Accepted Manuscript Title: What is the outcome of cancer patients admitted to the ICU after cardiac arrest? Results from a multicenter study Authors: B. Champigneulle MD S. Merceron MD V. Lemiale MD G. Geri MD D. Mokart MD F. Bruneel MD F. Vincent MD P. Perez MD J. Mayaux MD A. Cariou MD, PhD E. Azoulay MD, PhD PII: DOI: Reference:
S0300-9572(15)00164-1 http://dx.doi.org/doi:10.1016/j.resuscitation.2015.04.011 RESUS 6376
To appear in:
Resuscitation
Received date: Revised date: Accepted date:
16-2-2015 6-4-2015 11-4-2015
Please cite this article as: Champigneulle B, Merceron S, Lemiale V, Geri G, Mokart D, Bruneel F, Vincent F, Perez P, Mayaux J, Cariou A, Azoulay E, What is the outcome of cancer patients admitted to the ICU after cardiac arrest? Results from a multicenter study., Resuscitation (2015), http://dx.doi.org/10.1016/j.resuscitation.2015.04.011 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
What is the outcome of cancer patients admitted to the ICU after cardiac arrest?
2
Results from a multicenter study.
3 B. Champigneulle, MD1,8* ; S. Merceron, MD2 ; V. Lemiale, MD3 ; G. Geri, MD1,8,9 ; D.
5
Mokart, MD4 ; F. Bruneel, MD2 ; F. Vincent, MD5 ; P. Perez, MD6 ; J. Mayaux, MD7 ; A.
6
Cariou, MD-PhD1,8,9 ; E. Azoulay, MD-PhD3,10
cr
ip t
4
7
9
1
us
8
Medical Intensive Care Unit (ICU), Cochin University Hospital, Assistance Publique - Hôpitaux de Paris,
Paris, France
11
2
ICU, Mignot Hospital, Versailles, France
12
3
Medical ICU, Saint Louis University Hospital, Assistance Publique - Hôpitaux de Paris, Paris, France
13
4
Medical ICU, Paoli Calmette Institute, Marseille, France
14
5
Medical ICU, Avicenne University Hospital, Assistance Publique - Hôpitaux de Paris, Bobigny, France
15
6
Medical ICU, Nancy University Hospital, Nancy, France
16
7
Medical ICU, La Pitié-Salpêtrière University Hospital, Assistance Publique - Hôpitaux de Paris, Paris, France
17
8
Paris Descartes University, Sorbonne Paris Cité, Paris, France
18
9
INSERM U970 Sudden Death expertise Center, Paris Cardiovascular Research Center, Paris, France
19
10
M
d
te
Ac ce p
20
an
10
Paris 7 University, Paris, France
21
*Corresponding author: Benoit Champigneulle, Medical intensive care unit, Cochin
22
Hospital, Assistance Publique - Hôpitaux de Paris (AP-HP), 27 rue du Faubourg Saint-
23
Jacques, 75014 Paris, FRANCE
24
Phone: +33 158 412 501; Fax: +33 158 412 505
25
E-mail:
[email protected] 26
1
Page 1 of 24
26 Word count: 2856.
28
References: 30.
29
Tables: 3.
30
Figures: 2.
31
Abstract word count: 255.
32
Electronic supplementary data: Tables, 4; Figures, 2.
us
33
cr
ip t
27
Author contributions
35
BC, SM, VL, and GG contributed to conceive and design the study; collect, analyze, and
36
interpret the data; and write the manuscript.
37
AC and EA contributed to conceive and design the study and to write the manuscript.
38
DM, FB, FV, PP, and JM contributed to collect the data.
39
All authors read and approved the final version of the manuscript.
te
d
M
an
34
40
This original manuscript has not been published in all or in part and is not under
42
consideration for publication by another journal.
43
Ac ce p
41
44
Author disclosure: None of the authors has a financial relationship with a commercial entity
45
that
has
an
interest
in
the
subject
of
this
manuscript.
2
Page 2 of 24
46 Abstract
48
Aim: Low survival rate was previously described after cardiac arrest in cancer patients and
49
may challenge the appropriateness of intensive care unit (ICU) admission after return of
50
spontaneous circulation (ROSC). Objectives of this study were to report outcome and
51
characteristics of cancer patients admitted to the ICU after cardiac arrest.
52
Methods: A retrospective chart review in seven medical ICUs in France, in 2002-2012. We
53
studied consecutive patients with malignancies admitted after out-of-hospital cardiac arrest
54
(OHCA) or in-hospital cardiac arrest (IHCA).
55
Results: Of 133 included patients of whom 61% had solid tumors, 48 (36%) experienced
56
OHCA and 85 (64%) IHCA. Cardiac arrest was related to the malignancy or its treatment in
57
47% of patients. Therapeutic hypothermia was used in 51 (41%) patients. The ICU mortality
58
rate was 98/133 (74%). Main causes of ICU death were refractory shock or multiple organ
59
failure (n=64, 48%) and neurological injury (n=27, 20%); 42 (32%) patients died in ICU
60
after treatment-limitation decisions. Twenty-four (18%) patients were discharged alive from
61
the hospital. Overall 6-month survival rate was 14% (18/133, 95% confidence interval, 8%-
62
21%). Survival rates at ICU discharge and after 6 months did not differ significantly across
63
type of malignancy or between the OHCA and IHCA groups, and neither were they
64
significantly different from those in matched controls who had cardiac arrest but no
65
malignancy.
66
Conclusions: Even if low, the 6-month survival rate of 14% observed in cancer patients
67
admitted to the ICU after cardiac arrest and ROSC may support the admission of these
68
patients to the ICU and may warrant an initial full-code ICU management.
Ac ce p
te
d
M
an
us
cr
ip t
47
69 70
Keywords:
Cardiac
arrest;
Malignancy;
Prognosis;
Cancer.
3
Page 3 of 24
71 Conflicts of interests
73
None of the authors has a financial relationship with a commercial entity that has an interest
74
in the subject of this manuscript.
ip t
72
75 76
INTRODUCTION
us
78
cr
77
79
Survival after cardiac arrest remains low despite major advances in management
81
during resuscitation and after the return of spontaneous circulation (ROSC) [1–3]. Overall
82
hospital survival has been estimated at less than 10% after out-of-hospital cardiac arrest
83
(OHCA) and about 15% after in-hospital cardiac arrest (IHCA) [4–7]. Few data are available
84
on survival rates in vulnerable populations such as elderly individuals or patients with
85
malignancies [8–10]. Cancer patients were formerly considered as poor candidates to ICU
86
admission. However, substantial improvements in the survival of patients with malignancies
87
admitted to the ICU have been achieved in recent years [11–13]. These improvements may
88
be ascribable to progress in cancer treatments, better intensive care strategies, and closer
89
collaboration between oncologists and hematologists [14].
M
d
te
Ac ce p
90
an
80
Previous studies have assessed the management and outcomes of patients with
91
malignancies admitted to the ICU after cardiac arrest [10,15–19]. Survival rates were usually
92
low, particularly in the subgroups with metastatic solid tumors or hematological
93
malignancies [10,15–17]. These low rates may challenge the appropriateness of
94
cardiopulmonary resuscitation (CPR), the triage decision to ICU admission and other life-
95
supporting interventions after ROSC in cancer patients. However, most of these studies were
4
Page 4 of 24
96
done years ago and therefore fail to reflect recent advances in the ICU management of
97
cardiac arrest and patients with cancer. The primary aim of this study was to evaluate outcomes of patients with malignancies
99
admitted to the ICU after cardiac arrest followed by ROSC. Our secondary objectives were
100
to describe the epidemiology and ICU management of these patients. To meet these
101
objectives, we retrospectively reviewed the medical records of patients admitted to seven
102
French
ICUs.
Ac ce p
te
d
M
an
us
cr
ip t
98
5
Page 5 of 24
103 104
METHODS
105 Study design and patients
ip t
106
We retrospectively reviewed the medical records of patients admitted between 2002
108
and 2012 to seven medical ICUs involved in research on patients with malignancies (Groupe
109
de Recherche Respiratoire en Réanimation Onco-Hématologique, GRRROH) and having a
110
policy of broad ICU admission for patients with malignancies. We identified consecutive
111
patients aged 18 years or older who were admitted after OHCA or IHCA followed by ROSC
112
and who had solid or hematological malignancies. In the seven involved hospitals, cardiac
113
arrest patients were exclusively managed in medical ICU and patients with ROSC following
114
IHCA on the wards were systematically directly referred to the ICU. Exclusion criteria were
115
cardiac arrest in the ICU or during surgery and malignancy in remission for at least 5 years.
116
According to French law on retrospective and observational studies, neither ethics
117
committee approval nor informed consent was required for this study. Data used for the
118
study were collected in an anonymized file.
120 121
us
an
M
d
te
Ac ce p
119
cr
107
Patient management
International guidelines [2] were followed, particularly regarding the target
122
temperature during therapeutic hypothermia. Diagnostic and therapeutic procedures (e.g.,
123
coronary angiogram, percutaneous coronary intervention, thrombolysis, brain or chest
124
computed tomography) were performed at the discretion of the physicians in charge of each
125
patient. Renal replacement therapy (RRT) was used either continuously or discontinuously
126
depending on standard practice at each center. A thorough clinical neurological evaluation
127
was performed, followed by an electroencephalogram and/or somatosensory evoked
6
Page 6 of 24
potential recording if deemed appropriate. Treatment-limitation decisions were considered in
129
patients with a poor neurological prognosis or with refractory shock and multiple organ
130
failures with or without recurrent cardiac arrest (post-cardiac arrest syndrome). Criteria for
131
making treatment-limitation decisions were the same in all seven study ICUs and remained
132
unchanged throughout the study period (Figure S1 in the electronic supplement).
ip t
128
134
cr
133 Data acquisition
The medical charts were reviewed for information on the underlying malignancy,
136
characteristics of the cardiac arrest according to the Utstein style [20], and treatments used in
137
the ICU. The Knaus score was determined, as well as the Charlson comorbidity index with
138
and without the points related to the malignancy [21,22]. Poor baseline general health was
139
defined as a Knaus score C or D. Leukopenia was defined as a leukocyte count lower than
140
1000/mm3 [12]. Baseline status of the malignancy was categorized as uncontrolled disease
141
(newly diagnosed, recurrent, or progressive disease) or controlled disease (controlled or in
142
remission for less than 5 years) [23,24]. The possible contribution of the malignancy or its
143
treatment to the cardiac arrest was assessed based on the cause of cardiac arrest, after
144
reviewing case by three of the authors (BC, SM and VL). Vital status was determined by
145
querying the patient’s usual physician or public records office at ICU discharge and 6
146
months after ICU admission. The Cerebral Performance Categories (CPC) scale [25] was
147
used to assess neurological outcomes at ICU discharge; scores of 1 or 2 were classified as
148
good outcomes and scores of 3 to 5 as poor outcomes. Causes of death related to cardiac
149
arrest in the ICU were defined as previously described [26].
Ac ce p
te
d
M
an
us
135
150 151
Comparison with cardiac arrest patients free of known malignancies
7
Page 7 of 24
To further appreciate the survival rate of cardiac arrest patients with malignancies, a
153
comparison was performed with a control group of OHCA patients without known
154
malignancies issued from the French Parisian database, which was previously described
155
[26]. A crude comparison of both ICU and 6-months survival rate was performed between
156
unpaired patients with and without malignancies. This comparison was repeated by using
157
matched controls randomly selected from the database with a 1/1 ratio. For each patient with
158
malignancy (cases), a control was selected at random among individuals who met the
159
matching criteria, namely, age, gender, whether cardiac arrest occurred in a public place,
160
whether cardiac arrest was witnessed, whether CPR was delivered by a bystander, initial
161
cardiac rhythm, time from collapse to ROSC, and cause of cardiac arrest (probably cardiac
162
or other).
M
an
us
cr
ip t
152
163 Statistical analysis
d
164
Continuous variables were described as median [25th-75th percentiles] and compared
166
using the Mann-Whitney U test. Categorical variables were described as n (%) and
167
compared using the chi-square test or Fisher’s exact test as appropriate. Percentages were
168
calculated using the number of patients without missing data as the denominator. Variables
169
with more than 15% of missing data were identified. In subgroups of interest (solid versus
170
hematological malignancies and OHCA versus IHCA), 6-month survival was estimated by
171
plotting univariate Kaplan-Meier curves and compared using the log-rank test. P values
172
lower than 0.05 were considered significant. All statistical analyses were performed using
173
GraphPad Prism® v5.0 (GraphPad Software, San Diego, CA, USA) or STATA v11.0
174
(Lakeway Drive, TX, USA).
Ac ce p
te
165
175
8
Page 8 of 24
175 176
RESULTS
177 178
Study patients During the study period (2002-2012), 213 patients with malignancies were admitted to
180
the seven participating ICUs after OHCA or IHCA followed by ROSC. Among them, 133
181
were included in the study (Figure 1).
cr
ip t
179
Table 1 reports the main patient characteristics. Sites of solid malignancies were the
183
lung (n=27, 20%), prostate (n=12, 9%), urinary tract or kidney (n=8, 6%), head and neck
184
(n=8, 6%), breast (n=7, 5%), colon and rectum (n=6, 5%), and other (n=13, 10%). Spread of
185
solid malignancies was local and regional in 50 (38%) patients, metastatic in 27 patients
186
(20%), and unknown in 4 patients (3%). Hematological malignancies consisted of acute
187
leukemia (n=13, 10%), high-grade lymphoma (n=13, 10%), low-grade lymphoma (n=12,
188
9%), multiple myeloma (n=7, 5%), and other (n=7, 5%). Five (4%) patients with
189
hematological malignancies had received allogeneic stem cell transplantation. No significant
190
differences were apparent between the solid and hematological malignancy groups regarding
191
age, sex ratio, previous chronic health status, Charlson comorbidity index without
192
malignancy points, or disease status at admission (data not shown). The group with
193
hematological malignancies had a higher proportion of patients with chemotherapy in the
194
past month than did the group with solid malignancies (54% vs. 31%, P=0.01).
196
an
M
d
te
Ac ce p
195
us
182
Cardiac arrest characteristics (Table 2)
197
OHCA was more common in the solid malignancy group than in the hematological
198
malignancy group (47% vs. 19%, respectively; P