Prognostic Factors Predicting Survival in De Novo AML Patients Requiring Intensive Care.

Abstract

Acute myeloid leukemia (AML) is an acute life-threatening disease with variable clinical presentation. In this study, the percentage of patients (pts) with de novo AML requiring intensive care prior or during induction chemotherapy (ICT), as well as prognostic factors predicting survival in these pts were analyzed. A total of 471 consecutive pts (median age 62 years; range: 16–92) seen at the Vienna University Hospital between 1994–2006 were enrolled. In pts requiring critical care, simplified acute physiology score (SAPS) II as well as the need for invasive mechanical ventilation (IMV), vasopressor support (VP), and disease related markers were recorded at the intensive care unit (ICU). Eighty six percent (n=404) of all patients were eligible for ICT. Fifty four of these 404 patients (13.4%) required critical care prior or during ICT (median SAPS II 64, range 30–107), primarily due to respiratory failure (26 pts=48%) or life-threatening bleeding (12 pts=22%). Comparing ICU and non-ICU-pts with regard to disease-related markers, differences were found in white blood cell counts, WBC (ICU: 16.8 G/L; non-ICU: 11.9 G/L; p=0.084), whereas no differences were found regarding age, plt, and LDH. Forty pts received IMV (63%), and 32 VP (59%). The ICU survival rate was 41%. Significant prognostic factors with respect to ICU-survival were higher SAPS II scores (p<0.05), the need of IMV (p<0.05), and need of VP (p<0.05), whereas CRP, WBC, age, karyotype, or the time of admission to ICU (prior or during ICT) were not of prognostic significance. Survival was favourable in non-ICU-pts (median: 4.14 months; 22% at 8 years) compared to ICU-pts (median: 1.2 months; 9% at 8 years; p<0.05). Similar results were obtained when analyzing the overall survival, OS (non-ICU-pts: median: 4.1 months; 22% at 8 years; ICU-pts: median: 1.6 months; 12% at 8 years; p<0.05). Interestingly, the continuous complete remission, CCR (non-ICU-pts: 37% at 6 years; ICU-pts: 31% at 6 years; p>0.5) as well as OS of patients who had survived the first 28 days of therapy (non-ICU-pts: 29% at 6 years; ICU-pts: 20% at 6 years; p>0.5) did not differ significantly between ICU-pts and non-ICU-pts. With regard to OS, multivariate analysis revealed that ICU admission was an independent adverse prognostic parameter, as was a higher WBC, advanced age, higher LDH, or unfavourable karyotype. With regard to CCR, age and karyotype were independent prognostic variables, whereas ICU-admission was not of prognostic significance. In summary, 13% of pts with de novo AML eligible for ICT required critical care, primarily due to respiratory failure or bleeding. The probability of survival and OS of ICU-pts is inferior compared to non-ICU-pts. However, with regard to CCR and OS of pts surviving 28 days, no differences were observed between ICU-pts and non-ICU-pts. These observations favour the assumption that critical care should be considered in all de novo AML pts eligible for ICT.