Methods:
After institutional review board (IRB) approval, a prospective study analyzing immune response to COVID-19 vaccination in pediatric cancer patients was undertaken. Eligibility criteria included cancer patients aged 0-25 years of age actively receiving therapy, including chemotherapy, immunotherapy, and targeted therapy. We excluded patients who were completely vaccinated prior to their cancer diagnosis. Patients may have received any of the vaccines approved for emergency use by the U.S. Food and Drug Administration. This includes BNT162b2, developed by Pfizer and BioNTech, mRNA-1273, developed by Moderna. For the remainder of this paper, these vaccines will be referred to as Pfizer and Moderna. Blood samples were collected pre-vaccination, as well as 2, 4-6, and 8-12 weeks after completion of COVID vaccination. Completion was defined as completion of the two dose series. As a control group, we enrolled patients in our clinic aged 0-25 years who were not receiving cancer directed therapy, followed for a benign hematology issue, or were more than 6 months after completion of therapy for their cancer diagnosis. These participants had one sample drawn between 2 weeks and 6 months of completion of vaccination.
Demographic data was collected from the medical record, including age, gender, ethnicity, cancer diagnosis, personal history of COVID, current therapy, type of vaccine received, and dates of vaccination. Therapy details included if it was chemotherapy, immunotherapy, targeted therapy, or combination therapy. We also collected if therapy was intravenous, oral, or both. If available, we also collected complete blood count (CBC) prior to vaccination and CBCs within 4 days of all sample time points.
Once blood samples were obtained in clinic, they were transported to the lab and processed within two hours. In order to maximize T cell yield for analysis in patients who may have low lymphocyte counts due to therapy, participants blood samples were collected in Vacutainer® CPT™ Cell Preparation Tube (BD) and processed according to the manufacturer’s instructions. The cells were then re-suspended in 1ml of freezing medium (RPMI containing 40% FBS and 12.5% DMSO) and stored in liquid nitrogen.
To measure immunogenicity, we performed two assays. First, we measured neutralizing IgG antibodies, which has been shown to be highly predictive of immune protection from symptomatic SARS-CoV2 infection.10 This was done using the SARS-CoV2-Surrogate Virus Neutralization Test Kit (GenScript) according to the manufacturer’s instructions. All the samples were run in duplicates. The quantitative detection range of this assay is between 47 to 185U/ml. Titers ≤60.8 U/ml (lower 10% of the quantifiable detection range) was taken as inadequate B-cell response. Titers ≥60.8U/ml was taken as adequate B-cell response. The second assay measures interferon gamma (IFNγ) secretion, which is a marker of T cell response to viral antigens and another established method of measuring cellular immune response after vaccination.11 This is done using an enzyme-linked immunosorbent spot (ELISPOT) kit for IFNγ (Mabtech). Briefly, frozen peripheral blood mononuclear cells were thawed and then seeded onto anti-human IFNγ monoclonal antibody pre-coated plates at a concentration of 250,000 cells/well. These cells were then stimulated with SARS –CoV-2 spike peptide pools at a concentration of 2µg/ml by incubating for 36 hours in 5% CO incubator and sent to Mabtech, Inc. Ohio, USA, for quantitation. Results were reported as Spot forming units per million cells (SFU/106cells). SFU values ≤ Mean- standard deviation (i.e. ≤ 659 SFU/106 cells) of the control sample values were defined as an inadequate T-cell response.
When analyzing what clinical features may impact vaccine response, we utilized B cell response by neutralizing IgG antibodies to define our adequate and inadequate response groups. This was because this assay allowed for replicates to strengthen the validity of our results as compared to the T cell assay, which could be only done once. Statistical analysis was done using Graphpad Prism 8.0. When appropriate, unpaired t tests were obtained to evaluate for statistical significance, which was defined as p value <0.05.