Reconciliation of titer differences between SARS-CoV-2 neutralizing antibody assays
Neutralizing antibody measurements against SARS-CoV-2 are used as a correlate of protection for COVID-19 and inform public health decisions. However, there is no universally accepted standard neutralization assay, and different platforms often produce substantially different absolute titers. Understanding why assays disagree is essential for interpreting results across studies, comparing vaccine or infection-induced immunity, and for harmonizing laboratory approaches.
The analysis compared 28 samples across three commonly used neutralization platforms:
Each assay reports a neutralizing titer, but the biological substrates (live SARS-CoV-2 vs pseudotyped virus), readouts (cell death, foci, luminescence), and culture conditions differ between assays.
Despite methodological differences, titers across all three assays were highly correlated. Pairwise comparisons produced Spearman correlation coefficients greater than 0.95, with P < 0.0001 for all pairings. These results indicate consistent relative ranking of sample neutralizing activity between platforms: samples that score higher in one assay generally score higher in the others.
Although correlation was high, absolute titers differed significantly between assays (paired t-tests P < 0.001). Key quantitative findings reported were:
The IS and PV assays—both assessing the titer required to reduce foci or luminescence by 50%—had the smallest difference. IS titers were 1.2-fold higher than PV titers (95% confidence interval, CI: 1.2–1.2).
The CPE assay produced markedly lower titers compared with the IS assay: CPE titers were 5.5-fold lower (95% CI: 5.4–5.5) than IS titers.
These divergent absolute values show that although assays are concordant in relative terms, the same serum can yield substantially different numerical titers depending on the platform used.
The authors identify two primary assay elements that explain the observed absolute titer differences:
Inoculum framing: IS and PV assays effectively assess neutralization on a per virion basis—measuring the antibody titer needed to neutralize half of the infectious virions present during the assay. By contrast, the CPE assay behaves as a per well neutralization measurement where neutralization is assessed against the entire inoculum and half the time an entire well’s inoculum must be neutralized to prevent cytopathic effect. This distinction in whether neutralization is defined per infectious particle or per well changes how titers scale between assays.
Serum exposure during culture: In the CPE assay the serum is removed during the multi-day culture (a 5-day period in this study), which prevents antibodies from acting beyond the initial inoculum. Consequently, serum in the CPE assay can only neutralize the inoculum and cannot slow subsequent viral spread during culture. In contrast, assays where antibody remains present or where readouts capture reductions in reporter signal or foci formation can reflect both neutralization of initial virions and inhibitory effects during replication or spread.
Together, these procedural and conceptual differences—how an assay defines the neutralization target and whether antibodies remain present during viral replication—account for much of the systematic disparity in absolute titers between assays.
The findings underline two practical points for researchers, laboratories, and public health agencies:
High correlation of titers means different assays can reliably rank sera by neutralizing activity, but absolute values are not interchangeable without clear conversion or contextual understanding.
Harmonization efforts should focus on clarifying assay terminology (for example, distinguishing per virion versus per well neutralization), on standardized reporting of procedural elements (virus type, inoculum size, serum removal or retention during culture, and readout timing), and on identifying which assay elements must be aligned or accounted for when converting titers between platforms.
The abstract reports results based on 28 samples and three specific assay formats; detailed methodological parameters beyond those summarized were not provided in the abstract. The authors declare no conflict of interest. These results highlight assay elements that should be considered when comparing neutralization data, and they provide mechanistic insight that may guide standardization, but complete standardization will require broader field-wide efforts and consensus on definitions and procedures.