Risk Factors Determining the Outcome of Critically Ill Allogeneic Hematopoietic Stem Cell Transplantation Patients: Time to Step Down?

The outcome of patients undergoing allogeneic hematopoietic stem cell transplantation (HCT) has improved over the past decades. However, the outcome of critically ill patients undergoing HCT admitted to intensive care units (ICU) is still considered dismal. Factors that predict outcome of this patient cohort and allow a proper triage assessment before and after ICU admission are still under discussion with no existing clinical standards.

To identify accessible predictive factors, we retrospectively evaluated the outcomes of 127 HCT patients who required admission to the ICU between 2009 and 2016 out of 580 (21.9%) HCT patients during this period (Table 1) in our center. In general, 72 of 127 patients (56.7%) died on the ICU, which is in line with previous publications. Respiratory failure was the leading cause for ICU admission (70%). The median age was 50 y (+/-12.7) (range, 19-73y) and ICU admission mainly occurred after the engraftment period (≥30 days after HCT) in 107 patients. The median time from HCT to ICU transfer was 228 days (+/-419.8).

To better understand factors that predict the survival of HCT patients upon ICU admission, we assessed 73 clinical factors and parameters. In the according multivariate analysis, three out of 73 factors negatively impacted ICU mortality as well as overall survival (OS). These were septic shock (HR:3.27, p=0.002), SOFA score >9 (HR:5.93, p<0.004) and C-reactive protein (CRP) levels >200 mg/l (HR:2.1, p=0.003). HLA-identical donors positively impacted only on ICU survival (HR:0.39, p=0.005), suggesting a benefit in this particular patient cohort. In contrast to previous publications, invasive mechanical ventilation and renal replacement therapy were no independent prognostic factors in any multivariate analysis in our cohort (p=0.63 and p=0.56, respectively).

Twenty-eight (50.9%) out of 55 ICU-survivors died after transfer from the ICU to the general ward. Main causes of death were infections (53.6%) and respiratory failure (42.9%). Of note, 12 patients (42.9%) out of these 28 had to be readmitted to the ICU, suggesting the necessity of an ICU-to-Intermediate care step-down, allowing a close monitoring after ICU dismissal. Only 27 patients (21.3%) of the initial cohort were discharged from the hospital, resulting in a hospital mortality rate of 78.7% (100/127 patients) and translating into a 1-year overall survival (OS) rate of 7.8% for all patients in our cohort. Independent predictors for hospital mortality were multiorgan failure (HR:2.45, p=0.016), SOFA score >9 (HR:1.86, p=0.01) and C-reactive protein levels > 200 mg/l (HR:2.15, p=0.002).

In summary we identified a high SOFA score and a high CRP value as predictors for a negative OS of HCT patients who required admission to the ICU. Although a prospective validation is beneficial, both factors can be assessed routinely and potentially allow the identification of high risk HCT patients that require a timely admission to ICU. In addition, close monitoring through a stepdown to intermediate care after ICU should be considered to improve mortality of this fragile patient cohort.

View largeDownload slide

View largeDownload slide

Close modal