SARS-CoV-2 shedding and evolution in immunocompromised hosts during the Omicron period: a multicenter prospective analysis

Abstract

Background: Prolonged SARS-CoV-2 infections in immunocompromised hosts may predict or source the emergence of highly mutated variants. The types of immunosuppression placing patients at highest risk for prolonged infection and associated intrahost viral evolution remain unclear.

Methods: Adults aged ≥18 years were enrolled at 5 hospitals and followed from 4/11/2022 - 2/1/2023. Eligible patients were SARS-CoV-2-positive in the previous 14 days and had a moderate or severely immunocompromising condition or treatment. Nasal specimens were tested by rRT-PCR every 2-4 weeks until negative in consecutive specimens. Positive specimens underwent viral culture and whole genome sequencing. A Cox proportional hazards model was used to assess factors associated with duration of infection.

Results: We enrolled 150 patients with: B cell malignancy or anti-B cell therapy (n=18), solid organ or hematopoietic stem cell transplant (SOT/HSCT) (n=59), AIDS (n=5), non-B cell malignancy (n=23), and autoimmune/autoinflammatory conditions (n=45). Thirty-eight (25%) were rRT-PCR-positive and 12 (8%) were culture-positive ≥21 days after initial SARS-CoV-2 detection or illness onset. Patients with B cell dysfunction had longer duration of rRT-PCR-positivity compared to those with autoimmune/autoinflammatory conditions (aHR 0.32, 95% CI 0.15-0.64). Consensus (>50% frequency) spike mutations were identified in 5 individuals who were rRT-PCR-positive >56 days; 61% were in the receptor-binding domain (RBD). Mutations shared by multiple individuals were rare (<5%) in global circulation.

Conclusions: In this cohort, prolonged replication-competent Omicron SARS-CoV-2 infections were uncommon. Within-host evolutionary rates were similar across patients, but individuals with infections lasting >56 days accumulated spike mutations, which were distinct from those seen globally.

Keywords: COVID-19; SARS-CoV-2; evolution; immunocompromise; prolonged replication.

Figure 1.

Temporal dynamics of SARS-CoV-2 RNA viral load and culture positivity in 121 immunocompromised patients with SARS-CoV-2 infection. (A) Cycle threshold (Ct) values for total SARS-CoV-2 RNA and virus culture isolation over time in 121 patients by immunocompromised group. Open and closed circles indicate culture negative and positive specimens, respectively. (B) Kaplan-Meier survival curves showing time to last positive rRT-PCR test by immunocompromised group. p = 0.003 for difference in time to last positive specimen across groups.

Figure 2.

Within-host evolution of SARS-CoV-2 in 104 immunocompromised patients. (A) Stacked columns show percent of newly arising mutations identified at >2% frequency for specimens collected during the indicated time periods with mutation types color-coded: nonsynonymous (teal), noncoding (purple), stop codon (blue), synonymous (pink). Number of samples in each group is listed atop each bar; n = 93. (B) Genome-wide within-host divergence rate for individuals positive for SARS-CoV-2 by rRT-PCR for <21 days (n = 72) compared to those positive for ≥21 days (n = 16). Individuals in each group with rates of zero (e.g., no mutations identified ≥2% frequency in the final specimen) are not plotted given log transformation of y-axis and are indicated in parentheses at bottom of plot; however, these were included in the statistical analysis. Mann Whitney U test p=0.03 for differences in nonsynonymous rates and p=0.29 for differences in synonymous rates between prolonged and self-limited infections. Points are color-coded by immunocompromised group: B cell dysfunction, purple; SOT or HSCT, teal; AIDS, blue; non-B cell malignancy, pink; autoimmune/autoinflammatory, orange.

Figure 3.

De novo non-synonymous SARS-CoV-2 mutations in 65 immunocompromised hosts. (A) Mutations shared by the indicated number of individuals (y-axis), color coded by gene. Amino acid substitutions are labeled if shared by ≥5% (n=4) of patients. (B) Mutations in spike shared by the indicated number of individuals (y-axis), color coded by domain. Amino acid substitutions are labeled if shared by ≥5% (n=4) of patients. (C) Heatmaps of de novo nonsynonymous mutations in SARS-CoV-2 spike and their frequencies in five individuals with infections lasting >56 days and with ≥2 sequenced samples. Patients are color coded by immunocompromised group: B cell dysfunction, purple; SOT or HSCT, teal; AIDS, blue; non-B cell malignancy, pink; autoimmune/autoinflammatory, orange. Day of infection is indicated on the Y axis. EV138 received bebtelovimab on day 1, and EV022 received bebtelovimab at day 68. Mutations in the receptor binding domain are indicated by bold italics. Bisected squares indicate more than one codon mutation identified produced the same amino acid substitution. In EV022, shading at position 371 reflects the combined frequency of the 371F and 371P alleles.

Figure 4.

Mutations in 15 patients who received antiviral treatment. (A) Heatmap of de novo nonsynonymous mutations in SARS-CoV-2 spike among immunocompromised patients who received monoclonal antibody (bebtelovimab, sotrovimab, and/or tixagevimab/cilgavimab) and had a post-treatment sample that was sequenced (n=10). Sixteen patients had a post-treatment monoclonal antibody sample; of these, 10 had de novo non-synonymous mutations in spike. Patients are color coded by immunocompromised group: B cell dysfunction, purple; SOT or HSCT, teal; AIDS, blue; non-B cell malignancy, pink; autoimmune/autoinflammatory, orange. Monoclonal antibody received and treatment timepoints are denoted for each patient. Mutations in the receptor binding domain are shown in bold italics. Bisected squares indicate more than one codon mutation produced the same amino acid substitution. (B) Heatmap of mutations in SARS-CoV-2 nsp12 (RNA dependent RNA polymerase) among immunocompromised patients who received remdesivir and had a post-treatment sample (n=7). 17 patients had a post-treatment remdesivir sample; of these, 7 had de novo non-synonymous mutations in nsp12. In both (A) and (B), the day of infection is indicated to the left of heatmap, and the day of treatment is indicated to the right.