The End of Life for Patients with Acute Myeloid Leukemia (AML)- a Single Center Experience

Background

The diagnosis of acute myeloid leukemia (AML) is associated with a poor prognosis, particularly for patients older than age 65 and those with relapsed or refractory disease.  Little is known about the specific location where AML patients die and how disease status, therapeutic management, and symptoms influence the place of death.

Methods

We conducted a retrospective descriptive analysis of consecutive patients with AML who died at our institution between January 2007 and December 2012.  In addition to place of death, the variables of interest included: disease status at death, previous transplant, prior chemotherapy, presence of a dedicated caregiver, gender, and age at diagnosis. Disease status was classified as: new diagnosis, active disease, relapsed, or in remission. A new diagnosis was defined as a diagnosis within the prior 30 days. Active disease was defined as disease beyond 30 days and without prior remission.  Medical records were reviewed to identify whether palliative care service was consulted, and the range of symptoms reported during the last two weeks of life.  Categorical variables were compared among groups using chi-square and continuous variables with the Kruskal Wallis test. Results

We identified 166 patients with AML. The majority of study patients were white (90%) and male (55%) with a median age of 69 years (range 21-94 years).  Disease was characterized as de novoAML (49%), therapy related (14%), secondary to antecedent myelodysplasia (29%) or other hematologic disease (8%).  The median time from diagnosis to death was 5.5 months (range 0.7-50.4).  The majority (82%) had active AML at time of death. Other causes of death included transplant related complications (8%), sepsis (4%), bleeding (2%), or other (4%) causes.  Palliative care consultation occurred in 35% of the cases.  The consults typically were late in the course of illness with the median time from initial consult to time of death of 8 days (range 0-122 days).  Information on the location of death was available for 154 patients, of whom, 23 (15%) died in inpatient hospice, 41 (27%) at home, 48 (31%) died while hospitalized in a non-ICU bed and 42 (27%) died in either the Intensive Care Unit or Emergency Department (ED). Location of death was significantly associated with disease status at death (p< 0.0001, Figure 1), prior anthracycline/cytarabine based induction therapy (p=0.01), social support/dedicated caregiver (p=0.05), and age at diagnosis (p<0.0001).  Of the patients that died in the ICU/ED, a higher proportion were younger with median age of 61 years (range of 30-83). Of those with a new diagnosis of AML, 36% died in the ICU or  in the non-ICU hospital bed (33%) and 30% at home or inpatient hospice.  Patients in complete remission mostly died in the ICU or ED (80%) while those with relapsed disease died in a range of settings including at home (33%), hospital (33%), inpatient hospice (20%), or in the ICU/ED (15%). A higher proportion (23%) of those who had not received induction chemotherapy died in inpatient hospice than those who had (9%). Conversely, those that had received induction chemotherapy had a higher proportion of ICU/ED deaths (36%) than those that had not received induction (15%). A higher proportion of those with social support died at home/hospice (30% vs. 5%) and a lower proportion died in inpatient hospice (13% vs. 27%). The most commonly reported symptoms included pain (38%), delirium (29%) and bleeding (23%).  The 37 reported bleeding events included oral-nasal cavity bleeding (28%), soft tissue bleeding (5%), large volume pulmonary (8%), gastrointestinal (13%), and intra-cranial hemorrhage (38%).

Conclusions

The majority of AML patients are in a hospital setting at the time of death.   Important factors for the location of death include age at diagnosis, disease status, social support and prior induction chemotherapy. Symptoms at the end of life included pain, delirium and bleeding.  Palliative Care was not optimally utilized in the majority of cases.  Interventions are needed to improve symptom management and health care utilization at the end of life for patients with AML.