Predictors of Mortality in Patients Undergoing Autologous Stem Cell Transplant (ASCT) Admitted to the Intensive Care Unit: An Institutional Review of 1013 Transplant Patients over Five Years.

Abstract

Study Objective: To describe the outcomes of ASCT recipients transferred to the Intensive Care Unit (ICU), and identify predictors for mortality.Methods: Retrospective review of all ASCT recipients from Jan 2001-July 2006 who required ICU transfer up to 100 days post ASCT. Measurements and mainResults: Thirty-four of 1013 patients (3.3%) who underwent ASCT, were admitted to the ICU. The mean age at admission was 54.9 +/− 11.1 (range 28–71), 53% being female. Indications for ASCT included multiple myeloma (50%), amyloidosis (32%), or other malignancies (18%). Table 1 highlights the admission rate to the ICU by diagnosis. The primary admitting diagnosis in the ICU included sepsis (32%), cardiac related events (26%), or respiratory compromise (29%). Median days post ASCT was 10.0 days with a median in ICU stay of 4.0 days (range 1–37 days). Twenty patients (including all non-survivors) required mechanical ventilation for > 24 hours with a median duration of 3.0 days. Thirteen patients died (38%) in the ICU, with 11 dying of multi-system organ failure and 2 from cardiac arrest. Retrospectively collected parameters restricted to the first 24 hours of admission revealed that Sequential Organ Failure Assessment (SOFA) score (OR 1.30; CI95 1.09–1.64, P=0.003) and Acute Physiology and Chronic Health Evaluation (APACHE II) score (OR 1.43; CI95 1.14–2.16; P=0.0002) were statistically associated with mortality in univariate analysis. The variables predictive of mortality at 24 hours after admission are displayed in Table 2.Conclusion: ICU admission is uncommon, occuring in 3% of patients undergoing ASCT, of which 38% die (1% of total ASCTs). Admission is influenced by underlying diagnosis, with amyloid patients portending the highest risk. Mortality in ASCT patients admitted to the ICU can be predicted in the first 24 hours by specific assessment scores (SOFA and APACHE II); specific supportive care requirements: inotropic dependence, hemodialysis, and need for ventilation; and clinical findings of gram negative sepsis or > 2 organ failure. Patients with febrile neutropenia had a low risk of mortality (possibly due to aggressive antibiotic use, growth factors, and rapid engraftment post ASCT). These results may assist clinical decision making regarding the continuation of intensive care delivered 24 hours after admission.Percentage Admission Rate by Diagnosis (n = 1013)DiagnosisASCT (#)ICU Admission (#)/ (%)Non-survivors (#)Multiple Myeloma61517 / (2.8%)6Non-Hodgkin's Lymphoma1992/ (1.0%)1Hodgkin's Lymphoma1121 / (0.9%)0Amyloidosis3911/ (28.2%)6Acute Myeloid Leukemia171/ (5.9%)0Other (Germ Cell Tumour, Waldenstrom's Macroglobuliemia, POEMS)312/ (6.4%)0Variables Predictive of Mortality at 24 hours after AdmissionVariable PredictorsNumber of PatientsSurvivors (n = 21)Non-survivors (n = 13)P-valueFebrile Neutropenia1513 (62%)2 (15%)0.013Failure of > 2 organs209 (43%)11 (85%)0.030Mechanical Ventilation209 (43%)11 (85%)0.030Inotropic Support > 4 hours103 (14%)7 (54%)0.022Hemodialysis124 (19%)8 (62%)0.025Gram Negative Infection61 (5%)5 (42%)0.016