Improved Concordance of Minimal Residual Disease Measurements By Quantitative PCR and 10-Color Flow Cytometry in Pediatric Acute Lymphoblastic Leukemia

Introduction:

Detection of minimal residual disease (MRD) after induction and consolidation therapy is highly predictive of outcome for childhood acute lymphoblastic leukaemia (ALL) and is used to identify high risk patients in most current ALL clinical trials. Two methods broadly applicable for MRD analysis in ALL cases are real-time quantitative PCR based detection of unique immunoglobulin and T-cell receptor gene rearrangements (Ig/TCR PCR-MRD) and the multi-parameter flow cytometry based quantitation of Leukemia Associated  Immunophenotypes (LAIP Flow-MRD).   We compared the two techniques using samples from patients referred for PCR-MRD analysis initially   using 4-tube 4-colour flow and more recently 1-tube 10-color flow.

Methods:

 Newly diagnosed consented ALL patients enrolled on ANZCHOG ALL8 (2002-2011) or AIEOP-BFM ALL 2009 (2012-2014) had duplicate bone marrow aspirates, collected at diagnosis, day 15, day 33 and day 79, and analysed by PCR-MRD and Flow-MRD techniques. PCR-MRD analysis utilized clone specific primers and generic probes for Ig/TCR rearrangements according to EuroMRD guidelines. Flow-MRD which detects levels of aberrant combinations of cell-surface proteins using fluorescently labelled antibodies was performed until 2009 with 4-tube 4-colour flow before we adopted a 1-tube approach (9-colour for BCP-ALL and 10-colour T-ALL) based on the AIEOP-BFM harmonised protocol for 2012-2014.  

Results:

 Our early comparison showed a relatively poor correlation of 4-colour Flow-MRD results with PCR-MRD (Spearman rank correlation coefficient rho = 0.516, n= 267) for patients enrolled at a single centre on ANZCHOG ALL8 in 2002-2009. Only the PCR-MRD results were used for the MRD risk-adapted stratification for patients on this trial. Flow-MRD for subsequent patients on this trial (2010-11) was improved by using more antibodies and adopting a single tube approach. In our current trial, day 15 Flow-MRD results are used for the early identification of low risk patients for a randomized treatment reduction. In bone marrow samples from patients enrolled on this trial, the correlation of the PCR-MRD and Flow-MRD methods is high when considered for all time points (rho = 0.803 n=418; Figure 1).  In the same set of patient samples, the concordance between 2 different PCR markers based on different rearrangements was even better (rho = 0.929, n=390). A comparison of time points found that the best correlation between the two methods was observed at day 15 when MRD is often higher and the bone marrow is not regenerating (Table 1). Both PCR and 10-colour flow enabled MRD to be performed for 94% of ALL patients, and only one patient did not have a sensitive MRD assay. 

 

Conclusion:

The adoption of new approaches to measurement of Flow-MRD, using a single tube and 10-colors, for ALL patients has greatly improved the concordance of Flow-MRD and PCR-MRD results. It is not surprising given the different nature of the techniques that the correlation of results produced by two different markers for PCR-MRD is higher than that with Flow. However we conclude that these two methods can now be used interchangeably at day 15 in BFM-style protocols for ALL patients.  The concordance at later time points is weaker and warrants investigation in the whole trial cohort to enable effects of ALL subtype and patient outcomes to be evaluated.

 

Figure 1. Comparison of MRD levels measured by 1-tube 10-color Flow MRD versus PCR MRD (left) or by two different PCR Ig/TCR MRD markers (right) in the 418 and 390 paired measurements in the same set of patients.

 

 

Table 1. Concordance of MRD levels at different time points in the same set of patients (Spearman’s Rank correlation coefficient rho).

 

	MRD by PCR first Ig/TCR marker versus MRD by 10-colour flow	MRD by first Ig/TCR PCR marker versus second Ig/TCR marker
All timepoints	0.803  (n=418)**	0.921  (n=390)**
Day 15	0.795  (n=155)**	0.950  (n=129)**
Day 33	0.417 (n=137)	0.826  (n=132)**
Day 79	0.383 (n=126)	0.842  (n=129)**

** Correlation is significant at the 0.01 level (2 tailed)

Support: NHMRC Australia APP1057746 and Tour De Cure Foundation