trial is appropriate in the development of a drug for accelerated approval, for example when there are significant concerns about the feasibility of a randomized controlled trial. Careful consideration should be taken in determining whether a single-arm trial is appropriate in a particular clinical and regulatory context. Regardless of the approach under consideration, FDA recommends early discussion with the Agency before initiating and, as appropriate, during the conduct of, a trial(s).
A. Randomized Controlled Clinical Trials to Support Accelerated Approval
Sponsors can conduct separate randomized controlled trials – one trial with an early endpoint (e.g., response rate) to support the accelerated approval of the drug and a second trial powered for a longer-term clinical endpoint (e.g., progression-free survival (PFS) or overall survival (OS)) to verify clinical benefit. Alternatively, sponsors could design a single randomized controlled trial to support accelerated approval, that is also powered for the longer-term clinical endpoint with follow-up in the same trial to verify clinical benefit (i.e., “one-trial” approach). Below are recommendations for addressing the design, conduct, and analyses of data for either two separate randomized controlled clinical trials or for using the “one-trial” approach for accelerated approval and to verify clinical benefit.
1.Considerations for Two Randomized Controlled Clinical Trials
• Waiting to initiate a randomized controlled confirmatory trial until after an accelerated approval has been granted can create challenges in enrolling participants due to the availability of the drug in clinical practice. Therefore, to help ensure the feasibility and timely completion of the trial intended to verify clinical benefit, FDA strongly recommends that this trial be well underway, if not fully enrolled, by the time of the accelerated approval action.
• To facilitate completion of the confirmatory trial, it may be acceptable to evaluate the drug in the same cancer type but in another line of therapy. For instance, for an accelerated approval granted for an indication in a refractory cancer setting, the confirmatory trial could be conducted in an earlier disease setting. This approach has the potential to provide access to effective drugs to patients with earlier-stage disease in which benefit may be greater, and it facilitates patient accrual when a drug has already received accelerated approval for a later-stage indication.
• Given the inherent and residual uncertainties regarding the clinical benefit of the drug at the time of accelerated approval, timely completion of the trial(s) intended to verify clinical benefit is critical. Confirmatory trials should be underway when the marketing application is submitted.
1. Considerations for a Single Randomized Controlled Trial to Support Accelerated Approval and to Verify Clinical Benefit
• If planning a “one-trial” approach that uses the same trial to potentially support 156 accelerated approval with longer term follow-up to verify clinical benefit, sponsors should carefully assess the available preliminary clinical data prior to initiating the trial. FDA recommends selection of an endpoint for accelerated approval that is appropriate and feasible to evaluate earlier in the disease and earlier during the conduct of the trial. Sponsors should also consider the natural history of the disease (e.g., indolent cancers), the mechanism of action of the investigational drug, the ability to reliably characterize measurable disease to assess response, and other context-specific factors in selecting the accelerated approval endpoint.
• Preserving the integrity of the trial is critical in assessing the feasibility and appropriateness of the “one-trial” approach because the evaluation of the data and subsequent regulatory action on an accelerated approval application may inadvertently introduce bias. In assessing the potential for bias, sponsors should consider factors such as the anticipated impact of crossover (if permitted); the preliminary data on the drug’s effects, including the toxicity profile, the treatment landscape, and the treatment used in the control arm, among other factors.
• Before initiating the trial, sponsors should consider and discuss with FDA whether based on the available preliminary clinical data, the expected effect on response rate or other early endpoint is of a sufficient magnitude to be reasonably likely to predict clinical benefit. Depending on the disease course, the intended population, and guidance from FDA, use of endpoints other than response rate could also be evaluated in a “one-trial” approach together with subsequent evaluation of clinical benefit endpoints.
• If the drug development program is intended to evaluate a combination regimen, sponsors should specify the approach for demonstrating the contribution of each component. Evidence should be provided to support the individual contribution of components to the claimed effect(s), which would generally come from multi-arm trials with interim analyses for futility or from the use of other adaptive trial design elements.
• Sponsors should carefully consider whether the results of the trial are adequate to support submission of an application. A requirement of accelerated approval is that the drug must demonstrate an effect on a surrogate endpoint or intermediate clinical endpoint that is reasonably likely to predict clinical benefit, and provide meaningful advantage over available therapy. Among the factors FDA considers in evaluating whether these requirements have been met are the statistical significance and clinical meaningfulness of the treatment effect demonstrated on the endpoint, other context-specific evidence supporting why the observed effect is likely to predict clinical benefit, and whether the control arm represents the appropriate available therapy.
• If the treatment landscape has evolved since initiation of the trial (e.g., the treatment on the control arm no longer reflects best available therapy), the decision regarding submission of an application for accelerated approval versus deferring submission of an application until the results to support traditional approval are available should be discussed with FDA. Ultimately, the determination of what constitutes available therapy is made at the time the regulatory decision is made rather than at the time the trial was initiated.
• The trial should be designed, executed, and analyzed in such a way as to ensure a robust assessment of the efficacy endpoints. The protocol should specify a plan to strongly control the overall false positive rate (type-I error) for the endpoint supporting accelerated approval and the endpoint supporting verification of clinical benefit.
• The trial sample size should be chosen so that it has adequate power to detect a clinically meaningful and statistically significant improvement in both the endpoints for accelerated approval (e.g., response rate) and verification of clinical benefit (e.g., PFS or OS). The trial design can incorporate adaptive design elements (e.g., sample size re-estimation). With an adaptive design, sponsors should consider the type I error control based on the context of the between-arm comparisons, address the operational issues that this approach may raise, and design the trial with timely completion of the trial as a paramount consideration. For additional information, refer to the guidance for industry Adaptive Designs for Clinical Trials of Drugs and Biologics (December 2019).
• For a response-based endpoint, the analysis to support accelerated approval could be based on a pre-specified number of initially randomized patients, while for a time-to-event endpoint, pre-specifying the number of events is appropriate; in each case, the sponsor should ensure a robust assessment and reliable estimation at the earlier analysis time point. Analyses of efficacy to support accelerated approval should be avoided until the trial is close to or fully enrolled to mitigate potential challenges in accrual if an accelerated approval is granted. General considerations for determining the adequacy of the overall response rate (ORR) data to support accelerated approval are described in Section B below.
• Measures should be in place to prevent circumstances that may jeopardize the trial results or trial integrity. For example, blinding of data for the endpoint supporting verification of clinical benefit should be maintained until the endpoint’s protocol-specified analysis time point is reached to ensure a robust assessment of this endpoint.
• In reviewing an application for accelerated approval, FDA’s safety assessment may include evaluating whether the available data suggest a potential for harm from treatment on the investigational arm (e.g., detrimental effects on clinical endpoints such as OS). FDA may request summary results of the analysis on survival data to support such an assessment as part of an application submission and may request updated survival results during the course of the review of the application. Sponsors should specify a plan that describes measures to maintain study blind for such an analysis.
B. Single-Arm Trials to Support Accelerated Approval
As described above, whether a single-arm trial is appropriate to support accelerated approval in a particular clinical and regulatory context should be discussed with FDA. This section outlines considerations for designing, conducting, and analyzing data from a single-arm trial intended to support accelerated approval when appropriate, and considerations for determining whether the data may be adequate for this purpose.
1. Study Efficacy Considerations
• Endpoints: In oncology, response rate is the most frequently used endpoint to support accelerated approval when the approval is based on data from single-arm trials. Appropriate criteria for assessing the response rate (e.g., ORR based on Response Evaluation Criteria in Solid Tumors [RECIST]19) should be used. In certain disease settings, measures of response other than ORR may be more appropriate to characterize efficacy (e.g., complete remission rate, major molecular response, etc.). Use of new response assessment criteria or modifications of established criteria should be supported by a strong underlying rationale and should be discussed with FDA at the trial design stage. Whenever possible, the method of assessing response used in the trial should be the same one used for product labeling.
• Available therapy: Accelerated approval is reserved for drugs that are expected to provide a meaningful advantage (including an efficacy advantage) over available treatment.To facilitate the demonstration of advantage over available therapies, sponsors should pre-specify the historical trial(s) that will serve as the basis for the comparison, and the rationale for the selected trial(s). The time frame for the trial(s), trial size, clinical and demographic characteristics of the trial population, and any potential bias in the assessment of response, are some of the factors to consider in evaluating the applicability of a historical trial. FDA recognizes that it may be challenging, particularly for drugs being developed in molecularly defined patient populations, to identify a historical trial; in such cases, it may be appropriate to provide data to demonstrate that the magnitude of the treatment effect in the molecularly defined subgroup is better than in the historical trial.
• Sample Size: A single-arm trial should be sized to permit adequate precision around the point estimate, provide robust estimation of the duration of response, and sufficiently describe the adverse event profile of the drug.
2. Trial Analysis Considerations
• When the efficacy endpoint is response rate, the adequacy of the result to support accelerated approval should be based on the magnitude and duration of response. Sponsors should consider the follow-up time necessary to adequately characterize the response rate and the durability of response in a particular disease setting (e.g., a rapidly progressing disease vs. an indolent disease). Statistical inferential procedures are not necessary to evaluate these endpoints in single-arm trials. In most cases, a minimum follow-up of six months after the response is needed for most of the responders to characterize durability of response. However, there may be instances where a longer minimum follow-up after response is necessary to adequately characterize clinical benefit. In some cases, FDA may request additional data on the durability of response during the review of an application.
• The trial sample size and analysis population for response should be pre-specified. Given the small size of most single-arm trials, the analysis population is generally expected to be the entire trial population. Patients who have received at least one dose of the study drug would then be included in the analysis population regardless of whether they have had the opportunity to respond due to short follow-up time. Multiple increases to the study sample size with repeated looks at the data in the absence of a pre-specified plan may introduce bias in the assessment of efficacy and should be avoided.
• To reduce the potential to introduce bias and to mitigate variance in the assessment of response, blinded independent central review (BICR) of the response assessment should be performed.
A BICR charter that includes procedures for adjudication should be made available to FDA as part of a marketing application.
• Generally, and in the appropriate clinical context, FDA has defined response rate as the sum of partial responses plus complete responses. When defined in this manner, response is a direct measure of a drug’s antitumor activity which can be evaluated in a single-arm study. Stable disease should not be a component of response rate. Likewise, measures such as clinical benefit rate (e.g., response rate + stable disease > 6 months) should not be used. Such measures can largely reflect the natural history of disease, whereas reduction in tumor size represents a direct therapeutic effect.
C. Confirmatory Trial Following Accelerated Approval
For drugs granted accelerated approval in oncology, postmarketing confirmatory trials have been required to verify and describe the anticipated clinical benefit. Such trials help address residual uncertainties regarding the relationship between the surrogate or intermediate endpoint to the ultimate clinical benefit. In order to minimize the duration of this uncertainty, FDA may require, as appropriate, that studies intended to verify clinical benefit be underway prior to approval, or within a specified time period after the date of approval, of the applicable product. Postmarketing trials must be carried out with due diligence, and in accordance with the postmarketing trial conditions specified by FDA, which may include enrollment targets, the study protocol, and milestones, including the target date of study completion. An advantage of the “one-trial” approach is that a separate confirmatory trial may not be necessary. However, when a single-arm trial supports the accelerated approval, and FDA requires a postmarketing trial to evaluate PFS or OS, a separate randomized controlled trial may be needed. Early discussions with FDA regarding the design and initiation of both the trial intended to support accelerated approval and the postmarketing trial are recommended to provide evidence of clinical benefit in an expeditious manner.
个人中心
我是园区
退出
