Consequences of rush to emergency use authorization of bamlanivimab

Until November 2020, treatment for nonhospitalized patients with acute SARS-CoV-2 infection was largely limited to supportive measures. This changed on November 9, 2020, following emergency use authorization (EUA) for use of bamlanivimab, a monoclonal antibody administered in an ambulatory care setting by one-time intravenous (IV) infusion, to treat adolescents and adults with mild to moderate SARS-CoV-2 infection at high risk for progressing to hospitalization.¹ Prior to the revocation of the bamlanivimab monotherapy EUA on April 16, 2021, due to ineffectiveness against SARS-CoV-2 variants, nearly 1 million doses were distributed.

On November 12, 2020, the Infectious Diseases Society of America (IDSA) expressed concern about using bamlanivimab prior to collection of “sufficient safety and efficacy data to be confident regarding” routine use. On November 18, IDSA recommended against the use of bamlanivimab in its updated COVID-19 treatment guidelines. Authors of the study on which the EUA was based expressed similar caution about the sufficiency of the existing data, noting that bamlanivimab “could become a useful treatment . . . in patients with recently diagnosed COVID-19,” but only “if results are confirmed in additional analyses.” ^2-4 Yet, after the EUA enabled widespread off-trial use of bamlanivimab, clinical trial enrollment was more difficult. Therefore, little additional rigorous evidence from clinical trials about bamlanivimab’s effectiveness was likely to or did in fact emerge. Meanwhile, reports of adverse events and other costs associated with use of the drug mounted.

Evidentiary basis for the bamlanivimab EUA

In contrast to the rigorous standards to which the US Food and Drug Administration (FDA) is held in approving new drugs, the EUA process provides ample discretion for FDA to permit experimental drugs to be widely used despite insufficient safety and efficacy data. Once an emergency is declared, the statutory standard for issuing an EUA is low. Under section 564 of the Federal Food, Drug, and Cosmetic Act (21 U.S.C. § 360bbb-3), FDA need only determine, “based on the totality of the scientific evidence available”—which may not include any randomized controlled trial (RCT) data—that it is “reasonable to believe” that an investigational intervention “may be effective” in addressing a serious or life-threatening disease for which there is no adequate alternative, and that it is also “reasonable to believe” that the drug’s “known and potential benefits . . . outweigh the known and potential risks.”

Like large-scale expanded access programs (eg, that for COVID-19 convalescent plasma), EUAs can render trials infeasible, leading to widespread use of investigational products with unclear risk-benefit profiles, which may remain permanently unclear. The EUA issued for bamlanivimab was based on a planned interim analysis of a phase 2 randomized trial comparing bamlanivimab to placebo use. The primary outcome was change in viral load at day 11, a surrogate marker for improvement in SARS-CoV-2 infection that, as investigators eventually concluded, is likely not a clinically meaningful endpoint. Secondary outcomes were hospitalizations and adverse effects through day 29.¹

In the Blaze-1 trial, patients were randomly assigned to use of a placebo or one of 3 bamlanivimab IV doses: 700 mg, 2,800 mg, or 7,000 mg. The data provided to FDA in support of the EUA reported outcomes for 101 patients who received bamlanivimab 700 mg (the dose authorized by the EUA) and 156 who received placebo (these latter participants included 143 whose outcomes were reported in the New England Journal of Medicine and an additional 13 whose outcomes manifested post publication). There was no statistically significant difference in SARS-CoV-2–related hospitalizations in the full 700 mg cohort versus the placebo group (1 of 101 vs 9 of 156, P = 0.09) or in a post hoc analysis of high-risk patients (1 of 46 vs 7 of 69, P = 0.14). Total adverse event rates were similar in the bamlanivimab and placebo groups (25 of 101 vs 41 of 156), with most events occurring during the infusion. There was one serious adverse event (upper abdominal pain) in the placebo group and none with use of bamlanivimab.¹

The EUA documentation noted 3 other serious adverse events among 1,350 participants in inpatient and nursing home studies who received any dose of bamlanivimab: difficulty breathing and flushing during infusion; allergic-type reactions during infusion; and flu-like symptoms 3.5 hours post infusion.¹

Pre- and post-EUA bamlanivimab safety data.

Bamlanivimab’s EUA was based on studies with limited sample sizes and limited time for any risks to manifest. Downing and colleagues⁵ found that 32% of novel therapeutics are issued safety warnings, issued black box warnings, or withdrawn from the market within 11 years post FDA approval, with biologic agents nearly twice as likely as conventional treatments to result in postmarketing safety events and therapies granted accelerated approval 2.2 times more likely to be associated with such events. In the case of an experimental drug that is authorized for use prior to even the phase 3 trials needed for FDA approval, the risk of postmarketing safety concerns is presumably higher.

An interrogation of the FDA Adverse Event Reporting System in March of 2021 revealed 3,395 adverse events in 1,045 patients treated with bamlanivimab (a mean [SD] of 3.2 [2.6] events per patient) reported between November 9 and December 31, 2020. About 81.5% of patients (852 of 1,045) had reactions classified as severe, resulting in 560 hospitalizations, 30 life-threatening reactions, and 33 deaths. Many of the adverse effects noted are common with use of most medications (eg, nausea, headache), while others overlap with COVID-19 symptoms (eg, dyspnea, pyrexia). There was one case of Steven’s-Johnson syndrome, 2 cases of rhabdomyolysis, and 3 cases of QT interval prolongation—all more severe and concerning adverse events that are strongly tied to medications.⁶ These events all point to a possible signal of serious untoward effects of bamlanivimab not previously reported in RCTs. Confirmation through additional reports, thorough case examination, and comparison with background incidence rates are needed to determine if these reported serious adverse events were directly associated with bamlanivimab use.

Non–safety-related consequences of the bamlanivimab EUA

In addition to clinical risks to patients, the 1,450,000 bamlanivimab doses ordered by the US government cost over $1.8 billion. Moreover, administering bamlanivimab requires considerable scarce resources. For example, healthcare facilities repurposed infusion space to accommodate SARS-CoV-2–positive patients for greater than 1-hour outpatient infusions and monitoring. Skilled staff were redeployed or hired to administer and monitor patients, increasing their exposure to SARS-COV-2 and requiring use of limited personal protective equipment. Additional logistical challenges included procurement, storage, and distribution of bamlanivimab and appropriate allocation of limited doses in a high-demand environment.

Conclusion

FDA has substantial discretion in deciding when to issue an EUA; it need not issue one as soon as the very low bar for EUA authorization is cleared. The agency exercised restraint in waiting to issue EUAs for SARS-COV-2 vaccines until extremely compelling phase 3 trial data from tens of thousands of people were available. We cannot say the same of its decision regarding bamlanivimab. A public health emergency does not justify departures from rigorous evaluation of experimental interventions; if anything, the sheer numbers of people whose welfare is affected by such decisions suggests that the opposite is true.⁷ It is reasonable for patients to assume a certain degree of risk with use of a drug proven to be effective, but bamlanivimab’s efficacy was always questionable and its safety profile was incompletely understood. Patients and providers were therefore left to have largely evidence-free conversations about bamlanivimab and utilize resources within a healthcare system that was already overburdened due to the COVID-19 pandemic. We urge FDA to issue future EUAs only after compiling higher-quality safety and efficacy data in order to reduce healthcare waste and prevent patient harm.

Disclosures

Dr. Wright has received grant funding from Pfizer Inc., Merck Sharp and Dohme, the National Institute of Drug Abuse, and the Patient-Centered Outcomes Research Institute. The other authors have declared no potential conflicts of interest.