Table 2.

Predictors for late relapse—multinominal logistic regression.

Variable	Odds Ratio ”	95% CI	p
Treatment prior to allo-SCT Upfront #/CR vs. No CR	3.402	1.140–10.155	0.028
ELN 2017//IPSS-R//CPSS at diagnosis (14,16,18) * Favorable/intermediate Unfavorable	2.777	0.755–10.213	0.124
Karyotype Not complex vs. Complex	3.038	0.695–13.278	0.140

^” Odds ratio represents odds of late relapse versus odds of early relapse among patients who experienced relapse. ^# upfront: patients who proceeded to transplantation without prior cytoreductive chemotherapy, defined as no MDS- or AML-MRC-specific therapy except for growth factors, transfusions and a short period (<1 week) of hydroxyurea. * very good/good IPSS-R, favorable ELN and low CPSS risk groups were combined as one favorable disease risk group; intermediate IPSS-R, intermediate ELN and intermediate I/II CPSS formed one intermediate disease risk group; and high/very high IPSS-R, adverse ELN and high CPSS were summarized as one unfavorable disease risk group. The disease risk of patients with MDS, AML-MRC and CMML was estimated using IPSS-R, ELN 2017 classification and CPSS, respectively, and the respective subgroups were summarized as described above. Allo-SCT, allogeneic hematopoietic stem-cell transplantation; CI, confidence interval; CPSS, CMML-specific scoring system; CR, complete Remission; int, intermediate; ELN, European Leukemia Net; IPSS-R, International Prognostic Scoring System—Revised Version.