A longitudinal SARS-CoV-2 biorepository for COVID-19 survivors with and without post-acute sequelae

Since its initial identification in December 2019, severe acute respiratory syndrome-coronavirus-2 (SARS-CoV-2) has caused a global pandemic, killing millions and infecting over 100 million individuals worldwide [1, 2]. SARS-CoV-2 leaves behind millions of survivors, most of whom recover, but for many, symptoms persist [3‐7]. A broad range of symptoms can last over 6 months, and can include persistence of initial acute symptoms or development of new symptoms [3, 8, 9]. Persistent or new symptoms affecting COVID-19 survivors are named post-acute sequelae of COVID-19 (PASC) [9, 10]. Risk factors for developing PASC are currently being defined globally. A systematic and calculated approach to investigating the course of SARS-CoV-2 infection is needed to understand varied susceptibility to severe SARS-CoV-2 disease and COVID-19 sequelae. Scientists and industry have expressed the importance of a biorepository as an essential building block for critical research [11‐17]. Biobanking efforts for SARS-CoV-2 are underway globally, yet most are comprised of limited specimen types collected from participants at a single time point [16, 18‐21]. The Northern Colorado SARS-CoV-2 Biorepository (NoCo-COBIO) is uniquely multifaceted and contains longitudinal clinical data, electronic medical records (EMR), demographics, quality of life surveys, and biological specimens of stool, saliva, nasopharyngeal swab specimens, plasma, serum, peripheral blood mononuclear cells (PBMCs), and breastmilk if available, from each SARS-CoV-2-infected participant. The biospecimens are immensely valuable to longitudinally evaluate mechanisms underlying persistent symptoms and PASC. This clinically integrated-biorepository and COVID-19 survivorship project has been successful in securing patient compliance to long-term follow-up visits with the study team, and is a powerful resource for global collaboration and co-creation of statistically-powered study designs for assessing molecular, immune, and metabolic signatures associated with PASC in adults [22].

To date, 119 participants have been enrolled (Table 2). Thirty-one participants are classified as having had a severe initial infection (defined as oxygen requirement greater than 5 L/min), 30 experienced a moderate infection (defined as oxygen requirement of 1-5 L/min), and 58 experienced a mild or asymptomatic disease course (no oxygen required).

Sixty-six percent (N = 79) of all participants have completed 2 visits, 50% (N = 59) have completed 3 visits, and 33% (N = 39) have completed 4 visits. Including follow-ups, 301 visits have been conducted, and at 88% of these visits, participants were able to provide all samples requested. Participant visits are conducted at the time of enrollment (Visit 1-Baseline), and at 1 month (Visit 2), 3 months (Visit 3), and 6 months (Visit 4) from the time of enrollment. Nearly every visit or study encounter has yielded blood, nasopharyngeal, and saliva samples (99, 99, and 98% respectively), and 88% of visits yielded stool samples. The team typically processes 10 sample sets in a single day. The time to obtain, process, and store samples from 10 participants is approximately 50 personnel hours (5–7 people) and costs $1500 U.S. dollars per sample.

The participants included in the biorepository thus far represent a broad age range (18–82 years), and the elderly were more likely to have moderate or severe COVID-19 (Table 2). Seventy-one percent of participants are females, and 29% are males. Of participants that suffered severe disease, 69% are males. Nineteen participants identify as Hispanic or Latino, and of these, 15 were hospitalized with COVID-19. Obese participants in our cohort were also more likely to suffer from moderate or severe COVID-19. The average body mass index (BMI) of participants with mild disease was 26, in contrast to an average BMI of 36.5 in both the moderate and severe groups (Table 2).

PASC was defined herein as having at least one common persistent symptom, and has been previously described in the literature [3, 6, 27] for including fatigue or tiredness, dyspnea, difficulty breathing, or shortness of breath, chest pain, joint pain, and perceived cognitive impairment such as “brain fog”, absent-minded, forgetfulness, difficulty concentrating or confusion) at any follow-up visit. Of the 79 participants who have completed follow-up visits, 75 have completed the persistent symptoms questionnaire. Forty-nine percent of participants who followed up (N = 37 of 75) had PASC at any point during their follow-up visits, with symptom surveillance performed at up to 8 months from infection. Significantly more hospitalized participants (93%, N = 25 of 27) developed PASC compared to 23% of those not requiring hospitalization (N = 11 of 47) (p = 0.0001). Participants who had moderate and severe COVID-19 were equally as likely to develop PASC (92.3 and 92.9%, respectively). At 25–89 days post positive SARS-CoV-2 PCR, 55% of participants who attended clinic (N = 31 of 56) had at least one persistent symptom attributed to COVID-19 (Fig. 2a). At 90–174 days post positive SARS-CoV-2 PCR, 67% (N = 31 of 49) of participants attending clinic had persistent symptoms (Fig. 2b). Sixty-four percent of all participants (Fig. 2c), and 90% of hospitalized patients still reported symptoms after 175 post initial infection, with the most symptoms reported in this group being fatigue, difficulty concentrating, absent-minded or forgetful (also described as brain fog), ageusia, and anosmia. Participants with moderate disease reported more symptoms than participants with mild or severe disease, however those with severe disease were more likely to experience difficulty breathing and joint pain (Fig. 2d).

The individuals included in this biorepository also have a variety of underlying health conditions that have been associated with severe COVID-19, including, but not limited to, obesity, stroke, diabetes mellitus, liver disease, connective tissue disease, chronic obstructive pulmonary disease (COPD), congestive heart failure, and coronary artery disease [23, 24, 30‐33]. In our cohort, PASC is more prevalent in those with obesity and advanced age. This may be confounded by disease severity, given most participants with severe COVID-19 are obese and elderly, and a larger sample size is needed to correct for bias. When comparing those with and without PASC in participants with mild disease only, there was no significant difference in age, gender, BMI, or underlying health conditions. Interestingly, nearly all patients with PASC had normal physical exams.

SF-36 surveys were performed in 69 participants: 33 with PASC and 36 without PASC. The survey was performed at any point between day 21 and day 288 after positive SARS-CoV-2 PCR, with a mean of day 125. Participants with PASC had mean total score of 2019 out of a maximum score of 3600, while those without PASC had a higher mean total score of 2934. On average, participants with PASC scored lower in each of the 8 individual categories measured with SF-36 when compared to those without PASC. Questions in each of the 8 categories are averaged to give a maximum score of 100 for each category. The individual score differences between those with PASC and without are as follows: physical functioning (59 vs 92), role of limitations due to physical health (42 vs 77), role of limitations due to emotional problems (57 vs 85), energy and fatigue (38 vs 63), emotional well-being (71 vs 80), social functioning (70 vs 87), pain (66 vs 85), and general health (53 vs 79) (Fig. 3).

Although challenging to implement and establish, longitudinal SARS-CoV-2 biobanking efforts are a valuable investment and resource to expedite research and further our knowledge of COVID-19 and PASC [11, 22, 34]. In this cohort, obesity, age, male sex, and Hispanic/Latinx ethnicity were risk factors for severe COVID-19. These clinical characteristics have been previously associated with severe disease [35‐37], validating the NoCo-COBIO cohort. Hospitalized patients were also significantly more likely to suffer from PASC than non-hospitalized participants, similar to what has been previously reported for PASC [4, 5, 7]. The duration of PASC in this prospective cohort analysis is extensive, and we found that symptoms persisted in 89% of hospitalized participants at 90–174 days post infection, and in 90% after 174 days. Some of these participants required prolonged hospital stay, and many symptoms could be attributed to injuries or sequelae sustained from ICU stay alone [38]. However, those hospitalized with moderate disease (oxygen requirement of 1-5 L/min, not requiring ICU stay) were equally as likely to develop PASC as those who had severe disease (oxygen requirement of greater than 5 L/min). Twenty-three percent of our participants who were not hospitalized for COVID-19 also developed PASC. Although those who suffered mild COVID-19 are not at as high of a risk for developing PASC as those who were hospitalized, a staggering amount of people could be affected by PASC given high infection rates in the population. We have enrolled adults with no history of SARS-CoV-2 infection for matched biospecimens (N = 17). These non-infected individuals also completed the symptom questionnaire and did not report any symptoms. This is important, given some symptoms of PASC such as anxiety or fatigue could be perceived as attributable to social isolation and decreased activity during the pandemic. The adults with PASC had lower quality of life scores on SF-36 surveys when compared to those without PASC, consistent with previous research [39]. PASC could further burden our health care system during an already strained pandemic time, and continued prospective follow-up is essential to understand the pathophysiology underlying the physical difficulties and emotional complications of PASC [40, 41]. We aim to further analyze our cohort to find immunologic, metabolic, microbial, proteomic, and long-term clinical differences in those with and without PASC. Physical examinations and symptom surveillance over time increase the significance of this biorepository as it allows for depth during examination of those with post-viral sequelae. The administration of quality-of-life questionnaires is an additional unique aspect of this protocol, providing a structured tool to further assess those with PASC. This multifaceted approach to post-viral symptom surveillance is of important correlative and potential causal biological value with the linkages to innumerable data points and analytes that are under investigation from the specimens.

Though we were able to include individuals with Hispanic or Latinx ethnicity, the lack of racial diversity is one major shortcoming of the biorepository and we aim to enroll participants of various races to the biobank. Another shortcoming is the lack of a control group. The non-infected adults enrolled were healthier and younger than the COVID-19 infected participants. We do seek to enroll a participant-matched control group for age, BMI, sex and other co-morbidity characteristics. Additionally, the Cameron Peak wildfire was a major disturbance to air quality and outdoor activity in northern Colorado during the August–October 2020 portion of our data collection, which may have affected health perception and symptoms.

This biobank protocol with a manual of operations and standard operating procedures provides a template for other academic-clinical teams that reside outside of traditional metropolitan settings to replicate for ensuring compliance and follow-up for patient-matched, multi-matrix biospecimens. These stored sample sets are needed for improving diagnostic designs, advancing therapeutics research and development of SARS-CoV-2 vaccines, as well as understanding those with prolonged course of illness and those without PASC.

People have diverse and complex courses of illness from COVID-19 across organ systems. In this cohort, patients who were hospitalized for COVID-19 were more likely to have PASC than those not requiring hospitalization, yet the 23% of patients who were not hospitalized and developed PASC also merit research attention and quality of life assessments. This patient-matched, multi-matrix, longitudinal biorepository from COVID-19 survivors will allow for current and future research to better understand the pathophysiology of disease, prolonged symptoms and to identify targeted interventions to reduce risk for PASC.