Abstract
Studies on vaccines and infectious diseases are increasingly adding immune cytometry (for e.g., flow cytometry or ELISPOT assays) to the routine use of antibody titers as a way to report immunogenicity. We advocate that the classical presentation of cytometric data in terms of percentages should best be supplemented by an absolute cell count per ml. We do this by discussing a simple hypothetical example illustrating that without knowledge of the absolute cell count per ml detection competition will render a correct comparison between different samples impossible in all situations where an inter-individual or intra-individual (longitudinal) variation in peripheral blood mononuclear cell concentrations is present.
Dear Editor,
The increasing use of batch analysis in vaccine trials necessitates the collection, freezing and thawing of peripheral blood mononuclear cells (PBMC). In a recent update of the experimental caveats inherent to flow cytometric analysis in clinical trials, Maecker et al.Citation1 briefly stated that the use of absolute counts of cell subpopulations per volume-unit instead of the percentages of cell subpopulations within a parent population were “sometimes desirable” in specific situations (such as HIV infection). However, for instance in the field of vaccine trials, our experience is that standard reporting practice on flow cytometric and ELISPOT-based studies involves inter- and intra-individual immunological comparisons based on percentages, while absolute counts are not obtained. Stewart et al. introduced the addition of beads to samples to calculate absolute cell numbers,Citation2 however this method is not readily transferable to the analysis of Ficoll-separated PBMC.
We would like to draw attention to more expansive reasons to routinely report absolute counts in addition to percentages, and advocate it to be used as standard practice for all quantitative immunological assays in studies using separated PBMC, where possible.
In a hypothetical comparison of blood samples obtained in ideal situations, we assume for the sake of simplicity that only two types of cells, A and B, are present. Four tubes with absolute (‘true’) A and B cell counts are shown in . By flow cytometric acquisition of 200,000 events (cells), we can derive the observed counts and percentages as shown. Tubes 1 and 2 would be indistinguishable in the flow cytometric analysis, although tube 2 has higher concentrations of both cell types. Although tubes 3 and 4 differ in the same way from tube 1 in the flow cytometric comparison, the absolute cell count in tube 3 shows an elevation only in the number of B cells, whereas in tube 4 there is only a decline in A cells. Actually, only the flow cytometric comparison between tubes 2 and 3 can be seen as a correct reflection of the comparison in absolute cell counts.
Table 1. A hypothetical example illustrating the difference in interpretation between the absolute cell count per tube and the cytometric count
This example illustrates that without knowledge of the absolute cell count per tube (or the absolute number of parent cells in a volume-unit), detection competition makes a correct comparison of the different samples impossible. Percentages only show us a difference between samples, but as a sole means of assessment they do not provide the direction of the difference as illustrated by the examples in the table. We advocate, therefore, that it is necessary in all instances to present (or at least document) an absolute cell count per unit of volume in addition to the percentages, because not only recognizable perturbations due to an infection can distort the absolute counts, but inter-individual variation itself and intra-individual variation in situations of long-term follow-up also cause a bias which cannot be corrected for when only percentages are assessed. The latter could be ‘iatrogenic’ induced by vaccination due to a shift in an unexamined cell population.
One could argue that background variation, for e.g., diurnal or seasonal variation, of the absolute cell counts would interfere with the correct interpretation of absolute cell counts. We note however that a sound study protocol should be able to minimize the effects of background variation.
Our advocacy for absolute counts also applies to ELISPOT assays, where the number of spot-forming cells is related to the number of input PBMC with no reference to absolute counts of cells. At the same token, the determination of antibody titers (expressed per plasma or serum volume) also is inevitably biased since the percentage of plasma in whole blood differs with age, fluid state and other physiologic factors therefore yielding a presentation of the absolute antibody titer per blood tube.
In conclusion, due to the inter- and intra-individual variation in the number of PBMC per ml blood we strongly recommend the addition of the absolute cell count per ml blood to the existing routine of reporting percentages in cytometric analyses for vaccine trials, particularly when quantitative assessments are envisaged. This simple change in practice would reduce the ambiguity in the interpretation of reported findings from cytometric studies, and thus enable more appropriate and timely research building on such reports (e.g., statistical and mathematical models).
| Abbreviations: | ||
| ELISPOT | = | Enzyme-linked immunosorbent spot |
| PBMC | = | peripheral blood mononuclear cells |
Acknowledgments
B.O. has received financial support from the Research Foundation Flanders (FWO).
Disclosure of Potential Conflicts of Interest
B.O., V.V.T. and P.B. report no conflicts of interest. P.V.D. acts as (principal) investigator for vaccine trials PVD acts as (principal) investigator for vaccine trials conducted on behalf of the University of Antwerp, conducted on behalf of the University of Antwerp, for which the University obtains research grants from vaccine manufacturers. P.V.D. receives no personal remuneration for this work.
References
- Maecker HT, McCoy JP Jr., Amos M, Elliott J, Gaigalas A, Wang L, et al, FOCIS Human Immunophenotyping Consortium. A model for harmonizing flow cytometry in clinical trials. Nat Immunol 2010; 11:975 - 8; http://dx.doi.org/10.1038/ni1110-975; PMID: 20959798
- Stewart CC, Steinkamp JA. Quantitation of cell concentration using the flow cytometer. Cytometry 1982; 2:238 - 43; http://dx.doi.org/10.1002/cyto.990020407; PMID: 7056133