Table 4. Articles describing breakthrough following SARS-CoV-2 infection along with individual or aggregate humoral measures#.
| First author, publication year (study country) | Study design, population | Vaccines included in study and number of doses | Number of cases reported | Lineage of breakthrough infection | Time from last vaccine dose to antibody test* (days) | Antibody assay and target, isotype (cut-off) | Pre- breakthrough antibody level* | Time from antibody test* to breakthrough infection (days) | Statistical association |
|---|---|---|---|---|---|---|---|---|---|
| Strafella, 2021 (Italy) | Case report, HCW | Pfizer, 2 doses | 1 | B.1.1.7 | 26 | Euroimmun Anti-Sars-CoV-2, IgG Anti-S1, IgA Anti-S1, IgM Anti-N (Lubeck, Germany) (cut-off: ≥1.1) | IgG: 10.47 ratio units | 26 | None reported |
| IgA: 3.58 ratio units | |||||||||
| IgM: 0.2 ratio units | |||||||||
| 26 | Roche Elecsys Anti-Sars-CoV-2 Total anti-RBD (Basel, Switzerland) (cut-off: >0.8 BAU/ml) | 978.7 U/ml | 26 | None reported | |||||
| Schulte, 2021 (Germany) | Case report, HCW | Pfizer, 2 doses | 1** | B.1.525 | 9 | Roche, Total Ig, S1 (Basel, Switzerland) (cut-off not provided) | >250 U/mL | 45 | None reported |
| Gilbert, 2021 (USA) (Please see Table 5 for additional evidence) | Nested case-cohort within an RCT, vaccine trial participants | Moderna, 2 doses | 55 (text) or 46 (Table 1) | Not provided | ≤81 | MSD anti-S, IgG (Rockville, USA) (cut-off: >10.8424 IU/mL) | GMC of 1890 (95% CI 1449, 2465) IU/mL among cases, 2652 (95% CI 2457, 2863) IU/mL among non-cases. | Not provided | GMC ratio of cases/non-cases = 0.71 (95% CI 0.54, 0.94) |
| Cox regression to estimate association between risk of COVID-19 and anti-S IgG level (per 10-fold increase). HR = 0.66 (95% CI 0.50, 0.88). | |||||||||
| 34% decrease in risk for every 10-fold increase of Anti-S IgG | |||||||||
| ≤81 | MSD anti-RBD, IgG (Rockville, USA)(cut-off: >14.0858 IU/mL) | GMC of 2744 (95% CI 2056, 3664) IU/mL among cases, 3937 (95% CI 3668, 4227) IU/mL among non-cases | Not provided | GMC ratio of cases/non-cases 0.70 (95% CI 0.52, 0.94) | |||||
| Cox regression to estimate association between risk of COVID-19 and anti-RBD IgG level (per 10-fold increase). HR = 0.57 (95% CI 0.40, 0.82). | |||||||||
| 43% decrease in risk for every 10-fold increase of Anti-RBD IgG | |||||||||
| ≤81 | Pseudoneutralization assay with ID50 calibrated against WHO IS, neutralizing antibodies (no cut-off reported) | GMT of 160 (95% CI 117, 220) ID50 titre among cases, 247 (95% CI 231, 264) ID50 titre among non-cases. | Not provided | GMT ratio of cases/non-cases = 0.65 (95% CI 0.47–0.90) | |||||
| Cox regression to estimate association between risk of COVID-19 and neutralizing antibody level (per 10-fold increase). HR = 0.42 (95% CI 0.27, 0.65). | |||||||||
| 58% decrease in risk for every 10-fold increase of neutralizing antibodies | |||||||||
| Pseudoneutralization assay with ID80 calibrated against WHO IS, neutralizing antibodies (no cut-off reported) | GMT of 332 (95% CI 248, 444) ID80 titre among cases, 478 (95% CI 450, 508) ID80 titre among non-cases. | GMT ratio of cases/non-cases = 0.69 (95% CI 0.52, 0.93) | |||||||
| Cox regression to estimate association between risk of COVID-19 and neutralizing antibody level (per 10-fold increase). | |||||||||
| HR = 0.35 (95% CI 0.20, 0.61). | |||||||||
| 65% decrease in risk for every 10-fold increase of neutralizing antibodies | |||||||||
| Feng, 2021 (UK) | Cohort study secondary analysis of clinical trial data | AstraZeneca | 171** | Mostly B.1.1.7 and B.1.177 | 14–42 | MSD anti-S, IgG, (Rockville, USA) (no cut-off reported) | Median of 30501 (95% CI 16088, 49529) AU/mL for cases, and 33945 (95% CI 18450, 59260) AU/mL for non-cases | Not provided | Generalized additive model to estimate risk of symptomatic COVID-19. |
| Difference between median antibody levels for cases and non-cases: p>0.05 | |||||||||
| Risk was inversely correlated to anti-spike IgG (p = 0.003), | |||||||||
| There was no association between risk of asymptomatic COVID-19 and anti-spike IgG | |||||||||
| 14–42 | MSD Anti-RBD, IgG (Rockville, USA) (no cut-off reported) | Median of 40884 (95% CI 20871, 62934) AU/mL for cases, 45693 (95% CI 24009, 82432) AU/mL for non-cases | Not provided | Difference between median antibody levels for cases and non-cases: p>0.05 | |||||
| Risk was inversely correlated to anti-RBD IgG (p = 0.018). | |||||||||
| There was no association between risk of asymptomatic COVID-19 and anti-RBD IgG | |||||||||
| 14–42 | Microneutralization assay, neutralizing antibodies (no cut-off reported) | Median titre of 206 (95% CI 124, 331) for cases, 184 (95% CI 101, 344) for non-cases | Not provided. Median follow up period of 53 days (IQR 29,81), starting 7 days after blood draw. | Difference between median antibody levels for cases and non-cases: p>0.05 | |||||
| Risk was inversely correlated to microneutralization titre (p<0.001). | |||||||||
| There was no association between risk of asymptomatic COVID-19 and neutralizing antibodies | |||||||||
| Bergwerk, 2021 (Israel) | Case-control study, HCW | Pfizer, 2 doses | 22** | B.1.1.7 was identified in 85% of breakthroughcases, similar to community prevalence at the time | Median of 36 days (breakthrough infections), median of 35 days (controls) | Beckman Coulter, anti-S1 (Brea, USA)(no cut-off provided) | Case predicted anti-S IgG GMT: 11.2 (95% CI 5.3, 23.9); Control predicted GMT: 21.8 (95% CI 18.6,25.52) | Within a week of breakthrough for cases. Controls were matched to cases by time between second vaccine dose and serology test | Ratio of cases/control GMT: 0.514 (95% CI 0.282, 0.937) |
| Linear regression to assess correlation between Ct value of cases and neutralizing antibody level during peri-infection period. | |||||||||
| Slope = 171.2 (95% CI 62.9, 279.4). | |||||||||
| Median of 36 days (breakthrough infections), median of 35 days (controls) | Pseudoneutralization assay | Case predicted GMT: 192.8 (95% CI 67.6, 549.8); Control predicted GMT: 533.7 (95% CI 408.1, 698.0) | Within a week of breakthrough for cases. Controls were matched to cases by time between second vaccine dose and serology test | Ratio of cases/control GMT: 0.361 (95% CI 0.165, 0.787) | |||||
| Michos, 2021 (Greece) | Cohort study, HCW | Pfizer, 2 doses | 2 | Not provided | One month | GenScript cPass SARS-CoV-2 Neutralization antibody detection kit (Piscataway, USA) | 90 and 95% neutralization | ~10 days | None reported |
| Yamamoto, 2021 (Japan) | Case control study, HCW | Pfizer, 2 doses | 17 | 5 of 17 reported to be Delta | Median of 63 (IQR 43–69) days for cases; 62 (IQR 40–69) days for controls | Abbott Advise Dx SARS-CoV-2 IgG II (Sligo, Ireland), anti-RBD, (no cutoff provided) | Case predicted GMC: 5129 (95% CI 3881, 6779); Control predicted GMC: 6274 (95% CI 5017,7847) | 55 (45–64) days | Ratio of cases/control GMC: 0.82 (95% CI 0.65, 1.02), p = 0.07 |
| Median of 63 (43–69) days for cases; Median of 62 (40–69) days for controls | Roche Elecsys Anti-SARS-CoV-2 (Basel, Switzerland), Spike total antibody, (no cutoff provided) | Case predicted GMC: 1144 (95% CI 802,1632); Control predicted GMC: 1208 (95% CI 1053–1385) | 55 (45–64) days | Ratio of cases/control GMC: 0.95 (95% CI 0.70, 1.27), p = 0.72 | |||||
| Median of 63 (43–69) days for cases; Median of 62 (40–69) days for controls | PRNT/neutralization test (SARS-CoV-2 ancestral, Alpha and Delta strains) | Ancestral strain: case predicted GMT: 405 (95% CI 327,501); Control predicted GMT: 408 (320,520) | 55 (45–64) days | Ratio of cases/control GMT: 0.99 (95% CI 0.74, 1.34), p = 0.96 | |||||
| Alpha: Case predicted GMT: 116 (95% CI 80,169); Control predicted GMT: 122 (95% CI 96,155) | Ratio of cases/control GMT: 0.95 (95% CI 0.71, 1.28), p = 0.76 | ||||||||
| Delta: Case predicted GMT: 123 (95% CI 85, 177); Control predicted GMT: 135 (95% CI 108, 170) | Ratio of cases/control GMT: 0.91 (95% CI 0.61, 1.34), p = 0.63 |
#—Assay results from each study were included for every antibody type (i.e.–anti-S, anti-N, anti-RBD) and isotype (i.e.–IgG, IgM, IgA) measured. In instances where more than one assay target was used to measure the same antibody target in the same study (i.e. both PRNT and pseudoneutralization results, or anti-S results from two different assays), we included only one of these results. Full data extraction for every study can be provided upon request.
*- If more than one test result was provided, the result closest in time to re-infection is presented.
**—In these studies, other breakthrough infections were reported as well, but with no accompanying temporal and laboratory data
Definitions: anti-S = anti-spike, anti-N = anti-nucleocapsid, anti-RBD = anti-receptor binding domain, PRNT = plaque reduction neutralization test, LDT = laboratory-determined test, ELISA = enzyme-linked immunosorbent assay, AU = arbitrary units, OD = optical density, IRR = increased relative risk, HCW = health care worker, LTC = long term care, GMC = geometric mean concentration, GMT = geometric mean titre, 95% CI = 95% confidence interval, ID50 = infectious dose titer 50, WHO IS = World Health Organization SARS-CoV-2 antibody International Standard, HR = hazard ratio, RCT = randomized controlled trial, MSD = Mesoscale Discovery