Alectinib (Alecensa; Hoffman-La Roche/Genentech USA, Inc.), an oral tyrosine kinase inhibitor (TKI) that targets anaplastic lymphoma kinase (ALK) and ret proto-oncogene kinase, is indicated for the treatment of patients with ALK-positive metastatic non-small cell lung cancer (NSCLC) as detected by a Food and Drug Administration (FDA)-approved test.¹ The drug was approved in the United States on December 11, 2015, and the accompanying U.S. Product Information (USPI) included the adverse reaction of anemia, without specific mention of acquired abnormal erythrocyte morphology or hemolytic anemia as was recently described.^{2, 3} We, therefore, investigated the finding of hemolytic anemia reported with alectinib to the U.S. Food and Drug Administration Adverse Event Reporting System (FAERS) and in the literature to determine if regulatory action was required.

We searched (June 4, 2021) for reports of alectinib submitted to FAERS, a passive surveillance system, between December 11, 2015, and June 3, 2021, with a Preferred Term included in the Standardized Medical Dictionary for Regulatory Activities Query of “hemolytic disorders.” We also searched PubMed and Embase databases (June 25, 2021) for reports of hemolytic anemia and alectinib.

We identified and manually reviewed 60 FAERS reports. From our literature search, we retrieved 38 publications. Five of these publications were also reported to FAERS and included the reports from Australia by Kuzich et al and Yuan et al.^{2, 3} The remaining 33 publications did not provide case-level information, nor did any study specifically assess alectinib-associated hemolytic anemia. We created a case series of reports meeting our prespecified case definition of hemolytic anemia. A category I case was defined as a report with a diagnosis of hemolytic anemia and a positive direct antiglobulin test or decreased haptoglobin, or at least two of the following: corrected reticulocyte count >2%, increased total and/or indirect bilirubin, increased lactate dehydrogenase not attributed to underlying malignancy, abnormal red blood cell morphology (e.g., schistocytes, spherocytes, bite cells), hemoglobinuria, increased plasma free hemoglobin. Reports of hemolytic anemia with alectinib that did not meet these criteria were considered category II cases with a lower level of diagnostic evidence. We applied the World Health Organization-Uppsala Monitoring Centre causality criteria to our cases, thereby requiring evidence of a temporal relationship of hemolytic anemia following exposure to alectinib. Reports that were duplicates, did not meet our case definition, or had an unlikely causal relationship were excluded. Descriptive statistics were calculated using SAS version 9.4 (Cary, NC, USA).

We identified 31 cases of hemolytic anemia and assessed causality as probable in 10 (32%) cases and possible in 21 (68%) cases (Table 1). All probable cases met our category I case definition. Among the 20 cases that included information on time to onset, the median time from start of alectinib to first episode of hemolytic anemia was 56 days, with 75% of cases occurring within 90 days (inclusive) of drug exposure. Among cases reporting direct antiglobulin test (DAT) results, testing was uniformly negative. Most cases reported alectinib discontinuation, interruption, and/or dose reduction in response to hemolysis. Twelve cases described a positive dechallenge, and six cases noted a positive rechallenge. In two positive rechallenge cases, hemolytic anemia reoccurred after 20 and 84 days.

Anemia is prevalent in cancer populations, particularly in those undergoing cancer-directed treatment, but drug-induced hemolytic anemia (DIHA) is rare, with an estimated incidence of one event per 1–2 million individuals.⁴ Anemia is a labeled adverse reaction for alectinib, reported in 62% of alectinib-treated patients with previously untreated ALK-positive metastatic NSCLC compared to 36% of crizotinib-treated patients.¹ Based on the temporal association of hemolytic anemia and alectinib, resolution or improvement of hemolysis parameters with drug interruption (dechallenge), and recurrence of hemolytic anemia with drug reexposure (rechallenge), our case series suggests a plausible drug-event relationship.

In their retrospective case series, Kuzich and colleagues identified 43 patients treated with alectinib at their institution during 2013–2019 and reported erythrocyte membrane abnormalities in 95% of 20 patients with peripheral blood smears available.² They identified anemia in a majority of patients after start of alectinib and were able to confirm absence of abnormal red cell morphology prior to start of alectinib in a subset of these individuals. The authors had limited laboratory data on formal hemolysis assessments, but among four patients with DAT results, all reported a negative DAT, similar to the findings in our cases series. In retrospect, indirect support for hemolysis is also suggested by findings in the phase 3 ALEX trial that compared alectinib to crizotinib in previously untreated ALK-positive NSCLC, whereby a higher frequency of anemia (20% vs. 5%) and hyperbilirubinemia (15% vs. 1%) was reported in alectinib-treated than crizotinib-treated patients, respectively.⁵ Notably, both TKIs are associated with hepatotoxicity, but contrary to the bilirubin pattern observed, there was a lower occurrence of increased alanine transaminase (15% vs. 30%) and increased aspartate transaminase (14% vs. 25%) among individuals treated with alectinib compared to crizotinib, respectively,⁵ suggesting that hyperbilirubinemia might not be attributed solely to hepatotoxicity.

DIHA is typically mediated by immunologic mechanisms and infrequently occurs through oxidative mechanisms.^{4, 6} Although DAT-negative results following alectinib exposure suggest a nonimmune mechanism, 5%–10% of autoimmune hemolytic anemias are DAT-negative due to low levels of immunoglobulin G (IgG) antibodies or complement, non-IgG antibodies, low-affinity antibodies removed from the red blood cell surface during washing procedures, assay sensitivity, or other factors.⁶ However, reduced eosin-5-maleimide staining described in alectinib-treated patients, an atypical finding in autoimmune hemolytic anemia, suggests that alectinib induces a nonimmune erythrocyte membranopathy.² Additional investigation is needed to better understand the mechanism(s) underlying alectinib-related hemolysis, which may also provide insights into at-risk populations.

Despite the limitations of passive surveillance systems, including variable report quality, underreporting, and duplicate reporting, we provide convincing evidence of an association of hemolytic anemia with alectinib exposure. To this end, the alectinib USPI was recently updated to include hemolytic anemia under Warnings and Precautions and recommendations for dose modifications.¹ Although at-risk individuals and mechanisms of hemolysis with alectinib are undefined, heightened awareness of the potential for this adverse drug reaction will enable early diagnosis and facilitate appropriate management.

Alectinib (Alecensa; Hoffman-La Roche/Genentech USA, Inc.) 是一种靶向间变性淋巴瘤激酶 (ALK) 和 ret 原癌基因激酶的口服酪氨酸激酶抑制剂 (TKI)，适用于治疗ALK阳性患者经食品和药物管理局 (FDA) 批准的测试检测到的转移性非小细胞肺癌 (NSCLC)。¹该药物于 2015 年 12 月 11 日在美国获得批准，随附的美国产品信息 (USPI) 包括贫血的不良反应，但没有具体提及最近描述的获得性红细胞形态异常或溶血性贫血。^2、3因此，我们调查了美国食品和药物管理局不良事件报告系统 (FAERS) 和文献中使用艾乐替尼报告的溶血性贫血的发现，以确定是否需要采取监管行动。

我们搜索了（2021 年 6 月 4 日）在 2015 年 12 月 11 日至 2021 年 6 月 3 日期间提交给被动监测系统 FAERS 的艾乐替尼报告，其首选术语包含在监管活动标准化医学词典中的“溶血性”查询障碍。” 我们还搜索了 PubMed 和 Embase 数据库（2021 年 6 月 25 日），以查找溶血性贫血和艾乐替尼的报告。

我们确定并手动审查了 60 份 FAERS 报告。从我们的文献检索中，我们检索到了 38 篇出版物。其中五篇出版物也向 FAERS 报告，其中包括来自澳大利亚的 Kuzich 等人和 Yuan 等人的报告。^2、3其余 33 篇出版物没有提供病例级信息，也没有任何研究专门评估艾乐替尼相关的溶血性贫血。我们创建了一系列病例报告，符合我们预先指定的溶血性贫血病例定义。I 类病例被定义为报告诊断为溶血性贫血和直接抗球蛋白试验阳性或结合珠蛋白降低，或以下至少两项：校正的网织红细胞计数 > 2%、总胆红素和/或间接胆红素升高、乳酸升高脱氢酶不归因于潜在的恶性肿瘤、红细胞形态异常（如裂红细胞、球形红细胞、咬细胞）、血红蛋白尿、血浆游离血红蛋白升高。不符合这些标准的艾乐替尼溶血性贫血报告被认为是诊断证据水平较低的 II 类病例。我们将世界卫生组织-乌普萨拉监测中心的因果关系标准应用于我们的病例，因此需要证据证明接触艾乐替尼后溶血性贫血的时间关系。重复的、不符合我们的病例定义或具有不太可能的因果关系的报告被排除在外。使用 SAS 版本 9.4 (Cary, NC, USA) 计算描述性统计。或有不太可能的因果关系被排除在外。使用 SAS 版本 9.4 (Cary, NC, USA) 计算描述性统计。或有不太可能的因果关系被排除在外。使用 SAS 版本 9.4 (Cary, NC, USA) 计算描述性统计。

我们确定了 31 例溶血性贫血病例，并评估了 10 例 (32%) 病例和 21 例 (68%) 病例可能的因果关系（表 1）。所有可能的病例都符合我们的第一类病例定义。在包含发病时间信息的 20 例病例中，从开始使用艾乐替尼到首次出现溶血性贫血的中位时间为 56 天，其中 75% 的病例发生在药物暴露的 90 天（含）内。在报告直接抗球蛋白检测 (DAT) 结果的病例中，检测结果一致为阴性。大多数病例报告艾乐替尼因溶血而停药、中断和/或减少剂量。12 例描述了积极的去挑战，6 例指出了积极的再挑战。在两个阳性再激发病例中，溶血性贫血在 20 天和 84 天后再次发生。

贫血在癌症人群中很普遍，特别是在接受癌症治疗的人群中，但药物性溶血性贫血 (DIHA) 很少见，估计每 1-2 百万人中发生一次事件。⁴贫血是艾乐替尼标记的不良反应，62% 的艾乐替尼治疗的既往未治疗的ALK阳性转移性非小细胞肺癌患者报告了贫血，而克唑替尼治疗的患者为 36%。¹基于溶血性贫血与艾乐替尼之间的时间关联、溶血参数的消退或改善与药物中断（解除挑战）以及溶血性贫血与药物再次暴露（再次挑战）的关系，我们的案例系列表明了一种似是而非的药物事件关系。

在他们的回顾性病例系列中，Kuzich 及其同事确定了 2013-2019 年期间在其机构接受艾乐替尼治疗的 43 名患者，并报告了 20 名外周血涂片可用患者中 95% 的红细胞膜异常。²他们发现大多数患者在开始使用艾乐替尼后出现贫血，并且能够在这些个体中确认在艾乐替尼开始之前没有异常红细胞形态。作者在正式溶血评估方面的实验室数据有限，但在四名 DAT 结果患者中，均报告 DAT 阴性，与我们病例系列中的发现相似。回想起来，3 期 ALEX 试验的结果也暗示了对溶血的间接支持，该试验将艾乐替尼与克唑替尼在先前未治疗的情况下进行了比较ALK阳性 NSCLC，据报道，在艾乐替尼治疗的患者中，贫血（20% 对 5%）和高胆红素血症（15% 对 1%）的发生率分别高于克唑替尼治疗的患者。⁵值得注意的是，两种 TKI 都与肝毒性相关，但与观察到的胆红素模式相反，在接受治疗的个体中，丙氨酸转氨酶升高（15% 对 30%）和天冬氨酸转氨酶升高（14% 对 25%）的发生率较低艾乐替尼与克唑替尼相比，⁵表明高胆红素血症可能不仅仅归因于肝毒性。

DIHA 通常由免疫机制介导，很少通过氧化机制发生。^{4, 6}尽管艾乐替尼暴露后的 DAT 阴性结果表明存在非免疫机制，但由于免疫球蛋白 G (IgG) 抗体或补体、非 IgG 抗体、低-在洗涤过程、检测灵敏度或其他因素期间从红细胞表面去除的亲和抗体。⁶然而，在艾乐替尼治疗的患者中描述的 eosin-5-maleimide 染色减少，这是自身免疫性溶血性贫血的非典型发现，这表明艾乐替尼诱导了非免疫性红细胞膜病。²需要进一步调查以更好地了解艾乐替尼相关溶血的潜在机制，这也可能为高危人群提供见解。

尽管被动监测系统存在局限性，包括报告质量参差不齐、报告不足和重复报告，但我们提供了溶血性贫血与艾乐替尼暴露相关的令人信服的证据。为此，最近更新了艾乐替尼 USPI，将溶血性贫血纳入警告和注意事项以及剂量调整建议下。¹虽然未定义艾乐替尼的高危个体和溶血机制，但提高对该药物不良反应可能性的认识将有助于早期诊断并促进适当的管理。