Study population and setting
This was a retrospective cohort study examining the genomic characteristics and clinical outcomes of B.1.1.7 infection among patients admitted on or before Dec 20, 2020 at two hospitals in the United Kingdom (the University College London hospital (UCLH) and North Middlesex University hospital (NMUH)). Samples which were positive for SARS-CoV-2 infection collected from these patients were sequenced for the presence of variant of concern (VOC), (B.1.1.7). A total of 496 patients had positive PCR tests for SARS-CoV-2. Next generation sequencing (NGS) was conducted on 339 (68%). Of these, 141 (42%) were non-B.1.1.7 variants and 198 (58%) belonged to the B.1.1.7 lineage. Poisson regression with robust estimation was used to calculate the prevalence rate (PR) of death or severe disease over 28 days following symptom onset or PCR positive test among initially asymptomatic subjects. Models were adjusted for hospital, age (45 years or less, 45-59, 60-74, 74 years or more), sex, ethnicity (white or others), and presence of comorbidities (0, 1, or 2 or more comorbidities according to the Charlson Comorbidity index). Severe disease was defined as a score of 6 or more on the WHO ordinal scale of respiratory infection severity (range 0-10: 0 indicates asymptomatic disease, 6 and above indicates requirement of oxygen with positive pressure support, and 10 indicates death). Sixty-five percent of participants had a treatment escalation plan (advanced directive) that restricted the amount of organ support they would receive.
Summary of Main Findings
The median age of participants was 60 years (49% male, 42% with available ethnicity were non-white. While there was no difference between those who had and didn’t have next generation sequencing (NGS) by age, sex, ethnicity, or hospital admission, the proportion of death and severe disease was higher among those who had NGS compared to those who did not (37% vs 24%, respectively). Patients who died or had severe disease were more likely to be older in age and have comorbidities. In both unadjusted and adjusted models, there was no significant difference in the proportion of death or severe COVID-19 comparing those infected with the B.1.1.7 variant with those infected with a non-B.1.1.7 variant (non-VOC) (PR 1.02, 95%CI: 0.76-1.38). This result remained consistent within subgroups of age, sex, ethnicity, and in sensitivity analyses (after excluding individuals with any missing data or those with a treatment escalation plan, and when including subjects with a positive PCR test for SARS-CoV-2 prior to hospitalization). Among those with Ct (cycle threshold) values available (65 patients), B.1.1.7 samples had a significantly lower median Ct value than non-VOC samples (p value 0.0030) and a higher median genomic read depth (p value 0.003), suggesting higher viral loads among those infected with B.1.1.7. These trends persisted in various sensitivity analyses. Patients with the B.1.1.7 variant experienced a longer time from symptom onset to hospital admission compared to non-VOC patients (median of 6 VS 4 days), but this difference was not statistically significant. Samples from 34 patients with prolonged viral shedding (more than 21 days) and from 32 patients treated with remdesivir revealed no evidence of VOC mutations. Genomic diversity was lower in samples with the B.1.1.7 variant compared to other variants, suggesting easy transmissibility of a new virus strain with a common ancestor.
- This is one of the largest studies to investigate clinical outcomes of COVID-19 infection and describe the genomic characteristics among UK variants of concern compared to non-UK variants using a well-defined source population of COVID-19 patients who sought medical attention in two large hospitals in London.
Approximately half of the study population had missing BMI data, which may have biased the results if the distribution of BMI values was different by viral lineage, as BMI is associated with more severe disease and worse COVID-19 outcomes. Also, patients with sample available for next generation sequencing (NGS) were sicker compared to those who did not, which could have resulted in selection bias if the proportion of both variants who had severe disease and had NGS were different from those who did not have NGS.
The study adds critical information on the risk of death, COVID-19 disease severity, and viral load among those infected with the B.1.1.7 variant using patient data from two large hospitals in London.
This review was posted on: 28 May 2021