HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Stephen J. Rooney Robert S. Bonser Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Harrington, D. K. Articles by Bonser, R. S. Search for Related Content PubMed PubMed Citation Articles by Harrington, D. K. Articles by Bonser, R. S. Related Collections Great vessels

Ann Thorac Surg 2004;78:596-601
© 2004 The Society of Thoracic Surgeons

---

Original article: cardiovascular

Nonneurologic morbidity and profound hypothermia in aortic surgery

^a Cardiothoracic Surgical Unit, University Hospital Birmingham, Queen Elizabeth Medical Centre, Birmingham, United Kingdom

Accepted for publication January 21, 2004.

^* Address reprint requests to Dr Bonser, Cardiothoracic Surgical Unit, University Hospital Birmingham, Queen Elizabeth Medical Centre, Birmingham B15 2TH, UK
e-mail: robert.bonser{at}uhb.nhs.uk

	Abstract

Top Abstract Introduction Patients and methods Results Comment References

 
BACKGROUND: Use of profoundly hypothermic cardiopulmonary bypass may increase the risk of postoperative bleeding and lung and renal dysfunction. The aim of this study was to analyze postoperative blood loss and indices of pulmonary and renal dysfunction in patients undergoing proximal aortic surgery with and without the use of profound hypothermia to determine risk factors for nonneurologic morbidity.

METHODS: Risk factors for blood loss, transfusion requirement, and pulmonary and renal dysfunction were studied in 116 patients undergoing thoracic aortic surgery with profoundly or moderately hypothermic cardiopulmonary bypass.

RESULTS: Overall mortality was 8.6%. Mean (± standard deviation) cardiopulmonary bypass times were 191 ± 53 minutes (profoundly hypothermic group) and 131 ± 48 minutes (moderately hypothermic group; p < 0.0001). The incidence of blood loss more than 1 L or resternotomy for bleeding was 25% (29 patients). Fifteen patients (12.9%) experienced postoperative pulmonary dysfunction, and 25 patients (21.6%) had postoperative renal dysfunction. Forty-one patients (35.3%) had a prolonged intensive therapy unit length of stay. Multivariate analysis demonstrated that prolonged cardiopulmonary bypass time was the only predictor of postoperative hemorrhage and resternotomy for bleeding (p = 0.03). Increased intensive therapy unit length of stay was predicted by total arch replacement (p = 0.01) and low 6-hour ratio of partial pressure of arterial oxygen to inspired fraction of oxygen (p = 0.05). Increased preoperative creatinine (p = 0.002) and emergency status (p = 0.015) predicted postoperative renal dysfunction. Low 6-hour ratio of partial pressure of arterial oxygen to inspired fraction of oxygen was predicted by increased preoperative creatinine (p = 0.03) and prolonged cardiopulmonary bypass time (p = 0.03).

CONCLUSIONS: Profound hypothermia may cause a coagulopathy, but procedure extent is the primary determinant of postoperative bleeding. Profoundly hypothermic cardiopulmonary bypass does not appear to be a risk factor for renal or early pulmonary dysfunction or intensive therapy unit length of stay.

	Introduction

Top Abstract Introduction Patients and methods Results Comment References

 
Nonneurologic morbidity after cardiac surgery, including hemorrhage and pulmonary and renal dysfunction, is an important risk factor for subsequent mortality and prolonged hospitalization [1–5]. Coagulopathy associated with cardiopulmonary bypass (CPB) may be related to a number of factors, including anticoagulation, systemic inflammatory response, and use of hypothermia. Aortic disease may also in itself predispose to coagulopathy [6, 7]. Ruptured aortic aneurysm, aortic dissection, and infections such as endocarditis are all associated with an increased risk of disseminated intravascular coagulation and altered hemostasis [8, 9]. This is initially the result of activation of tissue thromboplastin as a consequence of exposure to interrupted intravascular surfaces [8, 9]. In addition, aortic surgery creates a number of high-pressure suture lines, often results in longer CPB times than simple valve replacement or coronary artery bypass grafting, and uses prosthetic materials that may predispose toward transgraft hemorrhage [10, 11]. End-organ damage after CPB may occur as a result of a number of factors, including endothelial dysfunction and the systemic inflammatory response, hypoperfusion, and low cardiac output state. This is often manifest as pulmonary or renal dysfunction [3, 4].

Profound hypothermia with circulatory arrest (HCA) has increasingly been used as a method of cerebral protection in aortic surgery, and its neurologic consequences have been well investigated. Profound hypothermic circulatory arrest allows repair of a dissection or aneurysm using an open distal anastomosis technique, which eliminates the need to clamp a fragile or atheromatous aorta [12]. The use of profound HCA, however, is also believed to be associated with significant nonneurologic morbidity such as increased perioperative hemorrhage and postoperative pulmonary and renal dysfunction [11, 13].

A coagulopathy associated with profoundly hypothermic CPB (PHCPB) may be related to a number of factors [8, 14]. Procedures requiring HCA generally have extended CPB times. Profound hypothermia results in alterations to both platelet function and morphology, possibly mediated through prostaglandin inhibition, prolonging bleeding time [8]. Enzymes in the coagulation cascade necessary for effective hemostasis are adversely affected by hypothermia, which retards the formation of a fibrin clot. This is evidenced by increases in activated partial thromboplastin time, prothrombin time, and thrombin time. In addition, there may also be deleterious effects on fibrinolysis. Incomplete rewarming either as a brain protection adjunct or as a consequence of temperature decrease after separation of CPB may also be expected to increase bleeding time and coagulopathy.

We hypothesized that the use of profound hypothermia per se would be a risk factor in the development of nonneurologic morbidity after proximal aortic surgery. We examined this by comparing two cohorts of patients receiving aortic root replacement with or without profound hypothermia for the use of circulatory arrest.

	Patients and methods

Top Abstract Introduction Patients and methods Results Comment References

 
A retrospective analysis of 116 consecutive proximal aortic procedures performed by a single surgical team between 1994 and 2000 was undertaken. Profoundly hypothermic CPB was used to reconstruct the aortic arch or perform an open distal anastomosis in 70 patients.

Bleeding morbidity was assessed by five variables. These were incidence of resternotomy for bleeding or tamponade, chest tube drainage more than 1 L at 12 hours postoperatively or on return to theater if this was sooner, chest drainage at 12 hours postoperatively, blood product administration, and operating room closure time. The 12-hour point was chosen as most patients with significant hemorrhage were declared by this time and a number of patients had had chest tubes removed before the 24-hour point. The criteria for returning to theater were blood loss more than 10 mL/kg in the first hour postoperatively, continued blood loss more than 3 mL/kg in the ensuing hours, or hemodynamic instability suggestive of tamponade. Closure time was defined as end CPB to time leaving the operating room as documented by theater records.

We defined early pulmonary morbidity as impaired early gas exchange (defined as the ratio of the partial pressure of arterial oxygen [in millimeters of mercury] to the fraction of inspired oxygen [PaO₂/FiO₂] less than 200 6 hours postoperatively).

Renal morbidity was indicated by peak creatinine levels more than 200 µmol/L, and the requirement for renal replacement therapy.

We undertook univariate analysis to assess risk factors for bleeding, renal, and pulmonary morbidity. Assessed variables were age, CPB time, PHCPB, moderately hypothermic CPB (MHCPB), emergency status, presence of acute dissection, infective endocarditis, redo operation, type of graft (prosthetic versus homograft), total arch replacement, use of aprotinin, temperature on return to intensive therapy unit (ITU), blood loss more than 1 L, preoperative serum creatinine, closure time, and 6-hour PaO₂/FiO₂ ratio. Multivariate logistic regression analysis was then performed to identify predictors of postoperative blood loss, ITU length of stay more than 2 days, total transfusion requirement, peak creatinine more than 200 µmol/L, and 6-hour PaO₂/FiO₂ ratio less than 200.

Anesthetic and perfusion management
Anesthesia was induced using etomidate, fentanyl, and pancuronium and maintained using intravenous propofol and alfentanil. Cardiopulmonary bypass management included a nonpulsatile roller pump, arterial line filter, membrane oxygenator, and -stat pH management for all patients. Cold blood or cold crystalloid cardioplegia was used for myocardial protection. In addition, patients undergoing PHCPB received 100 mg of dexamethasone and 1 g/kg mannitol approximately 20 minutes before arrest. Topical head-cooling was also used to maintain cranial hypothermia. Circulatory arrest was performed at 15°C, and at all times the gradient between the water bath and nasopharyngeal temperature was less than 10°C. Moderately hypothermic CPB patients were cooled to 28°C. After discontinuation of CPB and securing of surgical hemostasis, the operative site was loosely packed with gauze swabs while protamine was administered. Profoundly hypothermic CPB patients would also receive platelets and fresh-frozen plasma at this time. Their use in MHCPB patients was according to the anesthesiologist's discretion. Profoundly hypothermic CPB patients received 1 to 4 U of fresh-frozen plasma with concurrent hemofiltration during rewarming on CPB. The majority of patients received aprotinin. In MHCPB cases aprotinin commenced at the start of CPB. The full protocol entailed 1 million units intravenously, plus 2 million units into the CPB pump at the beginning of CPB, followed by 0.5 million units per hour intravenously. In PHCPB cases using a modified protocol, loading and infusion commenced after the arrest period with 2 million units into the CPB pump, then 0.5 million units per hour intravenously [9]. The full protocol was used in 42 patients undergoing PHCPB.

Operative technique
A proximal aortic cannulation site was made if possible; otherwise, common femoral arterial return was used. Once the patient was on CPB, core cooling depended on the anticipated need for HCA to conduct an arch repair or construct an open distal anastomosis. Cardioplegia was instilled in antegrade fashion by means of the aortic root or coronary ostia. In the situation in which an allograft conduit was used, the proximal annular anastomosis was performed using simple 4-0 polypropylene interrupted sutures. For prosthetic graft replacement, an everting mattress suture technique was used, with a multifilament 2-0 braided-nylon, polytetrafluoroethylene (Teflon)-buttressed suture. Coronary reimplantation was performed using a modified button technique, with a continuous monofilament 5-0 polypropylene suture. In five cases these anastomoses were buttressed with thin strips of polytetrafluoroethylene or pericardium (autologous or bovine). The left anastomosis was performed first. The anastomoses were then tested with pressurized cardioplegia to inspect for hemorrhage. All patients received 5 mL of fibrin glue (Tisseal, Baxter Hyland Immuno, Norfolk, UK), applied around the aortic root anastomoses once adequate hemostasis had been achieved on pressurized testing. The right coronary anastomosis was fashioned either then or after the distal anastomosis was performed to allow a more accurate placement once the graft was in position. The distal anastomosis was performed using a continuous buttressed suture technique after transection of the aorta in either an open or closed fashion.

Statistical analysis
Means and standard deviations are reported for all variables. The relationship of multiple risk factors to postoperative bleeding and resternotomy, prolonged ITU stay, postoperative renal dysfunction, and early postoperative pulmonary dysfunction was investigated by univariate and multivariate analyses. Univariate analysis was performed using ² or Fisher's exact tests for categorical data and Student's t tests as appropriate. Nonnormally distributed data were transformed. Multivariate forward stepwise logistic regression was performed. Statistical significance was assumed if p was less than 0.05. The cutoff point for entry into the multivariate analysis was p = 0.1; therefore, p values greater than this are not reported in the univariate analysis. The statistical software package SPSS 11.0 (SPSS Inc, Chicago, IL) was used for all calculations using a standard personal computer.

	Results

Top Abstract Introduction Patients and methods Results Comment References

 
There were 46 patients in the MHCPB group and 70 patients in the PHCPB group. Demographic variables are shown in Table 1. Table 2 shows the distribution of procedures performed. Prosthesis selection was determined by patient age and cause of disease. Elderly patients and those with infective endocarditis underwent homograft root replacement primarily.

There were 10 deaths (7 PHCPB, 3 MHCPB; p = 0.74), resulting in mortality rates of 10% and 6.5%, respectively. Causes of death were multiorgan failure in 6 patients, respiratory failure in 2 patients, primary myocardial failure in 1 patient, and late rupture of a descending and thoracoabdominal aortic aneurysm in 1 patient who had undergone arch replacement. There were four neurologic deficits (3 PHCPB, 1 MHCPB).

Mean 12-hour blood loss was 741 ± 700 mL (range 165 to 4,475 mL). The resternotomy rate was 14.7% (17 patients). A technical source for bleeding was identified in 10 patients. The overall incidence of blood loss more than 1 L or return to theater was 25% (29 patients).

Table 3 shows other potential indices of postoperative morbidity. The incidence of lung dysfunction was 12.9% (15 patients), and renal dysfunction occurred in 21.6% (25 patients). Seven patients (5 PHCPB) required renal replacement therapy. Profoundly hypothermic CPB patients had a significantly lower core temperature than MHCPB patients on return to ITU (33.5° ± 1.35°C versus 34.2° ± 1.38°C; p = 0.006). Forty-one patients (35.3%) had an ITU length of stay more than 2 days.

	Comment

Top Abstract Introduction Patients and methods Results Comment References

When analyzed using multivariate stepwise regression analysis the use of profound hypothermia per se was not associated with any of the indices of morbidity studied. Prolonged CPB time was associated with both increased postoperative blood loss and total transfusion requirement. Total arch replacement was a significant predictor of prolonged ITU length of stay. This suggests that it is the increased technical complexity of procedures such as arch replacement that is the key factor that dictates risk of excess hemorrhage. Although profound hypothermia and arch surgery produce a greater propensity for bleeding, the effect of this on patient outcomes can be at least partially counterbalanced by intraoperative strategies to create blood-tight anastomoses and preempt the coagulopathy of profound hypothermia. Overall, the results from this study do compare with other groups [12, 17, 18].

As well as profound hypothermia, the effect of circulatory arrest inducing stasis and contact with a foreign extracorporeal surface may predispose to the formation of microvascular thromboses. These are thought to be a cause of further end-organ damage resulting from the use of HCA [8]. Widespread vascular endothelial dysfunction has been shown to occur after deep HCA, and this may explain any increase in both pulmonary and renal dysfunction compared with those after standard cardiac surgical procedures with use of profound hypothermia [19].

The influence of profound hypothermia on renal and pulmonary function has not been extensively investigated in humans [19, 20]. Significant morbidity has been reported, however, in a number of series [11, 13]. On multivariate logistic regression analysis the predictors of both postoperative renal and early lung dysfunction did not include the use of profound hypothermia. Peak creatinine levels and requirement for renal replacement therapy are widely used as an index of renal dysfunction [10, 21]. Pulmonary dysfunction is often measured by PaO₂/FiO₂ ratio in critical illness [21]. The 6-hour point was chosen to allow for postoperative stability to be attained, but to give a reflection of any perioperative damage sustained, rather than postoperative complications, which were most likely to occur later.

A number of technical aspects are important to minimize potential hemorrhage. First, meticulous primary suture application is vital owing to the subsequent technical inaccessibility of anastomoses. In addition, the use of protein-impregnated grafts has aided the reduction of transgraft hemorrhage [10]. Other intraoperative strategies to improve hemostasis include suture-line buttressing with autologous or xenobiotic pericardium and examination of anastomoses using pressurized cardioplegia or perfusion before their removal from view within the surgical field. Despite these adjuncts, the resternotomy rate for bleeding was high, and most resternotomies identified a bleeding point requiring additional suture placement. Whether these sites of hemorrhage would have been likely to seal spontaneously with normal hemostatic function is unknown. The rate of resternotomy for hemorrhage in other series of aortic surgery varies from 8% to 16.8%, thus comparing with our own [10, 11, 16, 22, 23].

The main limitations of this study were, first, that it was retrospective. Second, there were a number of sources of potential confounding. The aprotinin protocols were not standardized, with the full protocol largely being used for earlier cases and the modified protocol being adopted more recently. Although a subanalysis (not presented) comparing full and modified protocols demonstrated a lower blood loss in the full protocol group, we were persuaded toward the modified protocol because of concerns regarding a possible prothrombotic effect of aprotinin during a period of circulatory stasis [9, 14, 22, 24]. We also acknowledge the perioperative differences between HCA and non-HCA cases, including that of preemptive blood product supplementation and also the use of differing graft materials.

In conclusion, after proximal aortic surgery, prolonged CPB time was the primary determinant of increased bleeding risk. Profound hypothermia was not associated with increased postoperative hemorrhage or renal or early pulmonary dysfunction. Intraoperative hemostatic strategies including the preemptive supplementation of blood products may partially counterbalance the coagulopathy caused by profound hypothermia, thus allowing comparable morbidity outcomes. The interaction between profound HCA and its subsequent morbidity requires further study to be more completely understood.

	References

Top Abstract Introduction Patients and methods Results Comment References

 

1. Moulton M., Creswell L., Mackey M., Cox J., Rosenbloom M. Reexploration for bleeding is a risk factor for adverse outcomes after cardiac operations. J Thorac Cardiovasc Surg 1996;111:1037-1046.[Abstract/Free Full Text]
2. Munoz J., Birkmeyer N., Dacey L., et al. Trends in rates of reexploration for hemorrhage after coronary artery bypass surgery. Ann Thorac Surg 1999;68:1321-1325.[Abstract/Free Full Text]
3. Levy J., Tanaka K. Inflammatory response to cardiopulmonary bypass. Ann Thorac Surg 2003;75(Suppl):S715-720.[Abstract/Free Full Text]
4. Zanardo G., Michielon P., Paccagnella A., et al. Acute renal failure in the patient undergoing cardiac operation. J Thorac Cardiovasc Surg 1994;107:1489-1495.[Abstract/Free Full Text]
5. Ng C., Wan S., Yim A., Arifi A. Pulmonary dysfunction after cardiac surgery. Chest 2002;121:1269-1277.[Abstract/Free Full Text]
6. Fisher D., Yawn D., Crawford E. Preoperative disseminated intravascular coagulation associated with aortic aneurysms. Arch Surg 1983;118:1252-1255.[Abstract/Free Full Text]
7. Cate J., Timmers H., Becker A. Coagulopathy in ruptured or dissecting aortic aneurysms. Am J Med 1975;59:171-176.[Medline]
8. Wilde J. Hematological consequences of profound hypothermic circulatory arrest and aortic dissection. J Card Surg 1997;12(Suppl):201-206.
9. Rooney S., Pagano D., Bognolo G., Wong C., Bonser R. Aprotinin in aortic surgery requiring profound hypothermia and circulatory arrest. Eur J Cardiothoracic Surg 1997;11:373-378.[Abstract]
10. Langley S., Rooney S., Dalrymple-Hay M., et al. Replacement of the proximal aorta and aortic valve using a composite bileaflet prosthesis and gelatin-impregnated polyester graft (Carbo-Seal): early results. J Thorac Cardiovasc Surg 1999;118:1014-1020.[Abstract/Free Full Text]
11. Crawford E., Svensson L., Coselli J., Safi H., Hess K. Surgical treatment of aneurysm and/or dissection of the ascending aorta, transverse aortic arch, and ascending aorta and transverse aortic arch. J Thorac Cardiovasc Surg 1989;98:659-674.[Abstract]
12. Shapira O., Aldea G., Cutter S., Fitzgerald C., Lazar H., Shemin R. Improved clinical outcomes after operation of the proximal aorta: a 10-year experience. Ann Thorac Surg 1999;67:1030-1037.[Abstract/Free Full Text]
13. Okita Y., Ando M., Minatoya K., Kitamura S., Takamoto S., Nakajima N. Predictive factors for mortality and cerebral complications in arteriosclerotic aneurysm of the aortic arch. Ann Thorac Surg 1999;67:72-78.[Abstract/Free Full Text]
14. Westaby S. Coagulation disturbances in profound hypothermia: the influence of anti-fibrinolytic therapy. Semin Thorac Cardiovasc Surg 1997;9:246-256.[Medline]
15. Pagano D., Carey J., Patel R., et al. Retrograde cerebral perfusion: clinical experience in emergency and elective aortic operations. Ann Thorac Surg 1995;59:393-397.[Abstract/Free Full Text]
16. Dossche K., Schepens M., Morshuis W., de la Riviere A., Knaepen P., Vermeulen F. A 23-year experience with composite valve graft replacement of the aortic root. Ann Thorac Surg 1999;67:1070-1077.[Abstract/Free Full Text]
17. Svensson L., Crawford E., Hess K., Coselli J., Safi H. Composite valve graft replacement of the proximal aorta: comparison of techniques in 348 patients. Ann Thor Surg 1992;54:427-439.[Abstract]
18. Ergin M., Galla J., Lansman S., Quintana C., Bodian C., Griepp R. Hypothermic circulatory arrest in operations on the thoracic aorta. J Thorac Cardiovasc Surg 1994;107:788-799.[Abstract/Free Full Text]
19. Cooper W., Duarte I., Thourani V., et al. Hypothermic circulatory arrest causes multisystem vascular endothelial dysfunction and apoptosis. Ann Thorac Surg 2000;69:696-702.[Abstract/Free Full Text]
20. Kirshbom P., Jacobs M., Tsui S., et al. Effects of cardiopulmo-nary bypass and circulatory arrest on endothelium-dependent vasodilation in the lung. J Thorac Cardiovasc Surg 1996;111:1248-1256.[Abstract/Free Full Text]
21. Waydhas C., Nast-Kolb D., Trupka A., et al. Posttraumatic inflammatory response, secondary operations, and late multiple organ failure. J Trauma 1996;40:624-631.[Medline]
22. Kouchoukos N., Wareing T., Murphy S., Perrillo J. Sixteen-year experience with aortic root replacement. Ann Surg 1991;214:308-318.[Medline]
23. Westaby S., Katsumata T., Vaccari G. Aortic root replacement with coronary button re-implantation: low risk and predictable outcome. Eur J Cardiothoracic Surg 2000;17:259-265.[Abstract/Free Full Text]
24. Hornick P., Taylor K. Is aprotinin safe when used in the context of profound hypothermia and circulatory arrest? A literary review. J Heart Valve Dis 1995;4:669-673.[Medline]

This article has been cited by other articles:

				H.-y. Luo, K.-j. Hu, J.-y. Zhou, and C.-s. Wang Analysis of the risk factors of postoperative respiratory dysfunction of type A aortic dissection and lung protection Perfusion, May 1, 2009; 24(3): 199 - 202. [Abstract] [PDF]

				J. P. Schwartz, M. Bakhos, A. Patel, S. Botkin, and S. Neragi-Miandoab Impact of pre-existing conditions, age and the length of cardiopulmonary bypass on postoperative outcome after repair of the ascending aorta and aortic arch for aortic aneurysms and dissections Interactive CardioVascular and Thoracic Surgery, October 1, 2008; 7(5): 850 - 854. [Abstract] [Full Text] [PDF]

				I. Dorotta, P. Kimball-Jones, and R. Applegate II Deep hypothermia and circulatory arrest in adults. Seminars in Cardiothoracic and Vascular Anesthesia, March 1, 2007; 11(1): 66 - 76. [Abstract] [PDF]

				H. Kamiya, C. Hagl, I. Kropivnitskaya, D. Bothig, K. Kallenbach, N. Khaladj, A. Martens, A. Haverich, and M. Karck The safety of moderate hypothermic lower body circulatory arrest with selective cerebral perfusion: A propensity score analysis J. Thorac. Cardiovasc. Surg., February 1, 2007; 133(2): 501 - 509. [Abstract] [Full Text] [PDF]

				A. Della Corte, M. Scardone, G. Romano, C. Amarelli, A. Biondi, L. S. De Santo, M. De Feo, G. Nappi, and M. Cotrufo Aortic Arch Surgery: Thoracoabdominal Perfusion During Antegrade Cerebral Perfusion May Reduce Postoperative Morbidity Ann. Thorac. Surg., April 1, 2006; 81(4): 1358 - 1364. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Stephen J. Rooney Robert S. Bonser Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Harrington, D. K. Articles by Bonser, R. S. Search for Related Content PubMed PubMed Citation Articles by Harrington, D. K. Articles by Bonser, R. S. Related Collections Great vessels