Sickle cell disease (SCD) is the most common inherited hematologic disorder in the United States (US).¹ Until 2017, hydroxyurea (HU) was the only approved disease modifying therapy. Three new drugs have been approved in the recent years: l-glutamine, voxelotor, and crizanlizumab.² Importantly, <200 individuals were treated with each medication at the approved dose prior to approval. Further, these clinical trials restricted the inclusion of people with SCD who had significant cardio renal disease. Thus, little is known about these potentially disease-modifying medications in severe SCD. One of these medications, voxelotor directly alters the red blood cell (RBC) by increasing the oxygen affinity of hemoglobin (Hb) through allosteric modification, thereby reducing the concentration of deoxygenated hemoglobin S (HbS), and preventing sickling.³ The phase II-III trial that led to the approval of voxelotor (HOPE trial), demonstrated that voxelotor increases Hb levels and reduces hemolysis in >60% of patients at the 1500 mg dose.⁴ The trial met the primary end point of a positive Hb response of >1 g/dl above baseline at week 24 but failed to meet any of the key secondary clinical endpoints that were initially required for drug approval. Common adverse events reported during the HOPE trial and the 72-week follow-up study included headache, diarrhea, and nausea.⁵

This is a 27-year-old Black woman with SCD (HbSS) complicated by frequent vaso-occlusive crisis (VOC), chronic kidney disease, pulmonary hypertension, chronic pain, systemic hypertension, and significant RBC allo-immunization with a history of delayed hemolytic transfusion reactions (DHTRs). Prior to starting voxelotor, she was on erythropoietin (EPO) 40 000 units (2–3 times/week) and HU 500 mg daily (low dose due to nausea). Her pre-voxelotor labs (baseline): Hb 7.5 g/dl, absolute reticulocyte count (ARC) 327/mm³, lactate dehydrogenase (LDH) 736 units/L, creatinine (Cr) 0.8 mg/dl, urine albumin-creatinine ratio 2760 mg/g, total bilirubin (T. bili) 1.3 mg/dl, brain natriuretic peptide (BNP) was 586 pg/ml, and Hb electrophoresis Hb S 92.9, Hb F 3.7, and Hb A2 3.4. Baseline vital signs (VS): oxygen saturation (SaO₂) 97% on room air, blood pressure (BP) 133/93, and heart rate (HR) of 94. Her most recent trans-thoracic echocardiogram (TTE) showed left ventricular ejection fraction (LVEF) 55–60%, a tricuspid regurgitant jet velocity (TRJV) of 2.95 m/s and right ventricular systolic pressure (RVSP) elevated at 40–50 mm Hg. To reduce the frequency of EPO injections and potentially improve her pulmonary hypertension, she was started on voxelotor 1000 mg daily (dose chosen to optimize her pill burden) while continuing HU and EPO injections. EPO was stopped on day 13 when Hb was 9.9 g/dl, LDH 460 units/L, ARC 21/mm³, T. bili 1.2 mg/dl, and O₂sat 98% on room air. On day 48, her Hb was 10.4 g/dl, T. bili 1.3 mg/dl, and CR 0.7 mg/dl.

Initial event (day 59): She presented to the SCD day hospital with acute VOC. She reported missing voxelotor for 3 days. She was in moderate distress (VS HR 129 bpm, BP 192/113 mm Hg, and O₂ Sat 83% on room air). Laboratories: Hb 5.9 g/dl, ARC 298/mm³, LDH 1014 units/L, Cr 1.3 mg/dl, T. Bili 1.4 mg/dl, and BNP 2777 pg/ml. Chest radiograph revealed mild pulmonary edema. She was admitted for management of acute heart failure, hypoxic respiratory failure, and pain control. The patient could not be transfused due to history of DHTR, so she was restarted on EPO 40 000 units daily in conjunction with voxelotor and treatment with intravenous iron. On day 65, her Hb began to increase but she remained hypoxic. Her TTE showed LVEF 40%, TRJV of 3.81 m/s, and RVSP of 73 mm Hg. Right heart catheterization revealed mean pulmonary artery pressure of 52 mm Hg with preserved cardiac index. She was not treated with vasopressor medications, hemodialysis, or transfusion but recovered with above noted interventions. At the time of discharge (day 74), her Hb was 9.2 g/dl, and her O₂sat was >93% on room air.

Following her discharge, she resumed EPO 40 000 units weekly until Hb >7 g/dl. The events leading to the previous hospitalization were discussed in detail including her abrupt discontinuation of voxelotor. She insisted she would remain diligent to the current regimen (HU 500 mg daily, voxelotor 1000 mg daily, and EPO weekly) and Table 1 shows consistent VS and labs for several weeks. She continued to require intermittent treatment for VOC (consistent with baseline).

Second event: On day 112, she re-presented to the SCD day hospital with VOC after missing 4 days of voxelotor and HU. She was noted to be lethargic, unable to stay alert despite significant pain. She was in acute distress (HR 124, RR 24, and SaO₂ of 60%) and was immediately started on oxygen (SaO₂ improved to 93% on 5 L by face mask). Laboratories: Hb 5.4 g/dl, ARC 134 cells/mm³, LDH 2093 units/L, Cr 1.5 mg/dl, BNP 1166 pg/ml, and elevated troponin 354 ng/L (no ST-T wave changes on electrocardiogram). She was admitted to the intensive care unit where she remained abnormally somnolent despite complaining of diffuse pain and she was treated with high flow oxygen as well as EPO 40 000 units daily, voxelotor, and iron. A repeat TTE re-demonstrated similar of elevated TRJV of 3.74 m/s and RVSP of 71 mm Hg. She was continued on HU, voxelotor, and EPO, and discharged once she recovered (day 125). Following discharge, she returned to her “on medication” baseline. She asked to continue voxelotor and was instructed that a subsequent event would require us to stop the medication. She has not had any further episodes in the 7 months follow-up period.

The limitations of a rare disease include the expedited approval of medications that are evaluated in small patient populations often representing the “healthiest” of patients. In the case of voxelotor, only 88 individuals had ever received the approved dose (1500 mg) at the time of approval. Furthermore, only 64 patients continued to receive the medication after the initial data cut-off date. Thus, while the medication was noted to be efficacious, only a very small number of people with SCD had received the approved dose for >24 weeks.

Although the need for disease modifying therapies outweighs the risks of these expedited approvals, additional post-approval monitoring should be mandated. In the case of SCD, many adults are followed by primary care physicians or community oncologists instead of by SCD specialists due to the paucity of adult-focused specialists. However, new medications can be prescribed by any provider regardless of SCD-specific knowledge or training.

This patient has severe, complicated SCD with few available treatment options. Voxelotor has not been systematically evaluated in patients with moderate–severe pulmonary hypertension or kidney disease. Thus, she was monitored weekly after initiation of voxelotor with laboratory and clinical assessments. Her response to therapy was positive and her visits were extended to monthly assessments. The abrupt (and accidental) discontinuation of therapy on two separate occasions led to acute worsening of anemia, increase in LDH, acute kidney injury, severe hypoxia, and heart failure. In both cases, the patient had taken a weekend vacation in which she had neglected to bring her medication with her as it was in a different location (pill box) than her other medications. There was no other cause of her acute decline outside of neglecting her voxelotor was identified and she was able to recover with resumption of medication (in conjunction with EPO and intravenous iron), further implicating its role in these acute events. While the timing of the events does not prove causality of the hyperhemolysis to the cessation of the medication (voxelotor), the repeated occurrences of these events enhance the potential for a causal (instead of coincidental) relationship.

The pathophysiologic mechanism underlying these events is unclear. Because voxelotor increases the Hb oxygen affinity, there have been concerns voiced about poor tissue oxygen extraction leading to cellular hypoxia⁶; but without in vivo validation. This was not thought to be the cause of these events. However, as the medication can increase some individuals' baseline Hb by 25%–30%⁵ we hypothesize that the abrupt cessation of voxelotor results in more available Hb that may hemolyze. We hypothesize that patients with poor cardiopulmonary reserve cannot tolerate the brisk hemolysis concomitant with a significant change in oxygen carrying capacity, resulting in multi-organ dysfunction. We believe this problem was not seen during the trial as it likely occurs only in high-risk patients with cardio-renal disease that precluded their inclusion. Similarly, the authors believe this rebound response would not occur in those individuals without significant, baseline cardiopulmonary complications of their SCD. Notably, this patient does have improved hematological assessments with reduced hemolysis when stable on the medication and is able to maintain normal oxygen saturation on room air (though this is a measure of Hb oxygenation and not tissue oxygenation). However, it is unclear if the benefit is worth the additional risk for this individual patient. While other medications are known to have substantial complications from their abrupt discontinuation, many are studied in larger populations with more data available detailing the risks of stopping the medication as well as the treatment in those scenarios. In this case, of a rare disease population, this information is not yet available but deserves further study.

We can draw several conclusions from these events. (1) Medications given expedited approval should be monitored more carefully due to the limited exposure during a clinical trial. (2) While it is not feasible to limit prescribing of medications such as voxelotor to only SCD specialists (given the paucity of available specialists), it is clear that more oversight is needed as well as additional education for prescribing physicians and patients on the potential risks and benefits of newly approved medications understanding the small population of individuals who have been previously exposed until sufficient post-approval evaluation can be completed. The paucity of adult SCD specialists is noted but prescribing physicians can partner with SCD specialists if needed. (3) Highest concern: Expensive medications such as voxelotor cannot be maintained hospital formularies. Hence, it is important for patients/families to bring their medication when admitted to the hospital to avoid abrupt discontinuation. Further, if their disease course worsens while being hospitalized, the patients may be diagnosed with a “severe vaso-occlusive crisis” without recognizing the potential impact of this medication's withdrawal. Thus, physicians treating individuals with SCD should be made aware of this potential complication. In addition, hospitals should have the means for providing the medication to prevent consequences of abrupt withdrawal.

镰状细胞病 (SCD) 是美国 (US) 最常见的遗传性血液病。¹直到 2017 年，羟基脲 (HU) 是唯一获批的疾病改善疗法。近年来已批准三种新药：l-谷氨酰胺、voxelotor和crizanlizumab。²重要的是，在批准之前，<200 人接受了每种药物在批准剂量下的治疗。此外，这些临床试验限制了患有严重心肾疾病的 SCD 患者的纳入。因此，对于严重 SCD 中这些潜在的疾病改善药物知之甚少。其中一种药物，voxelotor 通过变构修饰增加血红蛋白 (Hb) 的氧亲和力直接改变红细胞 (RBC)，从而降低脱氧血红蛋白 S (HbS) 的浓度，并防止镰状化。³导致 voxelotor 获批的 II-III 期试验（HOPE 试验）表明，在 1500 mg 剂量下，voxelotor 增加了 >60% 的患者的 Hb 水平并减少了溶血。⁴该试验在第 24 周达到了高于基线 > 1 g/dl 的阳性 Hb 反应的主要终点，但未能满足药物批准最初所需的任何关键次要临床终点。HOPE 试验和 72 周随访研究期间报告的常见不良事件包括头痛、腹泻和恶心。⁵

这是一名患有 SCD (HbSS) 的 27 岁黑人女性，并发频繁的血管闭塞危象 (VOC)、慢性肾病、肺动脉高压、慢性疼痛、全身性高血压和显着的红细胞同种异体免疫，并有延迟免疫史溶血性输血反应（DHTR）。在开始使用 voxelotor 之前，她服用促红细胞生成素 (EPO) 40 000 单位（每周 2-3 次）和每天 500 毫克的 HU（由于恶心而低剂量）。她的前体素实验室（基线）：Hb 7.5 g/dl，绝对网织红细胞计数 (ARC) 327/mm ³，乳酸脱氢酶（LDH）736 units/L，肌酐（Cr）0.8 mg/dl，尿白蛋白肌酐比2760 mg/g，总胆红素（T. bili）1.3 mg/dl，脑利钠肽（BNP）为586 pg/ml，以及 Hb 电泳 Hb S 92.9、Hb F 3.7 和 Hb A2 3.4。基线生命体征 (VS)：氧饱和度 (SaO ₂) 97% 在室内空气条件下，血压 (BP) 133/93，心率 (HR) 94。她最近的经胸超声心动图 (TTE) 显示左心室射血分数 (LVEF) 55-60%，三尖瓣反流射流速度 (TRJV) 为 2.95 m/s，右心室收缩压 (RVSP) 在 40-50 mm Hg 升高。为了减少 EPO 注射频率并可能改善她的肺动脉高压，她开始每天服用 1000 mg voxelotor（选择剂量以优化她的药丸负担），同时继续 HU 和 EPO 注射。在第 13 天，当 Hb 为 9.9 g/dl、LDH 460 单位/L、ARC 21/mm ³、T. bili 1.2 mg/dl 和 O ₂在室内空气中达到 98% 时停止 EPO。第 48 天，她的 Hb 为 10.4 g/dl，T. bili 为 1.3 mg/dl，CR 为 0.7 mg/dl。

初始事件（第 59 天）：她因急性 VOC 到 SCD 日间医院就诊。她报告失去了 3 天的 voxelotor。她处于中度窘迫状态（VS HR 129 bpm，BP 192/113 mm Hg，室内空气中 O ₂ Sat 83%）。实验室：Hb 5.9 g/dl，ARC 298/mm ³、LDH 1014 单位/L、Cr 1.3 mg/dl、T. Bili 1.4 mg/dl 和 BNP 2777 pg/ml。胸片显示轻度肺水肿。她因治疗急性心力衰竭、缺氧性呼吸衰竭和疼痛控制而入院。由于 DHTR 病史，该患者无法输血，因此她每天重新开始使用 EPO 40 000 单位，同时使用 voxelotor 和静脉铁剂治疗。第 65 天，她的 Hb 开始升高，但仍处于缺氧状态。她的 TTE 显示 LVEF 为 40%，TRJV 为 3.81 m/s，RVSP 为 73 mm Hg。右心导管检查显示平均肺动脉压为 52 mm Hg，心脏指数保持不变。她没有接受血管加压药物、血液透析或输血治疗，但通过上述干预措施康复。出院时（第 74 天），她的 Hb 为 9.2 g/dl，她的 O ₂坐在室内空气中> 93％。

出院后，她每周恢复 EPO 40 000 单位，直到 Hb >7 g/dl。详细讨论了导致先前住院的事件，包括她突然停止使用 voxelotor。她坚称她将继续坚持目前的治疗方案（每天 500 毫克 HU，每天 1000 毫克 voxelotor 和每周一次 EPO），表 1 显示了几周内一致的 VS 和实验室。她继续要求对 VOC 进行间歇性治疗（与基线一致）。

第二个事件：在第 112 天，她在错过了 4 天的 voxelotor 和 HU 后，再次因 VOC 到 SCD 日间医院就诊。她被注意到昏昏欲睡，尽管疼痛严重，但无法保持警觉。她处于急性窘迫状态（HR 124，RR 24，SaO ₂为 60%）并立即开始吸氧（面罩 5 L 后， SaO _{2提高到 93%）。}实验室：Hb 5.4 g/dl，ARC 134 细胞/mm ³，LDH 2093 单位/L，Cr 1.5 mg/dl，BNP 1166 pg/ml，肌钙蛋白升高 354 ng/L（心电图上无 ST-T 波变化）。她被送入重症监护病房，尽管抱怨弥漫性疼痛，她仍然异常嗜睡，她接受了高流量氧气以及每天 40 000 单位的 EPO、voxelotor 和铁剂的治疗。重复 TTE 再次证明 TRJV 升高 3.74 m/s 和 RVSP 升高 71 mm Hg。她继续使用 HU、voxelotor 和 EPO，并在她康复后（第 125 天）出院。出院后，她回到了“服药”基线。她要求继续使用voxelotor，并被告知后续事件将要求我们停止服药。在 7 个月的随访期间，她没有任何进一步的发作。

罕见病的局限性包括加快批准在通常代表“最健康”患者的小患者群体中评估的药物。就 voxelotor 而言，在批准时只有 88 人曾接受过批准的剂量（1500 毫克）。此外，在最初的数据截止日期之后，只有 64 名患者继续接受药物治疗。因此，虽然注意到该药物有效，但只有极少数 SCD 患者接受了批准的剂量超过 24 周。

尽管对疾病修饰疗法的需求超过了这些加速批准的风险，但应强制要求进行额外的批准后监测。在 SCD 的情况下，由于缺乏以成人为中心的专家，​​许多成年人被初级保健医生或社区肿瘤学家而不是 SCD 专家跟踪。但是，任何提供者都可以开新药，无论 SCD 的特定知识或培训如何。

该患者患有严重、复杂的 SCD，几乎没有可用的治疗选择。Voxelotor 尚未在中重度肺动脉高压或肾病患者中进行系统评估。因此，在开始使用 voxelotor 后每周对她进行监测，并进行实验室和临床评估。她对治疗的反应是积极的，她的访问扩展到每月评估。在两个不同的情况下突然（和意外）停止治疗导致贫血急性恶化、LDH 增加、急性肾损伤、严重缺氧和心力衰竭。在这两种情况下，患者都在周末度假，因为与其他药物位于不同的位置（药盒），因此她忽略了随身携带药物。除了忽略她的 voxelotor 之外，没有其他原因导致她的急性衰退，并且她能够通过恢复药物治疗（结合 EPO 和静脉铁剂）恢复，进一步暗示了它在这些急性事件中的作用。虽然事件的发生时间并不能证明高溶血与停药（voxelotor）的因果关系，但这些事件的反复发生增强了因果关系（而不是巧合）的可能性。

这些事件背后的病理生理机制尚不清楚。由于 voxelotor 增加了 Hb 氧亲和力，因此有人担心组织氧提取不良会导致细胞缺氧⁶；但没有体内验证。这被认为不是这些事件的原因。然而，由于药物可使某些人的基线 Hb 增加 25%–30% ⁵我们假设 voxelotor 的突然停止导致更多可用的 Hb 可能溶血。我们假设心肺储备较差的患者不能耐受伴随氧气承载能力显着变化的快速溶血，导致多器官功能障碍。我们认为这个问题在试验期间没有出现，因为它可能只发生在排除他们纳入的心肾疾病的高危患者身上。同样，作者认为这种反弹反应不会发生在那些没有显着的 SCD 基线心肺并发症的人身上。尤其，该患者的血液学评估确实有所改善，当药物稳定时溶血减少，并且能够在室内空气中维持正常的氧饱和度（尽管这是衡量 Hb 氧合而不是组织氧合的指标）。然而，目前尚不清楚该益处是否值得为该个体患者增加额外风险。虽然已知其他药物会因突然停药而产生严重并发症，但许多药物在更大的人群中进行了研究，有更多可用数据详细说明了停药的风险以及在这些情况下的治疗。在这种情况下，对于罕见病人群，该信息尚不可用，但值得进一步研究。虽然已知其他药物会因突然停药而产生严重并发症，但许多药物在更大的人群中进行了研究，有更多可用数据详细说明了停药的风险以及在这些情况下的治疗。在这种情况下，对于罕见病人群，该信息尚不可用，但值得进一步研究。虽然已知其他药物会因突然停药而产生严重并发症，但许多药物在更大的人群中进行了研究，有更多可用数据详细说明了停药的风险以及在这些情况下的治疗。在这种情况下，对于罕见病人群，该信息尚不可用，但值得进一步研究。

我们可以从这些事件中得出几个结论。(1) 由于临床试验期间暴露有限，应更仔细地监测给予加急批准的药物。(2) 虽然将 voxelotor 等药物的处方仅限于 SCD 专家是不可行的（鉴于缺乏可用的专家），但显然需要更多的监督以及对开处方的医生和患者进行额外的教育以了解潜在的了解新批准药物的风险和益处，了解之前接触过的少数个体，直到可以完成足够的批准后评估。注意到成人 SCD 专家的缺乏，但如果需要，处方医生可以与 SCD 专家合作。(3) 最受关注：voxelotor 等昂贵的药物无法保存在医院处方集中。因此，患者/家人在入院时带上他们的药物以避免突然停药是很重要的。此外，如果他们的病程在住院期间恶化，患者可能会被诊断为“严重的血管闭塞危象”，而没有意识到这种药物停药的潜在影响。因此，治疗 SCD 患者的医生应该意识到这种潜在的并发症。此外，医院应该有提供药物的手段，以防止突然停药的后果。如果他们的病程在住院期间恶化，患者可能会被诊断为“严重的血管闭塞危象”，而没有意识到停药的潜在影响。因此，治疗 SCD 患者的医生应该意识到这种潜在的并发症。此外，医院应该有提供药物的手段，以防止突然停药的后果。如果他们的病程在住院期间恶化，患者可能会被诊断为“严重的血管闭塞危象”，而没有意识到停药的潜在影响。因此，治疗 SCD 患者的医生应该意识到这种潜在的并发症。此外，医院应该有提供药物的手段，以防止突然停药的后果。