Comparison of single versus double platform analysis for CD34+ hematopoietic stem cell enumeration using International Society of Hematotherapy and Graft Engineering protocol and consensus among hematopathologists

INTRODUCTION

The International Society of Hematotherapy and Graft Engineering (ISHAGE) provides guidelines for enumeration of CD34+ hematopoietic stem cells (HSC) by flow cytometer during HSC transplantation (HSCT).^[1] This utilizes antigenic and light scatter properties such as CD34 bright expression, CD45 dim expression, low side scatter, and low-to-moderate forward scatter with respect to lymphocytes as defining criteria for HSC identification.^[1,2]

There are two methods, namely single platform (SP) and double platform (DP).^[1,2] The known calibrated number of fluorescent beads is added to the sample and a single instrument (FCM) calculates and provides the final CD34+ cell count/µL in SP. In DP, the values obtained from both a conventional hematology analyzer (total white blood cell count) and FCM (number of CD34+ events) are used to estimate the CD34+ cell count/µL. As expected, in DP method, there is increased variability between the actual CD34+ count and the calculated value.^[3] Studies have shown that the SP is recommended but expensive while DP is more cost effective.^[4-8] In this study, our primary aim was to look for the degree of correlation between SP and DP and to look for cost–benefit analysis.

There was a lesser understanding among hematopathologists, wider technical variations and 46% were doing tests incorrectly in the European study^[9] and very similar study was lacking in India. Therefore, our secondary objective was to analyze the level of agreement between three hematopathologists for the standardized reporting.

MATERIAL AND METHODS

This was a retrospective study conducted from January 2019 to December 2021. All patients who underwent HSCT in whom FCM based CD34 enumeration was performed were included. Two milliliters of pre-apheresis and post-apheresis samples were received in an ethylenediaminetetraacetic acid tube and processed within 2 h in the laboratory.

The baseline clinical characteristics, factors affecting the yield, and the interference by hematogones in all these samples were reported and published by our team.^[10,11] Total leukocyte count (TLC) was noted from SYSMEX XN-1000 hematology analyzer. If TLC was very high, sample was diluted to get at least 30,000–40,000 cells/µL and the appropriate dilution was noted. ln our laboratory, we used STEM KIT™ (IM3630) from Beckman Coulter Limited and stain-lyse-no wash technique was used. In all the cases, the test was done in duplicate and we had followed the manufacturer’s instruction for performing the tests using SP method.

The 3-laser 10-color NAVIOS flow cytometer (Beckman Coulter) was used and the total number of events acquired was 1,00,000 using NAVIOS acquisition software. In all the cases, originally only SP method using ISHAGE protocol was performed and list mode data (LMD) files were analyzed using KALUZA version 2.1 software. The average value obtained from both the tubes was multiplied by the appropriate dilution factor to obtain the final CD34+ HSC. We had followed only SP for the reporting and release of reports in hospital information system. The lymphocytes (test) and CD45-Fluorescein isothiocyanate (FITC)/Control - Phycoerythrin (CTRL-PE) (Isoclonic) (kit) were used as controls.

There were three reporting hematopathologists with different levels of experience in hematology (Prabhu. M (PM) – 10 years, Rakhee Kar (RK) – 18 years, and Debdatta Basu (DB) – 30 years). The initial values reported by three hematopathologists were taken as “original SP analysis.” The LMD were retrieved and re-analyzed using SP method (“re-analyzed SP”) and we did not use tubes to process DP individually in all these cases. Only the original LMD files were used for the proxy DP method, and to avoid the inter-observer variability, only re-analyzed SP was compared with DP analysis for the correlation. The calculation used was as follows: Absolute CD34+ cells ×10⁶ = (No. of CD34+ cells × beads concentration)/No. of beads for the SP method [Figure 1] and absolute CD34+ cells ×10⁶ = (TLC × 10⁹/L × average no. of CD34+ events from FCM × 1000)/Average no. of CD45+ events from FCM for the DP method [Figure 2]. Spearman’s correlation coefficient (R²) was used to evaluate the correlation between “re-analyzed SP” and DP. The median coefficient of variation (CV) was obtained for preapheresis and post-apheresis. The CV <10% and R² >0.7 were considered as good degree of correlation between these methods.

Close

Figure 1:

Single platform flow cytometry gating strategy for CD34+ stem cell enumeration: (a) Data acquisition versus time (b) SSC versus 7-AAD viability plot (c) SSC versus CD45 plot (R1 gate) (d) SSC versus CD34 plot (R2 gate) (e) SSC versus CD45 plot (R3 gate) (f) SSC versus FSC plot (R4 gate) (g) SSC versus FSC plot (R5 gate) (h) CD34 versus CD45 plot (i) Beads. FITC: Fluorescein isothiocyanate, PE: Phycoerythrin, INT: Integer, SSC: Side scatter, 7-AAD: 7-aminoactinomycin D, FSC: Forward scatter

Export to PPT

Close

Figure 2:

Double platform flow cytometry gating strategy for CD34+ stem cell enumeration: (a) SSC versus CD 45 plot (R1 gate) (b) SSC versus CD34 plot (R2 gate) (c) SSC versus CD45 plot (R3 gate) (d) SSC versus FSC plot (R4 gate) (e) SSC versus FSC plot (R5 gate) (f) CD34 versus CD45 plot. SSC: Side scatter, FSC: Forward scatter, INT: Integer, FITC: Fluorescein isothiocyanate, PE: Phycoerythrin.

Export to PPT

Bias in this study was prevented by choosing the files randomly without assessing either SP or DP values obtained for the CD34 enumeration. This was carefully done by the first author and issued the data files for re-analysis to the corresponding author. All the participating hematopathologists were blinded to the original report. Second, the degree of agreement between “original SP analysis” and “re-analyzed SP” among the three hematopathologists (PM, DB, and RK) was analyzed using Kappa coefficient and value of >0.2 was considered as “fair” degree of agreement. All the analysis was done using the Statistical Package for the Social Sciences (SPSS) version 26 (SPSS Inc., Chicago, IL, USA).

RESULTS

In this study, totally, 78 cases were recruited and their details were discussed elsewhere.^[10,11] Table 1 and Figure 3 indicate a fair degree of correlation between the “re-analyzed SP” and DP with significant positive correlation. It was also evaluated that cost per test by the SP and DP was Rs. 1,950/- and Rs. 650/-, respectively, indicating the huge difference between the two methods [Table 2]. Similarly, there was fair degree of agreement among the different reporting hematopathologists [Table 1].

Close

Figure 3:

Correlation between (a) pre-apheresis and (b) post-apheresis re-analyzed single platform and double platform method.

Export to PPT

DISCUSSION

The ISHAGE provides guidelines using SP and DP with their own advantages and limitations. Naithani et al.^[4] found that the two methods on regression analysis had R: 0.967 and P < 0.0001 with median values of CD34 cell count being 498.5 (range: 1–6902) for SP and 494.5 (range: 0–8796) for DP. Similarly, Keeney et al.^[7] and Ngoma et al.^[8] also showed that comparison of methods revealed a high R² with no statistically significant difference or bias between the SP and DP (P > 0.05).

Recently, Berhili et al.^[12] compared the performance of SP and DP using the Sysmex XN-1000 hematology analyzer and analyzed 86 samples from multiple myeloma patients planned for autologous HSCT. They had specifically used nucleated red blood channel in a Sysmex XN-1000 analyzer in the enumeration of CD34 HSC using DP method and compared with the reference SP method. They found that both SP and DP were effective, yielded similar results, and concluded that DP method is simple and affordable where the Sysmex XN-1000 analyzer is available.

Our study findings were exactly similar and in concordance with the above studies and also assure the utility of DP method in resource-limited settings with the experienced hematopathologists. However, unlike our study, Murugesan et al.^[6] in their study found that the mean CD34+ count in the SP was higher than that in the DP.

The European study^[9] had demonstrated that only 57% of participating laboratories were using the ISHAGE protocol properly. They concluded that there was a fundamental lack of understanding and need to monitor the participants rigorously. Similarly, in our country, Tata Medical Center in Mumbai is conducting the interlaboratory comparative program. However, the results of such program are not available. Therefore, we wanted to streamline the technical aspects, gating strategies, reporting, and release of results among our three hematopathologists. Our study demonstrated that there was a fair degree of agreement among three hematopathologists having different levels of experience. This was a significant finding as no such comparison has been reported in the literature from our country. It is warranted that through the survey-based Delphi method or through multicenter collaborative study is required in the near future to improve the agreement between different laboratories, especially in our country.

Both methods, if validated in the laboratory, can be used for CD34+ HSC enumeration; however, SP is recommended but more expensive as controls/beads to be additionally purchased while DP method is cheaper and might be tried in cost-constrained setting but has the drawback of increased variability.

CONCLUSION

SP is recommended but more expensive while DP is cheaper and useful in cost-constrained setting. It is possible to achieve a fair degree of agreement among hematopathologists.