- Email: [email protected]
Circulatory arrest versus cerebral perfusion during pulmonary endarterectomy surgery (PEACOG): a randomised controlled trial
Articles
Circulatory arrest versus cerebral perfusion during pulmonary endarterectomy surgery (PEACOG): a randomised controlled trial Alain Vuylsteke*, Linda Sharples, Gill Charman, John Kneeshaw, Steven Tsui, John Dunning, Ella Wheaton, Andrew Klein, Joseph Arrowsmith, Roger Hall, David Jenkins*
Summary Background For some surgical procedures to be done, a patient’s blood circulation needs to be stopped. In such situations, the maintenance of blood flow to the brain is perceived beneficial even in the presence of deep hypothermia. We aimed to assess the benefits of the maintenance of antegrade cerebral perfusion (ACP) compared with deep hypothermic circulatory arrest (DHCA).
Lancet 2011; 378: 1379–87
Methods Patients aged 18–80 years undergoing pulmonary endarterectomy surgery in a UK centre (Papworth Hospital, Cambridge) were randomly assigned with a computer generated sequence to receive either DHCA for periods of up to 20 min at 20°C or ACP (1:1 ratio). The primary endpoint was change in cognitive function at 12 weeks after surgery, as assessed by the trail-making A and B tests, the Rey auditory verbal learning test, and the grooved pegboard test. Patients and assessors were masked to treatment allocation. Primary analysis was by intention to treat. The trial is registered with Current Controlled Trials, number ISRCTN84972261.
See Comment page 1359
Findings We enrolled 74 of 196 screened patients (35 to receive DHCA and 39 to receive ACP). Nine patients crossed over from ACP to DHCA to allow complete endarterectomy. At 12 weeks, the mean difference between the two groups in Z scores (the change in cognitive function score from baseline divided by the baseline SD) for the three main cognitive tests was 0·14 (95% CI –0·14 to 0·42; p=0·33) for the trail-making A and B tests, –0·06 (–0·38 to 0·25; p=0·69) for the Rey auditory verbal learning test, and 0·01 (–0·26 to 0·29; p=0·92) for the grooved pegboard test. All patients showed improvement in cognitive function at 12 weeks. We recorded no significant difference in adverse events between the two groups. At 12 weeks, two patients had died (one in each group). Interpretation Cognitive function is not impaired by either ACP or DHCA. We recommend circulatory arrest as the optimum modality for patients undergoing pulmonary endarterectomy surgery.
This online publication has been corrected. The corrected version first appeared at thelancet.com on November 25, 2011
*Authors contributed equally Papworth Hospital NHS Foundation Trust, Cambridge, UK (A Vuylsteke FRCA, G Charman RN, J Kneeshaw FRCA, S Tsui FRCS, J Dunning FRCS, A Klein FRCA, J Arrowsmith FRCA, R Hall FRCA, D Jenkins FRCS); and Medical Research Council, Biostatistics Unit, Cambridge, UK (L Sharples PhD, E Wheaton MSc) Correspondence to: Dr Alain Vuylsteke, Department of Anaesthesia and Intensive Care, Papworth Hospital, Cambridge CB23 3RE, UK [email protected]
Funding J P Moulton Charitable Foundation.
Introduction Chronic thromboembolic pulmonary hypertension (CTEPH) was first described in the UK in 1951.1 It is now more widely recognised and develops in up to 3·8% of patients after acute pulmonary embolism.2 CTEPH leads to functional impairment and confers a poor prognosis,3 but many patients can be cured by pulmonary endarterectomy (PEA), having substantial improvement in symptoms and survival.4–6 For PEA to be successful, a complete surgical endarterectomy should be done, which requires a bloodless operative field to allow for precise dissection and maximum clearance of obstructive material. The most widely used technique is deep hypothermic circulatory arrest (DHCA), which allows complete cessation of blood flow and therefore optimum operating conditions. PEA surgery with DHCA leads to a reproducible reduction in pulmonary artery pressure and low in-hospital mortality.7 Because periods of DHCA are usually limited to 20 min, permanent neurological injury is rare, although complications have been reported.8 Detailed assessment of the effect of www.thelancet.com Vol 378 October 15, 2011
PEA with DHCA on cognitive function has not been reported. Some surgical groups have suggested that PEA is possible with a lesser degree of hypothermia (>28°C) and reduced periods of DHCA9,10 or partial circulatory arrest with continuous antegrade cerebral perfusion (ACP) to the brain.11 None of these alternative techniques that aim to decrease the potential morbidity from DHCA have been assessed in a randomised manner. These techniques have received criticism because of the potential risk of incomplete endarterectomy and a less successful operation.9 A fundamental compromise exists between the provision of sufficient operating time with a clear field to allow a complete endarterectomy and minimisation of the period that a patient’s brain is not perfused. Other surgical procedures, mainly on the aorta, need cessation of blood circulation and use the same techniques to protect the brain while enabling surgery. This prospective, randomised blinded study was done to compare cognitive and clinical outcomes in a homogeneous cohort of adult patients with CTEPH undergoing PEA surgery randomly allocated to either DHCA or ACP. 1379
Articles
Methods Study population All patients aged 18–80 years who were referred to the UK national centre for PEA surgery (Papworth Hospital, Cambridge) between December, 2006, and March, 2009, were eligible to be included in the study. The institution had equal experience with both methods of cerebral management for PEA surgery when the trial started. Patients were excluded if they could not complete cognitive functions tests (eg, sight or hearing impairments, physical barriers such as severe arthritis), were not fluent in English, were having other concomitant surgery, had previous cardiac surgery, or had a medical history of stroke or a psychiatric disorder. The ethics committee (Peterborough and Fenland; 06/Q0106/11) approved the trial protocol. Written informed consent was obtained from all eligible patients. The trial is registered with Current Controlled Trials, number ISRCTN84972261.
Procedures Anaesthetic and surgical methods were standardised in this pulmonary endarterectomy and cognition
196 assessed for eligibility 14 anticipated difficulty in completing cognitive function tests 6 not fluent in English 2 unable to gain informed consent 29 concomitant surgery 12 had medical history of stroke or psychiatric disorder 4 had redo sternotomy (including redo pulmonary thromboendarterectomy) 8 younger than 18 years or older than 80 years 6 had unacceptable risk 10 declined to participate 2 study staff not available 20 unsuitable for surgery after screening 6 declined surgery 1 died while waiting operation 2 screened but trial recruitment ended before day of operation 74 randomly allocated to treatment group
35 allocated to DHCA
Randomisation and masking
39 allocated to ACP (30 received only ACP)
1 death (in hospital) 1 withdrawal from study
33 followed up at 12 weeks 1 death 2 lost to follow-up
30 followed up at 12 months
1 death 1 withdrawal from study
37 followed up at 12 weeks
1 lost to follow-up
36 followed up at 12 months
Figure 1: Trial profile DHCA=deep hypothermic circulatory arrest. ACP=antegrade cerebral perfusion.
1380
(PEACOG) study. All patients received 1 mg/kg prednisolone 12 h before surgery and 1 g methylprednisolone on induction. Fentanyl and midazolam were given for induction of anaesthesia, pancuronium was used for neuromuscular blockade, and propofol by intravenous infusion was used for maintenance of anaesthesia. Standard monitoring was used as described elsewhere.11 All patients were cooled to a core (bladder) temperature of 20°C on cardiopulmonary bypass, with the same management of pH and rates of cooling and rewarming in both groups. Thiopental 1 g was given intravenously before the first of any DHCA period. During surgery, near-infrared spectroscopy was measured continuously with cerebral oximetry (Cerebral Oximeter, Somanetics; Troy, MI, USA) with probes applied to each side of the patient’s forehead. A modified cooling jacket was applied around the patient’s head (Polar Care 500 and WrapOn Polar Pad, Breg Inc; Carlsbad, CA, USA). The cooling jacket consists of a small electric pump circulating water at 7–10°C to wrap-on pads placed around the patient’s head. The endarterectomy dissection was done with the technique as described in detail elsewhere.12 The DHCA technique was based on a protocol that has proven successful in more than 2500 patients.13 Continuous periods of DHCA were limited to a maximum of 20 min, but were repeated as necessary. A 10-min reperfusion period was given between DHCA periods. The ACP technique was adapted from that used during aortic surgery and is described in detail elsewhere.11 In this technique, the aortic arch is isolated by crossclamping, which interrupts the circulation to the descending aorta and bronchial artery branches, collateralising the pulmonary circulation. Perfusion was maintained in the innominate and left carotid arteries with a mean pressure of 40–50 mm Hg recorded at the right radial artery. If the operative field was inadequate for complete endarterectomy dissection, the surgeon was permitted to use DHCA.
A computer-generated random allocation sequence (produced by an independent randomisation centre; Papworth Hospital Research and Development Unit, Cambridge, UK) was used to allocate patients to treatment group in a 1:1 ratio. On the morning of surgery, a member of the surgical team who was unmasked to intervention telephoned the randomisation centre to register the patient due to have surgery and confirm that consent had been obtained, and was then given the allocated surgical technique. Patients, individuals in the intensive care or ward clinical team, and individuals collecting either research or routine clinical outcome measures were masked to treatment allocation. Details of the surgical procedure were stored in a sealed envelope. The operating team collated data pertaining to the intraoperative procedure into a www.thelancet.com Vol 378 October 15, 2011
Articles
restricted access electronic file that was set up specifically for the study.
Study endpoints The primary endpoint was change in cognitive function at 12 weeks after PEA surgery. Predefined secondary outcomes were as follows: change in cognitive function at 52 weeks after surgery; change in cognitive function as measured by the Cambridge Neuropsychological Test Automated Battery (CANTAB) system14—a computerised test used to assess cognitive function—at 12 weeks, and 52 weeks after surgery; number of patients with substantial changes in cognitive function at 12 weeks; change in mean pulmonary artery pressure (mPAP) and pulmonary vascular resistance (PVR) at 12 weeks after surgery; quality of life as estimated by the Cambridge Pulmonary Hypertension Outcome Review (CAMPHOR)15 and a 6-min walk test at 12 weeks and 52 weeks; the intraoperative cerebral oxygen saturation changes as measured by nearinfrared spectroscopy; and duration of mechanical ventilation, length of stay in intensive care, and length of stay in hospital. All outcomes were defined before patients were enrolled. Adverse events were recorded prospectively and reviewed at quarterly steering group meetings by team members who were masked to treatment allocation. The first 44 patients in the study received the antifibrinolytic aprotinin as part of the protocol. Tranexamic acid was given to all other patients after the withdrawal of aprotinin from sale by the manufacturer. An independent observer who was masked to treatment allocation assessed cognitive function using a battery of conventional cognitive tests:16 the trail-making A and B tests, the Rey auditory verbal learning test, and the grooved pegboard test (webappendix 1). These three tests were used to define the number of patients needed and the primary outcome. Alongside the conventional battery of tests, we used a computerised battery of neuropsychological tests (CANTAB) that consisted of six tests reporting 24 measures that assessed attention, verbal recognition, and visual memory.14 Right heart catheterisation was done before referral for PEA and at 12 weeks after PEA by physicians who were masked to treatment allocation. Physicians and nurses masked to the randomisation group measured quality of life using CAMPHOR, a disease-specific instrument for the assessment of patient-reported symptoms, functioning, and quality of life in pulmonary arterial hypertension.15 Patients were admitted to the same intensive-care unit postoperatively and staff followed a well defined protocol with set variables for patients’ clinical management. Anticoagulation was restarted from the second postoperative day. Patients were subsequently admitted to the surgical ward. Physiotherapists judged when a patient was fit for hospital discharge. www.thelancet.com Vol 378 October 15, 2011
Deep hypothermic circulatory arrest (n=35)
Antegrade cerebral perfusion (n=39)
Clinical characteristics Sex Male
19 (54%)
16 (41%)
Female
16 (46%)
23 (59%)
33 (94%)
38 (97%)
Ethnic origin White British Other Age (years)
2 (6%) 53·5 (16·8)
1 (3%) 50·7 (13·7)
Weight (kg)
85·4 (19·0)
87·6 (19·7)
Height (cm)
170·3 (11·0)
168·3 (10·0)
29·4 (5·5)
30·9 (6·6)
1
0 (0%)
0 (0%)
2
3 (9%)
1 (3%)
3
21 (60%)
32 (82%)
4
11 (31%)
6 (16%)
Body-mass index (kg per m²) Clinical history New York Heart Association classification17*
History of depression Yes
3 (9%)
4 (10%)
No
32 (91%)
35 (90%)
49 (10)
46 (13)
Right heart catheter results Pulmonary artery pressure (mm Hg) Cardiac output (L per min) Pulmonary vascular resistance (dynes/s per cm⁵) Central oxygen venous saturation (%)
4·1 (1·1) 794 (333) 61·6 (8·4)
4·6 (1·2) 679 (333) 63·1 (7·4)
Pulmonary function tests results Forced expiratory volume in 1 s (% predicted)
85·4 (19·0)
87·5 (20·1)
Transfer factor of the lung for carbon monoxide (% predicted)
65·1 (17·6)
68·5 (15·4)
6-min walk distance (m)
309 (109)
308 (113)
Data are mean (SD) or n (%). *Grouped into two categories (score 1–3 and 4) to compute Fisher’s exact test.
Table 1: Baseline characteristics
Statistical analysis All analyses were done on an intention-to-treat basis. A secondary per-protocol analysis was done because some patients crossed over from ACP to DHCA. The results of the tests were standardised with a Z score defined as the change in cognitive function test score from baseline divided by the baseline standard deviation. For tests with more than one dimension, the mean Z score per test was used. The minimum clinically significant difference between the groups was defined before the trial began as a difference in mean Z score equal to or greater than one. Sample size established to give 90% power to detect a clinically significant difference in any one of the three cognitive function test scores of one times the baseline standard deviation, with a two-sided p value of 0·01 (for an overall type-1 error of 3%). The estimated sample size was 35 patients per group (70 patients in total). Differences in baseline measurements and duration of stay in the intensive-care unit and hospital between the
See Online for webappendix
1381
Articles
DHCA and ACP groups were assessed with the t test, Fisher’s exact test, and the Mann-Whitney U test, as appropriate. Differences in the primary endpoint measurements were formally assessed with multivariate ANOVA (MANOVA), with the Z scores for three cognitive function tests (trail-making A and B, Rey auditory verbal learning, or grooved pegboard) as response variables, and the method (DHCA or ACP) as an independent factor. Models adjusting for use of aprotinin were also fitted, but the treatment effect was not changed and these data are not presented. Secondary outcomes such as mean pulmonary artery pressure (mPAP), pulmonary vascular resistance (PVR), and 6-min walk distance at 12 and 52 weeks were compared between the two groups by analysis of covariance (ANCOVA) with group as a factor and baseline measurement as a covariate. Adverse events were A
B
Z score
1
1
0
0
–1
–1
–2
–2
–3
–3
Z score
C
D
3
3
2
2 1
1
0
0
–1 –1
F
E 2
0·5 0
Z score
1
–0·5 0 –1 –1
–1·5 –2
–2 DHCA
ACP
DHCA
ACP
Figure 2: Cognitive function at 12 weeks and 52 weeks DHCA=deep hypothermic circulatory arrest. ACP=antegrade cerebral perfusion. Z score for the trail-making test at (A) 12 weeks and (B) 52 weeks. Z score for the Rey auditory verbal learning test at (C) 12 weeks and (D) 52 weeks. Z score for the grooved pegboard test at (E) 12 weeks and (F) 52 weeks. The box range is IQR, the horizontal line within the box is the median, the whiskers are the range of the measurements with the exception of outliers, which are plotted as points.
1382
tabulated and compared. Poisson regression was used to compare relative risks. We used STATA (version 11) for statistical analysis.
Role of the funding source The sponsor of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. AV and DJ had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Results 74 of 196 screened patients were randomly allocated to one of the two treatment groups (figure 1). Overall inhospital mortality for the whole cohort was 1·4% (1 of 74 patients) and survival at 1 year was 96% (71 of 74 patients). Characteristics of patients at baseline were much the same between the two groups (table 1). We recorded no difference in cognitive outcome between treatment groups (figure 2, 3). For the primary outcome at 12 weeks, the mean differences between the Z scores were 0·14 (95% CI –0·14 to 0·42; p=0·33) for the trail-making test, –0·06 (–0·38 to 0·25; p=0·69) for the Rey auditory verbal learning test, and 0·01 (–0·26 to 0·29; p=0·92) for the grooved pegboard test. Not only were the differences not statistically significant, but the limits of the 95% CIs for the mean difference between the groups was less than our definition of the minimum clinically significant difference. On average, both groups showed improvement in cognitive function at 12 weeks and 52 weeks after surgery (figure 2). Only one patient (in the DHCA group), who suffered a postoperative stroke, had a clinically significant cognitive function deficit at 12 weeks and 52 weeks. This patient was excluded from the primary outcome analysis because of extreme values in the trail-making and grooved pegboard tests related to physical limitation. We also did a prespecified per-protocol analysis to compare patients who had received DHCA (n=44) with patients who had not (n=30); differences in Z scores between the two groups at 12 weeks were 0·23 (95% CI –0·05 to 0·51; p=0·11) in the trail-making test, 0·12 (–0·20 to 0·44; p=0·45) for the Rey auditory verbal learning test, and 0·10 (–0·18 to 0·38; p=0·47) for the grooved pegboard test. Differences between the groups were smaller at 52 weeks—in the intention-to-treat analysis, the difference was 0·003 (95% CI –0·19 to 0·20; p=0·97) in the trail-making test, –0·001 (–0·37 to 0·36; p=0·99) in the Rey auditory verbal learning test, and 0·006 (–0·24 to 0·25; p=0·96) for the grooved pegboard test. We recorded no differences between the groups in any of the 24 scales of the six CANTAB tests (figure 3). Although the proportion of patients in the ACP group who scored the maximum 24 in the verbal recognition memory test was greater than was that of patients in the DHCA group at 52 weeks, a higher proportion of patients in the ACP group scored 24 at www.thelancet.com Vol 378 October 15, 2011
100
30
90 80 70
Rapid visual information processing (total hits)
Reaction time (ms)
800
600
400
20
10
200
0
2000
80
90 80 70 60 50 40 30 20 10 0
1500 1000 500 0
Verbal recognition memory immediate test (% of patients who scored 24)
60
Spatial working memory (total errors)
Match-to-sample visual research (movement time; ms)
Pattern recognition memory (% correct)
Articles
60 40 20 0
DHCA ACP Baseline
DHCA ACP 12 weeks
DHCA ACP 52 weeks
DHCA ACP Baseline
DHCA ACP 12 weeks
78 59
54
59
54
41
DHCA ACP 52 weeks
DHCA ACP Baseline
DHCA ACP 12 weeks
DHCA ACP 52 weeks
Figure 3: Cambridge Neuropsychological Test Automated Battery system results DHCA=deep hypothermic circulatory arrest. ACP=antegrade cerebral perfusion. For box and whisker plots, the box range is IQR, the horizontal line within the box is the median, the whiskers are the range of the measurements with the exception of outliers, which are plotted as points. The bar chart plots the percentage of patients scoring the maximum 24 on the verbal recognition memory test.
Deep hypothermic circulatory arrest (n=35)
Antegrade cerebral perfusion (n=39)*
p value
Surgery details Total time in operating room (min)
463·6 (54·8)
453·3 (53·2)
0·41
Duration of cardiopulmonary bypass (min)
325·7 (49·4)
326·4 (49·2)
0·95
Duration of total myocardial ischaemic time (min)
51·5 (29·8)
0·09
Duration of surgery on right pulmonary artery with ACP (min)
62·5 (24·8) ··
21·8 (18·4)
··
Duration of surgery on left pulmonary artery with ACP (min)
··
17·1 (13·3)
··
Duration of surgery on right pulmonary artery with DHCA (min)
19·7 (8·6)
3·1 (7·0)†
··
Duration of surgery on left pulmonary artery with DHCA (min)
16·3 (8·4)
3·2 (7·2)†
··
Cerebral saturation values‡ Right-sided cerebral saturation at baseline
77·3 (12·1)
73·6 (10·5)
0·19
Right-sided cerebral saturation, lowest value
37·4 (9·0)
53·6 (12·1)
<0·001
Right-sided cerebral saturation, after sternal closure
66·6 (13·0)
65·4 (9·0)
0·65
Left-sided cerebral saturation, baseline
76·4 (12·3)
77·1 (8·6)
0·77
Left-sided cerebral saturation, lowest value
37·8 (9·1)
54·2 (13·9)
<0·0001
Left-sided cerebral saturation, after sternal closure
67·7 (12·0)
67·8 (10·3)
0·97
0·25§
Pulmonary disease type (Jamieson classification) Right pulmonary artery disease 1
8 (23%)
15 (38%)
2
22 (63%)
16 (41%)
3
4 (11%)
7 (18%)
4
1 (3%)
1 (3%)
Left pulmonary artery disease 1
6 (17%)
8 (21%)
2
22 (63%)
20 (51%)
3
5 (14%)
8 (21%)
4
2 (6%)
3 (8%)
0·81§
Intensive care data Duration of mechanical ventilation in intensive care (h)¶|| Duration of stay in intensive care (days)
20 (16–28)
19 (16–22)
0·81
3 (2–5)
3 (2–5)
0·88 (Continues on next page)
www.thelancet.com Vol 378 October 15, 2011
1383
Articles
Deep hypothermic circulatory arrest (n=35)
Antegrade cerebral perfusion (n=39)*
p value
(Continued from previous page) Duration of stay in hospital (days)
14 (10–21)
15 (11–22)
0·35
46·3 (11·7)
48·2 (14·5)
0·56
619·8 (236·3)
642·1 (309·1)
0·73
0·01§
Perioperative haemodynamic data Mean pulmonary artery pressure before cardiopulmonary bypass (mm Hg) Mean pulmonary vascular resistance (assuming pulmonary capillary wedge pressure of 10 mm Hg) before cardiopulmonary bypass (dynes/s per cm⁵)|| Inotropic support greater than dopamine at 5 μg/kg per min Yes
0 (0%)
7 (18%)
No
35 (100%)
32 (82%)
Mean pulmonary artery pressure on first postoperative day (mm Hg) Mean pulmonary vascular resistance (assuming pulmonary capillary wedge pressure of 10 mm Hg) on first postoperative day (dynes/s per cm⁵)|| Mean pulmonary artery pressure at week 12 (mm Hg)** Mean pulmonary vascular resistance (assuming pulmonary capillary wedge pressure of 10 mm Hg) at week 12 (dynes/s per cm⁵)**
20·6 (4·7) 223 (127) 28·1 (9·2) 316 (197)
23·4 (7·6) 260 (187) 27·8 (10·6) 338 (316)
0·10 0·40 0·70 0·37
Camphor results†† Symptoms score at baseline
14 (11–18)
14 (9–19)
0·91
Symptoms score at 12 weeks
4 (1–9)
2 (1–7)
0·30
Symptoms score at 52 weeks
4 (1–11)
1 (0–3)
0·03
12 (7–17)
13 (7–17)
0·58
3 (1–10)
0·23
Activity score at baseline Activity score at 12 weeks
7 (2–10)
Activity score at 52 weeks
5 (1–12)
1·5 (0–5)
0·09
11 (6–18)
10 (8–17)
0·93
Quality-of-life score at 12 weeks
3 (1–9)
4 (1–9)
0·97
Quality-of-life score at 52 weeks
4 (0–9)
1 (0–6)
0·26
Quality-of-life score at baseline
Walk distance 6-min walk distance (m) at 12 weeks§§
348 (126)
371 (104)
0·31
6-min walk distance (m) at 52 weeks¶¶
373 (113)
410 (120)
0·14
New York Heart Association classification17 Number in class at 12 weeks
1·00
1
21
23
··
2
10
12
··
3
1
2
··
4
1
0
··
1
17
23
2
13
11
··
3
0
1
··
4
0
0
··
Number in class at 52 weeks
0·61 ··
Data are mean (SD), n (%), or median (IQR) unless otherwise stated. *One patient with missing data—data for antegrade cerebral perfusion (ACP) are therefore for maximum 38 patients. †Nine patients randomly allocated to ACP also received deep hypothermic circulatory arrest (DHCA). The mean (SD) DHCA time for those nine patients was 13·3 min (8·7) for the right side and 13·7 min (9·2) for the left side. ‡Some cerebral saturation information is missing because the pads stopped recording. Records were obtained for 30 patients in ACP group and 36 patients in DHCA group (baseline), 33 patients in ACP group and 37 patients in DHCA group (lowest value), 30 patients in ACP group and 33 patients in DHCA group (after sternal closure) patients. §Fisher’s exact test. ¶One patient (in ACP group) had long intubation time (792 h) and had to be transferred to a neurological unit. One patient (in DHCA group) had missing information because he was transfered to another hospital—this patient was lost to follow-up at 3 months because of brain injury. Data from two patients missing from the ACP group because one was withdrawn by the investigators and one died. After the first extubation after surgery, six patients had to be intubated again (four from the DHCA group and two from the ACP group; these were also the cooling techniques that the patient actually received). The four patients in the DHCA group were intubated 1, 5, 11, and 17 days after the extubation date, respectively. The two patients in the ACP group were intubated 3 days and 15 days after extubation, respectively. ||Some haemodynamic information is missing because of pulmonary artery catheter problems—records were obtained for 35/37 and 35/36 patients (ACP/DHCA; n). **Some patients refused the procedure at 12 weeks— records were obtained for 32 patients in the ACP group and 37 patients in the DHCA group. ††Results were obtained for 35/39 at baseline, 33/37 at 12 weeks, 27/26 at 52 weeks (activity score), and 27/25 at 52 weeks (quality of life; ACP/DHCA; n). §§Data missing for one patient (in DHCA group) because they had a cerebral vascular incident and for one patient (in ACP group) because of a baseline mobility problem. ¶¶Data missing for one patient (in DHCA group) because they had a cerebral vascular incident, and for two patients (in ACP group) because of a baseline mobility problem and because follow-up was done by post.
Table 2: Operative data and secondary outcome measures
1384
www.thelancet.com Vol 378 October 15, 2011
Articles
Deep hypothermic Antegrade circulatory arrest cerebral perfusion Serious adverse events Arrhythmia
0
1
Central vascular accident or transient ischaemic attacks
2
1
Death
2
1
Pericardial effusions
3
1
Pulmonary infection
1
0
Re-exploration for postoperative bleeding
1
0 0
Acute renal failure
1
Lung reperfusion injury
1
3
Right heart failure
0
2
Haemoglobin less than 8 g/dL 0
1
Others
2
2
Neurological adverse events (mild or moderate) Confusion or mood alteration 5
4
Transient ischaemic attack
1
1
Dizziness
2
1
Headache
3
0
Tremor
4
5
Sensory disturbances
2
2
Table 3: Adverse events
baseline, meaning that the adjusted differences were not significantly different (odds ratio 1·2, 95% CIs 0·29–4·30; p=0·87). Haemodynamic improvement after surgery was much the same between the two groups. We recorded no differences in the results of the 6-min walk distance test, New York Heart Association class,17 duration of stay, or CAMPHOR scores at either 12 weeks or 52 weeks (table 2). We recorded 78 adverse events in 29 patients in the DHCA group and 105 events in 32 patients in the ACP group over 52 weeks. The relative risk for individuals in the ACP group compared with those in the DHCA group was 1·22 (95% CI 0·91–1·64; p=0·18). We recorded no difference between groups in the number of serious adverse events—ie, events that were life-threatening, required admission to hospital as an inpatient, extended the length of an individual’s stay in hospital, or resulted in death or persistent or substantial disability. We recorded 13 serious adverse events (which led to death or an extended stay in hospital) in the DHCA group and 12 in the ACP group (table 3).
Discussion Our findings show no benefit in the maintenance of brain perfusion over temporary periods of DHCA of up to 20 min at 20°C (total cumulative time 36 min) in patients undergoing PEA. Our study shows that although perfusion of the brain might be safe, no evidence exists that it is better than DHCA, because we www.thelancet.com Vol 378 October 15, 2011
did not detect any difference between the groups in cognitive function measurements. Perfusion of the brain is a laudable goal, but should not detract from the importance of surgical success. This study lends credence to previous concerns that the avoidance of DHCA might lead to suboptimum endarterectomy,18 because nine patients (23%) crossed over from ACP to DHCA to allow complete endarterectomy. We recorded equivalent clinical outcomes in both groups, and did not detect any advantages with the use of ACP—the more technically demanding technique. DHCA at 20°C should be the standard method for PEA surgery. Postoperative cognitive dysfunction is a decrease in several neuropsychological domains such as memory, executive functioning, and speed of processing. It was assessed with the tests recommended in an international consensus statement16 and a newer computer-based method (CANTAB).14 The frequency with which postoperative cognitive dysfunction occurs is unclear19 and might not be as prevalent after cardiac surgery as has previously been reported.20,21 The two groups of patients enrolled in this study were much the same in regards to preoperative or intraoperative precipitating factors22,23 and a postoperative decrease in cognitive function would have been expected in both groups. This absence of difference is surprising because the groups had substantially different patterns of cerebral oxygen saturation during surgery. The use of near-infrared spectroscopy is increasing in patients undergoing cardiac surgery, despite the scarcity of objective evidence of benefit and the absence of defined safe saturation thresholds at different temperatures. The average cognitive function test score improved in both groups at 12 weeks and 52 weeks. We suggest that the postoperative improvement in cardiac output, combined with an improvement in quality of life, might partly explain the overall improvement in cognitive function. Induction of hypothermia and DHCA is the standard technique to reduce metabolic activity and the risk of ischaemic cerebral damage during aortic and other types of cardiac surgery when interruption of blood flow to the brain is necessary. The inverse relation between temperature and neurological complications has led to the idea of a safe duration of total circulatory arrest.24 By comparison with DHCA, ACP should be better at maintaining intracranial hypothermia, providing oxygen and metabolite substrates, and removing toxic metabolites. These arguments have been used to justify the development of alternative methods such as antegrade and retrograde perfusion,25 but debate about their relative merits is ongoing and opinion is polarised.26 We know of no other randomised controlled trial in adult patients to address this issue (panel). In this study, ACP proved unsuitable for nine patients who eventually needed DHCA. However, we recorded no difference in outcomes when comparing patients who had DHCA with those who had not. 1385
Articles
Panel: Research in context Systematic review We know of no other prospective controlled trial to compare the cognitive outcome in adult patients randomly allocated to receive either deep hypothermic circulatory arrest (DHCA) or maintained brain perfusion. A Medline search with search term “DHCA” and “deep hypothermic circulatory arrest” produced no previous reports of randomised controlled trial comparing DHCA with brain perfusion in adult patients (the search had no date or language restrictions and the last search was done on Sept 10, 2011). The findings of a comprehensive review27 from 2010 confirmed that no previous trials had been done—we also assessed referenced studies in this review. Interpretation In this study to assess cognitive function in patients undergoing pulmonary endarterectomy, we report that cognitive function is not impaired and actually improves, whether or not the brain was perfused. We recorded no cognitive or other outcome benefits from the maintenance of cerebral perfusion compared with periods of DHCA of up to 20 min (cumulative 36 min) at 20°C. Our findings suggest that circulatory arrest is the optimum modality for patients undergoing pulmonary endarterectomy surgery.
This study could inform the debate in patients undergoing aortic surgery because we have conclusively proved that a continuous period of up to 20 min of DHCA at 20°C (total cumulative time of 36 min) is well tolerated and that no benefit exists in the maintenance of cerebral perfusion in similar conditions. The main limitation of this study is the small size and the risk of a type II error, although the study was powered for the primary outcome measure of cognitive function change at 12 weeks and the CIs were all within one SD, which was specified before the trial began as the minimum clinically important difference. Results might vary with the selection of different cognitive function tests.20 By using the CANTAB battery of tests, we investigated cognitive domains repeatedly suggested as important— learning and memory, attention, and executive functioning.28 The advantage of CANTAB is that it explores performance and is less dependent on the tester, although in this study the same person did all assessments.22 In the setting of PEA, maintenance of cerebral perfusion throughout the procedure adds to the complexity of an operation and confers no additional benefits. Whether these findings can be extrapolated for other types of surgery in which a choice exists between the use of DHCA or selective ACP is not known. Contributors AV and DJ had the idea for the study, supervised research staff, masked or unmasked assessors, led the writing of the paper. DJ, LS, GC, JK, ST, JD, JA, and DJ designed the protocol. AV, LS, and DJ applied for funding. AV, LS, JK, ST, JD, EW, AK, JA, RH, and DJ interpreted the results. GC,
1386
JK, ST, JD, and DJ recruited patients. LS and GC analysed the data. GC, JK, ST, JD, AK, JA, RH, and DJ collected the data. LS, GC, JK, ST, JD, EW, AK, JA, and RH reviewed the manuscript. LS and EW compiled the statistical report. AV is a member of the steering group. GC did the cognitive tests and coordinated the study. JK coordinated the theatre staff. EW analysed cognitive data. Conflicts of interest AK has received honorarium from Covidien, owner of Somanetics for lectures on the use of near-infrared spectroscopy. All other authors declare that they have no conflicts of interest. Acknowledgments Johanna Armstrong helped with writing the protocol and obtaining funding. Ella Wheaton did statistical analysis during a Clinical Trial Fellowship located within the Medical Research Council Biostatistics Clinical Trials Methodology Hub and funded by the National Institute of Health Research. Ray Latimer was involved in discussing the original protocol and supported the need for a study. We thank Stephanie Cadour for her contribution to the analysis of the CANTAB test results. We thank all staff at Papworth Hospital and the regional designated pulmonary hypertension referral hospitals for their care of the study patients. References 1 Petch CP. Cor pulmonale from recurrent pulmonary embolism. Lancet 1951; 1: 1346–47. 2 Pengo V, Lensing AW, Prins MH, et al. Incidence of chronic thromboembolic pulmonary hypertension after pulmonary embolism. N Engl J Med 2004; 350: 2257–64. 3 Piazza G, Goldhaber SZ. Chronic thromboembolic pulmonary hypertension. N Engl J Med 2011; 364: 351–60. 4 Condliffe R, Kiely DG, Gibbs JS, et al. Improved outcomes in medically and surgically treated chronic thromboembolic pulmonary hypertension. Am J Respir Crit Care Med 2008; 177: 1122–27. 5 Freed DH, Thomson BM, Berman M, et al. Survival after pulmonary thromboendarterectomy: effect of residual pulmonary hypertension. J Thorac Cardiovasc Surg 2010; 141: 383–87. 6 Mayer E, Jenkins D, Lindner J, et al. Surgical management and outcome of patients with chronic thromboembolic pulmonary hypertension: results from an international prospective registry. J Thorac Cardiovasc Surg 2011; 141: 702–10. 7 Jamieson SW, Kapelanski DP, Sakakibara N, et al. Pulmonary endarterectomy: experience and lessons learned in 1,500 cases. Ann Thorac Surg 2003; 76: 1457–62. 8 Surie S, Tijssen MA, Biervliet JD, et al. Chorea in adults following pulmonary endarterectomy. Mov Disord 2010; 25: 1101–04. 9 Macchiarini P, Kamiya H, Hagl C, et al. Pulmonary endarterectomy for chronic thromboembolic pulmonary hypertension: is deep hypothermia required? Eur J Cardiothorac Surg 2006; 30: 237–41. 10 Mikus PM, Mikus E, Martin-Suarez S, et al. Pulmonary endarterectomy: an alternative to circulatory arrest and deep hypothermia: mid-term results. Eur J Cardiothorac Surg 2008; 34: 159–63. 11 Thomson B, Tsui SS, Dunning J, et al. Pulmonary endarterectomy is possible and effective without the use of complete circulatory arrest—the UK experience in over 150 patients. Eur J Cardiothorac Surg 2008; 33: 157–63. 12 Thistlethwaite PA, Kaneko K, Madani MM, Jamieson SW. Technique and outcomes of pulmonary endarterectomy surgery. Ann Thorac Cardiovasc Surg 2008; 14: 274–82. 13 Jamieson SW, Kapelanski DP. Pulmonary endarterectomy. Curr Probl Surg 2000; 37: 165–252. 14 Sahakian BJ, Owen AM. Computerized assessment in neuropsychiatry using CANTAB: discussion paper. J R Soc Med 1992; 85: 399–402. 15 McKenna SP, Doughty N, Meads DM, Doward LC, Pepke-Zaba J. The Cambridge Pulmonary Hypertension Outcome Review (CAMPHOR): a measure of health-related quality of life and quality of life for patients with pulmonary hypertension. Qual Life Res 2006; 15: 103–15. 16 Murkin JM, Newman SP, Stump DA, Blumenthal JA. Statement of consensus on assessment of neurobehavioral outcomes after cardiac surgery. Ann Thorac Surg 1995; 59: 1289–95.
www.thelancet.com Vol 378 October 15, 2011
Articles
17
18 19 20
21
22
The Criteria Committee of the New York Heart Association. Nomenclature and criteria for diagnosis of diseases of the heart and great vessels (9th edn). Boston, MA: Little, Brown and Co, 1994: 253–56. Jamieson SW, Madani M. Invited commentary. Ann Thorac Surg 2008; 86: 1267. Selwood A, Orrell M. Long term cognitive dysfunction in older people after non-cardiac surgery. BMJ 2004; 328: 120–21. Rudolph JL, Schreiber KA, Culley DJ, et al. Measurement of post-operative cognitive dysfunction after cardiac surgery: a systematic review. Acta Anaesthesiol Scand 2010; 54: 663–77. Newman MF, Kirchner JL, Phillips-Bute B, et al. Longitudinal assessment of neurocognitive function after coronary-artery bypass surgery. N Engl J Med 2001; 344: 395–402. Funder KS, Steinmetz J, Rasmussen LS. Methodological issues of postoperative cognitive dysfunction research. Semin Cardiothorac Vasc Anesth 2010; 14: 119–22.
www.thelancet.com Vol 378 October 15, 2011
23
24
25 26 27
28
Moller JT, Cluitmans P, Rasmussen LS, et al. Long-term postoperative cognitive dysfunction in the elderly ISPOCD1 study: ISPOCD investigators—International Study of Post-Operative Cognitive Dysfunction. Lancet 1998; 351: 857–61. Kirklin JW, Baratt-Boyes BG. Hypothermia, circulatory arrest, and cardiopulmonary bypass. In: Kirklin JW, Barratt-Boyes BG, eds. Cardiac Surgery, 2nd edn. New York: Churchill Livingstone, 1993: 61–127. Bonser RS, Harrington DK. Editorial comment. Eur J Cardiothorac Surg 2005; 28: 102–03. Elefteriades JA. Editorial comment: what is the best method for brain protection in surgery of the aortic arch? Cardiol Clin 2010; 28: 403–04. Elefteriades JA. What is the best method for brain protection in surgery of the aortic arch? Straight DHCA. Cardiol Clin 2010; 28: 381–87. Bekker AY, Weeks EJ. Cognitive function after anaesthesia in the elderly. Best Pract Res Clin Anaesthesiol 2003; 17: 259–72.
1387
Circulatory arrest versus cerebral perfusion during pulmonary endarterectomy surgery (PEACOG): a randomised controlled trial Alain Vuylsteke*, Linda Sharples, Gill Charman, John Kneeshaw, Steven Tsui, John Dunning, Ella Wheaton, Andrew Klein, Joseph Arrowsmith, Roger Hall, David Jenkins*
Summary Background For some surgical procedures to be done, a patient’s blood circulation needs to be stopped. In such situations, the maintenance of blood flow to the brain is perceived beneficial even in the presence of deep hypothermia. We aimed to assess the benefits of the maintenance of antegrade cerebral perfusion (ACP) compared with deep hypothermic circulatory arrest (DHCA).
Lancet 2011; 378: 1379–87
Methods Patients aged 18–80 years undergoing pulmonary endarterectomy surgery in a UK centre (Papworth Hospital, Cambridge) were randomly assigned with a computer generated sequence to receive either DHCA for periods of up to 20 min at 20°C or ACP (1:1 ratio). The primary endpoint was change in cognitive function at 12 weeks after surgery, as assessed by the trail-making A and B tests, the Rey auditory verbal learning test, and the grooved pegboard test. Patients and assessors were masked to treatment allocation. Primary analysis was by intention to treat. The trial is registered with Current Controlled Trials, number ISRCTN84972261.
See Comment page 1359
Findings We enrolled 74 of 196 screened patients (35 to receive DHCA and 39 to receive ACP). Nine patients crossed over from ACP to DHCA to allow complete endarterectomy. At 12 weeks, the mean difference between the two groups in Z scores (the change in cognitive function score from baseline divided by the baseline SD) for the three main cognitive tests was 0·14 (95% CI –0·14 to 0·42; p=0·33) for the trail-making A and B tests, –0·06 (–0·38 to 0·25; p=0·69) for the Rey auditory verbal learning test, and 0·01 (–0·26 to 0·29; p=0·92) for the grooved pegboard test. All patients showed improvement in cognitive function at 12 weeks. We recorded no significant difference in adverse events between the two groups. At 12 weeks, two patients had died (one in each group). Interpretation Cognitive function is not impaired by either ACP or DHCA. We recommend circulatory arrest as the optimum modality for patients undergoing pulmonary endarterectomy surgery.
This online publication has been corrected. The corrected version first appeared at thelancet.com on November 25, 2011
*Authors contributed equally Papworth Hospital NHS Foundation Trust, Cambridge, UK (A Vuylsteke FRCA, G Charman RN, J Kneeshaw FRCA, S Tsui FRCS, J Dunning FRCS, A Klein FRCA, J Arrowsmith FRCA, R Hall FRCA, D Jenkins FRCS); and Medical Research Council, Biostatistics Unit, Cambridge, UK (L Sharples PhD, E Wheaton MSc) Correspondence to: Dr Alain Vuylsteke, Department of Anaesthesia and Intensive Care, Papworth Hospital, Cambridge CB23 3RE, UK [email protected]
Funding J P Moulton Charitable Foundation.
Introduction Chronic thromboembolic pulmonary hypertension (CTEPH) was first described in the UK in 1951.1 It is now more widely recognised and develops in up to 3·8% of patients after acute pulmonary embolism.2 CTEPH leads to functional impairment and confers a poor prognosis,3 but many patients can be cured by pulmonary endarterectomy (PEA), having substantial improvement in symptoms and survival.4–6 For PEA to be successful, a complete surgical endarterectomy should be done, which requires a bloodless operative field to allow for precise dissection and maximum clearance of obstructive material. The most widely used technique is deep hypothermic circulatory arrest (DHCA), which allows complete cessation of blood flow and therefore optimum operating conditions. PEA surgery with DHCA leads to a reproducible reduction in pulmonary artery pressure and low in-hospital mortality.7 Because periods of DHCA are usually limited to 20 min, permanent neurological injury is rare, although complications have been reported.8 Detailed assessment of the effect of www.thelancet.com Vol 378 October 15, 2011
PEA with DHCA on cognitive function has not been reported. Some surgical groups have suggested that PEA is possible with a lesser degree of hypothermia (>28°C) and reduced periods of DHCA9,10 or partial circulatory arrest with continuous antegrade cerebral perfusion (ACP) to the brain.11 None of these alternative techniques that aim to decrease the potential morbidity from DHCA have been assessed in a randomised manner. These techniques have received criticism because of the potential risk of incomplete endarterectomy and a less successful operation.9 A fundamental compromise exists between the provision of sufficient operating time with a clear field to allow a complete endarterectomy and minimisation of the period that a patient’s brain is not perfused. Other surgical procedures, mainly on the aorta, need cessation of blood circulation and use the same techniques to protect the brain while enabling surgery. This prospective, randomised blinded study was done to compare cognitive and clinical outcomes in a homogeneous cohort of adult patients with CTEPH undergoing PEA surgery randomly allocated to either DHCA or ACP. 1379
Articles
Methods Study population All patients aged 18–80 years who were referred to the UK national centre for PEA surgery (Papworth Hospital, Cambridge) between December, 2006, and March, 2009, were eligible to be included in the study. The institution had equal experience with both methods of cerebral management for PEA surgery when the trial started. Patients were excluded if they could not complete cognitive functions tests (eg, sight or hearing impairments, physical barriers such as severe arthritis), were not fluent in English, were having other concomitant surgery, had previous cardiac surgery, or had a medical history of stroke or a psychiatric disorder. The ethics committee (Peterborough and Fenland; 06/Q0106/11) approved the trial protocol. Written informed consent was obtained from all eligible patients. The trial is registered with Current Controlled Trials, number ISRCTN84972261.
Procedures Anaesthetic and surgical methods were standardised in this pulmonary endarterectomy and cognition
196 assessed for eligibility 14 anticipated difficulty in completing cognitive function tests 6 not fluent in English 2 unable to gain informed consent 29 concomitant surgery 12 had medical history of stroke or psychiatric disorder 4 had redo sternotomy (including redo pulmonary thromboendarterectomy) 8 younger than 18 years or older than 80 years 6 had unacceptable risk 10 declined to participate 2 study staff not available 20 unsuitable for surgery after screening 6 declined surgery 1 died while waiting operation 2 screened but trial recruitment ended before day of operation 74 randomly allocated to treatment group
35 allocated to DHCA
Randomisation and masking
39 allocated to ACP (30 received only ACP)
1 death (in hospital) 1 withdrawal from study
33 followed up at 12 weeks 1 death 2 lost to follow-up
30 followed up at 12 months
1 death 1 withdrawal from study
37 followed up at 12 weeks
1 lost to follow-up
36 followed up at 12 months
Figure 1: Trial profile DHCA=deep hypothermic circulatory arrest. ACP=antegrade cerebral perfusion.
1380
(PEACOG) study. All patients received 1 mg/kg prednisolone 12 h before surgery and 1 g methylprednisolone on induction. Fentanyl and midazolam were given for induction of anaesthesia, pancuronium was used for neuromuscular blockade, and propofol by intravenous infusion was used for maintenance of anaesthesia. Standard monitoring was used as described elsewhere.11 All patients were cooled to a core (bladder) temperature of 20°C on cardiopulmonary bypass, with the same management of pH and rates of cooling and rewarming in both groups. Thiopental 1 g was given intravenously before the first of any DHCA period. During surgery, near-infrared spectroscopy was measured continuously with cerebral oximetry (Cerebral Oximeter, Somanetics; Troy, MI, USA) with probes applied to each side of the patient’s forehead. A modified cooling jacket was applied around the patient’s head (Polar Care 500 and WrapOn Polar Pad, Breg Inc; Carlsbad, CA, USA). The cooling jacket consists of a small electric pump circulating water at 7–10°C to wrap-on pads placed around the patient’s head. The endarterectomy dissection was done with the technique as described in detail elsewhere.12 The DHCA technique was based on a protocol that has proven successful in more than 2500 patients.13 Continuous periods of DHCA were limited to a maximum of 20 min, but were repeated as necessary. A 10-min reperfusion period was given between DHCA periods. The ACP technique was adapted from that used during aortic surgery and is described in detail elsewhere.11 In this technique, the aortic arch is isolated by crossclamping, which interrupts the circulation to the descending aorta and bronchial artery branches, collateralising the pulmonary circulation. Perfusion was maintained in the innominate and left carotid arteries with a mean pressure of 40–50 mm Hg recorded at the right radial artery. If the operative field was inadequate for complete endarterectomy dissection, the surgeon was permitted to use DHCA.
A computer-generated random allocation sequence (produced by an independent randomisation centre; Papworth Hospital Research and Development Unit, Cambridge, UK) was used to allocate patients to treatment group in a 1:1 ratio. On the morning of surgery, a member of the surgical team who was unmasked to intervention telephoned the randomisation centre to register the patient due to have surgery and confirm that consent had been obtained, and was then given the allocated surgical technique. Patients, individuals in the intensive care or ward clinical team, and individuals collecting either research or routine clinical outcome measures were masked to treatment allocation. Details of the surgical procedure were stored in a sealed envelope. The operating team collated data pertaining to the intraoperative procedure into a www.thelancet.com Vol 378 October 15, 2011
Articles
restricted access electronic file that was set up specifically for the study.
Study endpoints The primary endpoint was change in cognitive function at 12 weeks after PEA surgery. Predefined secondary outcomes were as follows: change in cognitive function at 52 weeks after surgery; change in cognitive function as measured by the Cambridge Neuropsychological Test Automated Battery (CANTAB) system14—a computerised test used to assess cognitive function—at 12 weeks, and 52 weeks after surgery; number of patients with substantial changes in cognitive function at 12 weeks; change in mean pulmonary artery pressure (mPAP) and pulmonary vascular resistance (PVR) at 12 weeks after surgery; quality of life as estimated by the Cambridge Pulmonary Hypertension Outcome Review (CAMPHOR)15 and a 6-min walk test at 12 weeks and 52 weeks; the intraoperative cerebral oxygen saturation changes as measured by nearinfrared spectroscopy; and duration of mechanical ventilation, length of stay in intensive care, and length of stay in hospital. All outcomes were defined before patients were enrolled. Adverse events were recorded prospectively and reviewed at quarterly steering group meetings by team members who were masked to treatment allocation. The first 44 patients in the study received the antifibrinolytic aprotinin as part of the protocol. Tranexamic acid was given to all other patients after the withdrawal of aprotinin from sale by the manufacturer. An independent observer who was masked to treatment allocation assessed cognitive function using a battery of conventional cognitive tests:16 the trail-making A and B tests, the Rey auditory verbal learning test, and the grooved pegboard test (webappendix 1). These three tests were used to define the number of patients needed and the primary outcome. Alongside the conventional battery of tests, we used a computerised battery of neuropsychological tests (CANTAB) that consisted of six tests reporting 24 measures that assessed attention, verbal recognition, and visual memory.14 Right heart catheterisation was done before referral for PEA and at 12 weeks after PEA by physicians who were masked to treatment allocation. Physicians and nurses masked to the randomisation group measured quality of life using CAMPHOR, a disease-specific instrument for the assessment of patient-reported symptoms, functioning, and quality of life in pulmonary arterial hypertension.15 Patients were admitted to the same intensive-care unit postoperatively and staff followed a well defined protocol with set variables for patients’ clinical management. Anticoagulation was restarted from the second postoperative day. Patients were subsequently admitted to the surgical ward. Physiotherapists judged when a patient was fit for hospital discharge. www.thelancet.com Vol 378 October 15, 2011
Deep hypothermic circulatory arrest (n=35)
Antegrade cerebral perfusion (n=39)
Clinical characteristics Sex Male
19 (54%)
16 (41%)
Female
16 (46%)
23 (59%)
33 (94%)
38 (97%)
Ethnic origin White British Other Age (years)
2 (6%) 53·5 (16·8)
1 (3%) 50·7 (13·7)
Weight (kg)
85·4 (19·0)
87·6 (19·7)
Height (cm)
170·3 (11·0)
168·3 (10·0)
29·4 (5·5)
30·9 (6·6)
1
0 (0%)
0 (0%)
2
3 (9%)
1 (3%)
3
21 (60%)
32 (82%)
4
11 (31%)
6 (16%)
Body-mass index (kg per m²) Clinical history New York Heart Association classification17*
History of depression Yes
3 (9%)
4 (10%)
No
32 (91%)
35 (90%)
49 (10)
46 (13)
Right heart catheter results Pulmonary artery pressure (mm Hg) Cardiac output (L per min) Pulmonary vascular resistance (dynes/s per cm⁵) Central oxygen venous saturation (%)
4·1 (1·1) 794 (333) 61·6 (8·4)
4·6 (1·2) 679 (333) 63·1 (7·4)
Pulmonary function tests results Forced expiratory volume in 1 s (% predicted)
85·4 (19·0)
87·5 (20·1)
Transfer factor of the lung for carbon monoxide (% predicted)
65·1 (17·6)
68·5 (15·4)
6-min walk distance (m)
309 (109)
308 (113)
Data are mean (SD) or n (%). *Grouped into two categories (score 1–3 and 4) to compute Fisher’s exact test.
Table 1: Baseline characteristics
Statistical analysis All analyses were done on an intention-to-treat basis. A secondary per-protocol analysis was done because some patients crossed over from ACP to DHCA. The results of the tests were standardised with a Z score defined as the change in cognitive function test score from baseline divided by the baseline standard deviation. For tests with more than one dimension, the mean Z score per test was used. The minimum clinically significant difference between the groups was defined before the trial began as a difference in mean Z score equal to or greater than one. Sample size established to give 90% power to detect a clinically significant difference in any one of the three cognitive function test scores of one times the baseline standard deviation, with a two-sided p value of 0·01 (for an overall type-1 error of 3%). The estimated sample size was 35 patients per group (70 patients in total). Differences in baseline measurements and duration of stay in the intensive-care unit and hospital between the
See Online for webappendix
1381
Articles
DHCA and ACP groups were assessed with the t test, Fisher’s exact test, and the Mann-Whitney U test, as appropriate. Differences in the primary endpoint measurements were formally assessed with multivariate ANOVA (MANOVA), with the Z scores for three cognitive function tests (trail-making A and B, Rey auditory verbal learning, or grooved pegboard) as response variables, and the method (DHCA or ACP) as an independent factor. Models adjusting for use of aprotinin were also fitted, but the treatment effect was not changed and these data are not presented. Secondary outcomes such as mean pulmonary artery pressure (mPAP), pulmonary vascular resistance (PVR), and 6-min walk distance at 12 and 52 weeks were compared between the two groups by analysis of covariance (ANCOVA) with group as a factor and baseline measurement as a covariate. Adverse events were A
B
Z score
1
1
0
0
–1
–1
–2
–2
–3
–3
Z score
C
D
3
3
2
2 1
1
0
0
–1 –1
F
E 2
0·5 0
Z score
1
–0·5 0 –1 –1
–1·5 –2
–2 DHCA
ACP
DHCA
ACP
Figure 2: Cognitive function at 12 weeks and 52 weeks DHCA=deep hypothermic circulatory arrest. ACP=antegrade cerebral perfusion. Z score for the trail-making test at (A) 12 weeks and (B) 52 weeks. Z score for the Rey auditory verbal learning test at (C) 12 weeks and (D) 52 weeks. Z score for the grooved pegboard test at (E) 12 weeks and (F) 52 weeks. The box range is IQR, the horizontal line within the box is the median, the whiskers are the range of the measurements with the exception of outliers, which are plotted as points.
1382
tabulated and compared. Poisson regression was used to compare relative risks. We used STATA (version 11) for statistical analysis.
Role of the funding source The sponsor of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. AV and DJ had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Results 74 of 196 screened patients were randomly allocated to one of the two treatment groups (figure 1). Overall inhospital mortality for the whole cohort was 1·4% (1 of 74 patients) and survival at 1 year was 96% (71 of 74 patients). Characteristics of patients at baseline were much the same between the two groups (table 1). We recorded no difference in cognitive outcome between treatment groups (figure 2, 3). For the primary outcome at 12 weeks, the mean differences between the Z scores were 0·14 (95% CI –0·14 to 0·42; p=0·33) for the trail-making test, –0·06 (–0·38 to 0·25; p=0·69) for the Rey auditory verbal learning test, and 0·01 (–0·26 to 0·29; p=0·92) for the grooved pegboard test. Not only were the differences not statistically significant, but the limits of the 95% CIs for the mean difference between the groups was less than our definition of the minimum clinically significant difference. On average, both groups showed improvement in cognitive function at 12 weeks and 52 weeks after surgery (figure 2). Only one patient (in the DHCA group), who suffered a postoperative stroke, had a clinically significant cognitive function deficit at 12 weeks and 52 weeks. This patient was excluded from the primary outcome analysis because of extreme values in the trail-making and grooved pegboard tests related to physical limitation. We also did a prespecified per-protocol analysis to compare patients who had received DHCA (n=44) with patients who had not (n=30); differences in Z scores between the two groups at 12 weeks were 0·23 (95% CI –0·05 to 0·51; p=0·11) in the trail-making test, 0·12 (–0·20 to 0·44; p=0·45) for the Rey auditory verbal learning test, and 0·10 (–0·18 to 0·38; p=0·47) for the grooved pegboard test. Differences between the groups were smaller at 52 weeks—in the intention-to-treat analysis, the difference was 0·003 (95% CI –0·19 to 0·20; p=0·97) in the trail-making test, –0·001 (–0·37 to 0·36; p=0·99) in the Rey auditory verbal learning test, and 0·006 (–0·24 to 0·25; p=0·96) for the grooved pegboard test. We recorded no differences between the groups in any of the 24 scales of the six CANTAB tests (figure 3). Although the proportion of patients in the ACP group who scored the maximum 24 in the verbal recognition memory test was greater than was that of patients in the DHCA group at 52 weeks, a higher proportion of patients in the ACP group scored 24 at www.thelancet.com Vol 378 October 15, 2011
100
30
90 80 70
Rapid visual information processing (total hits)
Reaction time (ms)
800
600
400
20
10
200
0
2000
80
90 80 70 60 50 40 30 20 10 0
1500 1000 500 0
Verbal recognition memory immediate test (% of patients who scored 24)
60
Spatial working memory (total errors)
Match-to-sample visual research (movement time; ms)
Pattern recognition memory (% correct)
Articles
60 40 20 0
DHCA ACP Baseline
DHCA ACP 12 weeks
DHCA ACP 52 weeks
DHCA ACP Baseline
DHCA ACP 12 weeks
78 59
54
59
54
41
DHCA ACP 52 weeks
DHCA ACP Baseline
DHCA ACP 12 weeks
DHCA ACP 52 weeks
Figure 3: Cambridge Neuropsychological Test Automated Battery system results DHCA=deep hypothermic circulatory arrest. ACP=antegrade cerebral perfusion. For box and whisker plots, the box range is IQR, the horizontal line within the box is the median, the whiskers are the range of the measurements with the exception of outliers, which are plotted as points. The bar chart plots the percentage of patients scoring the maximum 24 on the verbal recognition memory test.
Deep hypothermic circulatory arrest (n=35)
Antegrade cerebral perfusion (n=39)*
p value
Surgery details Total time in operating room (min)
463·6 (54·8)
453·3 (53·2)
0·41
Duration of cardiopulmonary bypass (min)
325·7 (49·4)
326·4 (49·2)
0·95
Duration of total myocardial ischaemic time (min)
51·5 (29·8)
0·09
Duration of surgery on right pulmonary artery with ACP (min)
62·5 (24·8) ··
21·8 (18·4)
··
Duration of surgery on left pulmonary artery with ACP (min)
··
17·1 (13·3)
··
Duration of surgery on right pulmonary artery with DHCA (min)
19·7 (8·6)
3·1 (7·0)†
··
Duration of surgery on left pulmonary artery with DHCA (min)
16·3 (8·4)
3·2 (7·2)†
··
Cerebral saturation values‡ Right-sided cerebral saturation at baseline
77·3 (12·1)
73·6 (10·5)
0·19
Right-sided cerebral saturation, lowest value
37·4 (9·0)
53·6 (12·1)
<0·001
Right-sided cerebral saturation, after sternal closure
66·6 (13·0)
65·4 (9·0)
0·65
Left-sided cerebral saturation, baseline
76·4 (12·3)
77·1 (8·6)
0·77
Left-sided cerebral saturation, lowest value
37·8 (9·1)
54·2 (13·9)
<0·0001
Left-sided cerebral saturation, after sternal closure
67·7 (12·0)
67·8 (10·3)
0·97
0·25§
Pulmonary disease type (Jamieson classification) Right pulmonary artery disease 1
8 (23%)
15 (38%)
2
22 (63%)
16 (41%)
3
4 (11%)
7 (18%)
4
1 (3%)
1 (3%)
Left pulmonary artery disease 1
6 (17%)
8 (21%)
2
22 (63%)
20 (51%)
3
5 (14%)
8 (21%)
4
2 (6%)
3 (8%)
0·81§
Intensive care data Duration of mechanical ventilation in intensive care (h)¶|| Duration of stay in intensive care (days)
20 (16–28)
19 (16–22)
0·81
3 (2–5)
3 (2–5)
0·88 (Continues on next page)
www.thelancet.com Vol 378 October 15, 2011
1383
Articles
Deep hypothermic circulatory arrest (n=35)
Antegrade cerebral perfusion (n=39)*
p value
(Continued from previous page) Duration of stay in hospital (days)
14 (10–21)
15 (11–22)
0·35
46·3 (11·7)
48·2 (14·5)
0·56
619·8 (236·3)
642·1 (309·1)
0·73
0·01§
Perioperative haemodynamic data Mean pulmonary artery pressure before cardiopulmonary bypass (mm Hg) Mean pulmonary vascular resistance (assuming pulmonary capillary wedge pressure of 10 mm Hg) before cardiopulmonary bypass (dynes/s per cm⁵)|| Inotropic support greater than dopamine at 5 μg/kg per min Yes
0 (0%)
7 (18%)
No
35 (100%)
32 (82%)
Mean pulmonary artery pressure on first postoperative day (mm Hg) Mean pulmonary vascular resistance (assuming pulmonary capillary wedge pressure of 10 mm Hg) on first postoperative day (dynes/s per cm⁵)|| Mean pulmonary artery pressure at week 12 (mm Hg)** Mean pulmonary vascular resistance (assuming pulmonary capillary wedge pressure of 10 mm Hg) at week 12 (dynes/s per cm⁵)**
20·6 (4·7) 223 (127) 28·1 (9·2) 316 (197)
23·4 (7·6) 260 (187) 27·8 (10·6) 338 (316)
0·10 0·40 0·70 0·37
Camphor results†† Symptoms score at baseline
14 (11–18)
14 (9–19)
0·91
Symptoms score at 12 weeks
4 (1–9)
2 (1–7)
0·30
Symptoms score at 52 weeks
4 (1–11)
1 (0–3)
0·03
12 (7–17)
13 (7–17)
0·58
3 (1–10)
0·23
Activity score at baseline Activity score at 12 weeks
7 (2–10)
Activity score at 52 weeks
5 (1–12)
1·5 (0–5)
0·09
11 (6–18)
10 (8–17)
0·93
Quality-of-life score at 12 weeks
3 (1–9)
4 (1–9)
0·97
Quality-of-life score at 52 weeks
4 (0–9)
1 (0–6)
0·26
Quality-of-life score at baseline
Walk distance 6-min walk distance (m) at 12 weeks§§
348 (126)
371 (104)
0·31
6-min walk distance (m) at 52 weeks¶¶
373 (113)
410 (120)
0·14
New York Heart Association classification17 Number in class at 12 weeks
1·00
1
21
23
··
2
10
12
··
3
1
2
··
4
1
0
··
1
17
23
2
13
11
··
3
0
1
··
4
0
0
··
Number in class at 52 weeks
0·61 ··
Data are mean (SD), n (%), or median (IQR) unless otherwise stated. *One patient with missing data—data for antegrade cerebral perfusion (ACP) are therefore for maximum 38 patients. †Nine patients randomly allocated to ACP also received deep hypothermic circulatory arrest (DHCA). The mean (SD) DHCA time for those nine patients was 13·3 min (8·7) for the right side and 13·7 min (9·2) for the left side. ‡Some cerebral saturation information is missing because the pads stopped recording. Records were obtained for 30 patients in ACP group and 36 patients in DHCA group (baseline), 33 patients in ACP group and 37 patients in DHCA group (lowest value), 30 patients in ACP group and 33 patients in DHCA group (after sternal closure) patients. §Fisher’s exact test. ¶One patient (in ACP group) had long intubation time (792 h) and had to be transferred to a neurological unit. One patient (in DHCA group) had missing information because he was transfered to another hospital—this patient was lost to follow-up at 3 months because of brain injury. Data from two patients missing from the ACP group because one was withdrawn by the investigators and one died. After the first extubation after surgery, six patients had to be intubated again (four from the DHCA group and two from the ACP group; these were also the cooling techniques that the patient actually received). The four patients in the DHCA group were intubated 1, 5, 11, and 17 days after the extubation date, respectively. The two patients in the ACP group were intubated 3 days and 15 days after extubation, respectively. ||Some haemodynamic information is missing because of pulmonary artery catheter problems—records were obtained for 35/37 and 35/36 patients (ACP/DHCA; n). **Some patients refused the procedure at 12 weeks— records were obtained for 32 patients in the ACP group and 37 patients in the DHCA group. ††Results were obtained for 35/39 at baseline, 33/37 at 12 weeks, 27/26 at 52 weeks (activity score), and 27/25 at 52 weeks (quality of life; ACP/DHCA; n). §§Data missing for one patient (in DHCA group) because they had a cerebral vascular incident and for one patient (in ACP group) because of a baseline mobility problem. ¶¶Data missing for one patient (in DHCA group) because they had a cerebral vascular incident, and for two patients (in ACP group) because of a baseline mobility problem and because follow-up was done by post.
Table 2: Operative data and secondary outcome measures
1384
www.thelancet.com Vol 378 October 15, 2011
Articles
Deep hypothermic Antegrade circulatory arrest cerebral perfusion Serious adverse events Arrhythmia
0
1
Central vascular accident or transient ischaemic attacks
2
1
Death
2
1
Pericardial effusions
3
1
Pulmonary infection
1
0
Re-exploration for postoperative bleeding
1
0 0
Acute renal failure
1
Lung reperfusion injury
1
3
Right heart failure
0
2
Haemoglobin less than 8 g/dL 0
1
Others
2
2
Neurological adverse events (mild or moderate) Confusion or mood alteration 5
4
Transient ischaemic attack
1
1
Dizziness
2
1
Headache
3
0
Tremor
4
5
Sensory disturbances
2
2
Table 3: Adverse events
baseline, meaning that the adjusted differences were not significantly different (odds ratio 1·2, 95% CIs 0·29–4·30; p=0·87). Haemodynamic improvement after surgery was much the same between the two groups. We recorded no differences in the results of the 6-min walk distance test, New York Heart Association class,17 duration of stay, or CAMPHOR scores at either 12 weeks or 52 weeks (table 2). We recorded 78 adverse events in 29 patients in the DHCA group and 105 events in 32 patients in the ACP group over 52 weeks. The relative risk for individuals in the ACP group compared with those in the DHCA group was 1·22 (95% CI 0·91–1·64; p=0·18). We recorded no difference between groups in the number of serious adverse events—ie, events that were life-threatening, required admission to hospital as an inpatient, extended the length of an individual’s stay in hospital, or resulted in death or persistent or substantial disability. We recorded 13 serious adverse events (which led to death or an extended stay in hospital) in the DHCA group and 12 in the ACP group (table 3).
Discussion Our findings show no benefit in the maintenance of brain perfusion over temporary periods of DHCA of up to 20 min at 20°C (total cumulative time 36 min) in patients undergoing PEA. Our study shows that although perfusion of the brain might be safe, no evidence exists that it is better than DHCA, because we www.thelancet.com Vol 378 October 15, 2011
did not detect any difference between the groups in cognitive function measurements. Perfusion of the brain is a laudable goal, but should not detract from the importance of surgical success. This study lends credence to previous concerns that the avoidance of DHCA might lead to suboptimum endarterectomy,18 because nine patients (23%) crossed over from ACP to DHCA to allow complete endarterectomy. We recorded equivalent clinical outcomes in both groups, and did not detect any advantages with the use of ACP—the more technically demanding technique. DHCA at 20°C should be the standard method for PEA surgery. Postoperative cognitive dysfunction is a decrease in several neuropsychological domains such as memory, executive functioning, and speed of processing. It was assessed with the tests recommended in an international consensus statement16 and a newer computer-based method (CANTAB).14 The frequency with which postoperative cognitive dysfunction occurs is unclear19 and might not be as prevalent after cardiac surgery as has previously been reported.20,21 The two groups of patients enrolled in this study were much the same in regards to preoperative or intraoperative precipitating factors22,23 and a postoperative decrease in cognitive function would have been expected in both groups. This absence of difference is surprising because the groups had substantially different patterns of cerebral oxygen saturation during surgery. The use of near-infrared spectroscopy is increasing in patients undergoing cardiac surgery, despite the scarcity of objective evidence of benefit and the absence of defined safe saturation thresholds at different temperatures. The average cognitive function test score improved in both groups at 12 weeks and 52 weeks. We suggest that the postoperative improvement in cardiac output, combined with an improvement in quality of life, might partly explain the overall improvement in cognitive function. Induction of hypothermia and DHCA is the standard technique to reduce metabolic activity and the risk of ischaemic cerebral damage during aortic and other types of cardiac surgery when interruption of blood flow to the brain is necessary. The inverse relation between temperature and neurological complications has led to the idea of a safe duration of total circulatory arrest.24 By comparison with DHCA, ACP should be better at maintaining intracranial hypothermia, providing oxygen and metabolite substrates, and removing toxic metabolites. These arguments have been used to justify the development of alternative methods such as antegrade and retrograde perfusion,25 but debate about their relative merits is ongoing and opinion is polarised.26 We know of no other randomised controlled trial in adult patients to address this issue (panel). In this study, ACP proved unsuitable for nine patients who eventually needed DHCA. However, we recorded no difference in outcomes when comparing patients who had DHCA with those who had not. 1385
Articles
Panel: Research in context Systematic review We know of no other prospective controlled trial to compare the cognitive outcome in adult patients randomly allocated to receive either deep hypothermic circulatory arrest (DHCA) or maintained brain perfusion. A Medline search with search term “DHCA” and “deep hypothermic circulatory arrest” produced no previous reports of randomised controlled trial comparing DHCA with brain perfusion in adult patients (the search had no date or language restrictions and the last search was done on Sept 10, 2011). The findings of a comprehensive review27 from 2010 confirmed that no previous trials had been done—we also assessed referenced studies in this review. Interpretation In this study to assess cognitive function in patients undergoing pulmonary endarterectomy, we report that cognitive function is not impaired and actually improves, whether or not the brain was perfused. We recorded no cognitive or other outcome benefits from the maintenance of cerebral perfusion compared with periods of DHCA of up to 20 min (cumulative 36 min) at 20°C. Our findings suggest that circulatory arrest is the optimum modality for patients undergoing pulmonary endarterectomy surgery.
This study could inform the debate in patients undergoing aortic surgery because we have conclusively proved that a continuous period of up to 20 min of DHCA at 20°C (total cumulative time of 36 min) is well tolerated and that no benefit exists in the maintenance of cerebral perfusion in similar conditions. The main limitation of this study is the small size and the risk of a type II error, although the study was powered for the primary outcome measure of cognitive function change at 12 weeks and the CIs were all within one SD, which was specified before the trial began as the minimum clinically important difference. Results might vary with the selection of different cognitive function tests.20 By using the CANTAB battery of tests, we investigated cognitive domains repeatedly suggested as important— learning and memory, attention, and executive functioning.28 The advantage of CANTAB is that it explores performance and is less dependent on the tester, although in this study the same person did all assessments.22 In the setting of PEA, maintenance of cerebral perfusion throughout the procedure adds to the complexity of an operation and confers no additional benefits. Whether these findings can be extrapolated for other types of surgery in which a choice exists between the use of DHCA or selective ACP is not known. Contributors AV and DJ had the idea for the study, supervised research staff, masked or unmasked assessors, led the writing of the paper. DJ, LS, GC, JK, ST, JD, JA, and DJ designed the protocol. AV, LS, and DJ applied for funding. AV, LS, JK, ST, JD, EW, AK, JA, RH, and DJ interpreted the results. GC,
1386
JK, ST, JD, and DJ recruited patients. LS and GC analysed the data. GC, JK, ST, JD, AK, JA, RH, and DJ collected the data. LS, GC, JK, ST, JD, EW, AK, JA, and RH reviewed the manuscript. LS and EW compiled the statistical report. AV is a member of the steering group. GC did the cognitive tests and coordinated the study. JK coordinated the theatre staff. EW analysed cognitive data. Conflicts of interest AK has received honorarium from Covidien, owner of Somanetics for lectures on the use of near-infrared spectroscopy. All other authors declare that they have no conflicts of interest. Acknowledgments Johanna Armstrong helped with writing the protocol and obtaining funding. Ella Wheaton did statistical analysis during a Clinical Trial Fellowship located within the Medical Research Council Biostatistics Clinical Trials Methodology Hub and funded by the National Institute of Health Research. Ray Latimer was involved in discussing the original protocol and supported the need for a study. We thank Stephanie Cadour for her contribution to the analysis of the CANTAB test results. We thank all staff at Papworth Hospital and the regional designated pulmonary hypertension referral hospitals for their care of the study patients. References 1 Petch CP. Cor pulmonale from recurrent pulmonary embolism. Lancet 1951; 1: 1346–47. 2 Pengo V, Lensing AW, Prins MH, et al. Incidence of chronic thromboembolic pulmonary hypertension after pulmonary embolism. N Engl J Med 2004; 350: 2257–64. 3 Piazza G, Goldhaber SZ. Chronic thromboembolic pulmonary hypertension. N Engl J Med 2011; 364: 351–60. 4 Condliffe R, Kiely DG, Gibbs JS, et al. Improved outcomes in medically and surgically treated chronic thromboembolic pulmonary hypertension. Am J Respir Crit Care Med 2008; 177: 1122–27. 5 Freed DH, Thomson BM, Berman M, et al. Survival after pulmonary thromboendarterectomy: effect of residual pulmonary hypertension. J Thorac Cardiovasc Surg 2010; 141: 383–87. 6 Mayer E, Jenkins D, Lindner J, et al. Surgical management and outcome of patients with chronic thromboembolic pulmonary hypertension: results from an international prospective registry. J Thorac Cardiovasc Surg 2011; 141: 702–10. 7 Jamieson SW, Kapelanski DP, Sakakibara N, et al. Pulmonary endarterectomy: experience and lessons learned in 1,500 cases. Ann Thorac Surg 2003; 76: 1457–62. 8 Surie S, Tijssen MA, Biervliet JD, et al. Chorea in adults following pulmonary endarterectomy. Mov Disord 2010; 25: 1101–04. 9 Macchiarini P, Kamiya H, Hagl C, et al. Pulmonary endarterectomy for chronic thromboembolic pulmonary hypertension: is deep hypothermia required? Eur J Cardiothorac Surg 2006; 30: 237–41. 10 Mikus PM, Mikus E, Martin-Suarez S, et al. Pulmonary endarterectomy: an alternative to circulatory arrest and deep hypothermia: mid-term results. Eur J Cardiothorac Surg 2008; 34: 159–63. 11 Thomson B, Tsui SS, Dunning J, et al. Pulmonary endarterectomy is possible and effective without the use of complete circulatory arrest—the UK experience in over 150 patients. Eur J Cardiothorac Surg 2008; 33: 157–63. 12 Thistlethwaite PA, Kaneko K, Madani MM, Jamieson SW. Technique and outcomes of pulmonary endarterectomy surgery. Ann Thorac Cardiovasc Surg 2008; 14: 274–82. 13 Jamieson SW, Kapelanski DP. Pulmonary endarterectomy. Curr Probl Surg 2000; 37: 165–252. 14 Sahakian BJ, Owen AM. Computerized assessment in neuropsychiatry using CANTAB: discussion paper. J R Soc Med 1992; 85: 399–402. 15 McKenna SP, Doughty N, Meads DM, Doward LC, Pepke-Zaba J. The Cambridge Pulmonary Hypertension Outcome Review (CAMPHOR): a measure of health-related quality of life and quality of life for patients with pulmonary hypertension. Qual Life Res 2006; 15: 103–15. 16 Murkin JM, Newman SP, Stump DA, Blumenthal JA. Statement of consensus on assessment of neurobehavioral outcomes after cardiac surgery. Ann Thorac Surg 1995; 59: 1289–95.
www.thelancet.com Vol 378 October 15, 2011
Articles
17
18 19 20
21
22
The Criteria Committee of the New York Heart Association. Nomenclature and criteria for diagnosis of diseases of the heart and great vessels (9th edn). Boston, MA: Little, Brown and Co, 1994: 253–56. Jamieson SW, Madani M. Invited commentary. Ann Thorac Surg 2008; 86: 1267. Selwood A, Orrell M. Long term cognitive dysfunction in older people after non-cardiac surgery. BMJ 2004; 328: 120–21. Rudolph JL, Schreiber KA, Culley DJ, et al. Measurement of post-operative cognitive dysfunction after cardiac surgery: a systematic review. Acta Anaesthesiol Scand 2010; 54: 663–77. Newman MF, Kirchner JL, Phillips-Bute B, et al. Longitudinal assessment of neurocognitive function after coronary-artery bypass surgery. N Engl J Med 2001; 344: 395–402. Funder KS, Steinmetz J, Rasmussen LS. Methodological issues of postoperative cognitive dysfunction research. Semin Cardiothorac Vasc Anesth 2010; 14: 119–22.
www.thelancet.com Vol 378 October 15, 2011
23
24
25 26 27
28
Moller JT, Cluitmans P, Rasmussen LS, et al. Long-term postoperative cognitive dysfunction in the elderly ISPOCD1 study: ISPOCD investigators—International Study of Post-Operative Cognitive Dysfunction. Lancet 1998; 351: 857–61. Kirklin JW, Baratt-Boyes BG. Hypothermia, circulatory arrest, and cardiopulmonary bypass. In: Kirklin JW, Barratt-Boyes BG, eds. Cardiac Surgery, 2nd edn. New York: Churchill Livingstone, 1993: 61–127. Bonser RS, Harrington DK. Editorial comment. Eur J Cardiothorac Surg 2005; 28: 102–03. Elefteriades JA. Editorial comment: what is the best method for brain protection in surgery of the aortic arch? Cardiol Clin 2010; 28: 403–04. Elefteriades JA. What is the best method for brain protection in surgery of the aortic arch? Straight DHCA. Cardiol Clin 2010; 28: 381–87. Bekker AY, Weeks EJ. Cognitive function after anaesthesia in the elderly. Best Pract Res Clin Anaesthesiol 2003; 17: 259–72.
1387