TY - JOUR
T1 - Temporal trends and prognostic factors in critically ill adult patients with acute leukemia
T2 - an individual participant data meta-analysis
AU - Chean, Dara
AU - Dupont, Thibault
AU - Nates, Joseph L.
AU - Manjappachar, Nirmala K.
AU - Medic, Marijana
AU - Aygencel, Gulbin
AU - Orvain, Corentin
AU - Kouatchet, Achille
AU - Sheshadri, Ajay
AU - Haque, Tasmea K.
AU - Russell, Lene
AU - Pabst, Thomas
AU - Turkoglu, Melda
AU - Ostermann, Marlies
AU - Soares, Marcio
AU - Cetintepe, Tugba
AU - Buckstein, Rena
AU - Issa, Nahema
AU - Mokart, Djamel
AU - Caruso, Pedro
AU - Cornejo-Juarez, Patricia
AU - Schnell, David
AU - Lahmer, Tobias
AU - Fodil, Sofiane
AU - Lengliné, Etienne
AU - Luque-Paz, David
AU - Lemiale, Virginie
AU - Poole, Daniele
AU - Darmon, Michael
AU - Azoulay, Elie
N1 - Publisher Copyright:
© Springer-Verlag GmbH Germany, part of Springer Nature 2026.
PY - 2026
Y1 - 2026
N2 - Purpose: Critically ill patients with acute leukemia often require an intensive care unit (ICU) admission. As major therapeutic advances have been made during the last decades, the aim of this study was to assess temporal trends in ICU mortality, and identify prognostic factors to inform clinician decision-making. Methods: We conducted an individual participant data meta-analysis of studies including adults with acute leukemia admitted to the ICU. Patients with a history of allogeneic hematopoietic stem cell transplantation were excluded. Mixed-effects logistic regression models, accounting for center of ICU admission as a random variable, evaluated factors associated with ICU mortality, with particular focus on year of ICU admission, age (> 65 years) and invasive mechanical ventilation. Results: A total of 2003 patients from 55 ICUs across 19 countries were included (median age 58 years [IQR 44–67]; 72% acute myeloid leukemia [AML]; 64% admitted during induction chemotherapy). Invasive mechanical ventilation, vasopressors, and renal replacement therapy were required in 55%, 57%, and 21% of patients, respectively. Crude ICU mortality was 45% overall and 66% among ventilated patients. Age > 65 years (odds ratio (OR) 1.98 [95% CI 1.49–2.64]), diagnosis of AML (OR 1.70 [1.23–2.34]), admission during diagnosis or induction chemotherapy (OR 1.50 [1.08–2.07]), relapsed or refractory disease (OR 2.08 [1.36–3.21]), the need for mechanical ventilation (OR 6.46 [4.84–8.63]), and the need for other life-sustaining therapies (OR 2.21 [1.62–3.02]) were associated with increased ICU mortality. Year of ICU admission was associated with improved survival only among ventilated patients (OR per additional year 0.93 [0.93–0.93]). Conclusions: In this large international individual participant meta-analysis, survival of critically ill patients with acute leukemia improved over time, particularly among those requiring mechanical ventilation. Age and the need for mechanical ventilation and other life-sustaining therapies remain strong, independent predictors of ICU mortality. Future work should integrate frailty and functional assessments to refine prognostic stratification and guide treatment intensity in this complex population. Trial registration: The protocol was registered in PROSPERO (CRD420251046286). Visual abstract: (Figure presented.)
AB - Purpose: Critically ill patients with acute leukemia often require an intensive care unit (ICU) admission. As major therapeutic advances have been made during the last decades, the aim of this study was to assess temporal trends in ICU mortality, and identify prognostic factors to inform clinician decision-making. Methods: We conducted an individual participant data meta-analysis of studies including adults with acute leukemia admitted to the ICU. Patients with a history of allogeneic hematopoietic stem cell transplantation were excluded. Mixed-effects logistic regression models, accounting for center of ICU admission as a random variable, evaluated factors associated with ICU mortality, with particular focus on year of ICU admission, age (> 65 years) and invasive mechanical ventilation. Results: A total of 2003 patients from 55 ICUs across 19 countries were included (median age 58 years [IQR 44–67]; 72% acute myeloid leukemia [AML]; 64% admitted during induction chemotherapy). Invasive mechanical ventilation, vasopressors, and renal replacement therapy were required in 55%, 57%, and 21% of patients, respectively. Crude ICU mortality was 45% overall and 66% among ventilated patients. Age > 65 years (odds ratio (OR) 1.98 [95% CI 1.49–2.64]), diagnosis of AML (OR 1.70 [1.23–2.34]), admission during diagnosis or induction chemotherapy (OR 1.50 [1.08–2.07]), relapsed or refractory disease (OR 2.08 [1.36–3.21]), the need for mechanical ventilation (OR 6.46 [4.84–8.63]), and the need for other life-sustaining therapies (OR 2.21 [1.62–3.02]) were associated with increased ICU mortality. Year of ICU admission was associated with improved survival only among ventilated patients (OR per additional year 0.93 [0.93–0.93]). Conclusions: In this large international individual participant meta-analysis, survival of critically ill patients with acute leukemia improved over time, particularly among those requiring mechanical ventilation. Age and the need for mechanical ventilation and other life-sustaining therapies remain strong, independent predictors of ICU mortality. Future work should integrate frailty and functional assessments to refine prognostic stratification and guide treatment intensity in this complex population. Trial registration: The protocol was registered in PROSPERO (CRD420251046286). Visual abstract: (Figure presented.)
KW - Acute leukemia
KW - Acute lymphoid leukemia
KW - Acute myeloid leukemia
KW - Individual participant data meta-analysis
KW - Intensive care
KW - Mechanical ventilation
KW - Temporal trends
UR - https://www.scopus.com/pages/publications/105037886867
U2 - 10.1007/s00134-026-08449-8
DO - 10.1007/s00134-026-08449-8
M3 - Journal article
C2 - 42081104
AN - SCOPUS:105037886867
SN - 0342-4642
JO - Intensive Care Medicine
JF - Intensive Care Medicine
ER -