A Cursory Review of Casirivimab and Imdevimab for COVID-19
"Booksellers are the most valuable destination for the lonely, given the numbers of books that were written because authors couldn't find anyone to talk to."
- Alain de Botton
An important reason for treating both tuberculosis and HIV with multiple pharmacological agents is the ability of these infectious scourges to mutate and escape monotherapy – thereby selecting a new variant resistant to medicinal weaponry. While escape mutants might be considered a problem specific to small molecules like antimicrobials, it is also true for, comparatively large, monoclonal antibodies.
Given that mRNA viruses like SARS-CoV-2 readily mutate, it is argued that human monoclonal antibodies against the virus that causes COVID-19 be given as a ‘cocktail.’ That is, administering more than one antibody – each with high specificity to unique, non-overlapping epitopes of the viral spike protein – should reduce the chance of escape mutation.
Brief History of Casirivimab and Imdevimab
By late May of last year, researchers had employed parallel techniques to characterize human antibodies with high specificity for the SARS-CoV-2 spike protein. These approaches used humanized mice as well as B-cells extracted from convalescent humans to produce a diverse collection of antibodies which were then screened for synergistic effects against the receptor binding domain [RBD] of the SARS-CoV-2 spike protein. Four antibodies of particular interest were identified: REGN10987, REGN10933, REGN10989 and REGN10934. Notably, the former two bind distinct regions of the RBD such that are ‘non-competing.’ REGN10933 [casirivimab] broadly envelopes the ACE2 binding site; REGN10987 [imdevimab] adjacently binds the spike protein but with no overlap of the ACE2 binding site.
Thereafter, escape mutants to the aforementioned, individual monoclonal antibodies were generated in vitro. Importantly, mutations via this process readily occurred, often with a single amino acid substitution. Two of the in vitro mutants were comparable to known in vivo variants at the time, intimating that selective pressure from large-scale, individual, monoclonal antibody therapeutics could amplify resistant variants. Critically, the combination of casirivimab and imdevimab averted mutant selection – ostensibly because this event requires synchronous mutations in two distinct spike protein sites. Additionally, the combination of casirivimab and imdevimab maintained their neutralization potency against in vitro pseudoparticles expressing a variety of escape mutant spike proteins – including the E484K which is one of the concerning amino acid substitutions noted in the B.1.351 variant described in South Africa.
Months after the aforementioned publication, the combination of casirivimab and imdevimab earned world-wide recognition when this cocktail was administered to the former U.S. President when he – and many of his advisors – tested positive for SARS-CoV-2. After receiving these monoclonal antibodies, he was hospitalized and treated with corticosteroids for COVID-19.
Emerging Clinical Data
The combination of casirivimab and imdevimab – ‘REGN-COV2’ – has been studied in an on-going, double-blind, phase 1–3 trial. Data from the first 272 non-hospitalized COVID-19 patients was recently published in the New England Journal of Medicine; this represents patient data collected through early September, 2020.
Non-hospitalized patients were studied because the investigators sought patients in whom native humoural responses had not yet been activated. Indeed, this was a pre-specified sub-group analysis as patients were screened on enrollment for the presence of SARS-CoV-2 antibodies. The researchers hypothesized that those without endogenous antibodies would derive the most benefit from receiving the exogenous REGN-COV2 cocktail.
What They Did
Adult outpatients within 7 days of symptom onset and within 72 hours of positive, quantitative reverse transcriptase–polymerase-chain-reaction [RT-PCR] nasopharyngeal swab testing were enrolled. Treatment arms were randomly-assigned in a 1:1:1 ratio to receive a single intravenous infusion of either 2.4 g or 8 g of REGN-COV2 or placebo. The study’s clinical end-points were 1.) viral load change from day 1-7 and 2.) the percentage of patients who had at least one COVID-19 – related medically attended visit across 29 days of follow-up.
What They Found
There were about 90 patients allocated to each arm. The mean and median SARS-CoV-2 viral load was significantly higher in those who were antibody negative upon enrollment. As well, there was an inverse association between antibody titre and viral load.
The patients in the pre-specified antibody negative group on enrollment had the most viral load reduction with the entire 95% confidence interval less than no-difference as compared to placebo. As well, post-hoc subgroup analysis found the greatest reduction in viral load in those with the highest level on enrollment. The greatest reduction occurred by 72 hours.
Only 6 patients randomized to placebo suffered a medically-attended visit secondary to worsening COVID-19 symptoms. Of interest, 1 of those 6 was initially antibody positive. In totality, 6% of the placebo group and 3% of the combined REGN-COV2 group had a medically-attended visit within 29 days. When confined to those who were antibody negative at the outset, these numbers changed to 15% and 6%, respectively. There was no difference between REGN-COV2 and placebo in the clinical end-point when restricted to those who were antibody positive on enrollment.
The therapeutic rationale for REGN-COV2 is almost the converse of that given for corticosteroids. That is, the RECOVERY investigators hypothesized that immunosuppression would be most beneficial in patients with symptoms for at least a week. In other words, because viral replication extends from a few days prior to symptoms to roughly a week beyond, this time - in theory - is when immunosuppression might be most detrimental. However, beyond a week of symptoms, when the threat switches from the virus to the immune response, corticosteroids calm the latter. Thus, it is unlikely that the intensivist should consider REGN-COV2 as a therapeutic option as early, antibody negative patients are less likely to end up in the ICU. It is interesting to note, however, that in the aforementioned patients specifically-selected for symptoms less than 7 days, there was no association between symptom duration and endogenous antibody response.
Most importantly, for the intensivist, is the use limitation noted by the Food and Drug Administration [FDA] that monoclonal antibodies should not be administered to inpatients outside of a clinical trial. Indeed, the FDA specifically states:
“Benefit of treatment with casirivimab and imdevimab has not been observed in patients hospitalized due to COVID-19. Monoclonal antibodies, such as casirivimab and imdevimab, may be associated with worse clinical outcomes when administered to hospitalized patients requiring high flow oxygen or mechanical ventilation with COVID-19.”
While there is no specific evidence that hypoxemia is worsened by REGN-COV2, the EUA makes it quite clear that monoclonal antibodies not be standard of care for critically-ill COVID-19 patients. Instead, this antibody mixture is reserved for mild-to-moderate COVID-19 patients at high-risk for progression to severe disease. Risk factors for progression can be found here.
Lastly, the EUA reports additional REGN-COV2 data with 266, 267 and 266 patients having been randomized to low versus high-dose REGN-COV2 and placebo, respectively. The data published in NEJM comes from an interim analysis that halted in early September. In totality, the extended results were similar with reduced viral load and a 3.7% absolute risk reduction of the primary clinical end-point.