The impact of SARS-CoV-2 infection on infants and children is not well-defined. Children are typically asymptomatic or mildly symptomatic during the acute infection, although some can develop significant complications requiring intensive care. In order to capture the full range of SARS-CoV-2 infection in the pediatric population, a COVID-19 biospecimen collection study was designed and implemented, including patients seen in urgent care clinics and hospital settings, neonates born to SARS-CoV-2-infected mothers, and asymptomatic children. Each study population required specific tailoring of study conduct to effectively and efficiently collect critical samples.
Cornerstones of the biorepository included open dialogue between research and clinical team members, a sensitivity to procedures required for specimen collection in children, and clear documentation of study participation and sample collection. Close communication and collaborations with the adult COVID-19 biorepository enable paralleled recruitment efforts and processing procedures and ensured consistency and harmonization across patient cohorts, facilitating high-quality comparisons between patient groups and with adult cohorts. Central to the operation, physician-scientists in Pediatrics, Neonatology, Medicine-Pediatrics, and Obstetrics-Gynecology harmonized sample collection protocols, established clinical connections and provided clinical and scientific context to COVID-19-related research in the neonatal and pediatric population.
Ethical and Biosafety Review Board Processes:
To establish a Pediatric COVID-19 biorepository during the surge of COVID-19 cases locally, protocols were rapidly submitted to our Institutional Biosafety Committee (IBC, MGH IBC#2020B000061) and the Institutional Review Board (IRB, MGH IRB#2020P000955) for approval. During this initial wave of the COVID-19 pandemic in late March 2020, COVID-19 research proposals were prioritized by the IBC and IRB. A biosafety protocol was submitted to the IBC to transition from a Biosafety Level 2 (BSL2) to an enhanced BSL2+ laboratory environment, allowing collection, processing, and storage of SARS-CoV-2-infected samples. IBC approval was obtained within two weeks. An expedited IRB review facilitated IRB approval two days following submission. Figure 1 displays a timeline of study activity relative to the community surge of COVID-19 cases in Massachusetts.
Patient enrollment and sample collection:
Eligible participants in all cohorts were identified by screening outpatient clinic schedules or hospital admission lists, then discussing potential patients with care team members. If appropriate, parents/guardians or patients (if >18 years of age) were called by phone to introduce the study and, if interested in participation, an informed consent was completed by phone. Participants or their parent/guardian also selected which biospecimens they would provide to the biorepository. Assent was completed by phone, when possible, with parents/guardian present for children 7-17 years of age. In-person consent/assent was waived by the IRB to avoid close contact between patients and research staff and to abide by the social distancing measures implemented by the state. Witness requirements were also waived as a result of the restricted visitor hospital policy due to the pandemic. One copy of the signed consent form, and assent form if appropriate, was emailed to enrolled patients or parents/guardians, and another was uploaded into the electronic medical record flagging the patient as a research participant, clearly documenting research participation for the clinical care teams facilitating sample collection. Paper copies were not provided to participants due to the COVID-19 restrictions. Upon enrollment, participants were assigned a unique study ID number using REDCap, a secure, centralized online database platform that allows simultaneous recruitment at multiple sites without risking assigning the same number to multiple patients.
In order to include pediatric and neonatal patients from a range of clinical presentations for COVID-19, we established 4 cohorts of patients from ages 0-25 years, reflecting the ages of patients cared for by the pediatric teams during the surge of COVID-19 cases locally: 1) Pediatric patients with mild-moderate COVID-19, presenting to the MGH COVID-19 urgent care clinics, 2) Pediatric patients with severe COVID-19 or MIS-C requiring hospitalization, 3) Newborns born to mothers infected with SARS-CoV-2 at any point during their pregnancy and infants born to non-infected mothers, and 4) asymptomatic children presenting to their well-visits during the pandemic.
Each cohort presented unique challenges and required tailored strategies for a successful recruitment. In the pediatric urgent care units, challenges in enrollment included variability in patient volume and frequent rotation of nurses, medical assistants, and physicians. Research coordinators had to be flexible to adapt to the frequently changing workflow within the clinic. For enrollment in the hospitalized cohort, the research team remained attentive to new admission lists and promptly completed enrollment protocols in order to obtain specimens prior to the initiation of treatment, as interventions such as intravenous immunoglobulin or steroids would interfere with the natural immune responses to SARS-CoV-2 infection. Specimen collection from the hospitalized cohort was coordinated with clinical laboratory collections to minimize blood draws and collection procedures. For enrollment of newborns, the research coordinators coordinated with maternal arm of the adult COVID-19 biorepository to allow simultaneous enrollment of mother and newborns prior to birth facilitating the recruitment process and limiting non-clinical interactions with the mother during the perinatal period. Additionally, blood volumes collected for newborns were minimized to <1ml obtained by heal stick and coordinated with blood collection for newborn screening. Enrolling well-visits was challenging as blood draws are not routinely obtained in all ages and children are often unwilling to undergo voluntary venipuncture. In each cohort, the research staff was mindful of the physical and emotional stress the care teams were enduring while caring for patients with COVID-19 and therefore sought to minimize disruptions in clinical care. Figure 2 provides a schematic of the recruitment strategy.
1) Pediatric patients with mild-moderate COVID-19
As most children did not require hospital-level care, significant efforts were made to enroll patients in the outpatient setting. COVID-19 screening clinics, called Respiratory Infection Control clinics, were established at Massachusetts General Hospital. As COVID-19 symptoms are non-specific and current diagnostic reporting is time-delayed, all patients presenting to the pediatric COVID-19 screening clinics were eligible to participate in the biorepository. During the surge of COVID-19 cases locally, young adults up through 25 years of age were seen in the pediatric COVID-19 clinics.
For patients seen in the Respiratory Infection Control clinics, the CRC called eligible participants via telephone after acquiring approval from the lead physicians in the clinics. After acquiring verbal consent over the telephone, the CRC alerted the clinical team of patient enrollment and the clinical teams obtained specimens for research. Participants who provided informed consent could give nasopharyngeal, oropharyngeal swabs, and/or blood. Stool and urine were not collected given time limitations of clinic visits and patient flow patterns established to minimize potential COVID-19 exposures to clinical staff. Blood was collected into one tube with an EDTA anticoagulant (EDTA tube) (BD), one serum separator tube (SST) (BD), and a PAXgene RNA tube (BD). Blood volumes varied, depending on the age and weight of the patient, in accordance with limits established by the IRB. The aerosolizing procedure of collecting nasopharyngeal and oropharyngeal swabs into 15mL falcon tubes, containing 3mL phosphate buffered saline (PBS) (Gibco), was performed by clinical team members wearing N95 mask, face shield, protective outer gown, and disposable gloves.
2) Patients with severe COVID-19 or Multisystem Inflammatory Syndrome in Children requiring hospitalization
Pediatric patients who were hospitalized with suspicion of SARS-CoV-2 exposure and/or symptoms concerning for SARS-CoV-2 infection or MIS-C were identified by members of the research team, who subsequently requested approval from the clinical team to approach the patient. A member of the research team contacted the patient and family via phone to obtain informed consent, as described above, and coordinated with both the clinical and TCRC teams for specimen collection. Sample collection was pre-planned with the clinical teams via emails and occurred every 2-3 days for this cohort of patients. Hospitalized patients could opt to provide urine, stool, sputum, or if intubated, tracheal aspirates, in addition to blood, nasopharyngeal and oropharyngeal swabs. Phlebotomy was aligned with clinical blood draws, when feasible, although participants had an option to undergo a separate venipuncture for research purposes. Blood was collected into an EDTA tube, an SST tube, and a PAXgene RNA tube. Repeat samples were collected on alternating days, as feasible. Based on daily coordination between the research and clinical teams, discarded blood from clinical labs were also obtained from hospitalized patients.
3) Newborns born to mothers with and without SARS-CoV-2 infection
Pregnant women with confirmed SARS-CoV-2 infection followed in the MGH obstetrics practice, presenting to the Labor and Delivery (L&D) Unit, or hospitalized for SARS-CoV-2 illness, were approached to enroll their infant in the Pediatric COVID-19 Biorepository following birth in collaboration with the maternal arm of the Adult COVID-19 biorepository. When universal screening for SARS-CoV-2 infection was initiated on all pregnant women admitted to L&D, asymptomatic SARS-CoV-2 positive patients were identified and offered enrollment. Women who tested negative for SARS-CoV-2 were also approached as a control group. The pregnant mothers were simultaneously offered enrollment in the companion Obstetric COVID-19 Biorepository, which included collection of placental biopsies, umbilical cord blood, and other maternal samples. The clinical team assessed the patient’s interest in the Biorepository, then a member of the research staff contacted the patient via telephone to obtain informed consent. Parents could opt to have newborn blood, nasopharyngeal and oropharyngeal swabs, urine, stool, and (if intubated) tracheal aspirates collected. All samples were collected in the clinical setting by the clinical team members to accommodate COVID-19 infection control guidelines, minimizing the risk of SARS-CoV-2 transmission and limit personal protective equipment (PPE) use.
Blood was collected via heel stick between 24-36 hours of life, simultaneously with the heel stick for clinical newborn screening, into two EDTA microtainer tubes (BD). Research nasopharyngeal and/or oropharyngeal swabs were obtained after 24 hours of life, batched at the time of the nasopharyngeal swab for SARS-CoV-2 testing, if performed clinically. Stool and urine were collected on day of life 0 and 2. Stool was collected directly from the diaper. Urine was collected by placing cotton balls in the diaper, then transferring the urine-soaked cotton balls into a specimen cup for transport. If the recruited newborn was intubated for clinical indications, tracheal aspirates were collected at the time of clinical suctioning.
4) Asymptomatic children presenting to their well-visit during the COVID-19 pandemic
Children presenting for their 2-, 3-, or 4-year annual well-child visit with their pediatrician for planned phlebotomy were eligible to participate in this cohort. Eligible patients were identified by study staff and clinicians. If appropriate, researchers contacted the parents via telephone prior to their visit to explain the research and obtain informed consent. Blood and saliva were collected during their clinical phlebotomy. Saliva collection is not considered an aerosolizing procedure; thus, these specimens could be collected in clinic without the need for N95 mask use. The specimens were immediately transported to the laboratory for processing.
REDCap databases were used to record all study data, including: 1) An enrollment log serving as the decoding log - study ID numbers were assigned consecutively across all four patient groups; 2) A laboratory processing database with pertinent processing and freezer storage location information; 3) A chart review database, with demographic and clinical data, including COVID-19 exposures, SARS-CoV-2 polymerase chain reaction (PCR) results, symptoms, and outcomes; 4) A question-response database about COVID-19 exposures and risk factors, specifically for the well-visit cohort.
Specimen Transport and Processing:
In accordance with specimen transport guidelines, specimens were sealed in a leak-proof container labeled with subject’s study ID, then placed in a tight-sealed, biohazard-labeled, secondary container with a rigid outer container and lockable lid (e.g. Igloo cooler) for transport to the laboratory. The entire research team was properly trained on BSL2+ procedures, as required for handling SARS-CoV-2 specimens. A coordinated effort by research personnel enabled successful and efficient troubleshooting, and processing of high influx of samples to the lab during the acute rise of COVID-19 cases in the months of April-June 2020 (Figure 1). Scheduled shifts were implemented throughout the week to ensure the safety of all research staff and sample processing efficiency. Three laboratory roles were created: 1) blood processing technician with extensive technical skill required for blood cell isolation, 2) biospecimen processing technician fully trained in BL2+ enhancement protocols, and 3) specimen labeling, quality control, and sample storage staff. These roles optimized processing workflow, safety precautions, and resources (including staff resource). Paramount to the success of this biorepository included open communication via emails and the use of mobile group messaging outlets, frequent quality checks between staff regarding data and sample collection and processing, accessible leadership, and coordination with patients’ clinical care teams.
Blood samples were processed following BSL2 safety guidelines, with a lab coat, nitrile/latex gloves, and a face shield or safety goggles. All other samples, including nasopharyngeal and oropharyngeal swabs, sputum, saliva, tracheal aspirates, stool, and urine were processed following BSL2+ safety guidelines. BSL2+ safety precautions require all samples to be processed in a certified biosafety cabinet (BSC), class II A2, with intake airflow. Well-trained laboratory personnel handling infectious specimens were required to wear closed-front water impermeable gowns, double nitrile/latex gloves, sleeve covers, and a face shield. Outer gloves were removed when moving away from the BSC and replaced with a new glove when returning to work in the BSC.
Plasma: Blood samples collected in tubes with an EDTA anticoagulant were stored at room temperature until processed, within 24 hours of collection. Tubes were spun at 1000 g for 10 minutes with brake activated. Plasma was then collected, aliquoted, stored at -80°C, and logged in the REDCap database (Figure 3b).
Peripheral blood mononuclear cells (PBMCs): Immediately following the removal of plasma, samples with greater than 2mL initial volume were processed for PBMC isolation using a Ficoll density gradient. Briefly, blood was transferred into a 50mL conical tube, then diluted 1:1 with Hanks’ Balanced Salt Solution without calcium or magnesium (HBSS minus) (Gibco). This diluted blood was then gently layered on top of Ficoll-Paque Plus (GE Healthcare) at 2:1 ratio (2 volumes of blood diluted with HBSS minus to 1 volume Ficoll). Careful attention was made to avoid any mixing of blood with the Ficoll layer. The conical tube was then centrifuged at 1000 g for 30 minutes at room temperature with brake inactivated to allow adequate layering of cellular components. The cloudy ring below the plasma and above the Ficoll (i.e. the PBMC layer) was collected and transferred to a new 15mL conical tube, with HBSS minus added to bring the volume to 15mL (Figure 3a). This tube was then centrifuged at 330 g for 10 minutes, with high brake activated. The supernatant was removed, the PBMC pellet was again washed with HBSS minus, and then resuspended in 10mL HBSS minus for counting. Cell count was obtained by diluting 10µL of sample with 90µL of trypan blue, mixed, and sampled on a hemocytometer. Cells were then frozen in freshly-prepared freezing medium (RPMI 1640 Medium with 1% penicillin-streptomycin, L-glutamine, 1% sodium pyruvate, 1% non-essential amino-acids, and 20% Fetal Bovine Serum (FBS) (Sigma)) with 10% DMSO (Sigma) for a goal concentration range of 5-10 million cells/vial, placed in a chilled Mr. Frosty filled with isopropanol, then immediately placed at -80°C. Final concentration (5-10 million cells per 1mL of freezing medium) and number of aliquot vials were logged. The following day, PBMC cryovials were moved to a liquid nitrogen freezer for long term storage, and location was recorded in specimen log.
PBMCs were isolated within 24 hours of phlebotomy, although higher cell counts were obtained if isolated within 3-4 hours of collection. If less than 5mL blood was collected, a 15mL conical tube, rather than a 50mL conical tube could be used for Ficoll layering. Fresh freezing media were made throughout the day for each sample batch.
Neutrophils: Neutrophils were extracted from the red blood cell layer that remained following the collection of PBMCs (Figure 3a). Neutrophils were isolated using EasySep Direct Human Neutrophil Isolation Kit (StemCell Technologies). The remaining blood layer was incubated with EasySep Direct RapidSpheres and EasySep Direct Human Neutrophil Isolation Cocktail, then diluted in EasySep Buffer. Neutrophils were isolated by successive negative magnet selection using EasySep magnets, then counted using a hemocytometer and aliquoted into Eppendorf tubes for RNA extraction (1x105 cells/tube) or DNA analysis (5x106 cells/tube). Neutrophils designated for RNA extraction were resuspended in 100µL of RNA lysis buffer (TCL) (Qiagen) with 1% β-mercaptoethanol (Sigma), immediately stored at -80°C and logged. Neutrophils planned for DNA analysis were pelleted then directly stored at -80°C and logged.
For RNA extraction steps, a cleaning agent, such as RNaseZAP should be used to remove RNAse from the working surface. RNA lysis buffer should be newly made for each sample using a 10:1 TCL to β-mercaptoethanol ratio.
Serum: Serum samples were collected from blood drawn into serum separator tubes without any anticoagulant (BD). Blood was kept at room temperature, standing upright for 30-60 minutes, then spun at 1200 g for 10 minutes with brake activated. Serum was then collected, aliquoted, stored, and logged (Figure 3b).
Nasopharyngeal and Oropharyngeal Swabs: Swab samples were delivered in phosphate buffered saline (PBS). Samples were directly aliquoted into 1mL aliquots, then immediately stored at -80°C and logged (Figure 4a).
Sputum/Saliva: Samples collected into a collection cup were mixed well at 1:1 ratio with 500 mM DL-Dithiothreitol (DTT) (Sigma)/PBS solution according to CDC recommendations. Diluted samples were then divided into 1mL aliquots, volume permitting, immediately stored at -80°C and logged (Figure 4b).
Tracheal aspirates: Aspirates collected into a sterile collection cup were divided into 1mL aliquots (1mL/vial), immediately stored at -80°C and logged (Figure 4c).
Stool: Stool samples collected from a diaper or specimen cup were divided using a micro spatula, volume permitting, into cryovials with 1mL RNAlater (Invitrogen), empty cryovials without any additive/reagent, up to the 1.5mL tube mark, and cryovials with 1mL Buffered Glycerol Saline (Fisher). Stool samples were fully submerged in RNAlater or glycerol solution prior to immediate storage at –80°C. Samples were logged onto database (Figure 4d).
Urine: Urine samples collected with cotton balls placed inside baby diapers were transferred using forceps, to a 10mL syringe to dispense at most 1mL of fluid into cryovials and immediately stored at –80°C. Samples collected into a tube or a sterile collection cup were aliquoted into cryovials (at 1mL at most/vial) and immediately stored at –80°C (Figure 4e).
Supplies required for specimen collection and processing are listed in Supplemental Table 1. Sample labels, logging, storage, and quality control were performed by assigned lab #3 personnel.