Sensitivity and Applicability of Six-Color Flow Cytometry Is Comparable to ASO-Primer-Real-Time PCR (RQ-PCR) for Minimal Residual Disease (MRD) Monitoring in Adult Acute Lymphoblastic Leukemia (ALL) - A Comparative Analysis in 70 Patients from the German Multicenter Study Group for Adult Acute Lymphoblastic Leukemia (GMALL).

Although the quantification of MRD by PCR or flow cytometry (FC) is of proven prognostic value in ALL, both methods were never compared in adults. Since 3- to 4-color FC showed equal or inferior sensitivity to PCR in childhood ALL, we herein prospectively compared six-color FC (6C-FC) at initial diagnosis and during early follow-up in adult ALL to RQ-PCR as reference method.

Diagnostic samples from 70 consecutive ALL patients (pts) (22 T-, 48 B-lineage; 18 to 71 yrs) were screened by 14 different multiplex immungene consensus PCRs as published (Bruggemann et al., Blood, 2006) and simultaneously by 8 different 6C-FC combinations which included a total of 29 different antibodies. At least one of the consensus PCRs revealed monoclonality in 64/70 pts (91 %), whereas immunophenotypic aberrations could distinguish ALL from benign hematopoiesis in 67/70 cases (96 %). 6C-FC was positive in all PCR negative samples, while PCR was successful in all samples lacking immunophenotypic aberrations. The standardized 6C-FC method revealed a median of 4 immunophenotypic aberrations per pt (range 1 to 8). The most frequently observed aberrations in B-lineage ALL included CD58++ (76 %), CD11a++(50%), CD22++ (44 %), low CD38 (50%), and KORSA+ (35 %). Tdt+cytoplasmatic(cy)CD3+ hallmarked 86% of T-lineage ALL cases which were additionally characterized by CD7+ (100%), surface(s) CD3- (90%), CD99+ (52 %), and CD1a+ (38 %).

Up to now, RQ-PCR assays have been established for 38 pts (median sensitivity 10⁻⁴; range 10⁻⁵ to 10⁻²) thus allowing for comparisons of MRD assessments to 6C-FC in 155 follow-up samples. 110/155 (71 %) samples were collected during the first 4 therapy months. 6C-FC marker combinations for follow-up always included CD34/CD10/CD19/CD45 in B-lineage and TdT/cyCD3/sCD3/CD5 in T-lineage ALL. Stainings were individualized by adding 2 pt-specific markers each to up to 3 tubes. A total of 155 follow-up specimens comprised 13 FC-/RQ-PCR+ (8 %), 4 FC+/RQ-PCR- (3 %), 88 concordantly negative (57 %), and 50 concordantly positive (32 %) samples. Due to extremely low MRD, in 11/13 FC-/RQ-PCR+ samples (85 %) the results of RQ-PCR were qualitatively positive only, but not accurately quantifiable.

The minimum MRD level detectable by 6C-FC was 3x10⁻⁵, as proven by a concordantly positive RQ-PCR result. MRD levels obtained by both methods correlated well (Spearman r = 0.85; p < 0.0001) in follow-up samples that were positive both by RQ-PCR and by 6C-FC. The median ratio between 6C-FC and RQ-PCR MRD results was 0.48 (range 0.012 to 113). In 11% of samples the ratio between MRD levels obtained by the two methods differed more than tenfold.

Our novel standardized 6C-FC approach is highly sensitive for MRD assessments in adult ALL. Our results suggest the applicability of 6C-FC in a multicenter setting, an excellent specificity, a good quantitative correlation and a similar sensitivity when compared to RQ-PCR. Longer follow-up and the inclusion of more samples in particular from later time points are required to decide whether or not the two methods are of comparable clinical significance.