Establishment, implementation and utilization of a Human COVID-19 biospecimen biorepository at UCLA

Background: We have developed a semiautomated infrastructure to identify remnant clinical fluid samples to support specimen collection for UCLA Institute of Precision Health from patients who have opted in through the UCLA institutional Universal Consent. This system has allowed us to collect unique remnant fluid biosamples for subsequent DNA extraction and genotyping on over 30,000 patients to date. Here we describe the modification of this system to efficiently capture specimens from patients with SARS-CoV-2 causing coronavirus disease (COVID-19). Methods: We utilized a COVID-19 patient registry to identify potential sources of remnant biospecimens. This dynamically linked registry to UCLA Healthcare's electronic health record system includes all patients who have had at least one resulted COVID-19 PCR or Antibody IgG test. A COVID-19 remnant specimen report is generated three times each week by cross-referencing the COVID-19 patient registry to a listing of all clinical biospecimens scheduled for disposal for the next three days. This report is refined by selecting patients who have had a positive PCR or IgG test result and compared to our current inventory to identify specimens from previously unsampled patients. Additional filtering is performed to track the interval between positive test date and sample collection date, allowing for the collection of longitudinal plasma/serum samples. Through our Biomaterial Tracking and Management System (BTM, Daedalus Software Inc.), reports are generated daily to detect sample collection from onsite and offsite labs, to pull and courier samples to our repository. Whole blood samples are retrieved for DNA, serum, plasma and PBMC isolation. We also initiated the generation of an automated, weekly report from the EHR that alerts us to any autopsies from COVID-19 positive patients for potential remnant tissue samples. Results: From March to November, over 11,700 biospecimens have been collected from 1,300 unique patients, consisting of whole blood, DNA, PaxGene, PBMC, plasma, serum, and various tissue samples. Over 2,370 biospecimens have been released to approved study teams. Conclusions: We have successfully established a semiautomated infrastructure and workflow to capture annotated specimens from patients infected with SARS-CoV-2 that significantly reduces the chance that specimens may be missed. This allows for the efficient generation of a large number of specimens that researchers can rapidly obtain for study.