- Email: [email protected]
Impact of Cerebral Perfusion on Outcomes of Aortic Surgery: The Society of Thoracic Surgeons Adult Cardiac Surgery Database Analysis
Journal Pre-proof Impact of Cerebral Perfusion on Outcomes of Aortic Surgery: STS Adult Cardiac Database Analysis Shinobu Itagaki, MSc MD, Jo Chikwe, MD FRCS, Eric Sun, MD, Danny Chu, MD, Nana Toyoda, MD PhD, Natalia Egorova, PhD PII:
S0003-4975(19)31433-X
DOI:
https://doi.org/10.1016/j.athoracsur.2019.08.043
Reference:
ATS 33066
To appear in:
The Annals of Thoracic Surgery
Received Date: 14 April 2019 Revised Date:
17 July 2019
Accepted Date: 12 August 2019
Please cite this article as: Itagaki S, Chikwe J, Sun E, Chu D, Toyoda N, Egorova N, Impact of Cerebral Perfusion on Outcomes of Aortic Surgery: STS Adult Cardiac Database Analysis, The Annals of Thoracic Surgery (2019), doi: https://doi.org/10.1016/j.athoracsur.2019.08.043. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. 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. © 2019 by The Society of Thoracic Surgeons
Impact of Cerebral Perfusion on Outcomes of Aortic Surgery: STS Adult Cardiac Database Analysis Running head: Circulatory arrest cerebral protection
Shinobu Itagaki MSc MD,1 Jo Chikwe MD FRCS,1,2 Eric Sun MD,3 Danny Chu MD,4 Nana Toyoda MD PhD,1 Natalia Egorova PhD2;
1
Department of Cardiovascular Surgery, The Icahn School of Medicine at Mount Sinai, NY
2
Division of Cardiothoracic Surgery, The State University of New York, Stony Brook, NY
3
Department of Health Science and Policy, The Icahn School of Medicine at Mount Sinai, NY
4
Department of Cardiothoracic Surgery, University of Pittsburgh School of Medicine, PA
Corresponding author:
Joanna Chikwe MD
Correspondence:
Department of Cardiovascular Surgery, Mount Sinai Medical Center, 1190 Fifth Avenue, New York, NY 10029 USA Email: [email protected]
Abstract Background: Limited data informs cerebral protection during circulatory arrest: this study was designed to identify optimal approaches from a national clinical registry. Methods: A total of 7830 adults (mean age 63.1, standard deviation 13.1 years) who underwent hemi-arch (n=6891, 88.0%) or total arch (n=939, 12.0%) replacement with hypothermic circulatory arrest between 2014-2016, were identified from The Society of Thoracic Surgeons Database (version 2.81). Aortic dissections were excluded from the analysis. Multivariable logistic regression was used to adjust for 29 baseline and operative variables including demographics, comorbidity, surgery, and nadir temperature, comparing outcomes according to protection strategy. The primary end-point was a composite of 30-day and in-hospital mortality or major permanent neurological complications. Results: The rate of death or permanent neurological complication was 10.9% (n=850). Antegrade cerebral perfusion was most utilized (n=3369, 43%, median nadir temperature 23oC, median arrest time 30 minutes) compared to retrograde (n=1898, 24%, 20 oC, 24 minutes) and no cerebral perfusion (n=2563, 33%, 20 oC, 22 minutes). In multivariable analysis, deep hypothermia with antegrade (OR 0.65, 95% CI 0.52- 0.81) or retrograde (OR 0.57, 95% CI 0.45- 0.71) perfusion, and moderate hypothermia with antegrade perfusion (OR 0.61, 95% CI 0.46-0-.79) were associated with significant reductions in death and stroke compared to deep hypothermia without cerebral perfusion. Risk reduction was greatest in circulatory arrest >30 minutes. Conclusions: For patients without dissection and requiring >30 minutes circulatory arrest, optimal cerebral protection strategies are deep hypothermia with either antegrade or retrograde cerebral perfusion; or moderate hypothermia with antegrade cerebral perfusion. Words 250 (250)
Hypothermic circulatory arrest has been a standard component of aortic arch surgery, providing an extended time to operate in a bloodless field by reducing brain and visceral metabolic demand, since the technique was developed by R. Griepp et al in the mid-1970’s[1,2]. Hypothermic circulatory arrest is increasingly utilized in combination with retrograde cerebral perfusion, or with antegrade cerebral perfusion which may also be conducted with warmer circulatory arrest temperatures[3]. The primary rationale for antegrade cerebral perfusion is to minimize the adverse neurological sequalae of total circulatory arrest, extending the safe duration of circulatory arrest at warmer temperatures, and reducing the total time spent cooling and rewarming on cardiopulmonary bypass with the aim of minimizing coagulopathy and organ dysfunction[3]. In the absence of a robust evidence-base and consensus guideline recommendations, contemporary ascending aorta and arch surgery is characterized by wide variation in practice, particularly cerebral protection strategy, and clinical outcomes[2–5]. The current study was therefore designed to determine the optimal strategy for cerebral protection during elective ascending aorta and arch surgery from a national clinical registry.
Material and Methods Patient Population A retrospective cohort analysis was conducted on outcomes of 18,180 adult patients aged 18 years or older who underwent aortic arch or hemi-arch/ascending aortic replacement under circulatory arrest in the United States between 2014 and 2016. Patients were identified using the Society of Thoracic Surgeons Adult Cardiac Surgery Database (version 2.81) and stratified according to the adjunctive cerebral perfusion strategy utilized: either straight hypothermic circulatory arrest without cerebral perfusion, hypothermic circulatory arrest with antegrade cerebral perfusion, or hypothermic circulatory arrest with retrograde cerebral perfusion.
After excluding patients who presented with aortic dissections (43.1%, n=5941), had salvage surgery or who required cardiopulmonary resuscitation immediately prior to operation (2.3%, n=410), patients in whom both retrograde and antegrade cerebral perfusion were used, (2.9%, n=524), patients with a nadir body temperature less than 10oC (4.4%, n=805) or over 28oC during circulatory arrest (6.2%, n=1124), and patients with a total circulatory arrest time (with or without the adjunctive cerebral perfusion strategy) less than 10 minutes (11.8%, n=2114) or more than 150 minutes (0.6%, n=114), a total of 7830 patients remained who comprised the study cohort. Baseline comorbidities, surgical details, and operative outcomes were identified using variables provided in the Society of Thoracic Surgeons Adult Cardiac Database (Version 2.81).
This study was approved by the Society of Thoracic Surgeons, and the institutional review boards of the Icahn School of Medicine at Mount Sinai and The State University of New York at Stony Brook. The approval included a waiver of informed consent. The data for this research were provided by The Society of Thoracic Surgeons’ national Database participant User File Research program. Data analysis was performed at the investigators’ institutions.
Primary and Secondary Outcomes The primary outcomes were composite endpoints of in-hospital or 30-day mortality and irreversible major neurological complication, (permanent stroke, permanent paralysis or permanent encephalopathy due to intracranial bleed, emboli, or anoxic brain injury). Mortality or major neurological complications occurring at any time during the index admission, including after 30-days post-operatively was reported. Permanent stroke was defined according to Society of Thoracic Surgeons criteria as any confirmed neurological deficit of abrupt onset caused by a disturbance in blood supply to the brain that did not resolve within 24 hours. Transient ischemic attacks and reversible encephalopathy were not included in
the primary outcome. Secondary outcomes were the individual outcomes of in-hospital or 30-day mortality, irreversible neurological complications, and postoperative complications including prolonged ventilation, renal failure and length of hospital stay.
Statistical Analysis For descriptive analysis, continuous variables were reported as mean with standard deviation or median and interquartile range and categorical variables were reported as proportions. The Analysis of Variance or Kruskal-Wallis for continuous variables and Chi-squared test were used as appropriate to compare demographics of cerebral perfusion strategy groups. Multivariable logistic regression models were used to evaluate adjusted risk of each endpoint, controlling for the cerebral perfusion strategy (with straight hypothermic circulatory arrest without cerebral perfusion as the reference), age, gender, race, baseline comorbidities (including diabetes, dyslipidemia, dialysis, hypertension, lung disease, peripheral vascular disease, cerebrovascular accident, carotid disease, previous coronary artery bypass grafting, previous valve surgery, previous percutaneous coronary intervention, myocardial infarction, prior congestive heart failure), the mode of presentation (including cardiogenic shock, and surgical urgency status), and the surgical details (including total arch vs hemi-arch replacement, aortic cannulation sites, total circulatory arrest time, lowest body temperature and its measurement source). The total circulatory arrest time was incorporated in the model as a continuous variable first, and then as a categorical variable with 10-minute increment (10-<20 minutes, 20-<30 minutes, 30-<40minutes, 40-<50 minutes, and >/=50minutes). The nadir body temperature was incorporated in the model as a continuous variable, and then as a categorical variable (10-<20oC, 20-<24oC, and 24-<28oC), in line with the consensus definitions of moderate and deep hypothermia2. To validate the findings, sensitivity analyses were conducted using clinically relevant subgroups with multiple strata such as total arch vs non-total arch, different ranges of total circulatory arrest time, and different perfusion and temperature strategies. The main findings remained largely the
same and are provided in the supplemental data. All tests were two-tailed and an alpha level of 0.05 was considered statistically significant. All statistical analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC).
Results Study Population A total of 7830 patients underwent either total arch aortic replacement (n=939, 12.0%) or ascending aortic with hemi-arch replacement (n=6891, 88.0%) between 2014 and 2016. The mean age was 63.1 years and 65.3% of the patients were male. A total of 72.1%% (n=5646) of the patients underwent elective surgery. Baseline patient characteristics are summarized in Table 1.
Operative technique Operative approaches are summarized in Table 2. Of the 6891 patients who underwent hemi-arch or ascending aorta replacement with circulatory arrest, 39.5% (n=2721) received antegrade cerebral perfusion, 34.7% (n=2390) of patients underwent straight hypothermic circulatory arrest without cerebral perfusion, and 25.8% (n=1780) patients received retrograde cerebral perfusion (p<0.001). Overall, 12% (n=939) of patients underwent total arch replacement. Of these patients 69.0% (n=648) received antegrade cerebral perfusion, 18.4% (n=173) patients underwent straight hypothermic circulatory arrest without cerebral perfusion, and 12.6% (n=118) patients received retrograde cerebral perfusion (p<0.001). Overall, 30% (n=2342) of the patients had axillary or innominate arterial cannulation, which was more likely to be used with antegrade cerebral perfusion (56.1%, n=1891) compared to 12.4% (n=318) of patients undergoing straight hypothermic circulatory arrest without cerebral perfusion and 7.1% (n=134) undergoing retrograde cerebral perfusion (p<0.001). Femoral cannulation, used in 10.2% (n=800) of
patients overall, was the least utilized technique in patients who received antegrade cerebral perfusion (4.6%, n=156), compared to 17.2% (n=440) of patients undergoing straight hypothermic circulatory arrest without cerebral perfusion, and 10.8% (n=204) patients undergoing retrograde cerebral perfusion (p<0.001).
Antegrade cerebral perfusion was associated with the longest total circulatory arrest times (median 30 minutes [interquartile range (IQR) 18-37]) and the highest nadir body temperature (median 23oC [IQR 2026 oC]). Straight hypothermic circulatory arrest without cerebral perfusion was associated with the shortest circulatory arrest times (22 minutes [IQR 15-25] minutes) and the lowest nadir temperature (20 o
C [IQR 18-22]oC). The distributions of the total circulatory arrest time and the nadir body temperature in
each cerebral perfusion strategy are shown in Supplemental Figures 1 and 2, respectively. Overall, the temperature source was nasopharyngeal or tympanic in only 15.9% (n=1244) patients, with bladder or rectal temperature used in 56.6% (n=4435).
Mortality and Neurological Complications The primary and secondary outcomes are summarized in Table 3. The rate of death, permanent stroke or paraplegia was 10.9% (n=850). The rate of death, stroke or paraplegia was 14% (n=349) in patients who underwent straight hypothermic circulatory arrest without cerebral perfusion, 11% (n=355) in patients undergoing antegrade cerebral perfusion and 8% (n=146) in patients who underwent retrograde cerebral perfusion, respectively (p<0.001). Permanent neurological complications were observed in 8% (n= 195) of patients who underwent straight hypothermic circulatory arrest without cerebral perfusion), 6% (n=210) of patients undergoing antegrade cerebral perfusion and 5% (n=87) of patients after retrograde cerebral perfusion, respectively (p<0.001). In-hospital or 30-day mortality occurred in 9% (n=217), 6% (n=207), and 4% (n=83) respectively.
The relationship between unadjusted rates of primary outcomes and total circulatory arrest time are shown in Figure 1, and then stratified by cerebral perfusion strategy in Figure 2. In multivariable logistic regression, longer total arrest times were strongly associated with increasing risk of death or stroke regardless of cerebral perfusion strategy, but this effect was greatest in patients undergoing straight hypothermic circulatory arrest without cerebral perfusion (adjusted OR 1.33, 95% CI 1.19-1.49 per 10 minutes increment), followed by retrograde cerebral perfusion (adjusted OR 1.23, 95% CI 1.06-1.42), and least with antegrade cerebral perfusion (adjusted OR 1.16, 95% CI 1.07-1.25). Use of antegrade (adjusted OR 0.63, 95% CI 0.51-0.77), and retrograde cerebral perfusion (adjusted OR 0.56, 95% CI 0.45-0.70) were both associated with significant reductions in the risk of the primary endpoint of death or permanent neurological deficit compared with no cerebral perfusion.
The protective effect of the adjunctive cerebral perfusion strategy became significant starting in the subgroup of the circulatory arrest time of 20 minutes and longer, and was observed throughout the duration of circulatory arrest times with increasing protective effect with longer circulatory arrest times (Figure 2). The increased risk of death or stroke associated with increasing circulatory arrest times was most marked in patients after more than 20 minutes straight hypothermic circulatory arrest without cerebral perfusion circulatory arrest while for patients receiving either antegrade or retrograde cerebral perfusion, arrest time did not become a significant predictor of death or stroke until after 50 minutes of arrest time (Figure 3).
Nadir temperature was not associated with a significant difference in risk of primary endpoint overall, or in subgroup analysis. In multivariable analysis, deep hypothermia with antegrade (OR 0.65, 95% CI 0.520.81) or retrograde (OR 0.57, 95% CI 0.45- 0.71) perfusion, and moderate hypothermia with antegrade
perfusion (OR 0.61, 95% CI 0.46-0-.79) were associated with significant reductions in death and stroke compared to deep hypothermia without cerebral perfusion. Supplemental Data provides the results of subgroup analysis confirming these trends in patients who underwent total arch versus ascending aorta or hemi-arch replacement. In a subgroup analysis of patients where the cannulation site was specified (n=6482), femoral cannulation was associated with increased risk of the primary outcome compared to ascending aortic cannulation (adjusted OR 1.67, 95% CI 1.32-2.12), whereas the axillary cannulation was not.
Comment This analysis demonstrates firstly that contemporary elective ascending and aortic arch surgery in the United States remains associated with relatively high mortality and major neurological morbidity, and wide variation in practice and outcomes. Secondly, we confirm that in patients needing circulatory arrest longer than 30 minutes, deep hypothermia with antegrade or retrograde cerebral perfusion, or moderate hypothermia with antegrade cerebral perfusion appear safer strategies compared to deep hypothermic circulatory arrest without cerebral perfusion. Thirdly, we provide new insights into the interaction between increasing duration of circulatory arrest and cerebral protection strategy: longer circulatory arrest times were associated with increasing risk of mortality and stroke regardless of the cerebral protection strategy, but the negative effects of the longest circulatory arrest times are most effectively mitigated by adjunctive cerebral perfusion.
In this national registry, the operative mortality for patients undergoing elective ascending aorta or hemiarch surgery was 5.9%, and 9.9% after arch replacement, which compares to 2-5% achieved in single-centers series[6–15]. Poorer outcomes have been associated with low volume providers, which represent the majority of centers in the United States, where over 75% of the centers reporting to the STS
between 2010 and 2014 performed fewer than 6 aortic cases per year [16]. However, wide variation in cerebral protection strategy was observed even in high volume centers, suggesting that factors beyond choice of cerebral protection strategy, such as surgical expertise and team experience, contribute to improved outcomes[16]. However, in unselected contemporary clinical practice, our analysis showed that for most patients, either antegrade or retrograde cerebral perfusion was associated with superior outcomes compared to straight hypothermic circulatory arrest without cerebral perfusion.
Straight hypothermic circulatory arrest without cerebral perfusion has been the predominant cerebral protection strategy for aortic arch surgery since its introduction in the mid 1970’s by R. Griepp et al. In a recent analysis of aortic surgery in North America between 2011-2014, straight hypothermic circulatory arrest was the most predominant method of brain protection utilized in 35% of cases[2]. In our more recent study this proportion had dropped slightly to 32%. Patients operated on at temperatures below 20oC can tolerate 20-30 minutes of circulatory arrest which is sufficient for hemi-arch repair, although as few as 25% of these patients have been shown to achieve electrocerebral silence[17]. The extended cardiopulmonary bypass times required to achieve complete cooling and rewarming, and lower nadir temperatures, contribute to enzyme and end-organ dysfunction, and coagulopathy: which may have contributed to the shift observed in our analysis towards adjunctive strategies that allow extended circulatory arrest times at warmer temperatures.
Hypothermic circulatory arrest with retrograde cerebral perfusion, first described by Ueda et al in 1990, was used in 24% of patients in our analysis, for a median arrest time of 24 minutes with a median temperature nadir of 20oC[6]. The hypothesis that retrograde cerebral perfusion might safely extend circulatory arrest times by supplying oxygenated blood to cerebral territories has not been supported by porcine experiments indicating <0.1% of retrograde perfusate from the superior vena cava traverses
cerebral capillary beds[18]. Hypothermic circulatory arrest with antegrade cerebral perfusion has increasingly become the preferred protection strategy for complex arch surgery as the brain can be perfused during circulatory arrest, allowing longer safe circulatory arrest times and warmer nadir temperatures. The embolic risk of cannulating the innominate and carotid arteries may be partially mitigated by using axillary artery cannulation, which was used in 30% of patients overall, and in 56% of those undergoing antegrade cerebral perfusion in this analysis.
The novel finding of this study is that longer total arrest times were strongly associated with increasing risk of death or stroke regardless of cerebral perfusion strategy, but this effect was greatest in patients undergoing straight hypothermic circulatory arrest without cerebral perfusion (33% increased risk for every additional 10 minutes circulatory arrest), followed by retrograde cerebral perfusion (23% increased risk for every additional 10 minutes circulatory arrest), and least with antegrade cerebral perfusion (16% increased risk for every additional 10 minutes circulatory arrest). Cerebral perfusion (either antegrade or retrograde with deep or, antegrade with moderate hypothermia) appeared to mitigate adverse outcomes associated with circulatory arrest greater than 30 minutes significantly more effectively than the use of straight deep hypothermia alone. This threshold has not previously been reported, including in an earlier analysis of the STS database which stratified cerebral protection strategies by nadir temperature, rather than by duration of circulatory arrest. Our data supports the guideline recommendations for routine use of cerebral perfusion during hypothermic circulatory arrest for aortic arch surgery.
Limitations In this analysis, nadir body temperature was not a predictor of clinical outcomes. This finding should be interpreted with caution. Firstly, the duration of cooling, which impacts both the temperature and completeness of tissue cooling, was not available for analysis. Secondly, temperature source was missing
or coded as “other” in 27% of patients. Thirdly, there may be an interaction between nadir temperature, choice of perfusion strategy, and team experience - with increased use of moderate hypothermia by more experienced teams employing either antegrade cerebral perfusion, or no cerebral perfusion with shorter circulatory arrest times; we were unable to adjust for this variable as surgeon and institutional volume were not provided. Neither was information on detailed, quantitative neurological assessment, unilateral versus bilateral cerebral perfusion, or intra-operative changes to planned cerebral protection strategy. Additional factors which may affect the validity of this data are related to coding: a proportion of cases appeared to have incorrectly coded cerebral perfusion and total circulatory arrest times (for example, identical reported extended selective cerebral perfusion times and circulatory arrest times without cerebral perfusion), suggesting there may be utility in providing database coders with focused assistance in this area, as well as indicating a possible need for additional validation from patient records. Many of these limitations will be addressed by the latest iteration of the STS adult cardiac database, designed to improve the ability of researchers to determine optimal strategies in aortic surgery[19].
Conclusions In the United States, the practice and outcomes of aortic arch operation remain associated with wide variation, and high rates of mortality and permanent neurological complications. Our data suggest that in patients needing circulatory arrest longer than 30 minutes, deep hypothermia with antegrade or retrograde cerebral perfusion, or moderate hypothermia with antegrade cerebral perfusion appear safer strategies compared to deep hypothermic circulatory arrest alone. This data supports guideline recommendations for cerebral perfusion during hypothermic circulatory arrest for aortic arch surgery.
References [1] Griepp R, Stinson E, Hollingsworth J, Buehler D. Prosthetic replacement of the aortic arch. J Thorac Cardiovasc Surg 1975. doi:. [2] Englum BR, He X, Gulack BC, Ganapathi AM, Mathew JP, Brennan JM, et al. Hypothermia and cerebral protection strategies in aortic arch surgery: a comparative effectiveness analysis from the STS Adult Cardiac Surgery Database. Eur J Cardiothorac Surg 2017. doi:10.1093/ejcts/ezx133. [3] Yan TD, Bannon PG, Bavaria J, Coselli JS, Elefteriades JA, Griepp B, et al. Consensus on hypothermia in aortic arch surgery. Ann Cardiothorac Surg 2013;2:163–8. doi:10.3978/j.issn.2225319X.2013.03.03. [4] Yan TD, Tian DH, LeMaire SA, Misfeld M, Elefteriades JA, Chen EP, et al. The ARCH projects: Design and rationale (IAASSG 001). Eur J Cardio-Thoracic Surg 2014. doi:10.1093/ejcts/ezt520. [5] Erbel R, Aboyans V, Boileau C, Bossone E, Bartolomeo RD, Eggebrecht H, et al. 2014 ESC Guidelines on the diagnosis and treatment of aortic diseases: Document covering acute and chronic aortic diseases of the thoracic and abdominal aorta of the adult. Eur Heart J 2014. doi:10.1093/eurheartj/ehu281. [6] Ueda Y, Miki S, Kusuhara K, Okita Y, Tahata T, Yamanaka K. Surgical treatment of aneurysm or dissection involving the ascending aorta and aortic arch, utilizing circulatory arrest and retrograde cerebral perfusion. J Cardiovasc Surg (Torino) 1990. [7] Kazui T, Inoue N, Yamada O, Komatsu S. Selective cerebral perfusion during operation for aneurysms of the aortic arch: A reassessment. Ann Thorac Surg 1992. doi:10.1016/0003-4975(92)90767X. [8] Williams JB, Peterson ED, Zhao Y, O’Brien SM, Andersen ND, Miller DC, et al. Contemporary Results for Proximal Aortic Replacement in North America. J Am Coll Cardiol 2012. doi:10.1016/j.jacc.2012.06.023. [9] L.G. S, E.S. C, K.R. H, J.S. C, S. R, S.A. S, et al. Deep hypothermia with circulatory arrest: Determinants of stroke and early mortality in 656 patients. J Thorac Cardiovasc Surg 1993. [10] Ergin MA, Galla JD, Lansman s L, Quintana C, Bodian C, Griepp RB. Hypothermic circulatory arrest in operations on the thoracic aorta. Determinants of operative mortality and neurologic outcome. J Thorac Cardiovasc Surg 1994. doi:10.5555/uri:pii:S0022522394703345. [11] Ziganshin BA, Rajbanshi BG, Tranquilli M, Fang H, Rizzo JA, Elefteriades JA. Straight deep hypothermic circulatory arrest for cerebral protection during aortic arch surgery: Safe and effective. J. Thorac. Cardiovasc. Surg., 2014. doi:10.1016/j.jtcvs.2014.05.027. [12] Okita Y, Takamoto S, Ando M, Morota T, Matsukawa R, Kawashima Y, et al. Mortality and cerebral outcome in patients who underwent aortic arch operations using deep hypothermic circulatory arrest with retrograde cerebral perfusion: No relation of early death, stroke, and delirium to the duration of circulatory arrest. J Thorac Cardiovasc Surg 1998. doi:10.1016/S0022-5223(98)70451-9.
[13] Coselli JS, LeMaire SA. Experience with retrograde cerebral perfusion during proximal aortic surgery in 290 patients. J Card Surg 1997. [14] Girardi LN, Shavladze N, Sedrakyan A, Neragi-Miandoab S. Safety and efficacy of retrograde cerebral perfusion as an adjunct for cerebral protection during surgery on the aortic arch. J Thorac Cardiovasc Surg 2014. doi:10.1016/j.jtcvs.2014.07.024. [15] Preventza O, Coselli JS, Garcia A, Kashyap S, Akvan S, Simpson KH, et al. Moderate hypothermia at warmer temperatures is safe in elective proximal and total arch surgery: Results in 665 patients. J Thorac Cardiovasc Surg 2017. doi:10.1016/j.jtcvs.2016.09.044. [16] Hughes GC, Zhao Y, Rankin JS, Scarborough JE, O’Brien S, Bavaria JE, et al. Effects of institutional volumes on operative outcomes for aortic root replacement in North America. J Thorac Cardiovasc Surg 2013. doi:10.1016/j.jtcvs.2011.10.094. [17] M.M. S, a.T. C, a. P, G.P. K, T. P, S.J. W, et al. Deep hypothermic circulatory arrest: II. Changes in electroencephalogram and evoked potentials during rewarming. Ann Thorac Surg 2001;71:22–8. [18] Ehrlich MP, Hagl C, McCullough JN, Zhang N, Shiang H, Bodian C, et al. Retrograde cerebral perfusion provides negligible flow through brain capillaries in the pig. J Thorac Cardiovasc Surg 2001;122:331–8. doi:10.1067/mtc.2001.115244. [19] Bavaria JE, Fukuhara S, Desai ND. Thoracic aortic surgery enters the era of big data. Eur J Cardiothorac Surg 2017. doi:10.1093/ejcts/ezx225.
Table 1. Basic Demographics and Presentation, Overall and According to Cerebral Perfusion Strategy Overall
No Cerebral Perfusion
N=7,830
N=2,563
Antegrade Cerebral Perfusion N=3,369
Retrograde Cerebral Perfusion N=1,898
P-Value
Demographics Age (SD) Male
63.1 (13.1)
63.7 (12.8)
62.7 (13.0)
62.7 (13.7)
0.007
5,116 (65.3%)
1,666 (65.0%)
2,187 (64.9%)
1,263 (66.5%)
0.4
Race
<0.0001 White
6,512 (83.2%)
2,187 (85.3%)
2,726 (80.9%)
1,599 (84.3%)
Black
585 (7.5%)
164 (6.4%)
317 (9.4%)
104 (5.5%)
Others
733 (9.4%)
212 (8.3%)
326 (9.7%)
195 (10.3%)
28.5 (5.6)
28.7 (5.7)
28.4 (5.6)
28.4 (5.4)
0.04
Diabetes
1,197 (15.3%)
437 (17.1%)
500 (14.8%)
260 (13.7%)
0.006
Dyslipidemia
4,825 (61.6%)
1,637 (63.9%)
2,032 (60.3%)
1,156 (60.9%)
0.02
Body mass index (SD) Comorbidities
Dialysis Hypertension
107 (1.4%)
35 (1.4%)
52 (1.5%)
20 (1.1%)
0.3
6,125 (78.2%)
1,990 (77.6%)
2,704 (80.3%)
1,431 (75.4%)
0.0001
Lung disease
0.0003
None
6,072 (77.6%)
1,998 (78.0%)
2,541 (75.4%)
1,533 (80.8%)
Mild
1,125 (14.4%)
358 (14.0%)
524 (15.6%)
243 (12.8%)
633 (8.1%)
207 (8.1%)
304 (9.0%)
122 (6.4%)
Peripheral artery disease
1,457 (18.6%)
444 (17.3%)
650 (19.3%)
363 (19.1%)
Cerebrovascular accident
379 (4.8%)
131 (5.1%)
162 (4.8%)
86 (4.5%)
0.7
Carotid disease
282 (3.6%)
106 (4.1%)
112 (3.3%)
64 (3.4%)
0.2
Previous CABG
342 (4.4%)
135 (5.3%)
148 (4.4%)
59 (3.1%)
0.002
Moderate or severe
Previous valve surgery
0.1
1,058 (13.5%)
354 (13.8%)
483 (14.3%)
221 (11.6%)
0.02
Previous PCI
629 (8.0%)
228 (8.9%)
271 (8.0%)
130 (6.9%)
0.04
Myocardial infarction
908 (11.6%)
345 (13.5%)
377 (11.2%)
186 (9.8%)
0.0005
1,811 (23.1%)
564 (22.0%)
755 (22.4%)
492 (25.9%)
0.004
100 (1.28%)
40 (1.6%)
38 (1.1%)
22 (1.2%)
0.3
Prior heart failure Presentation Cardiogenic shock Urgency status
<0.0001
Elective
5,646 (72.1%)
1,826 (71.2%)
1,475 (77.7%)
2,345 (69.6%)
Urgent
1,776 (22.7%)
598 (23.3%)
842 (25.0%)
336 (17.7%)
408 (5.2%)
139 (5.4%)
182 (5.4%)
87 (4.6%)
Emergent
Table 2. Surgical Parameters, Overall and According to Cerebral Perfusion Strategy Overall
No Cerebral Perfusion
N=7,830
N=2,563
Antegrade Cerebral Perfusion N=3,369
Retrograde Cerebral Perfusion N=1,898
Surgery type Aortic arch Ascending/hemi-arch
<0.0001 939 (12.0%)
173 (6.8 %)
648 (19.2 %)
118 (6.2%)
6,891 (88.0%)
2,390 (93.3%)
2,721 (80.8%)
1,780 (93.8%)
Cannulation site Axillary/innominate artery
P-Value
<0.0001 2,343 (29.9%)
Femoral artery Others/unknown
318 (12.4%)
1,891 (56.1%)
134 (7.1%)
440 (17.2%)
156 (4.6%)
204 (10.8%)
1,805 (70.4%)
1,322 (39.2%)
1,560 (82.2%)
Total arrest time, minutes [IQR]
25.8 [16-30]
21.5 [15-25]
29.8 [18-37]
24.2 [16-28]
<0.0001
Nadir temperature, Celsius [IQR]
21.1 [18.024.0]
19.9 [18.022.1]
22.6 [19.925.7]
20.1 [18.022.3]
<0.0001
Temperature source
<0.0001
Nasopharyngeal/tympanic
1,244 (15.9%)
374 (14.6%)
529 (15.7%)
341 (18.0%)
Bladder/Rectal
4,435 (56.6%)
1,376 (53.7%)
1,965 (58.3%)
1,094 (57.6%)
Other/unknown
2,151 (27.5%)
813 (31.7%)
875 (26.0 %)
463 (24.4%)
Table 3. Surgical Outcomes According to the Cerebral Perfusion Strategy
Permanent neurological complications Stroke
Overall
No Cerebral Perfusion
Retrograde Cerebral Perfusion N=1,898
P-Value
N=2,563
Antegrade Cerebral Perfusion N=3,369
N=7,830 492 (6.3%)
195 (7.6%)
210 (6.2%)
87 (4.6%)
0.0002
419 (5.4%)
165 (6.4%)
179 (5.3%)
75 (4.0%)
0.001
Encephalopathy
127 (1.6%)
52 (2.0%)
52 (1.5%)
23 (1.2%)
0.09
Paralysis
50 (0.6%)
26 (1.0%)
21 (0.6%)
3 (0.2%)
0.002
Prolonged ventilation
1,690 (21.6%)
590 (23.0%)
768 (22.8%)
332 (17.5%)
<0.0001
Renal failure
390 (5.0%)
139 (5.4%)
174 (5.2%)
77 (4.1%)
0.09
Mortality
507 (6.5%)
217 (8.5%)
207 (6.1%)
83 (4.4%)
<0.0001
Mortality or Permanent neurological complications
850 (10.9%)
349 (13.6%)
355 (10.5%)
146 (7.7%)
<0.0001
Figure Legends Figure 1. Unadjusted incidence rates of primary endpoint (in-hospital or 30-day mortality, permanent stroke and paralysis) according to total circulatory arrest time. Figure 2. Unadjusted incidence rates of primary endpoint (in-hospital or 30-day mortality, permanent stroke and paralysis) according to the total circulatory arrest time for each brain protection strategy. Figure 3. Adjusted odds ratio for 30-day or in-hospital mortality, permanent stroke and paralysis, according to the total circulatory arrest time for the same cerebral perfusion strategy in patients.
S0003-4975(19)31433-X
DOI:
https://doi.org/10.1016/j.athoracsur.2019.08.043
Reference:
ATS 33066
To appear in:
The Annals of Thoracic Surgery
Received Date: 14 April 2019 Revised Date:
17 July 2019
Accepted Date: 12 August 2019
Please cite this article as: Itagaki S, Chikwe J, Sun E, Chu D, Toyoda N, Egorova N, Impact of Cerebral Perfusion on Outcomes of Aortic Surgery: STS Adult Cardiac Database Analysis, The Annals of Thoracic Surgery (2019), doi: https://doi.org/10.1016/j.athoracsur.2019.08.043. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. 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. © 2019 by The Society of Thoracic Surgeons
Impact of Cerebral Perfusion on Outcomes of Aortic Surgery: STS Adult Cardiac Database Analysis Running head: Circulatory arrest cerebral protection
Shinobu Itagaki MSc MD,1 Jo Chikwe MD FRCS,1,2 Eric Sun MD,3 Danny Chu MD,4 Nana Toyoda MD PhD,1 Natalia Egorova PhD2;
1
Department of Cardiovascular Surgery, The Icahn School of Medicine at Mount Sinai, NY
2
Division of Cardiothoracic Surgery, The State University of New York, Stony Brook, NY
3
Department of Health Science and Policy, The Icahn School of Medicine at Mount Sinai, NY
4
Department of Cardiothoracic Surgery, University of Pittsburgh School of Medicine, PA
Corresponding author:
Joanna Chikwe MD
Correspondence:
Department of Cardiovascular Surgery, Mount Sinai Medical Center, 1190 Fifth Avenue, New York, NY 10029 USA Email: [email protected]
Abstract Background: Limited data informs cerebral protection during circulatory arrest: this study was designed to identify optimal approaches from a national clinical registry. Methods: A total of 7830 adults (mean age 63.1, standard deviation 13.1 years) who underwent hemi-arch (n=6891, 88.0%) or total arch (n=939, 12.0%) replacement with hypothermic circulatory arrest between 2014-2016, were identified from The Society of Thoracic Surgeons Database (version 2.81). Aortic dissections were excluded from the analysis. Multivariable logistic regression was used to adjust for 29 baseline and operative variables including demographics, comorbidity, surgery, and nadir temperature, comparing outcomes according to protection strategy. The primary end-point was a composite of 30-day and in-hospital mortality or major permanent neurological complications. Results: The rate of death or permanent neurological complication was 10.9% (n=850). Antegrade cerebral perfusion was most utilized (n=3369, 43%, median nadir temperature 23oC, median arrest time 30 minutes) compared to retrograde (n=1898, 24%, 20 oC, 24 minutes) and no cerebral perfusion (n=2563, 33%, 20 oC, 22 minutes). In multivariable analysis, deep hypothermia with antegrade (OR 0.65, 95% CI 0.52- 0.81) or retrograde (OR 0.57, 95% CI 0.45- 0.71) perfusion, and moderate hypothermia with antegrade perfusion (OR 0.61, 95% CI 0.46-0-.79) were associated with significant reductions in death and stroke compared to deep hypothermia without cerebral perfusion. Risk reduction was greatest in circulatory arrest >30 minutes. Conclusions: For patients without dissection and requiring >30 minutes circulatory arrest, optimal cerebral protection strategies are deep hypothermia with either antegrade or retrograde cerebral perfusion; or moderate hypothermia with antegrade cerebral perfusion. Words 250 (250)
Hypothermic circulatory arrest has been a standard component of aortic arch surgery, providing an extended time to operate in a bloodless field by reducing brain and visceral metabolic demand, since the technique was developed by R. Griepp et al in the mid-1970’s[1,2]. Hypothermic circulatory arrest is increasingly utilized in combination with retrograde cerebral perfusion, or with antegrade cerebral perfusion which may also be conducted with warmer circulatory arrest temperatures[3]. The primary rationale for antegrade cerebral perfusion is to minimize the adverse neurological sequalae of total circulatory arrest, extending the safe duration of circulatory arrest at warmer temperatures, and reducing the total time spent cooling and rewarming on cardiopulmonary bypass with the aim of minimizing coagulopathy and organ dysfunction[3]. In the absence of a robust evidence-base and consensus guideline recommendations, contemporary ascending aorta and arch surgery is characterized by wide variation in practice, particularly cerebral protection strategy, and clinical outcomes[2–5]. The current study was therefore designed to determine the optimal strategy for cerebral protection during elective ascending aorta and arch surgery from a national clinical registry.
Material and Methods Patient Population A retrospective cohort analysis was conducted on outcomes of 18,180 adult patients aged 18 years or older who underwent aortic arch or hemi-arch/ascending aortic replacement under circulatory arrest in the United States between 2014 and 2016. Patients were identified using the Society of Thoracic Surgeons Adult Cardiac Surgery Database (version 2.81) and stratified according to the adjunctive cerebral perfusion strategy utilized: either straight hypothermic circulatory arrest without cerebral perfusion, hypothermic circulatory arrest with antegrade cerebral perfusion, or hypothermic circulatory arrest with retrograde cerebral perfusion.
After excluding patients who presented with aortic dissections (43.1%, n=5941), had salvage surgery or who required cardiopulmonary resuscitation immediately prior to operation (2.3%, n=410), patients in whom both retrograde and antegrade cerebral perfusion were used, (2.9%, n=524), patients with a nadir body temperature less than 10oC (4.4%, n=805) or over 28oC during circulatory arrest (6.2%, n=1124), and patients with a total circulatory arrest time (with or without the adjunctive cerebral perfusion strategy) less than 10 minutes (11.8%, n=2114) or more than 150 minutes (0.6%, n=114), a total of 7830 patients remained who comprised the study cohort. Baseline comorbidities, surgical details, and operative outcomes were identified using variables provided in the Society of Thoracic Surgeons Adult Cardiac Database (Version 2.81).
This study was approved by the Society of Thoracic Surgeons, and the institutional review boards of the Icahn School of Medicine at Mount Sinai and The State University of New York at Stony Brook. The approval included a waiver of informed consent. The data for this research were provided by The Society of Thoracic Surgeons’ national Database participant User File Research program. Data analysis was performed at the investigators’ institutions.
Primary and Secondary Outcomes The primary outcomes were composite endpoints of in-hospital or 30-day mortality and irreversible major neurological complication, (permanent stroke, permanent paralysis or permanent encephalopathy due to intracranial bleed, emboli, or anoxic brain injury). Mortality or major neurological complications occurring at any time during the index admission, including after 30-days post-operatively was reported. Permanent stroke was defined according to Society of Thoracic Surgeons criteria as any confirmed neurological deficit of abrupt onset caused by a disturbance in blood supply to the brain that did not resolve within 24 hours. Transient ischemic attacks and reversible encephalopathy were not included in
the primary outcome. Secondary outcomes were the individual outcomes of in-hospital or 30-day mortality, irreversible neurological complications, and postoperative complications including prolonged ventilation, renal failure and length of hospital stay.
Statistical Analysis For descriptive analysis, continuous variables were reported as mean with standard deviation or median and interquartile range and categorical variables were reported as proportions. The Analysis of Variance or Kruskal-Wallis for continuous variables and Chi-squared test were used as appropriate to compare demographics of cerebral perfusion strategy groups. Multivariable logistic regression models were used to evaluate adjusted risk of each endpoint, controlling for the cerebral perfusion strategy (with straight hypothermic circulatory arrest without cerebral perfusion as the reference), age, gender, race, baseline comorbidities (including diabetes, dyslipidemia, dialysis, hypertension, lung disease, peripheral vascular disease, cerebrovascular accident, carotid disease, previous coronary artery bypass grafting, previous valve surgery, previous percutaneous coronary intervention, myocardial infarction, prior congestive heart failure), the mode of presentation (including cardiogenic shock, and surgical urgency status), and the surgical details (including total arch vs hemi-arch replacement, aortic cannulation sites, total circulatory arrest time, lowest body temperature and its measurement source). The total circulatory arrest time was incorporated in the model as a continuous variable first, and then as a categorical variable with 10-minute increment (10-<20 minutes, 20-<30 minutes, 30-<40minutes, 40-<50 minutes, and >/=50minutes). The nadir body temperature was incorporated in the model as a continuous variable, and then as a categorical variable (10-<20oC, 20-<24oC, and 24-<28oC), in line with the consensus definitions of moderate and deep hypothermia2. To validate the findings, sensitivity analyses were conducted using clinically relevant subgroups with multiple strata such as total arch vs non-total arch, different ranges of total circulatory arrest time, and different perfusion and temperature strategies. The main findings remained largely the
same and are provided in the supplemental data. All tests were two-tailed and an alpha level of 0.05 was considered statistically significant. All statistical analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC).
Results Study Population A total of 7830 patients underwent either total arch aortic replacement (n=939, 12.0%) or ascending aortic with hemi-arch replacement (n=6891, 88.0%) between 2014 and 2016. The mean age was 63.1 years and 65.3% of the patients were male. A total of 72.1%% (n=5646) of the patients underwent elective surgery. Baseline patient characteristics are summarized in Table 1.
Operative technique Operative approaches are summarized in Table 2. Of the 6891 patients who underwent hemi-arch or ascending aorta replacement with circulatory arrest, 39.5% (n=2721) received antegrade cerebral perfusion, 34.7% (n=2390) of patients underwent straight hypothermic circulatory arrest without cerebral perfusion, and 25.8% (n=1780) patients received retrograde cerebral perfusion (p<0.001). Overall, 12% (n=939) of patients underwent total arch replacement. Of these patients 69.0% (n=648) received antegrade cerebral perfusion, 18.4% (n=173) patients underwent straight hypothermic circulatory arrest without cerebral perfusion, and 12.6% (n=118) patients received retrograde cerebral perfusion (p<0.001). Overall, 30% (n=2342) of the patients had axillary or innominate arterial cannulation, which was more likely to be used with antegrade cerebral perfusion (56.1%, n=1891) compared to 12.4% (n=318) of patients undergoing straight hypothermic circulatory arrest without cerebral perfusion and 7.1% (n=134) undergoing retrograde cerebral perfusion (p<0.001). Femoral cannulation, used in 10.2% (n=800) of
patients overall, was the least utilized technique in patients who received antegrade cerebral perfusion (4.6%, n=156), compared to 17.2% (n=440) of patients undergoing straight hypothermic circulatory arrest without cerebral perfusion, and 10.8% (n=204) patients undergoing retrograde cerebral perfusion (p<0.001).
Antegrade cerebral perfusion was associated with the longest total circulatory arrest times (median 30 minutes [interquartile range (IQR) 18-37]) and the highest nadir body temperature (median 23oC [IQR 2026 oC]). Straight hypothermic circulatory arrest without cerebral perfusion was associated with the shortest circulatory arrest times (22 minutes [IQR 15-25] minutes) and the lowest nadir temperature (20 o
C [IQR 18-22]oC). The distributions of the total circulatory arrest time and the nadir body temperature in
each cerebral perfusion strategy are shown in Supplemental Figures 1 and 2, respectively. Overall, the temperature source was nasopharyngeal or tympanic in only 15.9% (n=1244) patients, with bladder or rectal temperature used in 56.6% (n=4435).
Mortality and Neurological Complications The primary and secondary outcomes are summarized in Table 3. The rate of death, permanent stroke or paraplegia was 10.9% (n=850). The rate of death, stroke or paraplegia was 14% (n=349) in patients who underwent straight hypothermic circulatory arrest without cerebral perfusion, 11% (n=355) in patients undergoing antegrade cerebral perfusion and 8% (n=146) in patients who underwent retrograde cerebral perfusion, respectively (p<0.001). Permanent neurological complications were observed in 8% (n= 195) of patients who underwent straight hypothermic circulatory arrest without cerebral perfusion), 6% (n=210) of patients undergoing antegrade cerebral perfusion and 5% (n=87) of patients after retrograde cerebral perfusion, respectively (p<0.001). In-hospital or 30-day mortality occurred in 9% (n=217), 6% (n=207), and 4% (n=83) respectively.
The relationship between unadjusted rates of primary outcomes and total circulatory arrest time are shown in Figure 1, and then stratified by cerebral perfusion strategy in Figure 2. In multivariable logistic regression, longer total arrest times were strongly associated with increasing risk of death or stroke regardless of cerebral perfusion strategy, but this effect was greatest in patients undergoing straight hypothermic circulatory arrest without cerebral perfusion (adjusted OR 1.33, 95% CI 1.19-1.49 per 10 minutes increment), followed by retrograde cerebral perfusion (adjusted OR 1.23, 95% CI 1.06-1.42), and least with antegrade cerebral perfusion (adjusted OR 1.16, 95% CI 1.07-1.25). Use of antegrade (adjusted OR 0.63, 95% CI 0.51-0.77), and retrograde cerebral perfusion (adjusted OR 0.56, 95% CI 0.45-0.70) were both associated with significant reductions in the risk of the primary endpoint of death or permanent neurological deficit compared with no cerebral perfusion.
The protective effect of the adjunctive cerebral perfusion strategy became significant starting in the subgroup of the circulatory arrest time of 20 minutes and longer, and was observed throughout the duration of circulatory arrest times with increasing protective effect with longer circulatory arrest times (Figure 2). The increased risk of death or stroke associated with increasing circulatory arrest times was most marked in patients after more than 20 minutes straight hypothermic circulatory arrest without cerebral perfusion circulatory arrest while for patients receiving either antegrade or retrograde cerebral perfusion, arrest time did not become a significant predictor of death or stroke until after 50 minutes of arrest time (Figure 3).
Nadir temperature was not associated with a significant difference in risk of primary endpoint overall, or in subgroup analysis. In multivariable analysis, deep hypothermia with antegrade (OR 0.65, 95% CI 0.520.81) or retrograde (OR 0.57, 95% CI 0.45- 0.71) perfusion, and moderate hypothermia with antegrade
perfusion (OR 0.61, 95% CI 0.46-0-.79) were associated with significant reductions in death and stroke compared to deep hypothermia without cerebral perfusion. Supplemental Data provides the results of subgroup analysis confirming these trends in patients who underwent total arch versus ascending aorta or hemi-arch replacement. In a subgroup analysis of patients where the cannulation site was specified (n=6482), femoral cannulation was associated with increased risk of the primary outcome compared to ascending aortic cannulation (adjusted OR 1.67, 95% CI 1.32-2.12), whereas the axillary cannulation was not.
Comment This analysis demonstrates firstly that contemporary elective ascending and aortic arch surgery in the United States remains associated with relatively high mortality and major neurological morbidity, and wide variation in practice and outcomes. Secondly, we confirm that in patients needing circulatory arrest longer than 30 minutes, deep hypothermia with antegrade or retrograde cerebral perfusion, or moderate hypothermia with antegrade cerebral perfusion appear safer strategies compared to deep hypothermic circulatory arrest without cerebral perfusion. Thirdly, we provide new insights into the interaction between increasing duration of circulatory arrest and cerebral protection strategy: longer circulatory arrest times were associated with increasing risk of mortality and stroke regardless of the cerebral protection strategy, but the negative effects of the longest circulatory arrest times are most effectively mitigated by adjunctive cerebral perfusion.
In this national registry, the operative mortality for patients undergoing elective ascending aorta or hemiarch surgery was 5.9%, and 9.9% after arch replacement, which compares to 2-5% achieved in single-centers series[6–15]. Poorer outcomes have been associated with low volume providers, which represent the majority of centers in the United States, where over 75% of the centers reporting to the STS
between 2010 and 2014 performed fewer than 6 aortic cases per year [16]. However, wide variation in cerebral protection strategy was observed even in high volume centers, suggesting that factors beyond choice of cerebral protection strategy, such as surgical expertise and team experience, contribute to improved outcomes[16]. However, in unselected contemporary clinical practice, our analysis showed that for most patients, either antegrade or retrograde cerebral perfusion was associated with superior outcomes compared to straight hypothermic circulatory arrest without cerebral perfusion.
Straight hypothermic circulatory arrest without cerebral perfusion has been the predominant cerebral protection strategy for aortic arch surgery since its introduction in the mid 1970’s by R. Griepp et al. In a recent analysis of aortic surgery in North America between 2011-2014, straight hypothermic circulatory arrest was the most predominant method of brain protection utilized in 35% of cases[2]. In our more recent study this proportion had dropped slightly to 32%. Patients operated on at temperatures below 20oC can tolerate 20-30 minutes of circulatory arrest which is sufficient for hemi-arch repair, although as few as 25% of these patients have been shown to achieve electrocerebral silence[17]. The extended cardiopulmonary bypass times required to achieve complete cooling and rewarming, and lower nadir temperatures, contribute to enzyme and end-organ dysfunction, and coagulopathy: which may have contributed to the shift observed in our analysis towards adjunctive strategies that allow extended circulatory arrest times at warmer temperatures.
Hypothermic circulatory arrest with retrograde cerebral perfusion, first described by Ueda et al in 1990, was used in 24% of patients in our analysis, for a median arrest time of 24 minutes with a median temperature nadir of 20oC[6]. The hypothesis that retrograde cerebral perfusion might safely extend circulatory arrest times by supplying oxygenated blood to cerebral territories has not been supported by porcine experiments indicating <0.1% of retrograde perfusate from the superior vena cava traverses
cerebral capillary beds[18]. Hypothermic circulatory arrest with antegrade cerebral perfusion has increasingly become the preferred protection strategy for complex arch surgery as the brain can be perfused during circulatory arrest, allowing longer safe circulatory arrest times and warmer nadir temperatures. The embolic risk of cannulating the innominate and carotid arteries may be partially mitigated by using axillary artery cannulation, which was used in 30% of patients overall, and in 56% of those undergoing antegrade cerebral perfusion in this analysis.
The novel finding of this study is that longer total arrest times were strongly associated with increasing risk of death or stroke regardless of cerebral perfusion strategy, but this effect was greatest in patients undergoing straight hypothermic circulatory arrest without cerebral perfusion (33% increased risk for every additional 10 minutes circulatory arrest), followed by retrograde cerebral perfusion (23% increased risk for every additional 10 minutes circulatory arrest), and least with antegrade cerebral perfusion (16% increased risk for every additional 10 minutes circulatory arrest). Cerebral perfusion (either antegrade or retrograde with deep or, antegrade with moderate hypothermia) appeared to mitigate adverse outcomes associated with circulatory arrest greater than 30 minutes significantly more effectively than the use of straight deep hypothermia alone. This threshold has not previously been reported, including in an earlier analysis of the STS database which stratified cerebral protection strategies by nadir temperature, rather than by duration of circulatory arrest. Our data supports the guideline recommendations for routine use of cerebral perfusion during hypothermic circulatory arrest for aortic arch surgery.
Limitations In this analysis, nadir body temperature was not a predictor of clinical outcomes. This finding should be interpreted with caution. Firstly, the duration of cooling, which impacts both the temperature and completeness of tissue cooling, was not available for analysis. Secondly, temperature source was missing
or coded as “other” in 27% of patients. Thirdly, there may be an interaction between nadir temperature, choice of perfusion strategy, and team experience - with increased use of moderate hypothermia by more experienced teams employing either antegrade cerebral perfusion, or no cerebral perfusion with shorter circulatory arrest times; we were unable to adjust for this variable as surgeon and institutional volume were not provided. Neither was information on detailed, quantitative neurological assessment, unilateral versus bilateral cerebral perfusion, or intra-operative changes to planned cerebral protection strategy. Additional factors which may affect the validity of this data are related to coding: a proportion of cases appeared to have incorrectly coded cerebral perfusion and total circulatory arrest times (for example, identical reported extended selective cerebral perfusion times and circulatory arrest times without cerebral perfusion), suggesting there may be utility in providing database coders with focused assistance in this area, as well as indicating a possible need for additional validation from patient records. Many of these limitations will be addressed by the latest iteration of the STS adult cardiac database, designed to improve the ability of researchers to determine optimal strategies in aortic surgery[19].
Conclusions In the United States, the practice and outcomes of aortic arch operation remain associated with wide variation, and high rates of mortality and permanent neurological complications. Our data suggest that in patients needing circulatory arrest longer than 30 minutes, deep hypothermia with antegrade or retrograde cerebral perfusion, or moderate hypothermia with antegrade cerebral perfusion appear safer strategies compared to deep hypothermic circulatory arrest alone. This data supports guideline recommendations for cerebral perfusion during hypothermic circulatory arrest for aortic arch surgery.
References [1] Griepp R, Stinson E, Hollingsworth J, Buehler D. Prosthetic replacement of the aortic arch. J Thorac Cardiovasc Surg 1975. doi:. [2] Englum BR, He X, Gulack BC, Ganapathi AM, Mathew JP, Brennan JM, et al. Hypothermia and cerebral protection strategies in aortic arch surgery: a comparative effectiveness analysis from the STS Adult Cardiac Surgery Database. Eur J Cardiothorac Surg 2017. doi:10.1093/ejcts/ezx133. [3] Yan TD, Bannon PG, Bavaria J, Coselli JS, Elefteriades JA, Griepp B, et al. Consensus on hypothermia in aortic arch surgery. Ann Cardiothorac Surg 2013;2:163–8. doi:10.3978/j.issn.2225319X.2013.03.03. [4] Yan TD, Tian DH, LeMaire SA, Misfeld M, Elefteriades JA, Chen EP, et al. The ARCH projects: Design and rationale (IAASSG 001). Eur J Cardio-Thoracic Surg 2014. doi:10.1093/ejcts/ezt520. [5] Erbel R, Aboyans V, Boileau C, Bossone E, Bartolomeo RD, Eggebrecht H, et al. 2014 ESC Guidelines on the diagnosis and treatment of aortic diseases: Document covering acute and chronic aortic diseases of the thoracic and abdominal aorta of the adult. Eur Heart J 2014. doi:10.1093/eurheartj/ehu281. [6] Ueda Y, Miki S, Kusuhara K, Okita Y, Tahata T, Yamanaka K. Surgical treatment of aneurysm or dissection involving the ascending aorta and aortic arch, utilizing circulatory arrest and retrograde cerebral perfusion. J Cardiovasc Surg (Torino) 1990. [7] Kazui T, Inoue N, Yamada O, Komatsu S. Selective cerebral perfusion during operation for aneurysms of the aortic arch: A reassessment. Ann Thorac Surg 1992. doi:10.1016/0003-4975(92)90767X. [8] Williams JB, Peterson ED, Zhao Y, O’Brien SM, Andersen ND, Miller DC, et al. Contemporary Results for Proximal Aortic Replacement in North America. J Am Coll Cardiol 2012. doi:10.1016/j.jacc.2012.06.023. [9] L.G. S, E.S. C, K.R. H, J.S. C, S. R, S.A. S, et al. Deep hypothermia with circulatory arrest: Determinants of stroke and early mortality in 656 patients. J Thorac Cardiovasc Surg 1993. [10] Ergin MA, Galla JD, Lansman s L, Quintana C, Bodian C, Griepp RB. Hypothermic circulatory arrest in operations on the thoracic aorta. Determinants of operative mortality and neurologic outcome. J Thorac Cardiovasc Surg 1994. doi:10.5555/uri:pii:S0022522394703345. [11] Ziganshin BA, Rajbanshi BG, Tranquilli M, Fang H, Rizzo JA, Elefteriades JA. Straight deep hypothermic circulatory arrest for cerebral protection during aortic arch surgery: Safe and effective. J. Thorac. Cardiovasc. Surg., 2014. doi:10.1016/j.jtcvs.2014.05.027. [12] Okita Y, Takamoto S, Ando M, Morota T, Matsukawa R, Kawashima Y, et al. Mortality and cerebral outcome in patients who underwent aortic arch operations using deep hypothermic circulatory arrest with retrograde cerebral perfusion: No relation of early death, stroke, and delirium to the duration of circulatory arrest. J Thorac Cardiovasc Surg 1998. doi:10.1016/S0022-5223(98)70451-9.
[13] Coselli JS, LeMaire SA. Experience with retrograde cerebral perfusion during proximal aortic surgery in 290 patients. J Card Surg 1997. [14] Girardi LN, Shavladze N, Sedrakyan A, Neragi-Miandoab S. Safety and efficacy of retrograde cerebral perfusion as an adjunct for cerebral protection during surgery on the aortic arch. J Thorac Cardiovasc Surg 2014. doi:10.1016/j.jtcvs.2014.07.024. [15] Preventza O, Coselli JS, Garcia A, Kashyap S, Akvan S, Simpson KH, et al. Moderate hypothermia at warmer temperatures is safe in elective proximal and total arch surgery: Results in 665 patients. J Thorac Cardiovasc Surg 2017. doi:10.1016/j.jtcvs.2016.09.044. [16] Hughes GC, Zhao Y, Rankin JS, Scarborough JE, O’Brien S, Bavaria JE, et al. Effects of institutional volumes on operative outcomes for aortic root replacement in North America. J Thorac Cardiovasc Surg 2013. doi:10.1016/j.jtcvs.2011.10.094. [17] M.M. S, a.T. C, a. P, G.P. K, T. P, S.J. W, et al. Deep hypothermic circulatory arrest: II. Changes in electroencephalogram and evoked potentials during rewarming. Ann Thorac Surg 2001;71:22–8. [18] Ehrlich MP, Hagl C, McCullough JN, Zhang N, Shiang H, Bodian C, et al. Retrograde cerebral perfusion provides negligible flow through brain capillaries in the pig. J Thorac Cardiovasc Surg 2001;122:331–8. doi:10.1067/mtc.2001.115244. [19] Bavaria JE, Fukuhara S, Desai ND. Thoracic aortic surgery enters the era of big data. Eur J Cardiothorac Surg 2017. doi:10.1093/ejcts/ezx225.
Table 1. Basic Demographics and Presentation, Overall and According to Cerebral Perfusion Strategy Overall
No Cerebral Perfusion
N=7,830
N=2,563
Antegrade Cerebral Perfusion N=3,369
Retrograde Cerebral Perfusion N=1,898
P-Value
Demographics Age (SD) Male
63.1 (13.1)
63.7 (12.8)
62.7 (13.0)
62.7 (13.7)
0.007
5,116 (65.3%)
1,666 (65.0%)
2,187 (64.9%)
1,263 (66.5%)
0.4
Race
<0.0001 White
6,512 (83.2%)
2,187 (85.3%)
2,726 (80.9%)
1,599 (84.3%)
Black
585 (7.5%)
164 (6.4%)
317 (9.4%)
104 (5.5%)
Others
733 (9.4%)
212 (8.3%)
326 (9.7%)
195 (10.3%)
28.5 (5.6)
28.7 (5.7)
28.4 (5.6)
28.4 (5.4)
0.04
Diabetes
1,197 (15.3%)
437 (17.1%)
500 (14.8%)
260 (13.7%)
0.006
Dyslipidemia
4,825 (61.6%)
1,637 (63.9%)
2,032 (60.3%)
1,156 (60.9%)
0.02
Body mass index (SD) Comorbidities
Dialysis Hypertension
107 (1.4%)
35 (1.4%)
52 (1.5%)
20 (1.1%)
0.3
6,125 (78.2%)
1,990 (77.6%)
2,704 (80.3%)
1,431 (75.4%)
0.0001
Lung disease
0.0003
None
6,072 (77.6%)
1,998 (78.0%)
2,541 (75.4%)
1,533 (80.8%)
Mild
1,125 (14.4%)
358 (14.0%)
524 (15.6%)
243 (12.8%)
633 (8.1%)
207 (8.1%)
304 (9.0%)
122 (6.4%)
Peripheral artery disease
1,457 (18.6%)
444 (17.3%)
650 (19.3%)
363 (19.1%)
Cerebrovascular accident
379 (4.8%)
131 (5.1%)
162 (4.8%)
86 (4.5%)
0.7
Carotid disease
282 (3.6%)
106 (4.1%)
112 (3.3%)
64 (3.4%)
0.2
Previous CABG
342 (4.4%)
135 (5.3%)
148 (4.4%)
59 (3.1%)
0.002
Moderate or severe
Previous valve surgery
0.1
1,058 (13.5%)
354 (13.8%)
483 (14.3%)
221 (11.6%)
0.02
Previous PCI
629 (8.0%)
228 (8.9%)
271 (8.0%)
130 (6.9%)
0.04
Myocardial infarction
908 (11.6%)
345 (13.5%)
377 (11.2%)
186 (9.8%)
0.0005
1,811 (23.1%)
564 (22.0%)
755 (22.4%)
492 (25.9%)
0.004
100 (1.28%)
40 (1.6%)
38 (1.1%)
22 (1.2%)
0.3
Prior heart failure Presentation Cardiogenic shock Urgency status
<0.0001
Elective
5,646 (72.1%)
1,826 (71.2%)
1,475 (77.7%)
2,345 (69.6%)
Urgent
1,776 (22.7%)
598 (23.3%)
842 (25.0%)
336 (17.7%)
408 (5.2%)
139 (5.4%)
182 (5.4%)
87 (4.6%)
Emergent
Table 2. Surgical Parameters, Overall and According to Cerebral Perfusion Strategy Overall
No Cerebral Perfusion
N=7,830
N=2,563
Antegrade Cerebral Perfusion N=3,369
Retrograde Cerebral Perfusion N=1,898
Surgery type Aortic arch Ascending/hemi-arch
<0.0001 939 (12.0%)
173 (6.8 %)
648 (19.2 %)
118 (6.2%)
6,891 (88.0%)
2,390 (93.3%)
2,721 (80.8%)
1,780 (93.8%)
Cannulation site Axillary/innominate artery
P-Value
<0.0001 2,343 (29.9%)
Femoral artery Others/unknown
318 (12.4%)
1,891 (56.1%)
134 (7.1%)
440 (17.2%)
156 (4.6%)
204 (10.8%)
1,805 (70.4%)
1,322 (39.2%)
1,560 (82.2%)
Total arrest time, minutes [IQR]
25.8 [16-30]
21.5 [15-25]
29.8 [18-37]
24.2 [16-28]
<0.0001
Nadir temperature, Celsius [IQR]
21.1 [18.024.0]
19.9 [18.022.1]
22.6 [19.925.7]
20.1 [18.022.3]
<0.0001
Temperature source
<0.0001
Nasopharyngeal/tympanic
1,244 (15.9%)
374 (14.6%)
529 (15.7%)
341 (18.0%)
Bladder/Rectal
4,435 (56.6%)
1,376 (53.7%)
1,965 (58.3%)
1,094 (57.6%)
Other/unknown
2,151 (27.5%)
813 (31.7%)
875 (26.0 %)
463 (24.4%)
Table 3. Surgical Outcomes According to the Cerebral Perfusion Strategy
Permanent neurological complications Stroke
Overall
No Cerebral Perfusion
Retrograde Cerebral Perfusion N=1,898
P-Value
N=2,563
Antegrade Cerebral Perfusion N=3,369
N=7,830 492 (6.3%)
195 (7.6%)
210 (6.2%)
87 (4.6%)
0.0002
419 (5.4%)
165 (6.4%)
179 (5.3%)
75 (4.0%)
0.001
Encephalopathy
127 (1.6%)
52 (2.0%)
52 (1.5%)
23 (1.2%)
0.09
Paralysis
50 (0.6%)
26 (1.0%)
21 (0.6%)
3 (0.2%)
0.002
Prolonged ventilation
1,690 (21.6%)
590 (23.0%)
768 (22.8%)
332 (17.5%)
<0.0001
Renal failure
390 (5.0%)
139 (5.4%)
174 (5.2%)
77 (4.1%)
0.09
Mortality
507 (6.5%)
217 (8.5%)
207 (6.1%)
83 (4.4%)
<0.0001
Mortality or Permanent neurological complications
850 (10.9%)
349 (13.6%)
355 (10.5%)
146 (7.7%)
<0.0001
Figure Legends Figure 1. Unadjusted incidence rates of primary endpoint (in-hospital or 30-day mortality, permanent stroke and paralysis) according to total circulatory arrest time. Figure 2. Unadjusted incidence rates of primary endpoint (in-hospital or 30-day mortality, permanent stroke and paralysis) according to the total circulatory arrest time for each brain protection strategy. Figure 3. Adjusted odds ratio for 30-day or in-hospital mortality, permanent stroke and paralysis, according to the total circulatory arrest time for the same cerebral perfusion strategy in patients.