Is intensive care unit admission advisable in older patients with acute myeloid leukemia?

This month, the acute myeloid leukemia (AML) Global Portal (AGP) is focusing on the educational theme of treating unfit and elderly patients. Older patients (≥ 60 years old) with AML have a poor prognosis with a median survival of eight to 10 months. A subset of patients observe a clinical benefit from the use of intensive therapies, however these often require intensive care unit (ICU) admission due to complications such as septic shock.¹

Retrospective studies in younger patients (< 55 years old) with AML have shown survival rates after ICU admission (in-hospital, 90-day and one-year) in the range of 40–70%.^2,3 Data are lacking on the outcomes of older patients after ICU admission, but patient outcomes are assumed to be negative. Since older age and frailty are associated with higher morbidity and mortality, it is necessary to identify the risk factors associated with this increased risk, in order to aid clinician, patient and family decision-making.¹

In line with the monthly theme, this article will summarize the results of a study published by Samuel Slavin, Massachusetts General Hospital, Boston, US, and colleagues, in Cancer, which investigated the outcomes of older patients with AML who had been admitted to an ICU.¹

Study design and patient characteristics

• Retrospective analysis of 330 patients ≥60 years old, diagnosed with AML between 2005 and 2013 at two hospitals in Boston, US
• At the time of diagnosis, patients were receiving intensive induction or non-intensive therapy
  • Intensive induction (7+3 backbone) defined as 7+3 therapy with cytarabine and an anthracycline, or a modified version of this regimen in a trial with other agents added to the 7+3 backbone
  • Non-intensive therapy defined as hypomethylating agents, low-dose cytarabine, or single-agent therapy
• Patient characteristics used to identify risk factors for in-hospital mortality:
  • Demographics
  • Comorbidities (hematopoietic cell transplantation-specific comorbidity index (HCT-CI)
  • AML disease risk: European LeukemiaNet risk stratification
  • Performance status: Eastern Cooperative Oncology Group performance status (ECOG PS)
  • Number of life sustaining interventions
  • Categories for admission to an ICU: medical, surgical, cardiac or neurological

ICU admissions

In total, 96 patients (29%) were admitted to the ICU, with a median age of 67 (60–88) and a median HCT-CI of 2 (0–7). There was no difference in age, sex, race or ethnicity of those admitted to ICU and those not admitted. Table 1 shows the treatment stage that patients were admitted to the ICU and proportion alive at discharge, Table 2 shows the time from diagnosis to ICU admission and Table 3 shows the primary reason for admission and the outcomes of these patients.

Table 1. Treatment stage or response of patients admitted to ICU and proportion alive at hospital discharge (N= 96)

	At admission		At hospital discharge
Treatment stage / response	N= 96	Percentage (%)	Percentage (%)
Initial treatment for AML	52	54.2	54.7
Consolidation therapy	8	8.3	50
Complete remission (CR) and not receiving therapy	16	16.7	43.8
Second- or third-line therapy for relapsed/refractory (R/R) disease	19	19.8	26.3

 Table 2. Time from diagnosis to admission to ICU

Time from diagnosis to admission to ICU	N= 96	Percentage (%)
< 30 days	33	34.4
30–180 days	29	30.2
>180 days	26	27.1

Table 3. Primary reason for admission to ICU and proportion alive at hospital discharge

* At admission this includes; anaphylactic reaction, transfusion reaction, pulmonary embolism, disseminated intravascular coagulation and febrile neutropenia. At hospital discharge, within the other category, six patients with hemorrhagic shock were alive at discharge.
	At admission		At hospital discharge
Reason for admission	N= 96	Percentage (%)	N	Percentage (%)
Respiratory failure	37	39	12	32.4
Septic shock	27	28	11	40.7
Neurological compromise	9	9	2	22.2
Cardiogenic shock, myocardial infarction, arrhythmia	7	7	4	57.1
Other*	9	9	-	-

• Intensive supportive measures were required in many cases including; invasive ventilation (49%), vasopressors (47%) and renal replacement (11%)
• Median time from hospital admission to ICU admission: 4.5 days (0–49)
  • Within 24 hours of hospital admission: 31 (32%)
• Following ICU admission, patients remained alive at:
  • Hospital discharge: 45 (47%, Table 1 and Table 3) with 28 (62%) patients in CR
  • 90 days: 34 (35.4%), 86% patients in CR and/or continued to receive AML-directed therapy, 76% had an ECOG PS of 0 or 1
  • One year: 29 (30.2%)

Factors associated with in-hospital mortality post-ICU admission

• Factors associated with increased odds of in-hospital mortality are shown in Table 4
  • Patients requiring one life-sustaining therapy were more likely to be alive at discharge (61.5%) compared to patients needing ≥ 2 (22.9%)
  • Patients who had an prehospitalization ECOG PS scores < 2 were more likely to survive ICU (51.7%) compared to those with ECOG PS scores ≥ 2 (21.4%)
  • Patients who were admitted to the ICU soon after being diagnosed with AML had lower odds of in-hospital mortality compared to those who had an established diagnosis of AML

Table 4. Multivariate analysis: Factors independently associated with an increased odds-ratio (OR) of in-hospital mortality

	OR	CI (%)	Range	p value
Baseline ECOG score	2.76	95%	1.24–6.12	0.013
Requirement for ≥2 life-sustaining measures	12.39	95%	3.10–49.48	< 0.001

Limitations on the generalizability of the study

• Most of the patients in the study were white, educated patients receiving care in Boston, US
• ICU admission criteria were not standardized
• Retrospective study
• No assessment of frailty or patient-reported quality of life

Conclusion

Older patients with AML have high ICU admission rates, nearly double that of younger patients. Nearly half of patients survive to discharge and a minority survive up to 1 year after admission, however the rates of survival to discharge are similar to all adults with hematologic malignancies.

This study has shown that age alone should not preclude admission of an older patient with AML to the ICU, with patients who survived the ICU stay having minimal decline in performance status.

Medical decision-making for managing critical illness is complex, and a more accurate assessment of prognosis after critical illness is required in this population. Factors identified in this study to be associated with in-hospital mortality, such as ECOG status, are easy to record and practical to use in clinical decision-making.

The authors concluded it is important to provide critical care to this population.