371. Estimating SARS-CoV-2 Seroprevalence from Spent Blood Samples, January–March 2021

Abstract Background Measuring SARS-CoV-2 antibody prevalence in spent samples at serial time points can determine seropositivity in a diverse pool of individuals to inform understanding of trends as vaccinations are implemented. Methods Blood samples collected for clinical testing and then discarded ("spent samples") were obtained from the clinical laboratory of a medical center in Atlanta. A convenience sample of spent samples from both inpatients (medical/surgical floors, intensive care, obstetrics) and outpatients (clinics and ambulatory surgery) were collected one day per week from January-March 2021. Samples were matched to clinical data from the electronic medical record. In-house single dilution serological assays for SARS-CoV-2 receptor binding domain (RBD) and nucleocapsid (N) antibodies were developed and validated using pre-pandemic and PCR-confirmed COVID-19 patient serum and plasma samples (Figure 1). ELISA optical density (OD) cutoffs for seroconversion were chosen using receiver operating characteristic analysis with areas under the curve for all four assays greater than 0.95 after 14 days post symptom onset. IgG profiles were defined as natural infection (RBD and N positive) or vaccinated (RBD positive, N negative). Figure 1. Nucleocapsid serology assay validation Single dilution serological assays for SARS-CoV-2 nucleocapsid antibodies were validated using pre-pandemic and PCR-confirmed COVID-19 patient serum and plasma samples. ELISA optical density (OD) cutoffs for seroconversion were chosen using receiver operating characteristic (ROC) analysis with areas under the curve (AUC) for all four assays greater than 0.95 after 14 days post symptom onset. Results A total of 2406 samples were collected from 2132 unique patients. Median age was 58 years (IQR 40-70), with 766 (36%) ≥ 65 years. The majority were female (1173, 55%), and 1341 (63%) were Black. Median Elixhauser comorbidity index was 5 (IQR 2-9). 210 (9.9%) patients ever had SARS-CoV-2 detected by PCR, and 191 (9.0%) received a COVID-19 vaccine within the health system. Nearly half (1186/2406, 49.3%) of samples were collected from inpatient units, 586 (24.4%) from outpatient labs, 403 (16.8%) from the emergency department, and 231 (9.6%) from infusion centers. Overall, 17.0% had the IgG natural infection profile, while 16.2% had a vaccination profile. Prevalence estimates for IgG due to natural infection ranged from 24.0% in week 2 to 9.7% in week 5, and for IgG due to vaccine from 4.4% in week 2 to 32.0% in week 6 (Table, Figure 2). Table. SARS-CoV-2 antibody seropositivity by week of sample collection for spent routine blood chemistry samples. RBD = receptor binding domain. N = nucleocapsid. Seropositivity defined by enzyme-linked immunoassay (ELISA) optical density cutoffs selected using receiver operating characteristic analysis with areas under the curve (AUC) for all four assays greater than 0.95 after 14 days post symptom onset. IgG defined as positive if both RBD and N seropositive. Figure 2. RBD and Nucleocapsid seropositivity to differentiate natural infection vs. vaccination by week of sample collection. RBD = receptor binding domain. N = nucleocapsid. Seropositivity defined by enzyme-linked immunoassay (ELISA) optical density cutoffs selected using receiver operating characteristic analysis with areas under the curve (AUC) for all four assays greater than 0.95 after 14 days post symptom onset. Conclusion Estimated SARS-CoV-2 IgG seroprevalence among patients at a medical center from January-March 2021 was 17% by natural infection, and 16% by vaccination. Weekly trends likely reflect community spread and vaccine uptake. Disclosures Daniel Graciaa, MD, MPH, MSc, Critica, Inc (Consultant)


Session: P-16. COVID-19 Epidemiology and Screening
Background. Real-time reverse transcriptase PCR (rRT-PCR) has become the primary method for detection of SARS-CoV-2. Specific measurements of cycle threshold (Ct) values can give an estimate of viral load. Previous studies have shown temporal trends in Ct values, which could be used to predict the phase of the pandemic. This study's goal was to examine the relationship between Ct and disease severity, as well as Ct trends.
Methods. Testing was performed using the Abbott M2000 SARS-CoV-2 assay. Data was collected for 262 SARS-CoV-2 positive patients from March-May 2020. Kruskal-Wallis testing was performed to determine differences in median Ct based on age, gender, race and ethnicity. To determine relationship between symptom onset and clinical severity with Ct, linear and logistic regression were performed.
Results. The majority of the patients had mild to moderate disease. Average time since symptom onset was 5.9 days, and 92% were symptomatic. Figure 1 demonstrates the distribution of Ct by disease severity at time of testing. There was no significant difference in cycle threshold by sex, age, race or ethnicity. Figure 2 shows weekly mean cycle threshold by total new cases in Massachusetts to reflect temporal trend of Ct and cases. In the multivariable linear regression model, Ct increased with days since symptom onset (P< 0.001). Cycle threshold was inversely associated with disease severity in multivariable logistic regression though (OR 1.06, 95%CI 1.01-1.11, p=0.03), even when controlling for time since symptom onset.  Line represents mean Ct over time period included in this study overlaid on total new cases in Massachusetts. Lower Ct were seen in the course as cases were increasing which peaked as cases stabilized.
Conclusion. Cycle threshold increased with time since symptom onset, consistent with prior data showing increasing Ct from time since infection due to decreasing viral replication. This study showed an inverse relationship between cycle threshold and disease severity, which differs from previous studies which demonstrated higher odds of progression to severe disease and mortality with lower Ct. This finding may reflect the disease severity associated with the secondary inflammatory phase of SARS-CoV-2 seen later in the disease course, although there was only moderate correlation between Ct and time since symptom onset. Further research is needed to better understand the role of Ct in predicting clinical severity of SARS-CoV-2 infections.
Disclosures. Background. Measuring SARS-CoV-2 antibody prevalence in spent samples at serial time points can determine seropositivity in a diverse pool of individuals to inform understanding of trends as vaccinations are implemented.
Methods. Blood samples collected for clinical testing and then discarded ("spent samples") were obtained from the clinical laboratory of a medical center in Atlanta. A convenience sample of spent samples from both inpatients (medical/surgical floors, intensive care, obstetrics) and outpatients (clinics and ambulatory surgery) were collected one day per week from January-March 2021. Samples were matched to clinical data from the electronic medical record. In-house single dilution serological assays for SARS-CoV-2 receptor binding domain (RBD) and nucleocapsid (N) antibodies were developed and validated using pre-pandemic and PCR-confirmed COVID-19 patient serum and plasma samples (Figure 1). ELISA optical density (OD) cutoffs for seroconversion were chosen using receiver operating characteristic analysis with areas under the curve for all four assays greater than 0.95 after 14 days post symptom onset. IgG profiles were defined as natural infection (RBD and N positive) or vaccinated (RBD positive, N negative).

Figure 1. Nucleocapsid serology assay validation
Single dilution serological assays for SARS-CoV-2 nucleocapsid antibodies were validated using pre-pandemic and PCR-confirmed COVID-19 patient serum and plasma samples. ELISA optical density (OD) cutoffs for seroconversion were chosen using receiver operating characteristic (ROC) analysis with areas under the curve (AUC) for all four assays greater than 0.95 after 14 days post symptom onset.
Results. A total of 2406 samples were collected from 2132 unique patients. Median age was 58 years (IQR 40-70), with 766 (36%) ≥ 65 years. The majority were female (1173, 55%), and 1341 (63%) were Black. Median Elixhauser comorbidity index was 5 (IQR 2-9). 210 (9.9%) patients ever had SARS-CoV-2 detected by PCR, and 191 (9.0%) received a COVID-19 vaccine within the health system. Nearly half (1186/2406, 49.3%) of samples were collected from inpatient units, 586 (24.4%) from outpatient labs, 403 (16.8%) from the emergency department, and 231 (9.6%) from infusion centers. Overall, 17.0% had the IgG natural infection profile, while 16.2% had a vaccination profile. Prevalence estimates for IgG due to natural infection ranged from 24.0% in week 2 to 9.7% in week 5, and for IgG due to vaccine from 4.4% in week 2 to 32.0% in week 6 (Table, Figure 2). Table. SARS-CoV-2 antibody seropositivity by week of sample collection for spent routine blood chemistry samples. RBD = receptor binding domain. N = nucleocapsid. Seropositivity defined by enzyme-linked immunoassay (ELISA) optical density cutoffs selected using receiver operating characteristic analysis with areas under the curve (AUC) for all four assays greater than 0.95 after 14 days post symptom onset. IgG defined as positive if both RBD and N seropositive. RBD = receptor binding domain. N = nucleocapsid. Seropositivity defined by enzyme-linked immunoassay (ELISA) optical density cutoffs selected using receiver operating characteristic analysis with areas under the curve (AUC) for all four assays greater than 0.95 after 14 days post symptom onset.
Conclusion. Estimated SARS-CoV-2 IgG seroprevalence among patients at a medical center from January-March 2021 was 17% by natural infection, and 16% by vaccination. Weekly trends likely reflect community spread and vaccine uptake.
Disclosures. Daniel Graciaa, MD, MPH, MSc, Critica, Inc (Consultant) Background. Tissue donors are evaluated for communicable disease in order to minimize the risk of transmission to recipients. Although there are data suggesting SARS-CoV-2 viremia across a wide spectrum of illness, prevalence in deceased tissue donors and the potential for transplant transmission are unknown.

Detection of SARS-CoV-2 RNAemia in Deceased Tissue Donors
Methods. Eight tissue banks participated in a retrospective analysis of samples from eligible deceased tissue donors from Oct 2019 through June 2020, one participant in Canada and the remainder located in the United States. All four Census regions of the continental US and all major racial-ethnic groups were represented. EDTA or sodium citrate plasma aliquots were tested in singlicate with the Research Use Only Procleix SARS-CoV-2 Assay on the Procleix Panther System, which uses transcription-mediated nucleic acid amplification (TMA) technology for detection of the SARS-CoV-2 RNA. Plasma (or if unavailable, serum) aliquots were sent to Grifols for an alternate SARS-CoV-2 nucleic acid amplification (NAT) test to verify reactivity and also sent for antibody testing using the emergency use authorization Ortho VITROS Immunodiagnostic Products Anti-SARS-CoV-2 Total test. The VITROS assay uses immunometric technology for qualitative measurement of total antibody (IgG, IgA and IgM) to SARS-CoV-2. The proportion of donors with confirmed RNAemia (i.e., presence of SARS-CoV-2 RNA in plasma or serum) and 95% confidence intervals were computed.
Results. Of 3,455 donor samples with valid final results, 26 (0.76%) were initially positive for SARS-CoV-2 RNA; of these, 3 were confirmed by alternate NAT. Of donor samples collected in 2019 0.00% (95% CI: 0.00%,0.43%) were confirmed RNAemic, while of those collected in 2020, 0.12% (0.04%,0.34%) were confirmed RNAemic. One of 26 initial positive, and none of the three samples confirmed by alternate NAT, tested positive for anti-SARS-CoV-2 Spike antibodies by serology. Infectivity studies are pending on one sample with sufficient available volume.
Conclusion. The rate of SARS-CoV-2 RNAemia in deceased tissue donors is approximately 1 per 1,000, and it is unknown whether this RNAemia reflects the presence of infectious virus. Given these results, the risk of transmission through tissue is most likely to be low.