A recent study—ASC4FIRST¹—builds the case that the novel drug, asciminib, a BCR::ABL1 inhibitor, is superior to current tyrosine kinase inhibitors (TKIs) for the treatment of chronic phase (CP) chronic myeloid leukemia (CML). Some have even taken to social media to announce a post-imatinib era.

As the first TKI to be approved in oncology, imatinib has been a transformative oral anti-cancer drug, improving survival for patients with CP CML. Imatinib, and subsequent drugs—dasatinib, bosutinib, and nilotinib—have increased the life expectancy of patients diagnosed with CML to essentially the same length as the general population without CML.²

Will asciminib further improve upon existing TKIs? We consider this alongside three questions raised by the ASC4FIRST trial: does it establish superiority over second-generation TKIs (dasatinib, bosutinib, and nilotinib), does the improvement in major molecular milestones mean the drug will improve survival or quality of life, and what can we conclude about adverse effects in this open-label study?

A primary concern with the ASC4FIRST trial is its approach to comparing asciminib with other TKIs, as a combined entity. The trial's design included two primary comparisons: asciminib versus all TKIs (a combined group of imatinib and second-generation TKIs) and asciminib versus imatinib alone. However, a direct comparison between asciminib and second-generation TKIs was relegated to a “secondary objective” and “not compared […] as a primary objective.” This design choice raises critical questions about the validity and clinical relevance of the findings.

Combining imatinib and second-generation TKIs into a single control group undermines the distinct therapeutic profiles and efficacy of these drugs. It is well-established that second-generation TKIs, such as dasatinib and nilotinib, outperform imatinib in achieving significant molecular responses in CML patients.³ By lumping these agents together, the trial essentially sets up a comparison that is guaranteed to favor asciminib. This strategy, which we have called the use of “nested groups” as opposed to “adjacent groups,” is a common tactic in clinical trials which creates confusion about precisely which groups benefit or which comparisons are significant.^{4, 5} In this case it lacks clinical justification and can mislead stakeholders about the true efficacy of the investigational drug.

In the ASC4FIRST trial, the difference in the 48-week major molecular response (MMR) between asciminib and the combined TKI group (a nested group) was significant. Yet, the more relevant comparison—asciminib versus second-generation TKIs (omitting imatinib, an adjacent subgroup)—revealed no significant difference (66.0% vs. 57.8%, respectively).¹ This finding is crucial because it highlights that asciminib may not offer a substantial improvement over current second-generation TKIs. We have depicted this in Figure 1. Therefore, the trial's conclusion that asciminib is superior to imatinib does not mean it is superior to all available treatment options.

Another concern is the choice of the primary endpoint. The ASC4FIRST trial used the 48-week MMR as its primary measure of efficacy. While achieving MMR is an important milestone in the management of CML, its correlation with long-term clinical outcomes is not absolute.⁵ Molecular milestones like the 48-week MMR are often used in clinical trials due to their convenience and shorter timeline for assessment. However, these milestones are not definitive indicators of long-term survival or overall clinical benefit.

Data showing that switching or escalating therapy based on MMR improves outcomes like overall survival (OS) are lacking.⁶ MMR's importance originated from the IRIS study, which correlates MMR with progression-free survival and not OS or quality of life. Studies have not conclusively shown that MMR correlates with improved patient-centered outcomes. For instance, various analyses, such as those between second-generation TKIs and imatinib, revealed no significant difference in OS between patients who achieved MMR and those who did not.⁶

The superiority of second-generation TKIs is uncertain with respect to clinical outcomes, such as survival or quality of life. The National Comprehensive Cancer Network guidelines (version 1.2023) recommend second-generation TKIs as first-line therapy for patients with intermediate or high-risk Sokal or Euro scores. However, the reliance on Sokal and Euro scores for stratifying CML patients is problematic because these scores were developed from data on chemotherapy or interferon-alpha treatments and are not relevant in the TKI era. In the German CML Study IV there is no significant differences in cumulative incidence probabilities (CIPs) of death among different Sokol risk groups (Figure 3a in Pfirrmann et al.).⁷

In terms of efficacy, there is no evidence that the OS for second-generation TKIs is superior to imatinib. The ENESTnd study reported a 10-year OS of 88.3% for imatinib versus 90.3% for nilotinib (p = .40).⁸ Similarly, the DASISION study found 5-year OS rates of 90.0% for imatinib and 91.0% for dasatinib (p = .1192).^{9, 10} Furthermore, while treatment-free remission (TFR) rates might be high with second-generation TKIs, actual TFR rates from discontinuation trials are similar between imatinib, nilotinib, and dasatinib (approximately 50% for each).⁹ Considering the significantly higher cost, increased toxicity, and adverse effects leading to possibly higher treatment interruptions,⁸ we and others have argued that imatinib remains the preferred first-line treatment for all CML patients, regardless of their risk category.

While molecular milestones provide valuable insights, they should be interpreted within the broader context. This understanding is especially pertinent when evaluating new therapies that are poised to replace well-established treatments like imatinib.

The open-label nature of the ASC4FIRST trial introduces an additional concern. Participants might report side effects differently based on their knowledge of the treatment they are receiving, and researchers might unconsciously interpret data in a way that favors the investigational drug.

While open-label designs are sometimes necessary, especially in early-phase trials or when blinding is impractical, they also demand a critical evaluation of the reported outcomes. The ASC4FIRST trial's safety data, indicating fewer adverse events with asciminib compared with imatinib and second-generation TKIs, should be interpreted with caution. Without blinding, the potential for bias in adverse event reporting could be significant, and these findings should be corroborated with data from double-blind studies.

Double-blind trials, where neither the participants nor the researchers know who is receiving which treatment, are the gold standard for eliminating bias. They provide a more reliable assessment of both efficacy and safety, ensuring that the observed outcomes are attributable solely to the intervention. Therefore, future studies on asciminib should consider a double-blind design to validate the safety profile observed in the open-label ASC4FIRST trial.

The annual cost per patient for asciminib is nearly $300 000, while generic imatinib now can be obtained for less than $2000. The financial burden from choosing asciminib as the first-line therapy would be massive both for individual patients and health care systems.¹¹ Ideally, experts without financial ties to the company marketing asciminib are best capable of adjudicating the evidence.

The ASC4FIRST trial presents a view on the post-imatinib era, with promising data on asciminib as a new treatment for CML. However, the methodological concerns highlighted—nested group comparisons over adjacent groups, the validity of molecular milestones, and the open-label design—underscore we are nowhere close to moving on from imatinib. As the oncology community continues to explore new therapeutic options, it is imperative to ensure that study designs are robust and that endpoints are clinically meaningful. For the time being, we believe the era of imatinib is still here to stay.

最近的一项研究 ASC4FIRST¹ 建立了一个案例，即新药 asciminib（一种 BCR：：ABL1 抑制剂）在治疗慢性期 （CP） 慢性髓性白血病 （CML） 方面优于目前的酪氨酸激酶抑制剂 （TKI）。有些人甚至在社交媒体上宣布后伊马替尼时代。

作为首个获批用于肿瘤学的 TKI，伊马替尼已成为一种变革性的口服抗癌药物，可提高 CP CML 患者的生存率。伊马替尼和后续药物（达沙替尼、博舒替尼和尼洛替尼）将诊断为 CML 的患者的预期寿命延长至与未确诊 CML 的一般人群基本相同的长度。

asciminib 会进一步改善现有的 TKI 吗？我们将这一点与 ASC4FIRST 试验提出的三个问题一起考虑：它是否确立了优于第二代 TKI（达沙替尼、博舒替尼和尼洛替尼）的优势，主要分子里程碑的改善是否意味着该药物将提高生存率或生活质量，以及我们可以对这项开放标签研究中的不良反应得出什么结论？

ASC4FIRST 试验的一个主要问题是其将 asciminib 与其他 TKI 作为联合实体进行比较的方法。该试验的设计包括两个主要比较：asciminib 与所有 TKI （伊马替尼和第二代 TKI 的联合组） 以及 asciminib 与单独使用伊马替尼。然而，asciminib 和第二代 TKI 之间的直接比较被降级为“次要目标”和“未比较 [...]作为主要目标。这种设计选择引发了关于研究结果的有效性和临床相关性的关键问题。

将伊马替尼和第二代 TKI 联合到一个对照组中会破坏这些药物独特的治疗特征和疗效。众所周知，第二代 TKI，如达沙替尼和尼洛替尼，在 CML 患者中实现显著的分子反应方面优于伊马替尼。³ 通过将这些药物放在一起，该试验基本上建立了一个保证有利于 asciminib 的比较。这种策略，我们称之为使用 “嵌套组” 而不是 “相邻组”，是临床试验中的一种常见策略，它造成了对哪些组受益或哪些比较具有重要性的混淆。^4、5在这种情况下，它缺乏临床依据，并且可能会误导利益相关者关于研究药物的真正疗效。

在 ASC4FIRST 试验中，asciminib 与联合 TKI 组 （嵌套组） 之间 48 周主要分子反应 （MMR） 的差异显著。然而，更相关的比较——asciminib 与第二代 TKI（省略相邻亚组伊马替尼）——显示没有显著差异（分别为 66.0% 和 57.8%）。¹ 这一发现至关重要，因为它强调 asciminib 可能不会比目前的第二代 TKI 提供实质性改善。我们在图 1 中对此进行了描述。因此，该试验得出的结论是 asciminib 优于伊马替尼并不意味着它优于所有可用的治疗方案。

另一个问题是主要终点的选择。ASC4FIRST 试验使用 48 周 MMR 作为其疗效的主要衡量标准。虽然实现 MMR 是 CML 管理的一个重要里程碑，但它与长期临床结果的相关性并不是绝对的。⁵ 像 48 周 MMR 这样的分子里程碑经常用于临床试验，因为它们方便且评估时间短。然而，这些里程碑并不是长期生存率或整体临床获益的明确指标。

缺乏数据表明，基于 MMR 的转换或升级治疗可改善总生存期 （OS） 等结局。⁶ MMR 的重要性源于 IRIS 研究，该研究将 MMR 与无进展生存期相关，而不是 OS 或生活质量。研究尚未最终表明 MMR 与改善以患者为中心的结果相关。例如，各种分析，例如第二代 TKI 和伊马替尼之间的分析，显示达到 MMR 的患者和未达到 MMR 的患者之间的 OS 没有显著差异⁶。

第二代 TKI 在临床结局（例如生存率或生活质量）方面的优越性尚不确定。美国国家综合癌症网络指南（1.2023 版）建议将第二代 TKI 作为中度或高风险 Sokal 或 Euro 评分患者的一线治疗。然而，依赖 Sokal 和 Euro 评分对 CML 患者进行分层是有问题的，因为这些评分是根据化疗或干扰素-α 治疗的数据开发的，与 TKI 时代无关。在德国 CML 研究 IV 中，不同 Sokol 风险组之间的死亡累积发生率 （CIP） 没有显着差异（Pfirrmann 等人的图 3a）。⁷

在疗效方面，没有证据表明第二代 TKI 的 OS 优于伊马替尼。ENESTnd 研究报告称，伊马替尼的 10 年 OS 为 88.3%，而尼洛替尼为 90.3% （p = .40）。⁸ 同样，DASISION 研究发现伊马替尼的 5 年 OS 率为 90.0%，达沙替尼为 91.0% （p = .1192）。^9、10此外，虽然第二代 TKI 的无治疗缓解 （TFR） 率可能较高，但伊马替尼、尼洛替尼和达沙替尼之间停药试验的实际 TFR 率相似（各约 50%）。⁹ 考虑到成本明显升高、毒性增加和不良反应可能导致更高的治疗中断率，⁸ 我们和其他人认为，伊马替尼仍然是所有 CML 患者的首选一线治疗，无论其风险类别如何。

虽然分子里程碑提供了有价值的见解，但应在更广泛的背景下对其进行解释。在评估准备取代伊马替尼等成熟疗法的新疗法时，这种理解尤其相关。

ASC4FIRST 试验的开放标签性质引入了另一个问题。参与者可能会根据他们对所接受治疗的了解情况，以不同的方式报告副作用，并且研究人员可能会无意识地以有利于研究药物的方式解释数据。

虽然开放标签设计有时是必要的，尤其是在早期试验中或盲法不切实际时，但它们也需要对报告的结果进行批判性评估。ASC4FIRST 试验的安全性数据表明，与伊马替尼和第二代 TKI 相比，asciminib 的不良事件更少，应谨慎解释。如果不采用盲法，不良事件报告可能存在显著的偏倚，这些发现应与双盲研究的数据相印证。

双盲试验，参与者和研究人员都不知道谁在接受哪种治疗，是消除偏倚的黄金标准。它们对疗效和安全性提供了更可靠的评估，确保观察到的结局完全归因于干预。因此，未来关于 asciminib 的研究应考虑双盲设计，以验证在开放标签 ASC4FIRST 试验中观察到的安全性。

asciminib 的每位患者每年的成本接近 300 000 美元，而仿制药伊马替尼现在只需不到 2000 美元即可获得。选择 asciminib 作为一线治疗对个体患者和医疗保健系统来说都是巨大的经济负担。¹¹ 理想情况下，与销售 asciminib 的公司没有经济联系的专家最有能力裁决证据。

ASC4FIRST 试验提出了对后伊马替尼时代的看法，并提供了 asciminib 作为 CML 新疗法的有希望的数据。然而，强调的方法学问题——相邻组的嵌套组比较、分子里程碑的有效性和开放标签设计——强调我们离摆脱伊马替尼还很远。随着肿瘤学界不断探索新的治疗方案，必须确保研究设计稳健且终点具有临床意义。目前，我们相信伊马替尼的时代仍然存在。