Validation of a Flow Cytometric Scoring System for the Prognostication of the Myelodysplastic Syndromes: a Retrospective Cohort Study

Abstract 4012

The myelodysplastic syndromes (MDS) are a heterogeneous group of clonal hematopoietic stem cell disorders, characterized by cytopenia(s), dysplasia and a propensity to evolve into acute myeloid leukemia. The International Prognostic Scoring System (IPSS) and WHO-based prognostic scoring system provide prognostic information. However, even if patients are allocated in the same risk category their clinical course remains heterogeneous. Recent developments in the treatment of MDS require refinement of prognostication and identification of patients who might benefit from treatment with potentially disease modifying agents such as lenalidomide or azacitidine. Flow cytometry (FC) is emerging as a valuable technique for the diagnosis and prognosis of MDS. Recently, we demonstrated that flow cytometric analysis of BM in low and int-1 risk MDS is instrumental to identify clinically relevant subgroups. (Westers et al, Blood 2010) Previously, it was reported that a flow cytometric scoring system (FCSS) is predictive for worse outcome in MDS. (Wells et al, Blood 2003, van de Loosdrecht et al, Blood 2008) The FCSS is a scoring system that allows for a numerical display of immunophenotypic aberrancies in the (im)mature myelo-monocytic lineage. Scores are generated by enumerating abnormalities; a high score reflects a high number of aberrancies. The current study aimed to validate the FCSS for identification of prognostic subgroups in MDS. We analyzed aberrancies in (im)mature myelo-monocytic cells by FC in BM of 102 MDS patients, including 48 MDS patients from the previous cohort. The diagnoses according to WHO 2001 classification were RA(RS) n=19, RCMD(RS) n=54, RAEB-1 n=11, RAEB-2 n=13, MDS-U n=5 and also age-matched healthy volunteers (n=39) were included. The median age of MDS patients was 66 and of healthy volunteers 57. The FCSS in RA(RS) (median=3, range 1–6) patients was significantly higher compared with healthy controls (median=1, range 0–2, p<0.001). In contrast to our previous results the FCSS for RA(RS) and RCMD(RS) patients did not differ. This is a remarkable finding, since by morphology RA(RS) patients have unilineage dysplasia, in contrast to flow cytometric findings, where 84% (16/19) of the RA(RS) patients had two or more aberrancies in the (im)mature myelomonocytic compartment. The FCSS was higher in RAEB-1 (median=6, range 2–7) and RAEB-2 (median=6, range 4–8) compared with RCMD(RS) (median=3, range 0–6, p=0.02 and p<0.0001, respectively). Overall, the FCSS correlated significantly with WHO 2001 classification (p<0.0001). The FCSS showed a significant correlation with IPSS categories low, int-1, int-2 and high (p<0.0001). Remarkably, the FCSS was not correlated with cytogenetic risk categories low, intermediate and poor. The new German-Austrian Cytogenetic Prognostic Scoring System for MDS was also not correlated with the FCSS. This indicates that the FCSS and cytogenetics might provide separate prognostic information in MDS. Neutrophil granularity corresponding with side scatter by FC was significantly decreased in MDS patients compared with healthy volunteers (p<0.0001). In the RA(RS) and RCMD(RS) category, 40% (29/73) of patients expressed an aberrant marker such as CD5, CD7 and/or CD56 on myeloid progenitors. Transfusion data was available of 51 patients. Interestingly, the majority of MDS patients who were transfusion dependent or progressive, had aberrant expression of CD5, CD7 and/or CD56 on myeloid progenitors compared with MDS patients without aberrant marker expression (64% (16/25) vs 31% (8/26), respectively p=0.04). When the cumulative amount of all aberrancies in the (im)mature myelo-moncytic cells were taken into account, transfusion dependent patients had significantly more aberrancies than transfusion independent MDS patients, (median 6.5 vs 4, respectively, p=0.006). In conclusion, we here confirmed our previous findings in a larger cohort. The majority of RA(RS) patients already has multilineage dysplasia as detected by FC, which might be of prognostic relevance. Although the FCSS correlates with current prognostic systems, a striking heterogeneity remains within prognostic subgroups. Therefore, the FCSS and detection of aberrant myeloid progenitors can provide refined prognostication by identification of patients at risk for transfusion dependency and adverse clinical outcome, independent of current classification systems.