Detection of Minimal Residual Disease by Flow-Cytometry Otimizes Risk Stratification When Combined with Cytogenetics in AML.

Cytogenetic risk assessment at diagnosis is a major determinant of outcome in adult patients with AML. Nevertheless, the impact of this up-front prognosticator may be altered by additional factors intervening during and after treatment such as toxicity, speed of blast clearance and minimal residual disease (MRD), both the latters reflecting the quality of response. We have demonstrated that the persistence of given amounts of leukemic cells after consolidation, affects the prognosis of AML, independently from cytogenetics, leading to a faster leukemic relapse. In the present study, we aimed at proving that the integrated analysis of karyotype and MRD determination by multiparametric flow-cytometry (MPFC) may improve the risk stratification process of patients belonging to established cytogenetic groups. We analyzed 127 non-M3 AML cases entered into the EORTC/GIMEMA protocols AML10/AML12 (age <61yrs) or AML13/AML15/AML17 (age >61 yrs). By applying the maximally selected log-rank statistics, the threshold discriminating MRD negative from positive cases was set at 3.5 x10⁻⁴ residual leukemic cells, a level that distinguished, at the post-consolidation time-point, discrete subsets of patients with different prognosis. According to the MRC classification, 18/127 (14%) and 7/127 (6%) were classified into the favorable and unfavorable risk categories, respectively; 102/127 (80%) patients had an intermediate risk karyotype. As expected, having a favourable, intermediate or unfavourable karyotype was associated with a significant difference in terms of relapse rate (28%, 63%, 100%; p=0.005), 5-years overall survival (OS) (67%, 27%, 14%; p<0.001) and 5-years relapse free survival (RFS) (68%, 28%, 0%; p<0.001). These three groups, also differed for the rate of MRD negativity: 78% in the favourable group, 28% in the intermediate group, whereas none among unfavourable risk patients reached a MRD⁻ status (p<0.001). Therefore, we examined each cytogenetic group according to the MRD status and we observed that:

1. patients bearing favourable karyotypes and achieving MRD negativity (good-MRD⁻) had a better outcome in terms of 5-years OS (81%) and RFS (76%) as compared to those remaining MRD positive (good-MRD⁺) (5-years OS % and RFS 25% and 43%, respectively) (p=0.06).

2. also the patients with intermediate risk karyotype were dissected in two distinct categories by MRD status, MRD⁺ patients (Int-MRD⁺) with a very poor prognosis (5-yrs RFS and OS less than 20% for both) and MRD⁻ ones (Int-MRD⁻) (5-yrs RFS and OS 69% and 50%, respectively) (p<0.001).

Therefore, favorable karyotype does not guarantee by definition a good prognosis; in fact, good-MRD⁺ patients have a high relapse rate (50%) and a disappointing long term outcome; similarly, the outcome of Int-MRD⁺ patients was comparable to that of poor risk cytogenetics whereas Int-MRD⁻ patients fared as favorably as those in the good karyotype category. In conclusion, an appropriate risk assessment should result from the combination of pre-treatment and delayed parameters (cytogenetics plus post-consolidation MRD determination) rather than relies on the evaluation of sole up-front prognosticators; this approach would enable to outline in a more reliable manner the outcome of adult patients with AML and the relative post-remissional therapy.