ABSTRACT
Background: Omicron subvariants questioned the efficacy of the approved therapies for the early COVID-19. In vitro data show that remdesivir (RDV), molnupiravir (MLN), and nirmatrelvir/ritonavir (NMV/r) all retained activity against all sub-lineages, while poor neutralizing activity was observed for Sotrovimab (SOT) and Tixagevimab/cilgavimab (TIX/CIL). No data about the risk of clinical failure or even in vivo antiviral activity are available. Method(s): Single-center observational comparison study enrolling all consecutive patients (pts) seen for care with a confirmed SARS-CoV-2 Omicron diagnosis and who met the AIFA criteria for eligibility for treatment with RDV, MLN, NMV/r, TIX/CIL, or SOT. Treatment allocation was subject to drug availability, time from symptoms onset, and comorbidities. Nasopharyngeal swab (NPS) VL was measured on day 1 (D1) and D7 and was expressed by log2 cycle threshold (CT) scale. Comparisons between treatment groups were made by Chi-square, and Wilcoxon paired tests. Primary endpoint was D1-D7 VL variation. Potential decrease in VL and average treatment effect (ATE) were calculated from fitting marginal linear regression models weighted for calendar month of drug initiation, duration of symptoms, and immunodeficiency using NMV/r as the comparator trial arm. Result(s): A total of 971 pts received treatments (SOT 321, MLN 231, NMV/r 211, TIX/CIL 70, and RDV 138): female 457 (47%), median age 67 yrs (IQR 56-78), 93% vaccinated;12% with negative baseline serology. At D1, median time from symptoms onset was 3 days (IQR 2,4). 379 (39%) pts were infected with BA.1, 215 (22%) with BA.2, 372 with BA.4/5 (38%), and 5 with BQ.1 (0,5%). D1 mean viral load was 4.02 log2. Adjusted analysis (ATE) showed that NMV/r significantly reduced VL compared to all the other drugs in pts infected with all sublineages, (Fig.1A-B) while less evidence for a difference vs. TIX/CIL was seen in those infected with BA.2 (p=0.05) (Fig.1 C-D). Conclusion(s): In this analysis of in vivo early VL reductions, NMV/r appears to be the drug showing the greatest antiviral activity, regardless of the underlying subvariant, perhaps with the exception of TIX/CIL in people infected with BA.2 for which there was less evidence for a difference. In the Omicron era, due to the high prevalence of vaccinated people and in absence of clinical events, VL is one of the possible alternative endpoints which guarantees adequate statistical power. Fig 1 SARS-CoV-2 RNA levels at D1 and D7 in patients treated with Nirmatrelvir/ ritonavir, Sotrovimab, Molnupiravir, Remdesivir, and Tixagevimab/cilgavimab. Dot-plots showing the comparison of viral loads detected at D1 and D7 and the variation of RNA levels observed between the two time-points by intervention in (A) all patients treated with Nirmatrelvir/ritonavir (n=211), Sotrovimab (n=321), or Molnupiravir (n=231), or Remdesivir (n=138), or Tixagevimab/ cilgavimab (n=136);(C) patients with Omicron BA.2 infection treated with Nirmatrelvir/ritonavir (n=58), Sotrovimab (n=81), or Molnupiravir (n=21), or Remdesivir (n=37), or Tixagevimab/cilgavimab (n=18);(D) patients with Omicron BA.4/5 infection treated with Nirmatrelvir/ritonavir (n=102), Sotrovimab (n=92), or Molnupiravir (n=110), or Remdesivir (n=16), or Tixagevimab/cilgavimab (n=52). Viral RNA levels are expressed as log2 CT values. The horizontal dashed line represents the limit of detection (CT: 40.0), values >=40 are considered negative. Mean of log2 CT values, and SD are shown in the graph. Statistical analysis of the differences in viral loads by intervention as compared to Nirmatrelvir/ritonavir was performed by Mann-Whitney test. Potential decrease in VL and average treatment effect (ATE) were calculated from fitting marginal linear regression models weighted for calendar month of drug initiation, duration of symptoms, and immunodeficiency using NMV/r as the comparator trial arm. Results are shown (B) for patients infected with all Omicron sublineages and (D) for those infected with Omicron BA.2 sublineage.
ABSTRACT
Background: Few data are available about comparison of different monoclonal antibodies (MAbs) for COVID-19 in the real-world setting. We aim to compare effectiveness of bamlanivimab/etesevimab (BAM/ETE) versus (vs) casirivimab/imdevimab (CAS/IMD) and to estimate predictors of hospitalization/death. Methods: Observational analysis of all consecutive outpatients (pts) with mild/moderate COVID-19 enrolled within the AIFA access program in a single center in Rome, from March to October, 2021. At first baseline (BL) visit, RT-PCR from nasopharyngeal swab with cycle thereshold (CT) measurement and viral sequencing was performed. Pts received intravenous BAM/ETE (700/1400 mg) or CAS/IMD (1200/1200 mg) and were followed through day 30. Primary endpoint was hospitalization/death due to severe COVID-19 by day 30. Average treatment effect (ATE) in the multiplicative scale of the odds was the chosen estimand to compare the two treatments, adjusted for age, obesity, time from onset to infusion, median C-reactive protein (CRP), vaccination, variant of concern (VOC) and BL-CT. Predictors of clinical failure were explored by two different models of multivariable logistic regression. Results: 242 pts receiving BAM/ETE (n=76) or CAS/IMD (n=166) were included (male 54%;median age 65 yrs;median SpO2 97%;diabetes 12%;hypertension 40%;CVD 17%;COPD 26%;autoimmune diseases 12%;immunodeficiency 18%). Median time from symptoms onset to infusion was 4 days (IQR 3-6). No differences were observed between the two MAbs for BL characteristics except for BMI>35 (BAM/ETE 24%, CAS/IMD 12%), CRP (BAM/ETE 1.8, CAS/IMD 1.2), vaccination (BAM/ETE 26%, CAS/IMD 46%) and distribution of VOC (Alpha 46% BAM/ETE vs 22% CAS/IMD;Gamma 20% vs 7%;Delta 5% vs 55%). Proportion of patients with COVID-related hospitalization/death by day 30 was 12/76 (15.8%) for BAM/ETE and 6/166 (3.6%) for CAS/IMD. Estimate of causal effect of BAM/ETE exposure compared to CAS/IMD on primary end point by ATE is reported in Table 1a. Factors associated with an increased risk of clinical failure by fitting multivariable logistic regression were BMI >35 and P1/Gamma VOC;higher BL-CT was associated with a reduced risk (Table 1b-1c). Conclusion: In a real-life setting, receiving BAM/ETE was associated with a 4-fold higher risk of COVID-19 progression to hospitalization/death than CAS/IMD. SARS-CoV-2 P.1/Gamma, but not B.1617.2/Delta VOC, obesity and higher BL viral load also predicted an increased risk of clinical worsening.