Survival and Predictors of Outcome in Patients WithAcute Leukemia Admitted to the Intensive Care Unit

Snehal G. Thakkar, MD1

Alex Z. Fu, PhD2

John W. Sweetenham, MD1

Zachariah A. Mciver, DO3

Sanjay R. Mohan, MD3

Giridharan Ramsingh, MD3

Anjali S. Advani, MD1

Ronald Sobecks, MD1

Lisa Rybicki, MS2

Matt Kalaycio, MD1

Mikkael A. Sekeres, MD, MS1

1 Department of Hematologic Oncology and BloodDisorders, Cleveland Clinic Taussig Cancer Cen-ter, Cleveland, Ohio.

2 Department of Quantitative Health Sciences,Cleveland Clinic Taussig Cancer Center, Cleve-land, Ohio.

3 Department of Internal Medicine, ClevelandClinic, Cleveland, Ohio.

BACKGROUND. Predictors of outcome and rates of successful discharge have not

been defined for patients with acute leukemia admitted to intensive care units

(ICUs) in the US.

METHODS. This is a retrospective analysis of 90 patients with acute leukemia (no

history of bone marrow transplant) admitted to an ICU from 2001–2004. The pri-

mary endpoints were improvement and subsequent discharge from the ICU, dis-

charge from the hospital, and 2-month survival after hospital discharge.

Secondary endpoints were 6- and 12-month survival. Univariate and multivariate

logistic regression analyses were performed to identify factors predicting out-

come.

RESULTS. The median age of patients was 54 years and 48 (53%) were male. The

most common reason for ICU transfer for all patients was respiratory compro-

mise. The majority of all patients (68%) were eventually placed on ventilator sup-

port and approximately half required pressors. During the ICU course, 29

patients (32%) improved and subsequently resumed aggressive leukemia manage-

ment, and 24 patients (27%) survived to be discharged from the hospital. The 2-,

6-, and 12-month overall survival was 24 (27%), 16 (18%), and 14 (16%), respec-

tively. Higher APACHE II score, use of pressors, undergoing bone marrow trans-

plantation preparative regimen, and adverse cytogenetics predicted worse

outcome. Newly diagnosed leukemia, type of leukemia, or age did not.

CONCLUSIONS. One of 4 patients with acute leukemia survived an ICU admission

to be discharged from the hospital and were alive 2 months later. A diagnosis of

acute leukemia should not disqualify patients from an ICU admission. Cancer

2008;112:2233–40. � 2008 American Cancer Society.

KEYWORDS: acute leukemia, intensive care unit admission, survival.

D espite advances (eg, growth factors, newer antibiotics) and

steady improvements in our understanding of acute leukemias,

the prognosis for many patients with acute leukemia remains poor,

particularly for older adults.1–6 Approximately 16,000 new cases of

acute leukemia will be diagnosed in the US in 2006.7 Estimates of 5-

year disease-free survival range from 25% to 40% in younger

patients and 5% to 15% in older adults. Treatment-related mortality

from remission induction therapy approaches 25% for certain sub-

groups.4,6,8–11 Patients with acute leukemia are also at high risk for

treatment-related morbidities, and these complications may necessi-

tate admission to an intensive care unit (ICU).12

Patients with acute leukemia historically have had poor out-

comes after admission to the ICU.13–16 Patients with cancer as a

whole are discouraged from admission to ICUs based on published

recommendations supporting the poor outcome of patients with

cancer, and on the belief that limited resources should be reserved

Address for reprints: Mikkael A. Sekeres, MD,MS, Hematologic Oncology and Blood Disorders,Cleveland Clinic Taussig Cancer Center, 9500Euclid Ave., Desk R35, Cleveland, OH 44195;Fax: (216) 636-0636; E-mail: sekerem@ccf.org

Received October 9, 2007; revision receivedNovember 19, 2007; accepted November 27,2007.

ª 2008 American Cancer SocietyDOI 10.1002/cncr.23394Published online 17 March 2008 in Wiley InterScience (www.interscience.wiley.com).

2233

for patients with reversible medical conditions.17,18

Although predictors of outcome after an ICU stay for

the general population admitted to the ICU have

been published, these predictors have not been vali-

dated in patients with acute leukemia.19–21

The purpose of this study was to evaluate out-

comes and to determine predictors of outcome for

patients with acute leukemia admitted to the ICU.

MATERIALS AND METHODSThe Institutional Review Board at the Cleveland

Clinic initially approved this study in May 2005 and

an extension was approved in May 2006.

PatientsAll patients with a diagnosis of acute leukemia (acute

myeloid leukemia [AML] or acute lymphoblastic leu-

kemia [ALL] as defined by the World Health Organi-

zation) who had not received a bone marrow

transplantation (BMT) were enrolled if an ICU

admission at the Cleveland Clinic was part of the

hospital course.22 This included patients who were

newly diagnosed having not yet received chemother-

apy; patients undergoing or having just received

remission induction therapy; patients receiving or

having just received postremission therapy; patients

with recurring acute leukemia undergoing salvage

treatment; and patients beginning their preparative

regimens for BMT, but before stem cell infusions.

All patients with an AML diagnosis were treated

with an anthracycline- or anthracenedione-based

remission induction regimen. Postremission therapy

for younger patients (<55 years) consisted of at least

1 cycle of high-dose cytarabine (HiDAC), whereas for

older patients (�55 years) it included at least 1 cycle

of a low-dose cytarabine-based regimen. Patients

with ALL were treated with a vincristine/prednisone-

based remission induction regimen, followed by a

cytarabine-based intensification cycle, central nerv-

ous system prophylaxis that was either methotrexate

or radiation-based, and vincristine/prednisone-based

maintenance therapy. There were several BMT pre-

parative regimens, with the majority including busul-

fan and cyclophosphamide with or without

etoposide. Three of the preparative regimens also

included total body irradiation.

Description of the ICUApproximately 400 patients with acute leukemia are

admitted per year (approximately 130 with a new

acute leukemia diagnosis) to the Cleveland Clinic.

The medical ICU (MICU) is a closed 20-bed unit that

admits approximately 1400 patients per year. In addi-

tion, there are 5 other closed specialized units (surgical,

cardiothoracic, cardiac, heart failure, neurosurgical)

where overflow patients from the medical ICU can

be admitted.

The daytime MICU team consists of a senior

intensivist, 2 pulmonary/critical care fellows, and a

team of senior residents and interns. The nighttime

duties fall to a senior resident and intern who remain

in-house with pulmonary/critical care and leukemia

fellows available by phone. The nursing-staff-to-

patient ratio is 1:2. Patients with a leukemia diagno-

sis admitted to the ICU continue to be followed by

the primary leukemia service.

Data CollectionEvery patient with a diagnosis of acute leukemia

admitted directly to the ICU or to the leukemia ser-

vice between January, 2001 and February, 2005 was

identified. This list was then cross-referenced to the

MICU-tagged patient list available through the

Department of Pulmonary and Critical Care. This

was further corroborated by a documented substan-

tial increase in daily hospital charges, indicating

transfer from the leukemia floor to an intensive care

setting. A chart collection for those consecutive

patients admitted to the ICU was undertaken and

the written documentation in conjunction with the

electronic chart was the basis for the review. No

patients or families of patients were contacted.

Statistical AnalysisThe primary purpose of this study was to identify

prognostic factors for outcome. Successful outcomes

were defined as improvement in the ICU with subse-

quent ICU discharge and resumption of aggressive

leukemia management, 2-month, 6-month, and 1-

year survival measured from hospital admission

(Hospital Admission 5 Day 0). Patients were cen-

sored for death or ICU discharge to a palliative or

hospice setting. Each outcome assessed was meas-

ured against 10 prognostic variables. These variables

included age, white blood cell count (WBC) at hospi-

tal admission, lactate dehydrogenase (LDH) at admis-

sion, acute leukemia type, cytogenetics (AML, as

defined by CALGB; ALL, BCR/ABL translocation),

phase of treatment, hospital day of transfer to the

ICU, reason for transfer to the ICU, the use of press-

ors in the ICU, and APACHE II (Acute Physiology and

Chronic Health Evaluation) score.20,23 The APACHE II

score was calculated based on several factors, with

the total score predicting mortality (Table 1).21

To identify univariate correlates of outcome, a

logistic regression analysis was performed. A desired

outcome was defined as an odds ratio (OR) of greater

2234 CANCER May 15, 2008 / Volume 112 / Number 10

than 1. A similar logistic regression analysis was per-

formed to identify multivariate correlates of out-

come. There was a stepwise selection procedure for

the variables, which required a P < .20 for the vari-

able to enter the model, and then P < .05 for the

variable to retain itself within the model. A final

bootstrap analysis was performed to further support

the conclusions. The same multivariate logistic

regression was used to predict the survival outcomes

of patients with acute leukemia admitted to the ICU.

A decision model was developed for these patients

with pressor use, intubation, and length of ICU stay

information. The Cochran-Armitage trend test was

used to determine whether the percentage of death

in the ICU correlated with the predicted mortality

score (APACHE II, Table 1).

To better understand the impact of ICU care,

additional analyses were conducted comparing this

group of patients with a group of 1:1 age-matched

patients diagnosed with acute leukemia during the

same calendar years as the ICU patients, but who

were not admitted to the ICU. A logistical regression

was performed to compare the 2-, 6-, and 12-month

survival between these 2 groups after controlling for

baseline characteristics.

RESULTSPatient CharacteristicsNinety consecutive patients were identified as having

been admitted to the ICU and available for analysis.

The only leukemia patients excluded from analysis

were those patients who had already received the

stem cell infusion of their BMT regimen. Patient

characteristics are shown in Table 2. The median age

was 54 years and there was an approximate 1:1 ratio

of males to females. Seventy-four (82%) of the

reviewed patients had AML, 16 (18%) ALL. The distri-

bution of cytogenetic abnormalities was typical of

this population mix of younger and older adults: 74

(91%) of the AML patients had intermediate- or

poor-risk karyotypes, whereas 6 (38%) ALL patients

had a BCR-ABL translocation (the Philadelphia chro-

mosome).

Factors Associated With ICU AdmissionAdmission to the ICU was based on a comprehensive

review of the clinical scenario, including the wishes

of the patient. As criteria within the critical care liter-

ature defining set parameters for ICU admission of-

ten include no methodology, are not based on a

comprehensive literature review, or are based on an

author’s personal experience, and thus are not uni-

form, our institution does not follow set ICU admis-

sion criteria.24 Thus, the decision for ICU transfer

was based solely on the initial ICU consult. No

patient, however, was denied admission to the ICU

based on history of acute leukemia.

The most common reason for hospital admission

was fever or infection (Table 2). Twenty-seven (30%)

patients had a new diagnosis of acute leukemia and

were undergoing remission induction therapy.

Patients were transferred to the ICU a median of

8 days (range, 0–81) after hospital admission.

Forty patients (44%) were transferred to the ICU for

multiple reasons. Sixty-one required endotracheal

intubation with mechanical ventilatory support.

Approximately half (48%) required blood pressure

support with ionotropic therapies (pressors). The

median APACHE II score was 22 (range, 9–40) indi-

cating a predicted mortality of 40%.21

Outcome and Predictors of OutcomeTwenty-nine patients (32%) survived their ICU stay

to return to aggressive leukemia management,

whereas 24 patients (27%) survived to hospital dis-

charge. All but 6 patients were discharged to home.

Of the 6 patients not discharged to home, 5 were dis-

charged to a subacute nursing facility and 1 was dis-

charged to a regional community hospital near the

patient’s home. All 24 patients who survived to hos-

pital discharge were alive at 2 months. Six- and 12-

month survivals were 18% and 16%, respectively. Me-

dian survival for the entire population was 197 days

(Fig. 1). Patients with AML had a median survival of

199 days and those with ALL had a median survival

of 192 days. Of 11 patients with acute leukemia with

recurring disease, 7 (64%) died during their ICU stay.

Univariate analysis is shown in Table 3. The use

of pressors and a higher APACHE II score predicted

for shorter 2-, 6-, and 12-month survival. Only the

TABLE 1APACHE II Scoring System14

Apache II score* % Nonoperative mortality

0–4 4

5–9 6

10–14 12

15–19 22

20–24 40

25–29 51

30–34 71

�35 82

* The Acute Physiology Apache II Score is based on 14 variables (temperature, mean arterial pres-

sure, heart rate, respiratory rate, A-aPO2, PAO2, arterial pH, serum bicarbonate, serum sodium, serum

potassium, serum creatinine, hematocrit, white blood-cell count, Glasgow coma score). There is also

an Age Adjustment, Chronic Health Adjustment score.

Outcomes of Leukemia Patients in the ICU/Thakkar et al. 2235

TABLE 2Patient Characteristics

Characteristics No. %

Age, y, mean � SD/median (range) 52.6 � 12.3/54 (17–90)

Sex, men/women 48/42 53.3/46.7

Race

White 75 83.3

African American/other 13/2 14.4/2.2

Performance status, 0–1 to �2 67/23 74/26

WBC at admission

Mean � SD 36.5 � 67.3

Median (range) 7.8 (0.01–317.9)

LDH at admission, n 5 76

Mean � SD 618 � 749

Median (range) 305 (13–4624)

Growth factors used 19 21.1

Past medical history

DM 18 20.0

Cardiac, CAD/CHF/HTN 48 53.0

Other, RF/CVA/etc 47 52.2

None 38 42.2

Reason for hospital admissions

Fever or infection 32 35.6

New diagnosis 27 30.0

Relapse/BMTX prep 19 21.1

Other, incl WBC/SOB/etc 22 24.4

Diagnosis AML/ALL 74/16 82.2/17.8

If AML subtype n 5 74; M3/Non-M3 subtype 7/67 9.5/90.5

If AML, n 5 74

Cause unknown 50 67.6

MDS 20 27.0

Therapy 4 5.4

If ALL subtype n 5 16; Pre–B-cell/T-cell 13/3 81.3/18.7

If ALL

Philadelphia chromosome, n 5 16; Ph1/Ph2 6/6 37.5/37.5

Unknown 4 25.0

Cytogenetics

Good/Intermediate/Poor 7/36/38 8.6/44.4/46.9

Unknown 9 10.0

Type of treatment

Induction 27 30.0

BMT 19 21.1

Other 44 48.9

Days of transfer to ICU, mean � SD/median (range) 13.4 � 15.6/8 (0–81)

ICU admit chemotherapy, d. n 5 55, mean � SD/median (range) 15.8 � 13.2/14 (0–51)

Reason for transfer to ICU

Respiratory 57 63.3

Cardiovascular 12 13.3

Sepsis 51 56.7

Other 20 22.2

No. of reasons for transfer to ICU

0–1 50 55.5

2 37 41.1

3–4 3 3.4

Extubation (61 pts intubated) 25 41.0

ICU pressors 44 48.9

Length of ICU stay, d, n 5 89, mean � SD/median (range) 6.3 � 8.3/4 (1–60)

Outcome

Improved 29 32.2

Died in ICU/Died after transfer 51/10 56.7/11.1

(continued)

2236 CANCER May 15, 2008 / Volume 112 / Number 10

APACHE II score predicted ICU improvement. Age

and type of leukemia were not predictive of out-

come.

In multivariate analyses (Table 4), both type of

therapy (BMT preparative regimen) and increasing

APACHE II score predicted worse outcomes in terms

of ICU improvement and 2-month survival. With

regard to 6-month survival, cytogenetics and the use

of pressors predicted for outcome. One-year survival

was only influenced by increasing APACHE II score.

Applying the same multivariate regression models, a

decision model was developed to assist in predicting

the survival outcomes of patients with acute leuke-

mia admitted to the ICU with different pressors use,

intubation use, and length of ICU stay information

(Fig. 2).

Additional analysis identified 243 patients with

acute leukemia at the Cleveland Clinic who were not

admitted to the ICU. These patients were diagnosed

within the same calendar years as the 90 patients

who were admitted to the ICU. For those who were

not admitted to the ICU, the 2-, 6-, and 12-month sur-

vivals were 84%, 63%, and 40%, respectively. After 1:1

exact matching by age, 85 patient pairs were achieved

with a median age 54 years (range, 19–79). The

2-, 6-, and 12-month survival for the 85 ICU patients

remained the same (27%, 18%, and 16%, respec-

tively). The 2-, 6-, and 12-month survival for the 85

non-ICU matched patients were 85%, 69%, and 46%,

respectively. The logistic regressions showed that the

ICU patients had significantly worse 2-, 6-, and 12-

month survival than their non-ICU counterparts,

with ORs of 0.04 (P < .0001), 0.11 (P < .0001), and

0.36 (P 5 .014), respectively.

DISCUSSIONPatients with acute leukemia are thought to have

poor outcomes after admission to an ICU. Whereas

this bias may have been valid in the past, advances

in intensivist management and in supportive care for

leukemia patients have made denying ICU admission

to a patient purely on the basis of a leukemia diag-

nosis unreasonable.

This study demonstrates that 32% of patients

with acute leukemia will improve during an ICU

admission and be discharged from the ICU. One in 4

patients will survive to be discharged from the hospi-

tal and are alive 2 months after an ICU stay. In com-

TABLE 2(continued)

Characteristics No. %

ICU readmission during same admission 6 6.7

Alive at 2 mo 24 26.7

Alive at 6 mo 16 17.8

Alive at 12 mo 13 15.6

Apache score, mean � SD/median (range) 22.2 � 7.0/22 (9–40)

Mortality score

6 3 3.3

12 12 13.3

22 21 23.3

40 22 24.4

51 18 20.0

71 10 11.1

82 4 4.4

WBC indicates white blood-cell count; SD, standard deviation; LDH, lactate dehydrogenase; DM, diabetes mellitus; CAD, coronary artery disease; CHF, conges-

tive heart failure; HTN, hypertension; RF, renal failure; CVA, cerebrovascular accident; BMTX prep, bone marrow transplant preparative regimen; WBC, white

blood cell count elevation; SOB, shortness of breath; AML, acute myeloid leukemia; ALL, acute lymphoblastic leukemia; MDS, myelodysplastic syndrome; ICU,

intensive care unit.

FIGURE 1. Overall survival of patients who were admitted to the intensivecare unit (ICU) with a diagnosis of acute leukemia (N 5 90).

Outcomes of Leukemia Patients in the ICU/Thakkar et al. 2237

parison, the overall in-hospital mortality for all

patients admitted to the MICU during this timeframe

(4800 patients) was 30%. Whereas it appears that

patients with acute leukemia admitted to the ICU

have a worse overall survival compared with age-

matched controls with acute leukemia treated during

the same time period, their survival is not that differ-

ent from patients with other diseases admitted to the

same ICU. The use of limited ICU resources in this

potentially curable population was thus warranted.

The utility of aggressive ICU management of

patients with cancer has been debated. In particular,

patients with a diagnosis of leukemia have been

identified as a subgroup of patients with cancer with

a particularly poor outcome when intensive manage-

ment is required.25,26 In a study evaluating ICU

admission criteria, Thiery et al.27 reported the out-

comes of patients with cancer considered for the

ICU. Patients had a range of malignancies, including

61% with hematologic malignancies. Whereas 26.2%

of patients were denied ICU admission for being

considered ‘too sick to benefit,’ 26% of those patients

survived 30 days, and 21% of patients considered

TABLE 3Univariate Correlates of Outcomes by Logistic Regression

Improved in ICU Alive at 2 months Alive at 6 months Alive 12 months

POR 95% CI OR 95% CI OR 95% CI OR 95% CI

Age by 10 y : 0.94 0.71–1.26 0.85 0.62–1.16 0.90 0.63–1.29 0.97 0.66–1.43 .88

WBC admission

By 10 unit : 1.01 0.95–1.08 1.01 0.94–1.08 1.02 0.95–1.10 0.97 0.87–1.08 .59

301 to <30 1.27 0.48–3.35 1.54 0.55–4.26 1.93 0.61–6.14 1.16 0.32–4.19 .82

501 to <50 1.73 0.61–4.92 1.97 0.66–5.83 2.14 0.63–7.27 1.13 0.28–4.58 .87

1001 to <100 1.79 0.44–7.24 1.50 0.34–6.56 1.47 0.28–7.90 0.71 0.08–6.19 .76

LDH at admission

By 50 unit : 1.03 0.99–1.06 1.00 0.97–1.03 1.00 0.96–1.04 0.96 0.89–1.03 .26

Diagnosis

ALL/AML 0.65 0.19–2.24 0.61 0.16–2.38 0.66 0.13–3.26 0.81 0.16–4.05 .79

Cytogenetics

Good 1 Int to Poor 2.02 0.79–5.17 2.08 0.76–5.68 3.02 0.87–10.46 2.32 0.65–8.27 .19

Type of treatment

Induction/other 1.03 0.38–2.78 1.29 0.46–3.66 0.86 0.25–2.90 0.47 0.12–1.93 .30

BMT/other 0.33 0.08–1.30 0.30 0.06–1.52 0.21 0.03–1.79 0.21 0.03–1.79 .15

Ind/BMT 1 other 1.36 0.53–3.52 1.72 0.63–4.65 1.18 0.36–3.86 0.65 0.16–2.58 .54

Days of transfer to ICU

By 1 d : 0.98 0.95–1.01 0.99 0.96–1.02 0.97 0.93–1.02 0.98 0.94–1.03 .43

No.of reasons for transfer to ICU (of 5)

By 1 reason : 0.58 0.28–1.22 0.80 0.38–1.69 0.55 0.22–1.40 0.81 0.32–2.04 .65

ICU pressors

Yes/No 0.52 0.21–1.28 0.34* 0.13–0.94 0.12* 0.03–0.56 0.15 0.03–0.71 .02*

Apache score

By 5 point : 0.85* 0.78–0.93 0.83* 0.75–0.92 0.89* 0.81–0.98 0.88 0.80–0.98 .02*

ICU indicates intensive care unit; OR, odds ratio; CI, confidence interval; WBC, white blood-cell count; LDH, lactate dehydrogenase; ALL/AML, acute lymphoblastic leukemia/acute myeloid leukemia; BMT,

bone marrow preparative regimen; Int, intermediate.

* Statistical significance (P < .05).

TABLE 4Multivariate Analysis

OR 95% CI P

Model for improvement in ICU

Type of treatment

BMT/Induction 1 other 0.16 0.04–0.70 .01

Apache score

By 5 point : 0.83 0.75–0.91 <.0001

Model for living 2 mo

Type of treatment

BMT/Induction 1 other 0.11 0.02–0.65 .01

Apache score

By 5 point : 0.80 0.71–0.90 <.0001

Model for living 6 mo

Cytogenetics

Good 1 Int/Poor 5.39 1.37–21.2 .01

ICU pressors

Yes/No 0.10 0.02–0.54 .01

Model for living 12 mo

Apache score

By 5 point : 0.88 0.79–0.99 .02

OR indicates odds ratio; CI, confidence interval; ICU, intensive care unit; BMT, bone marrow pre-

parative regimen; Int, intermediate.

2238 CANCER May 15, 2008 / Volume 112 / Number 10

‘too well to benefit’ from an ICU admission died

within 30 days. It was concluded that enough inaccu-

racy existed in predicting ICU outcome that broader

admission criteria for patients with cancer being

considered for ICU admission should be considered.

Our study identified significant predictors of out-

come, including the treatment type and APACHE II

score (for earlier timepoints), and cytogenetics and

pressor use for 6- and 12-month survival. Those

patients who were being treated with a BMT prepara-

tive regimen had worse outcomes if they were

admitted to the ICU. This probably reflects the

increased intensity of the preparative regimen as

compared with typical induction therapy, and that

patients with acute leukemia who undergo a BMT

have already been exposed to multiple courses of

chemotherapy, and are more likely to be immuno-

suppressed and experience morbidities as a result.

This is the first study to validate the APACHE II

scoring system as a predictor of outcome in patients

with acute leukemia. This extends the finding that

the SAPS II has utility in this population.28 The use

of pressors in the ICU has already been correlated

with worse outcome in other reviews.29,30 Interest-

ingly, cytogenetic risk groups retained similar predic-

tive abilities in patients with acute leukemia

requiring an ICU stay, as they do in the acute leuke-

mia population as a whole.1,31 In these studies the 1-

year survival for those with poor-risk cytogenetics

was 40%, to almost 70% in those with good-risk cyto-

genetics. A similar dichotomy can be seen with our

ICU patients at 1 year between those with good-risk

cytogenetics (43% 1-year survival) and those with

poor risk (11% 1-year survival).

Other studies have also identified predictors of

outcome in patients with cancer admitted to the

ICU, including mechanical ventilation and older

age.32–35 Age was not a predictive factor in our study

and should not be used as an exclusion criterion for

ICU admission. That invasive mechanical ventilation

was not predictive reflects that most of our patients

required this modality, and thus it could not be ana-

lyzed separately.

On the basis of our results from the multivariate

analysis (Table 4) a decision model was developed to

assist in predicting the survival outcomes of patients

with acute leukemia admitted to the ICU (Fig. 2).

The differences between outcomes are useful both

clinically and statistically because of significant pre-

dictors form the multivariate analysis in Table 4. We

recognize that these survival estimates reflect our

single institution experience as well as only 90

patients, but it can serve as a template upon which

clinicians, patients, and families can discuss code

status as well as end-of-life decisions.

In conclusion, our review supports aggressive

ICU management of patients with acute leukemia.

Cytogenetics, APACHE II score, and the use of press-

ors predicted outcome. Mechanical ventilation,

increasing age, and WBC did not. It is reasonable to

expect almost 1 in 3 patients to improve after an

ICU admission to continue aggressive leukemia man-

agement and 1 in 4 to be alive 2 months after hospi-

tal discharge.

REFERENCES1. Byrd JC, Mrozek A, Dodge RK, et al. Pretreatment cytoge-

netic abnormalities are predictive of induction success, cu-

mulative incidence of relapse, and overall survival in adult

patients with de novo acute myeloid leukemia: results

from Cancer and Leukemia Group B (CALGB 8461). Blood.

2002;100:4325–4336.

FIGURE 2. Intensive care unit (ICU) outcomes decision model with the 2 months, 6 months, and 12 months chance of survival.

Outcomes of Leukemia Patients in the ICU/Thakkar et al. 2239

2. Baudard M, Beauchamp-Nicoud A, Delmer A, et al. Has

the prognosis of adult patients with acute myeloid leuke-

mia improved over years? A single institution experience of

784 consecutive patients over a 16-year period. Leukemia.

1999;13:1481–1490.

3. Pui CH, Evans WE. Treatment of acute lymphoblastic leu-

kemia. N Engl J Med. 2006;354:166–178.

4. Stone RM, Berg DT, George SL, et al. Granulocyte-macro-

phage colony-stimulating factor after initial chemotherapy

for elderly patients with primary acute myelogenous leuke-

mia. Cancer and Leukemia Group B. N Engl J Med.

1995;332:1671–167.

5. Goldstone AH, Burnett AK, Wheatley K, et al. Attempts to

improve treatment outcomes in acute myeloid leukemia

(AML) in older patients: the results of the United Kingdom

Medical Research Council AML11 trial. Blood. 2001;98:

1302–1311.

6. Larson RA, Dodge RK, Linker CA, et al. A randomized con-

trolled trial of filgrastim during remission induction and

consolidation chemotherapy for adults with acute lympho-

blastic leukemia: CALGB study 9111. Blood. 1998;92:1556–

1564.

7. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2006. CA

Cancer J Clin. 2006;56:106–130.

8. Sekeres MA, Stone RM, Zahrieh D, et al. Decision-making

and quality of life in older adults with acute myeloid leuke-

mia or advanced myelodysplastic syndrome. Leukemia.

2004;18:809–816.

9. Rowe JM, Neuberg D, Friedenberg W, et al. A phase 3 study

of 3 induction regimens and of priming with GM-CSF in

older adults with acute myeloid leukemia: a trial by the

Eastern Cooperative Oncology Group. Blood. 2004;103:479–

485.

10. Mayer RJ. Two steps forward in the treatment of colorectal

cancer. N Engl J Med. 2004;350:2406–248.

11. Hoelzer D, Thiel E, Loffler H, et al. Prognostic factors in a

multicenter study for treatment of acute lymphoblastic

leukemia in adults. Blood. 1988;71:123–131.

12. Darmon M, Thiery G, Ciroldi M, et al. Intensive care in

patients with newly diagnosed malignancies and a need

for cancer chemotherapy. Crit Care Med. 2005;33:2488–

2493.

13. Brunet F, Lanore JJ, Dhainaut JF, et al. Is intensive care jus-

tified for patients with haematological malignancies? Inten-

sive Care Med. 1990;16:291–297.

14. Carlon GC. Admitting patients with cancer to the intensive

care unit. Crit Care Clin. 1988;4:183–191.

15. Schuster D, Marion JM. Precedents for meaningful recovery

during treatment in a medical intensive care unit. Out-

come in patients with hematologic malignancy. Am J Med.

1983;75:402–408.

16. Rubenfeld GD, Crawford SW. Withdrawing life support

from mechanically ventilated recipients of bone marrow

transplants: a case for evidence-based guidelines. Ann

Intern Med. 1996;125:625–633.

17. Guidelines for intensive care unit admission, discharge,

and triage. Task Force of the American College of Critical

Care Medicine, Society of Critical Care Medicine. Crit Care

Med. 1999;27:633–638.

18. Consensus statement on the triage of critically ill patients.

Society of Critical Care Medicine Ethics Committee. JAMA.

1994;271:1200–1203.

19. Lemeshow S, Teres D, Klar J, et al. Mortality probability

models (MPM II) based on an international cohort of in-

tensive care unit patients. JAMA. 1993;270:2478–2486.

20. Knaus WA, Wagner DP, Draper EA, et al. The APACHE III

prognostic system. Risk prediction of hospital mortality

for critically ill hospitalized adults. Chest. 1991;100:1619–1636.

21. Knaus WA, Draper EA, Wagner DP, Zimmerman JE.

APACHE II: a severity of disease classification system. Crit

Care Med. 1985;13:818–829.

22. Jaffe ES, Harris NL, Diebold J, Muller-Hermelink HK. World

Health Organization classification of neoplastic diseases of

the hematopoietic and lymphoid tissues. A progress report.

Am J Clin Pathol. 1999;111(1 Suppl 1):S8–12.

23. Siro CA, Bastos PG, Knaus WA, Wagner DP. APACHE II

scores in the prediction of multiple organ failure syn-

drome. Arch Surg. 1991;126:528–529.

24. Giacomini M, Cook D, DeJean D, Shaw R, Gedge E. Deci-

sion tools for life support: a review and policy analysis.

Crit Care Med. 2006;34:864–870.

25. Tremblay LN, Hyland RH, Schouten BD, Hanley PJ. Sur-

vival of acute myelogenous leukemia patients requiring

intubation/ventilatory support. Clin Invest Med. 1995;18:

19–24.

26. Epner DE, White P, Krasnoff M, et al. Outcome of mechani-

cal ventilation for adults with hematologic malignancy.

J Investig Med. 1996;44:254–260.

27. Thiery G, Azoulay E, Darmon M, et al. Outcome of patients

with cancer considered for intensive care unit admission: a

hospital-wide prospective study. J Clin Oncol. 2005;23:

4406–4413.

28. Rabbat A, Chaoui D, Montani D, et al. Prognosis of patients

with acute myeloid leukaemia admitted to intensive care.

Br J Haematol. 2005;129:350–357.

29. Aggarwal A, Ong JP, Younossi ZM, et al. Predictors of mor-

tality and resource utilization in cirrhotic patients admitted

to the medical ICU. Chest. 2001;119:1489–1497.

30. Gildea TR, Cook WC, Nelson DR, et al. Predictors of long-

term mortality in patients with cirrhosis of the liver

admitted to a medical ICU. Chest. 2004;126:1598–1603.

31. Wetzler M, Dodge RK, Mrozek K, et al. Prospective karyo-

type analysis in adult acute lymphoblastic leukemia: the

cancer and leukemia Group B experience. Blood. 1999;93:

3983–3993.

32. Bellamy E, Oye RK. Admitting elderly patients to the ICU:

dilemmas and solutions. Geriatrics. 1987;42:61–63, 67–68.

33. Rabe C, Mey U, Paashaus M, et al. Outcome of patients

with acute myeloid leukemia and pulmonary infiltrates

requiring invasive mechanical ventilation-a retrospective

analysis. J Crit Care. 2004;19:29–35.

34. Azoulay E, Alberti C, Bornstain C, et al. Improved survival

in patients with cancer requiring mechanical ventilatory

support: impact of noninvasive mechanical ventilatory

support. Crit Care Med. 2001;29:519–525.

35. Schapira DV, Studnicki J, Bradham DD, Wolff P, Jarrett A.

Intensive care, survival, and expense of treating critically ill

patients with cancer. JAMA. 1993;269:783–786.

2240 CANCER May 15, 2008 / Volume 112 / Number 10