A 72-year-old immunocompromised man in France who had chronic lymphocytic leukemia associated with hypogammaglobinemia for 4 years experienced diarrhea, asthenia, fever, and cough associated with coronavirus disease (COVID-19). Although he had received 1 injection of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) mRNA vaccine (BNT162b2; Pfizer/BioNTech, https://www.pfizer.com) 20 days earlier, we confirmed a diagnosis of COVID-19 by using a semiquantitative SARS-CoV-2 reverse transcription PCR (RT-PCR) viral load assay. This assay showed a cycle threshold (C_t) value of 27 for a nasopharyngeal swab specimen. His most recent monoclonal antibody (mAb) chemotherapy treatment (venetoclax and rituximab) had been conducted 17 days earlier. Because of his immunocompromised status, treatment with bamlanivimab (LY-CoV555), a neutralizing IgG1 mAb, was initiated at day 0, 4 days after onset of symptoms (Table). The patient received an infusion of 700 mg in a single dose and was discharged.

Analysis of samples showed a high viral load in a nasopharyngeal swab specimen (C_t 20) and a blood sample (C_t 37) (Table). Three days after the mAb infusion, the patient’s symptoms worsened, and he was hospitalized in the Infectious Diseases Department at Grenoble Hospital (Grenoble, France) on day 6. The condition of the patient had deteriorated; he had an additional need for oxygen, which resulted in a convalescent-phase plasma transfusion on day 10.

After this treatment, the condition of the patient continued to deteriorate, and he was transferred to the intensive care unit on day 13. A high dose of corticosteroids was given on days 21‒26. This treatment resulted in an improvement of his respiratory condition, but the patient remained dependent on supplemental oxygen (6 L/min). The patient was discharged from the intensive care unit and returned to the infectious disease department on day 33, but still had a high viral load in nasopharyngeal swab specimens (C_t v20 on day 45).

Because of this persistent viral replication, the patient was given remdesivir on day 47 and this treatment was continued for 10 days (200 mg for 1 day, followed by 100 mg/d for 9 days). SARS-CoV-2 carriage in a nasopharyngeal swab specimen decreased during treatment, and the patient was discharged from the infectious disease department and transferred to a rehabilitation center. The nasopharyngeal swab specimen viral load became negative on day 61.

To monitor viral evolution, we performed a multiplex RT-PCR based on melting curve analyses with VirSNIP Kits (TIB Molbiol, https://www.tib-molbiol.de) to evaluate the presence of the S: E484K and S: N501Y mutations in SARS-CoV-2 variants. Three days after mAb treatment (day 3), RT-PCR results suggested the presence of S: N501Y and an absence of S:E484K on an nasopharyngeal swab specimen. On day 6, the S: N501Y mutation was still present but was also found associated with an undetermined mutation at position 484 (melting temperatures different from those of wild-type E and the mutated strain K). On day 11, we detected the S: N501Y mutation in a blood sample but found no mutation at position 484. No nasopharyngeal swab specimen or blood sample from before mAb administration was available for analysis and comparison.

Figure

Figure. Severe acute respiratory syndrome coronavirus 2 variants in immunocompromised patient in France given antibody monotherapy showing compartmentalization and variation of mutation frequency for the spike protein. Mutations of interest are...

We performed whole-genome sequencing on 12 clinical samples by using amplicon-based technology on the Ion Torrent Platform (ThermoFisher, https://www.thermofisher.com) according to the protocol of and plug-ins used by Sjaarda et al. (1). We confirmed results of this analysis by using the minimap2 program (2). This analysis detected clade 20I/501Y.V1, Alpha variant (Pangolin: B.1.1.7), on day 3 in nasopharyngeal swab specimens. Three days later (day 6), a novel mutation (G23012C, S: E484Q) appeared in nasopharyngeal swab specimens at frequency of 82%, which rapidly reached >99% (S: E484Q) 10 days after mAb treatment (Table; Figure). Eleven days after the mAb infusion, we detected an additional nucleotide mutation A23040G (S: Q493R) in only a blood sample at a frequency of 64%. This rate reached 76% at day 17 without any detection in nasopharyngeal swab specimens.

Clinical trials of monotherapy treatment for SARS-CoV-2 infection have shown that subsequent dynamic shifts in the viral population appear to be frequent (3,4). An in vitro model showed that E484 and Q493 are 2 amino acid mutations of the spike protein that are known to be critical for bamlanivimab binding (5,6). The S: E484Q mutation is a hotspot of escape and could reduce susceptibility to bamlanivimab by >1,000-fold (6) and S: Q493R by >6,666-fold (7). Use of bitherapy with bamlanivimab and etesevimab decreases the risk for emergence of drug-resistant variants (5,8). However, an escape mutation after use of this drug combination was recently described (7).

Our analysis identified signs of compartmentalized viral populations on the basis of sequences recovered in blood and nasopharyngeal swab samples (notably on day 17). Such a phenomenon has been reported in clinical trials (9,10). Further analysis is needed to distinguish genetic changes that occur in the primary viral population from apparent changes to clarify whether such escape mutants are enough to spread and persist in humans and how SARS-CoV-2 displays compartmentalized replication. Genomic surveillance for SARS-CoV-2 variants is encouraged for COVID-19 patients given mAbs as monotherapy or biotherapy.

Top