In December 2020, we established a prospective longitudinal study in Vancouver, Canada, to examine SARS-CoV-2 specific humoral immune responses following vaccination with BNT162b2 (Comirnaty; BioNTech/Pfizer) or mRNA-1273 (Spikevax; Moderna) in a cohort of adults aged 24-98 years [described in (25, 26)]. Serum and plasma were collected longitudinally up to 6 months following the third dose ( Figure 1A ). At each visit, serum was tested for the presence of SARS-CoV-2 anti-nucleocapsid (N) antibodies, which indicate seroconversion following infection, using the Elecsys Anti-SARS-CoV-2 assay on a Cobas e601 module analyzer (Roche Diagnostics). In addition to the case participant, immune measures from a comparison group of 124 participants of this cohort, who remained anti-N seronegative up until at least one month post-third vaccine dose, are included for context.

All participants or their authorized substitute decision makers provided written informed consent. This study was approved by the University of British Columbia/Providence Health Care and Simon Fraser University Research Ethics Boards (protocol H20-03906).

Diagnostic samples from the case participant’s two SARS-CoV-2 infections were tested at the St. Paul’s Hospital Virology Laboratory using the cobas^® SARS-CoV-2 Test which targets conserved regions within the Orf1a/b and E genes (Roche Diagnostics) followed by screening using a real-time reverse transcription (RT)-PCR based algorithm for SARS-CoV-2 lineage classification that is frequently updated to detect emerging variants (27, 28). Following this, the diagnostic samples were subjected to full-genome SARS-CoV-2 sequencing in two independent laboratories: the BC Centre for Disease Control, the provincial laboratory that performs all SARS-CoV-2 sequencing for epidemiological surveillance, and the BC Centre for Excellence in HIV/AIDS. Both laboratories use the Illumina platform. The SARS-CoV-2 full genome sequences for the participant’s BA.1.15 and subsequent BA.2 infections are available in GISAID (Accession IDs EPI_ISL_12767799 and EPI_ISL_12662303, respectively) and in Genbank (Accession Numbers OP237526 and OP237527).

We quantified anti-Spike Receptor Binding Domain (RBD) binding IgG concentrations in serum using the V-plex SARS-CoV-2 (IgG) ELISA kit (Panel 22, Meso Scale Diagnostics), which features wild-type and Omicron BA.1 RBD antigens. For a subset of participants, Anti-Spike binding IgG concentrations in serum were also quantified using the V-plex SARS-CoV-2 (IgG) ELISA kit (Panel 25, Meso Scale Diagnostics), which features full-length S antigens from wild-type, Omicron BA.1 and Omicron BA.2. This panel was used because, at the time of analysis, no reagents featuring Omicron BA.2 RBD were offered by the manufacturer. Both assays were performed on a Meso QuickPlex SQ120 instrument, with sera diluted 1:10000. Results are reported as WHO International Standard Units (BAU/mL), using the manufacturer-provided conversion factors.

We assessed the ability of serum antibodies to block the wild-type and Omicron BA.1 RBD-ACE2 receptor interaction by competition ELISA (Panel 22 V-plex SARS-CoV-2 [ACE2]; Meso Scale Diagnostics). For a subset of 28 participants, we also assessed the ability of serum antibodies to block the wild-type, BA.1 and BA.2 Spike-ACE2 receptor interaction using the same methods (Panel 25 V-plex SARS-CoV-2 [ACE2]). Both assays were performed on a Meso QuickPlex SQ120 instrument, with sera diluted 1:40. Results are reported as % ACE2 displacement.

Neutralizing activity in plasma was examined in live SARS-CoV-2 assays using a wild-type isolate (USA-WA1/2020; BEI Resources) and a local Omicron BA.1 isolate (GISAID Accession # EPI_ISL_9805779) on VeroE6-TMPRSS2 (JCRB-1819) target cells. Viral stock was adjusted to 50 TCID50/200 µl in Dulbecco’s Modified Eagle Medium in the presence of serial 2-fold plasma dilutions (from 1/20 to 1/2560), incubated at 4°C for 1 hour and added to target cells in 96-well plates in triplicate. Cultures were maintained at 37°C with 5% CO2 and the appearance of viral cytopathic effect (CPE) was recorded three days post-infection. Neutralizing activity is reported as the reciprocal of the highest plasma dilution able to prevent CPE in all triplicate wells. Samples exhibiting partial or no neutralization at 1/20 dilution were defined as below the limit of quantification (BLOQ).

Data visualization and analysis was conducted in Prism v9.2.0 (GraphPad). As this is a report of a single case where immune response data are contrasted to those observed in a COVID-19-naive comparison group, response magnitudes are characterized in detail, but no formal statistical tests were applied.

The participant was a frontline health care worker, approximately 30 years of age. The participant had no major health conditions, but did take medications for hypothyroidism and familial hypercholesterolemia, both of which were well-controlled. The participant received three doses of mRNA vaccine (all BNT162b2; 30 mcg) in late December 2020, early February 2021 and late October 2021 ( Figure 1A ). All blood samples collected up to one month following the third immunization were anti-N seronegative.

In early January 2022, ten weeks after the third immunization, the participant experienced moderate COVID-19 symptoms including sore throat, fatigue, congestion, body aches, severe headaches, loss of taste and smell, coughing, shortness of breath and nausea. Symptoms, primarily cough, intensified in the second week after diagnosis requiring corticosteroid therapy. By the third week, symptoms had subsided except for shortness of breath and fatigue, with minimal improvement from short- and long-acting bronchodilating agents. A saline gargle collected on January 10, 2022 tested positive on the cobas^® SARS-CoV-2 Test with a cycle threshold (Ct) value of 21 for both Orf1a/b and E gene targets. Real-time RT-PCR-based molecular screening identified the infection as Omicron BA.1, with subsequent full-genome viral sequencing confirming the specific lineage as BA.1.15.

In early April 2022, 13 weeks following the BA.1 infection (and 23 weeks following the third immunization) the participant experienced a profile of milder COVID-19 symptoms that differed compared to the first infection, consisting of a sore throat, fever, body aches, headaches, and diarrhea. No change in sense of taste or smell was noted. The participant noted persisting weakness, fatigue and mental fog, as well as severe long-term, treatment-resistant shortness of breath triggered by mild activities or exercises. A nasopharyngeal swab collected on April 9, 2022 tested positive on the cobas^® SARS-CoV-2 Test with Ct values of 24 (Orf1a/b) and 23 (E). This infection was identified as BA.2 by molecular screening and confirmed by full-genome viral sequencing.

We began by investigating the magnitude of the participant’s humoral immune responses following immunization, in context of a control group of 124 COVID-naïve individuals who were vaccinated during the same period. The comparison group was 74% female with a median age of 57 (Interquartile Range [IQR] 38-76) years. We quantified antibody responses to wild-type and Omicron BA.1 strains in the participant and the comparison group at one month after the second and third vaccine doses, as these time points should capture peak responses post-vaccination ( Figures 1B–D ).

One month post-second dose, the participant’s wild-type and BA.1-specific RBD IgG concentrations were 3.34 and 2.70 log₁₀ BAU/mL respectively, which were equivalent to the 53^rd and 50^th percentile values of the comparator cohort ( Figure 1B ). One month post-third dose, the participant’s wild-type and BA.1-specific RBD IgG concentrations had increased to 3.51 and 3.06 log₁₀ BAU/mL respectively, equivalent to the 30^th and 50^th percentile values of the cohort. Similarly, one month post-second dose, the participant’s ability to disrupt the interaction between the ACE2 receptor and the wild-type and BA.1 RBDs were 97% (57^th percentile) and 42% (55^th percentile) respectively ( Figure 1C ). At one month post-third dose, the participant’s wild-type- and BA.1-specific RBD-ACE2 displacement activities had increased to 99% (74^th percentile) and 76% (73^rd percentile) respectively. Finally, at one month post-second dose, the participant’s plasma neutralized wild-type and BA.1 SARS-CoV-2 at reciprocal dilutions of 320 and 20, which were equivalent to the 97^th and 76^th percentile values of the cohort. At one month post-third dose, the participant’s wild-type and BA.1 neutralization titers were 320 (78^th percentile) and 40 (59^th percentile) respectively. These results indicate that the participant’s overall vaccine responses were typical of the cohort, but nevertheless insufficient to prevent infection by BA.1 approximately six weeks later.

Seventeen days after testing positive with BA.1 (which coincided with a three-month post-third-dose study visit), the participant’s wild-type and BA.1-specific responses were boosted substantially, reaching the cohort 95th percentile for most measures at this time when immune responses had begun to decline in the comparator cohort ( Figures 1B–D ). The participant’s wild-type RBD IgG concentration increased to 4.15 log₁₀ BAU/mL, while their BA.1-specific RBD IgG concentration increased to 3.55 log₁₀ BAU/mL ( Figure 1B ). For context, these values would have placed the participant in the 94th and 91st percentiles of “peak” cohort values, measured at one month post-third vaccine dose. Similarly, the participant’s wild-type-specific RBD-ACE2 competition activity remained high at 99.9%, while their BA.1-specific RBD-ACE2 competition activity increased to 96.3%. For context, these values would be equivalent to the 99th percentiles of peak cohort values one month post-third vaccine dose ( Figure 1C ). The participant’s wild-type and BA.1-specific neutralization values held at 320 and 40, respectively, equivalent to the 78^th and 59^th percentiles of peak cohort values ( Figure 1D ). These results indicate that BA.1 infection boosted the participant’s humoral response, particularly in terms of binding antibody concentrations and ACE2 displacement activity. Nevertheless, this boost was insufficient to prevent reinfection by BA.2 approximately 10 weeks later.

Sixteen days after testing positive with BA.2 (which coincided with a six-month post-third-dose study visit), the participant’s wild-type-specific responses remained steady or declined slightly (e.g. RBD IgG) from prior measurements ( Figures 1B–D ). Nevertheless, most of the participant’s wild-type-specific responses remained at the cohort 100th percentile at this time point, which is unsurprising given that vaccine-induced responses had declined substantially over this time in the COVID-19 naive comparison group. For context, the participant’s wild-type-specific responses at this time point would have represented the 84^th (RBD IgG), 99^th (ACE2 competition) and 78^th (neutralization) percentiles of peak cohort values measured at one month post-third vaccine dose.

By contrast, the BA.2 reinfection had mixed effects on BA.1 responses. While the participant’s BA.1-specific RBD IgG concentration rose substantially to 4.01 log₁₀ BAU/mL (whereas the cohort median at this time point was nearly 2 log₁₀ lower), no change was seen for BA.1-specific RBD-ACE2 competition, and BA.1 neutralization increased only modestly ( Figures 1B–D ). The more pronounced impact of BA.2 reinfection was to extend the duration of BA.1-specific responses in the participant, who maintained an RBD-ACE2 competition activity of 95.7% (compared to the cohort median 29% at this time point) and a neutralization activity of 80 (compared to the cohort median BLOQ at this time point). For further context, the participant’s BA.1-specific responses at this time point would represent the 99^th (RBD IgG), 99^th (ACE2 competition) and 83^rd (neutralization) percentiles of peak cohort values measured at one month post-third vaccine dose. Nevertheless, despite BA.1 infection and BA.2 reinfection, the participant’s virus neutralization activity against BA.1 at this time point, which represented the highest activity measured during the study, remained 4-fold lower compared to that against the wild-type strain one month post-third vaccine dose ( Figure 1D ). The substantially weaker ability of even “hybrid” (vaccine and dual-infection induced) immune responses to neutralize Omicron compared to wild-type suggests that the participant may remain at risk of additional Omicron infection.

We next characterized BA.2-specific Spike IgG and ACE2 competition activities in the participant, compared to a subset of 28 control vaccinees (79% Female, median age 59 years) beginning one month following the third vaccine dose ( Figure 2 ). As these analyses focus on whole Spike (rather than RBD antigen), the corresponding wild-type and BA.1 Spike-specific responses are also shown for context ( Figure 2 ). We additionally confirmed the (strong) correlations between wild-type- and BA.1-specific RBD and Spike responses in these individuals (all p<0.0001; Figure S1 ). At one month post-third vaccine dose, the participant displayed wild-type, BA.1 and BA.2-specific Spike IgG concentrations of 3.76, 3.14 and 3.41 log₁₀ BAU/mL respectively ( Figure 2A ), and ACE2 competition activities of 99.4%, 50.6% and 64.3% respectively ( Figure 2B ). The participant’s wild-type and BA.2-specific Spike-specific IgG concentrations and ACE2 competition activities were broadly average compared to control vaccinees at this time (54^th and 68^th percentiles for IgG, respectively; 46^th and 54^th percentiles for ACE2 competition, respectively). The participant’s values for BA.1-specific IgG and BA.1 Spike-specific ACE2 competition however were slightly lower than the averages of the control group (37^th and 39^th percentiles, respectively).

Following BA.1 infection, the participant’s wild-type, BA.1 and BA.2-specific Spike IgG concentrations increased modestly, to 3.77, 3.17 and 3.44 log₁₀ BAU/mL, respectively ( Figure 2A ). Though the magnitude of these increases was not as pronounced as those observed in the RBD-based assays (shown in Figure 1B ), these values nevertheless placed the participant at or above the 85^th percentile compared to control vaccinees at this time point, when immune responses had begun to decline in the broader cohort. For context, these values would place the participant in the 57^th, 39^th, and 68^th percentile of peak cohort values measured at one month post-third vaccine dose. Similar to the ACE2 competition activities measured using RBD antigens (shown in Figure 1C ), the participant’s wild-type Spike-ACE2 competition activities remained high at 99.7%, while BA.1 and BA.2 Spike-ACE2 activities rose substantially to 87.9% and 89.3%, respectively ( Figure 2B ); values that represented the 71^st, 75^th, and 79^th percentiles of peak values for the control vaccinees at one month post-third vaccine dose.

Following BA.2 infection, the participant’s wild-type Spike IgG concentration declined slightly to 3.73 log₁₀ BAU/mL, whereas their BA.1 and BA.2-specific values increased slightly to 3.20 and 3.51 log₁₀ BAU/mL, respectively ( Figure 2A ). These trends were consistent with the anti-RBD IgG concentrations measured in the primary analysis (shown in Figure 1B ), though of a smaller magnitude. Similar to the ACE2 competition activities measured using RBD antigens (shown in Figure 1C ), the participant’s wild-type Spike-ACE2 competition activity remained high (99.5%) after BA.2 infection. BA.1 and BA.2 Spike-ACE2 activities increased, though only marginally, to 88.6% and 90% respectively ( Figure 2B ).

Together, these results confirm that the participant’s humoral responses to wild-type and Omicron variants were broadly average for a COVID-19-naive individuals one month post-third vaccine dose. While subsequent BA.1 infection boosted Omicron-specific immune responses (highlighted by an increase in BA.1 and BA.2 Spike-ACE2 competition activities), BA.2 reinfection did not substantially augment these activities further, but rather extended the duration of these responses.