Adverse Effect of Very Poor Cytogenetics and Monosomal Karyotype On Outcomes Following T-Deplete Reduced Intensity Conditioned Stem Cell Transplant for MDS and AML.

Abstract 3141

Cytogenetics are an important predictor of outcomes for patients diagnosed with MDS and AML. The new 5-group cytogenetic classification for MDS and identification of adverse prognosis associated with monosomal karyotype (MK) has enhanced risk stratification. Recent data have validated these factors to predict outcomes post-haematopoietic stem cell transplant (HSCT) in a multicentre study of predominantly T-replete transplantation. We sought to evaluate the 5-group cytogenetic classification and prognostic significance of MK at diagnosis and pre-HSCT in patients with MDS and AML secondary to MDS (MDS-AML) following T-cell deplete reduced intensity conditioned HSCT.

We reviewed karyotypic profiles in 267 patients transplanted between 1999 and December 2011. Conditioning regimens consisted of FBC (fludarabine 150mg/m² iv, busulphan 8mg/kg oral or 6.4 mg/kg iv, alemtuzumab 100mg iv) or FBATG (fludarabine 150mg/m² iv, busulphan 6.4 mg/kg iv, ATG 6mg/kg iv) in 190 (71%) and 18 (7%) of patients respectively. A double dose of busulphan was used in 50 (19%) FB2C and 9 (3%) FB2ATG cases. Cyclosporine was used for GVHD prophylaxis.

Diagnoses included RA/RCMD (n = 66, 25%), RAEB (n=65, 24%), MPN/MDS (n=22, 8%) and MDS-AML (n= 114, 43%). Using the IPSS, 176 (66%), 50 (19%) and 41 (15%) patients had good, intermediate and poor prognosis karyotypes respectively. Using the 5-group cytogenetic classification, patients were stratified: very good (n=3, 1%), good (n= 178, 67%), intermediate (n=46, 17%), poor (n=25, 9%) and very poor (n=15, 6%). MK was observed in 20 (8%) of patients.

Seventeen (6%) patients had >5% blasts and 42 (16%) had cytogenetic abnormalities detectable pre-HSCT. Those without karyotypic abnormality pre-HSCT had normal cytogenetics (n=185, 69%), no analysable metaphases (n=20, 8%) or unavailable results (n=20, 8%). Cytogenetic abnormalities pre-HSCT in the 42 patients revealed fewer very good (n=3, 7%) and good karyotypes (n=3, 7%) and a higher preponderance of intermediate (n=22, 52%), poor (n=8, 19%) or very poor (n=6, 14%) karyotypes. MK was present in 7 (17%) cases. Their conditioning regimens comprised FBC (n=23, 55%) or FBATG (n=2, 5%); 17 received double-dose busulphan regimens: FB2C (n=15, 36%) or FB2ATG (n=2, 5%).

Overall, 5 year OS, EFS (death from any cause, relapse and graft failure) and cumulative incidence of relapse (CIR) were 45%, 34% and 48% respectively. Graft failure occurred in 17 (7%) patients. Non-relapse mortality rates were 18% and 27% at 1y and 5y respectively.

Grouping by traditional IPSS cytogenetics at diagnosis did not significantly influence survival or relapse. Classification of cytogenetics at diagnosis into 5 groups demonstrated patients with very poor karyotype to have significantly inferior OS, EFS and CIR compared with the other 4 groups. None of the 15 cases with very poor karyotype are alive at 5y; 5y OS was 48%, 49% and 43% for very good/good, intermediate or poor risk groups respectively (p<0.01). Patients with MK had 5y OS of 13% vs 48% for those without MK (p<0.01). CIR was 100% and 68% for the very poor karyotype and MK patients respectively.

Persistent cytogenetic disease pre-HSCT was associated with significantly inferior 5y OS of 33%, vs 49% without (p=0.01); 5y EFS and CIR were 23% vs 37% (p=0.013) and 72% vs 43% (p<0.01) respectively. Patients with >5% BM blasts pre-HSCT had worse 5y OS at 32% vs 46% for those with <5% blasts (p=0.027) and higher 5y CIR (68% vs 47%, p<0.01).

MK, BM blasts >5% at HSCT and residual cytogenetic abnormalities pre-HSCT all retained significance on multivariate analysis with respect to OS (HR 1.97 [95% confidence interval (CI) 1.13–3.45], 1.92 [95% CI 1.07–3.43] and 1.33 [95% CI 1.02–1.73] respectively); very poor karyotype did not retain significance (with HR 1.08 [95% CI 0.82–1.41]).

These data demonstrate the adverse impact of very poor karyotype and MK at diagnosis on outcomes post-T-deplete HSCT for MDS and MDS-AML, emphasizing the effect of cytogenetics regardless of therapeutic intervention. In addition residual cytogenetic disease pre-transplant predicts inferior results despite higher intensity conditioning used in these patients. Future directions to improve outcomes for these poor prognostic groups should include prospective trials to determine the value of complete eradication of residual cytogenetic disease pre-HSCT and identification of novel strategies directed at preventing relapse.