Assessing the safety and pharmacokinetics of casirivimab and imdevimab (CAS+IMD) in a cohort of pregnant outpatients with COVID-19: results from an adaptive, multicentre, randomised, double-blind, phase 1/2/3 study.

Infection with SARS-CoV-2, the causative agent for COVID-19, causes a spectrum of disease, ranging from asymptomatic infection to respiratory failure, multiorgan dysfunction and death. 1 5 It was recognised early in the COVID-19 pandemic that SARS-CoV-2 infection during pregnancy may be associated with increased risk for severe outcomes, with initial reports showing higher risk of maternal mortality, intensive care unit admission, preterm birth (before 37 weeks) and neonatal care unit admission compared with non-pregnant or non-infected controls. 6 10 These findings were largely confirmed and expanded on in a meta-analysis comprised of more than 13 000 pregnant women, 11 highlighting the need for safe and effective therapeutic options for this population. To reduce the risk of severe disease in pregnancy, COVID-19 vaccination is recommended for all individuals who are pregnant or planning to become pregnant. For those who develop COVID-19 during pregnancy, clinical management is largely driven by the patient setting (ie, outpatient vs inpatient) and includes therapeutic considerations such as nirmatrelvir-ritonavir, remdesivir, dexamethasone, supplemental oxygen and venous thromboembolism prophylaxis. Monoclonal antibodies (mAbs) against SARS-CoV-2 have proven to be well-tolerated and effective in the treatment and prevention of COVID-19 in high-risk adults, serving as a critical component of the initial COVID-19 armamentarium. Anti-SARS-CoV-2 mAbs work by targeting the spike protein of SARS-CoV-2, thereby blocking its interaction with host receptor angiotensin-converting enzyme 2 to prevent entry of the virus into susceptible cells. In adults, a single intravenous dose of casirivimab plus imdevimab (CAS+IMD), a combination of two recombinant mAbs that bind simultaneously in a non-competing manner to the receptor-binding domain of the SARS-CoV-2 spike protein, 12 was well-tolerated and reduced COVID-19-related hospitalisations and all-cause death, reduced SARS-CoV-2 viral load and shortened the time to symptoms resolution compared with placebo. 13 15 However, the emergence of SARS-CoV-2 variants conferring a loss of activity for most previously approved or authorised anti-SARS-CoV-2 mAbs, including CAS+IMD, has limited the utility of anti-SARS-CoV-2 mAbs for the treatment and prevention of COVID-19. Several retrospective studies evaluating anti-SARS-CoV-2 mAbs in pregnant women have indicated favourable safety and tolerability results, with no increased adverse maternal, neonatal or obstetrical outcomes; 16 21 however, data from prospective clinical trials are limited. Given that next-generation anti-SARS-CoV-2 mAbs are currently being developed to address ongoing unmet needs in COVID-19, it is important that clinical trial data exploring the safety, tolerability and pharmacokinetics (PK) of anti-SARS-CoV-2 mAbs in pregnant participants be shared with the medical community to inform their role in this vulnerable population. Herein, we present the final safety, tolerability, clinical outcome, PK and immunogenicity data from the pregnant cohort of a phase 1/2/3 study, conducted prior to the widespread circulation of Omicron-lineage variants, evaluating CAS+IMD in outpatients with COVID-19 who had ≥1 risk factor for severe disease.

Study COV-2067 was an adaptive, multicentre, randomised, double-blind, placebo-controlled phase 1/2/3 study to evaluate the efficacy and safety of CAS+IMD in outpatients with ≥1 risk factor for severe COVID-19 ( NCT04425629 ; online supplemental figure S1 ). Details of the primary phase 3 analysis in non-pregnant, adult outpatients have been previously published. 15 Here, we describe the primary results from the pregnant participants enrolled during the phase 3 portion of the study. Pregnant participants were initially randomised 1:1:1 as part of the wider phase 3 study to a single intravenous dose of CAS+IMD 1200 mg, CAS+IMD 2400 mg or placebo; however, following implementation of a protocol amendment, all subsequent pregnant participants were enrolled in a separate cohort that did not include a placebo arm (ie, participants were randomised 1:1 to CAS+IMD 1200 mg or 2400 mg). To consolidate all phase 3 pregnant participants into a single analysis set, those randomised prior to the establishment of the pregnancy cohort were reclassified into the pregnancy cohort prior to unblinding. The pregnancy cohort continued to enrol participants 1:1 to CAS+IMD 1200 mg or 2400 mg until a further study change required that all participants receive the 1200 mg dose. Participants were enrolled between December 2020 and November 2021. For context, early 2021 marked the emergence of the Alpha variant (B.1.1.7), followed by emergence of the Delta variant (B.1.617.2) in the summer of 2021, and the Omicron variant (B.1.1.529) in December 2021. 22 Notably, in vitro studies showed that CAS+IMD retained activity against all SARS-CoV-2 variants of concern, including B.1.1.7, B.1.429, B.1.617 and E484K-containing variants, 23 24 until the emergence of the Omicron-lineage variants, against which CAS+IMD showed reduced potency, suggesting it was unlikely to be active. Eligible participants had to be pregnant and non-hospitalised at randomisation. All participants had confirmed COVID-19, with a positive SARS-CoV-2 test result received ≤72 hours before randomisation, and symptom onset ≤7 days before randomisation. The trimester of pregnancy at study entry was calculated based on the last reported menstrual period start date. For participants with a partial date for the last menstrual period start date (eg, missing day or month), the date was imputed as an estimated delivery date <280 days. All participants were followed up to day 169 (end of study). Considering the time period in which this study was being conducted (ie, the height of the COVID-19 pandemic), only targeted categories of treatment-emergent adverse events (TEAEs) were collected to provide the most relevant safety information for evaluating the safety and tolerability of CAS+IMD. Included were serious adverse events (SAEs) up to day 169, adverse events of special interest (AESIs) of grade ≥2 infusion-related reactions (IRRs) up to day 4 and grade ≥2 hypersensitivity reactions (HSRs) up to day 29, and, to contextualise clinical events and thus described under clinical outcomes, any TEAE that led to a medically attended visit (MAV) up to day 29. TEAEs voluntarily reported outside of the targeted categories were included in the safety analysis. Pregnancy outcomes and development outcomes in infants until 1 year of age were also collected. Clinical endpoints included the proportion of participants with ≥1 COVID-19-related hospitalisation or all-cause death through day 29, the proportion of participants with COVID-19-related MAVs (including hospitalisation, emergency room visits, urgent care visits and physician office/telemedicine visits), and time to resolution of symptoms consistent with COVID-19. The virologic endpoint was changed from baseline in viral load at each visit through day 29. PK endpoints included the concentrations of CAS and IMD in serum over time. Additionally, similarity in concentration at comparable doses at the end of infusion and at study day 29 between pregnant and non-pregnant adult populations was also investigated. Serum samples were analysed for total CAS and total IMD using a validated electrochemiluminescence (ECL) bridging immunoassay. 25 The lower limit of quantitation was 0.156 mg/L in neat serum. Immunogenicity was assessed by determining incidence and titre of anti-drug antibodies (ADAs) and incidence of neutralising antibodies (NAbs) to CAS and IMD. For serum sample analysis, validated ECL bridging immunoassays were used for monitoring ADAs and validated competitive ligand-binding assays were used for monitoring NAbs. The full analysis set (FAS) included all randomised pregnant participants; the safety analysis set included all randomised participants who received any study drug; and the PK analysis set (PKAS) included all participants who received any study drug and had at least one non-missing result following the first dose of the study drug (serum samples for drug concentrations from participants who received placebo treatment were not analysed). The ADA analysis sets included all treated participants who received any amount of study drug (active or placebo) and had ≥1 non-missing ADA result following dosing of the study drug; the ADA analysis sets were based on actual treatment received (as treated) rather than as randomised. The NAb analysis sets were defined as described for ADA analysis sets and required that participants either tested negative at all ADA sampling times or tested positive for ADAs with at least one non-missing NAb result after dosing of the study drug (active or placebo). The concentration-response analysis set (CR-seronegative modified FAS [mFAS]) consisted of the baseline RT-qPCR positive (mFAS), seronegative PKAS and placebo-treated participants who were included in the seronegative mFAS who had ≥1 non-missing baseline and one non-missing post dose viral load or clinical efficacy measurement. For the purposes of this concentration-response population, as samples from placebo participants were not assayed, all missing concentrations were imputed as 0. The proportion of participants with ≥1 COVID-19-related MAV or all-cause death, and the proportions of participants with each type of MAV, were summarised descriptively. No multiplicity adjustment was applied. No patients were directly involved in the design or development of plans for the implementation of the studies reported in this post hoc analysis. No patients participated in the interpretation or writing up of results.

In total, 82 pregnant participants were randomised and included in the FAS. Nearly all (n=79, 96.3%) participants were randomised to CAS+IMD (1200 mg intravenous, n=43; 2400 mg intravenous, n=36), with only three randomised to placebo, precluding any meaningful comparisons between treatment and control groups. Of the 82 randomised participants, two were not treated (one in the CAS+IMD 2400 mg arm and one in the placebo arm), 75 (91.5%) completed the study and seven discontinued due to either being lost to follow-up (n=4) or participant decision (n=3) ( online supplemental figure S2 ). The median age was 28 years (range 17–41 years, one participant was aged <18 years), 81.7% were non-Hispanic or Latino, 85.4% were White and 8.5% were Black or African American ( table 1 ). At randomisation, 12.2% were in the first trimester of pregnancy (weeks 1–12), 36.6% were in the second trimester (weeks 13–26) and 51.2% were in the third trimester (27 weeks to the end of pregnancy). By virtue of being pregnant, all participants had ≥1 risk factor for severe COVID-19 disease, with some having more than one risk factor, including 12.2% with chronic lung disease (including asthma), 7.3% with cardiovascular disease (including hypertension) and 3.7% with type 1 or 2 diabetes mellitus. The trimester of pregnancy at study entry was calculated based on the last menstrual period start date. For patients with a partial date for the last menstrual period start date (eg, missing of day or month), the date was imputed as estimated delivery date − —280 days; if the estimated delivery date was also missing, then no imputation would beas performed. CAS, casirivimab; FAS, full analysis set; IMD, imdevimab; NP, nasopharyngeal Given that eligibility criteria required participants to be unvaccinated with no history of a positive SARS-CoV-2 diagnostic test >72 hours prior to randomisation and symptom onset ≤7 days prior to randomisation, the majority (72.0%) of participants were SARS-CoV-2 seronegative at baseline. The remaining participants were SARS-CoV-2 seropositive (19.5%) or had a borderline or unknown serostatus result (8.5%). The median (Q1:Q3) time from symptom onset to randomisation was 3 (2.0:4.0) days, and the mean baseline viral load (SD) in pregnant participants was 6.00 (2.63) log 10 copies/mL. These results are consistent with observations in non-pregnant populations from the same study. 15 As expected, pregnant participants who were seronegative at baseline had a higher mean (SD) baseline viral load than those who were seropositive at baseline (6.58 [2.44] log 10 copies/mL vs 5.08 [1.97] log 10 copies/mL, respectively). Safety data are presented for the 80 pregnant participants who received any amount of study drug (including 78 participants who received CAS+IMD, and two who received placebo) ( table 2 ). TEAEs collected include TE SAEs, AESIs, and grade 3/4 TEAEs as well as ad hoc/voluntarily reported TEAEs by some sites. TEAEs deemed treatment related as per investigator assessment. Infusion interruption: the administration of the infusion was interrupted before being completed, but subsequently was re-started and the full planned dose was administered. Infusion discontinuation: the administration of the infusion was stopped before being completed, and the full planned dose was not administered. AESI, adverse event of special interest; CAS, casirivimab; IMD, imdevimab; MAV, medically attended visit; SAE, serious adverse event; SAF, safety analysis set; TE, treatment-emergent; TEAE, treatment-emergent adverse event For safety, no dose-effect was observed. AE categories were generally balanced across the CAS+IMD 1200 mg and 2400 mg treatment groups, with only minor numerical discrepancies which, when observed, were likely driven by the small sample size. Due to the small number of participants in the placebo group, no meaningful comparisons could be made between the CAS+IMD treatment groups and the placebo group. There were no TEAEs leading to death, study withdrawal, study infusion interruption or study infusion discontinuation, and no safety signals were observed with respect to pregnancy exposure or development outcomes. A total of 22 SAEs occurred during the course of the study ( table 3 ): 14 in 12 maternal participants and eight in the fetus/infant. Between day 1 and day 169, 11 maternal participants had 13 SAEs and there were three SAEs among three fetuses/infants. After day 169, one maternal participant had one SAE and five fetuses/infants had five SAEs. All but one of the SAEs were considered not related to the study treatment. The SAE considered related to study treatment was reported as a grade 4 event of spontaneous abortion experienced by a participant in the 1200 mg treatment group. Although this was reported as a spontaneous abortion, the event occurred at 29 weeks gestation (day 104; 103 days after receiving study treatment) and was an intrauterine death/stillbirth. This case was considered related to the study drug by the investigator due to the temporal relationship; however, there is no biologically plausible mechanism of action of CAS+IMD that could account for this. Moreover, the event was confounded by a medical history of endometriosis. CAS, casirivimab; IMD, imdevimab; SAE, serious adverse event; SAF, safety analysis set; TEAE, treatment-emergent adverse event There were seven AESIs, all of which were TEAEs that led to a MAV. There were no grade ≥2 HSRs or IRRs ( table 2 ). Six participants in the CAS+IMD 1200 mg group experienced eight grade 3 or 4 TEAEs (six were maternal and two were fetal/infant); a further two participants in the 2400 mg group experienced four grade 3 or 4 TEAEs (three were maternal and one was fetal/infant). There were 70 pregnancy reports available for review. Of the 70 pregnancies enrolled and randomised, five (7.1%) participants declined further participation/did not wish to provide any further details regarding the pregnancy. Two (2.9%) of the 70 reports noted elective abortions, and one (1.4%) noted an intrauterine death (discussed earlier). Of the remaining 62 reports, 55 (89%) were full-term deliveries (ie, baby born between 39 weeks, 0 days and 40 weeks, 6 days, as per definitions from The American College of Obstetricians and Gynaecologists [ACOG] and the Society for Maternal-Fetal Medicine [SMFM], including one reported as full-term with delivery week not specified) and seven (11%) were early-term deliveries (ie, baby born between 37 weeks, 0 days and 38 weeks, 6 days). A total of 35 (57%) deliveries were reported as vaginal, 25 (40%) involved a caesarean section, and for two (3%) the delivery method was unknown. Seven pregnancies reported issues in the baby, details of which are summarised in online supplemental table S1 . None of the reported issues were considered related to the study drug, as all were confounded by maternal medical history or considered due to an alternate aetiology. One-year infant developmental status was available for 33 of the 62 available pregnancy reports. The remaining 29 reports noted the participant was either lost to follow-up, declined participation, withdrew consent or could not participate due to site closure. Of the 33 reports where developmental status was available, 31 reported no issues. The remaining two case reports reporting issues at the 1-year infant developmental follow-up are summarised in table 4 . None of the developmental issues reported were considered related to the study drug due to confounding maternal history or implausible temporal association. As such, no safety signals related to the study drug were observed in infants during follow-up at 1 year of age. CAS, casirivimab; IMD, imdevimabIUGRintrauterine growth restrictionLMPlast menstrual period As previously noted, most pregnant participants received CAS+IMD, limiting the ability to make any meaningful conclusions with regard to treatment effect for CAS+IMD vs placebo. However, both CAS+IMD 1200 mg and 2400 mg treatments were associated with rapid reductions in viral loads. Patients treated with CAS+IMD had a median of approximately 100% viral load reduction at day 3, which was maintained at days 15 and 29. The reduction in viral load was similar to that observed in non-pregnant participants in the same study that had received CAS+IMD. 15 Among the 82 randomised participants, only two participants experienced a COVID-19-related MAV. One participant in the 2400 mg group had a COVID-19-related hospitalisation on day 1 (the day of dosing), and one participant in the 1200 mg group had a COVID-19-related emergency room visit on day 2. The low rates of COVID-19-related MAVs in the pregnant cohort were consistent with the rates previously reported in non-pregnant adults receiving CAS+IMD. 15 The median time to symptoms resolution was 10 days for the 1200 mg group and 14 days for the 2400 mg group; however, the small sample size limits the ability to make any meaningful dose response conclusions or comparisons to the results observed. CAS, IMD and CAS+IMD combined concentrations in serum displayed linear PK in pregnant participants ( online supplemental figure S3 ), reflected by a dose-proportional increase in concentration up to the last measurable concentration. There were no meaningful differences in concentrations of antibodies to day 120 across different trimesters of pregnancy at baseline ( figure 1 ). Concentrations of CAS and IMD at the end of the infusion and day 28 in the pregnant female population were similar to those observed in non-pregnant females and adult males ( online supplemental figure S4 ). Thus, due to the linear PK of CAS and IMD, similar PK is expected between pregnant females and non-pregnant adults. For the active treatments combined, 1.4% and 4.2% of participants had low titre, treatment-emergent ADAs against CAS and IMD, respectively. None of the participants with treatment-emergent ADAs against CAS and IMD were positive for NAbs ( online supplemental table S2 ) and no meaningful difference in concentrations of CAS or IMD was observed between participants with different ADA or NAb status (data not shown).

Pregnancy has been associated with increased risk for severe outcomes from COVID-19 infection. Randomised, controlled data informing the use of antiviral therapeutics for the treatment of pregnant patients with COVID-19 remain limited. Given the proven efficacy of anti-SARS-CoV-2 mAbs for the prevention and treatment of COVID-19 caused by susceptible variants, and ongoing efforts to develop next-generation SARS-CoV-2 mAbs to address emerging variants that have rendered previously approved/authorised mAbs ineffective, data illuminating the experience of anti-SARS-CoV-2 mAbs in pregnant patients still represents an important medical communication. Although real-world experiences with anti-SARS-CoV-2 mAb treatments in pregnant women have reported similar results to a non-pregnant population with no new safety concerns, 16 2126 prospective clinical data in this vulnerable and defined at-risk population remain limited. 27 28 This report presents the largest cohort of pregnant participants treated with anti-SARS-CoV-2 mAbs in any prospective study to date. Our study found that treatment of non-hospitalised pregnant participants with CAS+IMD was well-tolerated, with no deaths, and no grade ≥2 IRRs or HSRs, consistent with observations in high-risk non-pregnant adult outpatients with COVID-19. 15 Furthermore, a review of pregnancy outcomes did not identify any new safety signals with regard to CAS+IMD, either in the exposed pregnant women or their babies. There were no trends in the types of events observed in CAS+IMD-treated pregnant participants, and the rates of treatment-emergent ADAs against CAS and IMD were low and did not result in NAbs. SARS-CoV-2 infection during pregnancy has been shown to be associated with an increased risk of maternal mortality, pre-eclampsia, preterm birth (before 37 weeks), stillbirth, and intensive care unit and neonatal intensive care unit admission. 1129 32 Among the 62 pregnancy reports with complete delivery data, 89% were full-term deliveries (born between 39 weeks, 0 days and 40 weeks and 6 days, per ACOG and SMFM definitions) and 11% were early-term deliveries (born between 37 weeks, 0 days and 38 weeks, 6 days), similar to rates reported in the general population. Although two pregnancies (3%) reported issues in the fetus/neonate, they were confounded by maternal history or considered to be due to an alternate aetiology. There were no reported issues in infants at 1 year that were considered related to CAS+IMD. CAS+IMD displayed linear PK in pregnant participants, with similar concentrations of CAS and IMD observed across different trimesters of pregnancy at baseline. Concentrations of CAS and IMD at the end of infusion and study day 28 weresimilar in pregnant and non-pregnant adult outpatients at both the 1200 mg and 2400 mg intravenous doses. Together with linear PK of these antibodies, these data indicate that pregnancy status did not affect the PK of CAS and IMD. A similar response was also observed in nasopharyngeal viral load reduction in pregnant and non-pregnant participants for all selected doses which, coupled with similar clinical outcome data for both doses, supported the use of the 1200 mg intravenous dose of CAS+IMD in pregnant outpatients for the treatment of SARS-CoV-2 infection. This study had some limitations. The sample size was relatively small (particularly for symptoms analyses and follow-up in infants at 1 year) and included few participants in the placebo arm, limiting any meaningful conclusions with regard to treatment effect. In addition, the majority of participants in this cohort were seronegative at baseline. While this may be considered a limitation, as the results cannot be generalised to the current situation where >95% of the world’s population is seropositive, 33 it should be noted that, in the much larger adult cohort, CAS+IMD has been associated with clinical benefit regardless of baseline serum antibody status. 15 Furthermore, most participants in this cohort were White, non-Hispanic, limiting insights for CAS+IMD treatment in other racial and ethnic pregnant populations. However, a post hoc study by Levey et al investigating the safety and efficacy of CAS+IMD in a predominantly (95%) non-Hispanic Black or Hispanic pregnant population with COVID-19 also found the treatment to be safe. 20 In summary, this study affirms that CAS+IMD was well-tolerated in pregnant women at all stages of pregnancy, with no new safety signals for CAS+IMD identified in the mothers or their infants up to 1 year old. The data showed a similar safety, PK, and clinical outcome profile to that observed in non-pregnant adults, providing meaningful data to inform the development of next-generation mAb-based therapeutics for pregnant patients.