Accelerated drug approvals and patient trust: impact of voxelotor and crizanlizumab for sickle cell disease Open Access

Sickle cell disease (SCD) is a life-threatening inherited hemoglobinopathy that affects >100 000 Americans and millions worldwide. SCD was first described in 1910 and identified in 1949 as the first molecular disease, but owing in large part to the effects of structural racism, there has been tremendous stagnation in the advancement of sickle cell research, including new therapies. Because of the historical origins of the sickle mutation, most individuals with SCD are of African ancestry and/or Latina/Latino ethnicity in both the United States and Europe. The effects of systemic, structural, and interpersonal racism on scientific advancement and prioritization within the health care system have arguably affected the SCD population more than any other.¹ Finally, over the past decade, there has been an overdue surge in new sickle cell therapeutic options and clinical trials.^2,3 However, this has included both successes and failures, highlighting the importance of clear and transparent messaging to foster trust between patients and the multidimensional health care community.

Owing to a lack of prioritization in the scientific or drug development arena, there were no disease-modifying therapies developed for SCD for >100 years from its initial description. Hydroxyurea, the first true disease-modifying pharmacologic therapy, was serendipitously discovered to induce fetal hemoglobin, demonstrating a significant reduction in pain and other SCD complications,^4-6 and was approved for SCD by the US Food and Drug Administration (FDA) in 1998 for adults (but not until 2017 for children at least 2 years old) and granted orphan designation for “sickle cell syndrome” by the European Medicines Agency (EMA) in 2003. It would not be until July 2017, almost 20 years later, that another treatment, l-glutamine, received FDA approval for SCD; this was followed shortly by the FDA approval of 2 additional agents in November 2019. Voxelotor (brand name Oxbryta, produced by Global Blood Therapeutics, now Pfizer Inc) is a small-molecule oral medication that binds hemoglobin and increases oxygen affinity, decreasing the deoxy-hemoglobin S (HbS) fraction that can participate in HbS polymerization.⁷ Crizanlizumab (brand name Adakveo, produced by Novartis AG), a humanized monoclonal antibody administered intravenously once monthly, binds to P-selectin and theoretically reduces the risk of vaso-occlusion.⁸ These approvals brought much excitement and enthusiasm to the SCD patient and provider community, delivering hope to where there has been none for far too long. Unfortunately, the approval of these new therapies has not had the clinical success initially hoped for. Here, we describe the FDA’s accelerated approval process using the experiences of crizanlizumab and voxelotor. We highlight the critical need to build and retain trust with patients to allow for maximal uptake and clinical impact of new SCD therapies that will allow for improved quality and length of life for our patient population.

The FDA’s Expedited Programs for Serious Conditions, issued during the HIV/AIDs epidemic in 1992, were designed to facilitate the development and review of drugs to address unmet need for life-threatening serious, and particularly rare, conditions with the intent to maintain rigorous standards for safety and effectiveness.⁹ A “serious condition” is defined as any disease state that causes significant morbidity and directly affects the daily functioning of the affected individual. A rare condition is defined as one that affects fewer than 200 000 persons in the United States. Thus, SCD qualifies as both a serious and rare condition. The FDA Guidance for Industry document states as rationale for expedited programs that “the regulations specifically recognize that patients and physicians are generally willing to accept greater risks and side effects from treatment of life-threatening and severely debilitating diseases than they would for other diseases.”¹⁰ 

The FDA offers 4 related but distinct designations through their expedited programs: (1) fast track designation, (2) breakthrough therapy designation, (3) accelerated approval, and (4) priority review designation (Table 1). For fast track designation, the drug or treatment may be at any stage of development to be granted this designation as long as it provides evidence of activity to address an unmet medical need. Thus, for drugs in early development, evidence of activity in a nonclinical model, a mechanistic rationale, or pharmacologic data may be adequate to meet criteria. For those drugs in later development, clinical data should of course demonstrate the potential to address the unmet need. In contrast, for breakthrough therapy designation, the drug must have strong preliminary clinical evidence of a treatment effect that is substantially greater than that of available therapies, typically from phase 1 or 2 trials. These data are from 1 of 2 study end point types: (1) a surrogate end point or biomarker, such as a decline in viral load in a drug that targets HIV and therefore predicts disease suppression, and (2) an intermediate clinical end point that measures an effect before the development of irreversible morbidity or mortality, such as a treatment for multiple sclerosis shown to be effective for 6 months, with a lack of data regarding long-term effectiveness. Not all drugs that receive breakthrough therapy designation will then be eligible for accelerated approval. For accelerated approval, a study drug must have an effect on either of the 2 end points listed earlier that is substantially improved over the available standard of care; this can also include significantly decreased side effects or efficacy in a subpopulation of a disease. As part of the accelerated approval process, the study sponsor must perform confirmatory trials “with due diligence”; however, until recently, the timing of the confirmatory trial has not been explicitly defined. The accelerated approval process is designed to facilitate early collaboration between study sponsors and the FDA, allowing for efficient trial design and using the efficacy results of phase 2 studies for subsequent more formalized approval. Under priority review designation, the FDA commits to review the marketing application for a study drug within 6 months compared with the average 10 months for the standard approval pathway. Finally, in a separate and sometimes concurrent process, drugs may receive orphan drug designation if they target a rare disease or condition; this allows for certain incentives, including tax credits for qualified clinical trials, exemptions from user fees, and the potential for 7 years of market exclusivity after drug approval. Each of these expedited approval processes has been used to help bring new therapies to the SCD market in the past several years.

The phase 3 HOPE trial was an international, multicenter double-blind placebo-controlled trial that compared voxelotor 1500 mg daily to 900 mg daily with placebo in individuals (age 12-65 years) with all SCD genotypes.⁷ The primary outcome was not a decrease in acute vaso-occlusive event (VOE), but rather a hemoglobin response, defined as at least a 1 g/dL increase in hemoglobin, by week 24. The annualized incidence of painful VOEs was a secondary end point. In the intention-to-treat analysis, 51% of individuals in the voxelotor 1500 mg group and 33% of the voxelotor 900 mg reached the primary biomarker end point of hemoglobin response compared with only 7% of the placebo group. The mean increase in hemoglobin (1.1 g/dL) for those treated with 1500 mg was statistically significant compared with placebo (−0.1 g/dL). Importantly, the annualized VOE rate did not differ significantly among the 3 groups. There were also no reported significant differences in the number of non-SCD- or SCD-related adverse events among the 3 groups. The FDA granted accelerated approval on 25 November 2019, based specifically on the surrogate clinical end point of the increase in hemoglobin concentration. The FDA noted that continued approval “may be contingent upon verification and description of clinical benefit in confirmatory trial(s),”¹¹ but no clear details were provided on the specific confirmatory studies.

The SUSTAIN trial was a phase 2, multicenter, randomized, double-blind, placebo-controlled study to assess safety and efficacy of crizanlizumab compared with placebo. The study enrolled individuals (ages 16-65 years) with all SCD genotypes and randomized them to receive high-dose crizanlizumab (5 mg/kg), low-dose crizanlizumab (2.5 mg/kg), or placebo.⁸ The primary efficacy end point was the annualized rate of acute VOE at the end of 52 weeks of treatment. In the intention-to-treat analysis, there was a 45% lower VOE rate in the high-dose crizanlizumab group than placebo (1.6 VOE per year vs 3.0 VOE per year; P = .01). No difference was observed between the low-dose crizanlizumab group and placebo (2.0 vs 3.0 VOE per year; P = .18). The number of adverse events did not differ significantly among the 3 groups. Crizanlizumab received accelerated FDA approval on 15 November 2019. Similarly, no details were provided on the expected confirmatory studies for continued approval.

Unexpectedly for both the SCD patient and provider community, on 5 September 2024, through a press release published at 5 PM Eastern Standard Time (released overnight for patients and providers in Europe), Pfizer voluntarily withdrew voxelotor from global markets and discontinued all voxelotor clinical trials worldwide. The decision was based upon “the totality of clinical data that now indicates the overall benefit…no longer outweighs the risks.” Further details were vague, limited to new data suggesting an “imbalance in vaso-occlusive crises and fatal events which require further assessment.”¹² The EMA had begun a review of voxelotor in July 2024, after data emerged from 2 international phase 3 studies, GBT440-032 and GBT440-042, showing a greater number of deaths in participants receiving voxelotor than placebo.¹³ GBT440-032 evaluated the effect of voxelotor on transcranial doppler velocities in children aged 2 to 15 years. Of 236 children recruited from multinational sites, 8 children in the voxelotor arm died compared with 2 in the placebo arm. GBT440-042 assessed the effect of voxelotor on leg ulcers in individuals with SCD at least 12 years old in Nigeria, Kenya, and Brazil. Of 88 patients enrolled in the open-label extension trial, 8 died.¹³ These and other studies were subsequently paused. However, limited information was disseminated to the general public, including both patients and providers using this medication clinically. The EMA recommended suspension of the market authorization of voxelotor the day after Pfizer pulled the drug from the market, based on both the safety concerns and data emerging from registry studies showing higher rates of VOE in individuals after starting voxelotor.¹⁴ The FDA has yet to formally revoke approval of voxelotor with a press release on 26 September 2024, stating that they are conducting a safety review and will communicate any additional findings, if necessary. The number of patients on voxelotor at the time of its withdrawal has not been made public.

It remains unclear whether the deaths reported in the abovementioned trials were related to treatment with voxelotor, although at least half were complicated by infection, including malaria. However, speculation regarding the drug’s safety related to the mechanism of action had been broached as early as 2017, even before the completion of the HOPE trial.¹⁵ Voxelotor functions by binding the high-affinity oxygenated conformation (R) of HbS, thereby preventing the change to the low-affinity deoxygenated form (T) that is susceptible to polymerization. However, as Henry et al¹⁶ demonstrated via simulations of in vivo conditions, red cells treated with voxelotor only deliver more oxygen to tissues than untreated cells at the lowest partial pressures of oxygen. There were concerns that this higher oxygen affinity could paradoxically impede oxygen delivery to tissues and, therefore, has no effect on reducing pain or end organ damage.¹⁶ This concern is especially relevant to the cerebrovascular system in individuals with SCD,¹⁵ where increased oxygen delivery is a critical compensatory mechanism for the chronic hypoxemia and lower total hemoglobin concentration.^17,18 Ultimately, though, these explanations are only speculative; we await the additional release of information on the actual causes of death in the voxelotor arms and the reason for the drug’s withdrawal.

In October 2020, less than a year after the FDA's accelerated approval, Novartis released a statement warning that, of 3500 patients treated, postmarketing reports demonstrated 22 cases of the onset of severe pain within 24 hours of the first or second infusion.¹⁹ It was not clear at the time whether these reactions were immediate, immunoglobulin E– or immunoglobulin G–mediated hypersensitivity reactions²⁰ or a complement activation–related pseudoallergy.²¹ The painful event, which included musculoskeletal pain in the extremities and back, typically resolved within 3 days.¹⁹ In response, Novartis initially created a website for the reporting and publication of updated safety information, but this site has since been inactivated. These painful events and other infusion-related reactions are now recognized as potential side effects of crizanlizumab therapy and limit its use clinically for many patients.²² 

More recently, in January 2023, Novartis released a statement regarding preliminary results of the multinational phase 3 STAND trial, showing no significant difference in the annualized rate of VOE in crizanlizumab 5 mg/kg (2.5 VOE per year) or 7.5 mg/kg (1.9 VOE per year) compared with placebo (2.1 VOE per year) (P > .99).²³ Based on these data, the EMA’s Committee for Medicinal Products for Human Use recommended revoking the conditional marketing authorization (equivalent to the FDA’s accelerated approval) for crizanlizumab; these recommendations were adopted by the European Commission in August 2023, and crizanlizumab was subsequently removed from European markets; the number of patients on the drug at the time of withdrawal has not been disclosed. Crizanlizumab seems to have some efficacy at reducing the frequency of priapism in men with SCD given that the primary analysis of the SPARTAN trial, released in December 2023, showed a 46% reduction in the number of priapic events per patient from baseline after 26 weeks of therapy.²⁴ However, the full analysis of the 52-week trial period is still pending. Crizanlizumab remains available in the United States with no changes to FDA designation.

Recognizing the inequity in research funding and therapeutic discovery in the over 100 years since SCD was first described, the FDA, under the fifth reauthorization of the Prescription Drug User Fee Act in July 2012, committed to Patient-Focused Drug Development and selected 20 diseases to focus on over 5 years (2013-2017) as a means to receive direct patient input on available treatments. SCD was 1 of these 20 diseases, leading to a race by pharmaceutical companies to develop novel therapeutics for SCD. The results of this were the FDA-approved medications that came in rapid succession: l-glutamine, voxelotor, and crizanlizumab. However, it is worth noting that, despite being selected for this program, SCD continued to receive a third of the federal funding through 2018 of cystic fibrosis, which affects ∼1 in 2500 White individuals in the United States.^25,26 This was largely unchanged from 2004, at the time of the passing of the Sickle Cell Treatment Act, which expanded specialized sickle cell treatment programs and research support.²⁷ 

Although surrogate or biomarker end points for FDA approval are a means to move the needle in therapeutic availability for populations suffering from treatment disparities, such as in the case of voxelotor for SCD, this initial approval process requires timely and attentive monitoring to ensure outcome-based primary end points are met. The accelerated approval pathway of the FDA allows surrogate markers as primary efficacy end points in clinical trials for authorization of drugs for serious diseases, if they are “reasonably likely to predict clinical benefit.”²⁸ Sponsors that receive accelerated approval are required to conduct postmarketing trials to confirm the expected clinical benefit, but details of these requirements and potential consequences if they are not performed in a timely manner are unclear. An analysis completed by National Public Radio found that, from 1992 to 2021, 42% of confirmatory studies (50) took over 1 year to begin after the accelerated approval of the drug, with 16% taking at least 3 years to begin.²⁹ Thus, in actuality, on average, the time to withdrawal of a failed drug from initial accelerated approval is 46 months.³⁰ Previous studies have raised questions around the utility of using nonclinical end points, finding that surrogate markers used to inform FDA approval for these indications are not consistently correlated with clinical outcomes such as overall survival or quality of life.^31,32 In addition, accelerated approval may denote to the public a clinical benefit out of proportion to what the surrogate marker actually suggests.³³ In the case of voxelotor, an increase in hemoglobin, the accepted surrogate marker, did not correlate with a decrease in painful events in the clinical trials that led to approval or in follow-up assessment.^7,34 

The recent challenges with new SCD therapies should not result in despair or a sense of hopelessness for SCD providers or patients. Instead, these events create an opportunity to reanalyze, both within and across the SCD provider and patient communities, where we are and how we can provide the best care to our patients. In addition, the experiences with crizanlizumab and voxelotor draw attention to the benefits and limitations of the FDA’s current expedited approval process. The naming of SCD as 1 of the 20 diseases of focus under the Patient-Focused Drug Development Program in 2012 finally brought new potential therapies to SCD for the first time in 30 years. Unfortunately, voxelotor and crizanlizumab both fell victim to issues that have plagued the FDA’s accelerated approval system, including the use of surrogate markers for efficacy and lack of specified postapproval efficacy trials.^30-32 However, with the passage of the Food and Drug Omnibus Reform Act of 2022, we can be optimistic that future potential therapies will undergo more rigorous testing before, or soon after, FDA approval, thereby facilitating trust in our patient population. The 2022 Food and Drug Omnibus Reform Act stipulates that the FDA must specify conditions of a confirmatory trial, including the target completion date, at the time of its accelerated approval; furthermore, the postapproval study must start before or closely after approval of the drug. Finally, more oversight is placed on the sponsor, including progress reports every 6 months (an increase from annually) and implications for failing to conduct the postapproval study within the specified time.³⁵ These changes are especially important as related to SCD, given the recent FDA approval in late 2023 of 2 gene therapy products for SCD (Lyfgenia, Bluebird Bio Inc; Casgevy, Vertex) with several others in development.^36,37 

For a population that has experienced the systemic and medical inequities that have led to disparities in outcomes, transparency throughout the drug development process is essential to building and maintaining trust in research and in clinical care of patients with SCD. It can be argued that, in the instances of crizanlizumab and voxelotor, the regulatory and monitoring system did what it was intended to do, resulting in the removal of approval or cessation of manufacturing of these drugs. However, the delay in data availability and the way in which the information regarding voxelotor’s removal spread across the SCD community placed health care providers and their patients in a difficult position with little certainty on how to respond. Thus, there is no time like the present for health care providers, pharmaceutical companies, and regulatory bodies to talk transparently about the drug development and approval process and to actively engage individuals living with SCD into this process. Through these efforts, we can continue to expand the number of disease-modifying and transformative therapies for our patients, allowing those with SCD to live full and productive lives.

Contribution: P.T.M. developed the concept for the manuscript; and all authors wrote the manuscript and agreed on the final version of the manuscript.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Kristine Karkoska, The Vontz Center for Molecular Studies #3102, 3115 Eden Ave, Cincinnati, OH 45229; email: karkoska@mail.uc.edu.