Outcomes for Older Patients with Acute Myeloid Leukemia Admitted to the Intensive Care Unit

Introduction: Older adults (≥ 60 years) with acute myeloid leukemia (AML) generally have a poor prognosis and a high rate of health care utilization including intensive care unit (ICU) admissions. While the outcomes of older patients with advanced solid tumors admitted to the ICU are poor, the prognosis of older patients with AML has not been well studied. We sought to examine the outcomes of this population following an ICU admission and to identify predictors of mortality.

Methods: We conducted a retrospective analysis of 329 consecutive patients (≥60 years) diagnosed with AML between 5/1/2005 and 12/31/2011 at two tertiary care hospitals to examine the incidence and primary reason for ICU admission. We compared clinical characteristics between patients admitted to the ICU versus those not admitted to the ICU using Kruskal-Wallis rank tests and Chi square tests for continuous and categorical variables, respectively. We determined rates of ICU, hospital, and 30-day survival. We identified predictors of ICU and hospital mortality using purposeful modeling to assess the effect of age, gender, performance status, disease risk (based on SWOG risk stratification), induction strategy (intensive vs. non-intensive chemotherapy), receipt of stem cell transplantation (HCT), comorbidity score (HCT-comorbidity index), ICU length-of-stay, and multi-organ involvement.

Results: 25% (82/329) of patients in this cohort had 107 ICU admissions. Patients admitted to the ICU (n=82) were similar to those without an ICU admission (n=247) in terms of gender, race, marital status, education, disease risk, receipt of HCT, and performance status. However, they were younger (median 67 vs. 69, P= 0.01), had a higher HCT-comorbidity index (median 2 vs. 0, P = 0.03), and were more likely to have received intensive induction chemotherapy (72% vs. 56%, P=0.01). The primary reasons for ICU admission were respiratory failure (45%), hypotension (36%), and bleeding (17%). The median length of stay in the ICU was 3 (range 1-6) days. Among those admitted to the ICU, 66%, 52%, and 50% were alive at ICU discharge, hospital discharge, and 30 days post-ICU admission, respectively. Age, gender, performance status, comorbidity score, disease risk, induction strategy, and receipt of HCT were not associated with survival outcomes. In multivariable logistic regression analyses, multi-organ dysfunction was associated with higher ICU mortality (OR 5.3, 95% CI 2.0 - 14.3, P = 0.001) and hospitalization mortality (OR 3.3, 95% CI 1.4-8.3, P = 0.006). Longer stay in the ICU showed a trend towards higher ICU mortality (OR 1.1, 95% 1.0-1.1, P= 0.07).

Conclusion: A substantial proportion of older patients with AML are admitted to the ICU. However, approximately one-half survive and are discharged from the hospital. Multi-organ failure is strongly associated with higher mortality. Unlike older patients with advanced solid tumors, these findings suggest that critically ill older patients with AML who lack multi-organ dysfunction may benefit from ICU level of care and have good survival outcomes.