Refractory Spasm of Coronary Arteries and Grafted Conduits Following Isolated Coronary Artery Bypass Surgery

Page 1

AD

ULT

CA

RD

IAC

Refractory Spasm of Coronary Arteries andGrafted Conduits After Isolated Coronary ArteryBypass SurgeryRoberto Lorusso, MD, PhD, Elena Crudeli, MD, Fabiana Lucà, MD,Giuseppe De Cicco, MD, Enrico Vizzardi, MD, Antonio D’Aloia, MD,and Sandro Gelsomino, MD, PhD

Cardiac Surgery and Cardiology Units, Community Hospital, Brescia; and Cardiovascular Research Unit, Careggi Hospital,Florence, Italy

Background. Refractory vascular spasm (RVS) concom-itantly involving the entire coronary artery system andgrafted conduits after coronary artery bypass grafting(CABG) surgery is a rare, but dreadful event. No consen-sus exists in terms of appropriate management.

Methods. Between 1986 and 2009, 5,762 patients under-went isolated CABG at our institution, and 7 patientsexperienced RVS involving the coronary arteries andimplanted conduits. Mean age was 65.6 years and 3 werefemale. All patients received from 3 to 5 distal anastomo-ses, including use of the left internal mammary artery.During the same time period, 18 patients experiencedperioperative vasospasm of a single coronary artery or ofa grafted conduit.

Results. All diffuse RVS events occurred between 3and 8 hours after surgery. All patients had diffuse isch-emic-like electrocardiographic changes, and 5 patientsrapidly developed cardiogenic shock in the intensivecare unit. Angiography was quickly performed in allpatients and showed diffuse RVS involving either thenative coronary arteries or the anastomosed arterial and

venous conduits. The first 5 patients of this series died in

nity Hospital, Piazzale Spedali Civili, 1 25125 Brescia, Italy; e-mail:[email protected].

© 2012 by The Society of Thoracic SurgeonsPublished by Elsevier Inc

the catheterization lab due to rapidly evolving refractorycardiogenic shock and unresponsive cardiac arrest, de-spite intraaortic counterpulsation and aggressive phar-macologic interventions (selective vasodilators and sys-temic inotropes). In the last 2 patients, extracorporealmembrane oxygenation was quickly instituted (1 in thecatheterization lab, 1 in the operating room) and RVScould be successfully managed with complete resolutionof ongoing vasospasm. In the single vascular spasm,there was only 1 death for refractory cardiac arrest,whereas all the other patients were successfully treatedwith direct infusion of vasodilators.

Conclusions. Diffuse RVS after CABG is a rare butlethal condition. Our experience, although limited, indi-cates that in such cases an aggressive treatment, that is,prompt extracorporeal membrane oxygenation institu-tion and controlled cardiocirculatory assistance, repre-sents the preferred solution to face such a dramatic eventand may save patient lives.

(Ann Thorac Surg 2012;93:545–51)

© 2012 by The Society of Thoracic Surgeons

Refractory vascular spasm (RVS) concomitantly in-volving native coronary arteries and grafted con-

duits after coronary artery bypass grafting (CABG) is arare, but life-threatening complication with high morbid-ity and mortality [1–7]. Refractory vascular spasm hasbeen reported to occur between 0.8% and 1.3% of CABGprocedures [2, 3], although transient coronary or graftspasm has been shown to affect up to 11% of operatedpatients [8]. Usually, vascular spasm involves a singleimplanted graft or a single coronary artery [1–4] but afew reports have described a more diffuse form [5–13]that is usually characterized by extremely difficult man-agement [12–14] and may lead to dismal outcome [13].The mechanisms underlying this perioperative compli-cation are poorly understood. Vascular damage, high

Accepted for publication Sept 28, 2011.

Address correspondence to Dr Lorusso, Cardiac Surgery Unit, Commu-

levels of administered or endogenous vasoconstrictors,oxidative stress, and other predisposing factors havebeen claimed to be the determinants of such an event [3],but no conclusive agreement exists in this respect.Whether the isolated and the diffuse patterns of RVS maydiffer in terms of predisposing or causative factors is alsoanother unsolved issue. In terms of treatment, a fewoptions have been proposed, with variable success. Sys-temic intravenous infusion of vasodilators appeared of-ten to be insufficient to counteract vascular hyperreactiv-ity. Intracoronary or intragraft perfusion withvasodilators (mainly adenosine or calcium channel an-tagonists) therefore has been applied in an attempt tosolve the spasm of coronary arteries and CABG conduits[2–16]. In some cases, also, this therapy has shown to beinsufficient to face a progressively marked and refractorycardiocirculatory impairment induced by severe myocar-dial hypoperfusion and subsequent cardiocirculatory col-

lapse [1, 10–13]. The application of an intraaortic balloon

0003-4975/$36.00doi:10.1016/j.athoracsur.2011.09.078

Page 2

pc

aodaalAvig

gtct

ltfnwpaiatti

fcdop

Cds

Ca

546 LORUSSO ET AL Ann Thorac SurgCORONARY AND GRAFT SPASM AFTER CABG 2012;93:545–51

AD

ULT

CA

RD

IAC

pump (IABP) is obviously mandatory to increase coro-nary and graft flow and support the failing ventricle, butquick deterioration and cardiac arrest may yet develop,particularly in the condition of diffuse RVS [1, 5–9, 13].We reviewed our limited clinical experience in relation toRVS after CABG to describe the clinical scenario and thepatient outcome according to the type of therapy adoptedin this setting.

Material and Methods

Among 5,762 CABG procedures performed at our insti-tution (Cardiac Surgery Unit, Community Hospital,Brescia, Italy) from 1986 to 2009, there were 7 cases ofdiffuse RVS involving concomitantly the coronary arter-ies and the grafted conduits during the early postopera-tive period. During the same time period, there were 18patients (as demonstrated by postoperative angiography)who experienced perioperative vasospasm involving asingle native coronary artery, or a single grafted coronaryconduit. All patient charts and coronary angiographywere carefully reviewed to assess preoperative and peri-operative factors potentially linked to RVS occurrence.Ethical Committee approval was waved due to the retro-spective analysis of the study according to the nationallaw regulating observational retrospective studies (lawNo. 11960, released on July 13, 2004).

Preoperative and operative patient features are shownin Table 1. Patient groups did not differ in terms of

reoperative medications, medical history, or anatomicharacteristics of coronary artery disease.

In all cases of both groups, the left internal mammaryrtery (LIMA) was harvested as a pedicle conduit bypening the left pleura, and no case of suboptimal flow orimension was reported. All patients received distalnastomosis with LIMA on the left anterior descendingrtery and saphenous veins on the other branches of theeft coronary circulation or on the right coronary artery.ll coronary anastomoses were performed on the targetessel with stenosis affecting more than 75% of thenternal lumen at angiography. A few patients in bothroups had occluded coronary vessels (Table 1).Cold crystalloid cardioplegia (antegrade and retro-

rade) was used in all patients, with no difference be-ween groups in terms of total amount. Extracorporealirculation circuit and management did not differ be-ween groups.

Surgical procedures were uneventful, and no pro-onged total extracorporeal circulation and aortic clampimes were recorded in either group (Table 1). Weaningrom extracorporeal circulation was smoothly achieved inormothermia and with appropriate hemodynamics,ith no need for inotropic agents in all cases, even in theatients with preoperative low ejection fraction. Lactatesnd gas analysis parameters were within normal rangesntraoperatively and postoperatively until the appear-nce of electrocardiographic (ECG) changes related tohe onset of acute myocardial ischemia. Body tempera-ure was also within normal range in all the patients on

ntensive care unit arrival and beyond.

Results

In the patients with diffuse RVS, appearance of ECGischemic changes occurred within 8 hours from surgery(range, 3 to 8 hours; mean, 5.6 hours) and with rapidhemodynamic compromise in 5 cases. There was noapparent predisposing factor in terms of intraoperativeconduct or event, total blood loss, gas exchange, lactatelevels, pH, myocardial enzymes, systemic arterial pres-sure, or body temperature in the operative theatre or inthe intensive care unit. Two patients underwent trans-esophageal echocardiography in the intensive care unitwhile in stable hemodynamic conditions, which showed

Table 1. Preoperative and Operative Characteristics ofPatients With Isolated or Diffuse Spasm

Characteristic Isolated Spasm Diffuse Spasm

Number of patients 18 7Gender (male to female) 11:7 4:3Age (years) 67.2 � 6 65.6 � 7Angina at rest 10 3aHypertension 14 5bHyperlipidemia 11 4cDiabetes mellitus 8 3LV ejection fraction

� 0.30 3 10.30–0.50 9 4� 0.50 7 2

Target-vessel stenosis(vessel number)

75%–90% 11 37�90% 5 9100% 4 5

Left main stenosis 3 7CPB time (minutes) 61–104 68–95Aortic cross-clamp time

(minutes)33–57 36–48

Number of grafts(mean)

3.3 (2–5) 3.4 (3–5)

Site of vascular spasm All native andgraftedvessels

Right coronary 8Left anterior

descending1

Circumflex artery 2LIMA graft 3Vein graft 4

a Blood pressure � 140/90 or a history of high blood pressure or the needor antihypertensive medications. b Serum low-density lipoproteinholesterol � 140 mg/dL or treatment with a statin. c Symptoms ofiabetes mellitus plus casual plasma glucose concentration � 200 mg/dLr fasting plasma glucose level � 126 mg/dL or 2-hour postprandiallasma glucose � 200 mg/dL after a 75-g glucose load.

ontinuous variables are presented as mean � standard deviation;iscrete variables are presented as n. Nonparametric variables are pre-ented as median [interquartile range].

PB � cardiopulmonary bypass; LIMA � left internal mammaryrtery; LV � left ventricular.

akinesia originally localized and progressively involving

Page 3

pdpsehItp

ni

547Ann Thorac Surg LORUSSO ET AL2012;93:545–51 CORONARY AND GRAFT SPASM AFTER CABG

AD

ULT

CA

RD

IAC

a larger myocardial area, eventually leading to low car-diac output syndrome. All patients were immediatelytransferred to the catheterization lab for coronary an-giography. The IABP was inserted in the intensive careunit in 2 patients, and upon arrival in the catheterizationlab in the other 5 patients, respectively. In 2 patients, thehemodynamic conditions deteriorated just prior to thetransfer to the catheterization lab, while the other 5patients developed low output syndrome during coro-nary angiography.

In all patients, diffuse spasm (along the entire coronaryand graft courses) of the native coronary arteries and ofthe implanted grafts could be shown at angiography (Fig1). Besides IABP support, repeated boli of vasodilatorsagents (0.2 mg diluted in 2 mL of nitroglycerin, 0.2 mg ofverapamil, and adenosine) were selectively and repeat-edly infused in the coronary arteries and into the CABGconduits without resolution of the ongoing spasm. Ino-tropic drugs were also subsequently administered in allpatients, after selective vasodilative drug infusion, tosupport the heart and the circulation to counteract ongo-ing cardiogenic shock. Five patients rapidly deterioratedand died for refractory ventricular fibrillation and cardiacarrest despite several unsuccessful external ventriculardefibrillations and despite prolonged resuscitation ma-neuvers. Patient 6 had 2 episodes of ventricular fibrilla-tion treated by external defibrillation. After the secondepisode of ventricular fibrillation, which required pro-longed cardiopulmonary resuscitation, the patient waspromptly transferred to the operating room. As soon asthe chest was reopened, ventricular fibrillation reoc-curred. Immediate institution of extracorporeal mem-brane oxygenation (ECMO) was carried out by means ofaortic and right atrial cannulation. At visual inspection,native coronary arteries as well as the grafted conduitsappeared extremely spastic. While maintaining full car-diocirculatory support, progressive resolution of coro-nary spasm was macroscopically evident. As no pulsationwas still apparently present along the LIMA conduit, anew vein graft was deemed necessary and anastomosedto the left anterior descending artery on a beating heart.After coronary incision no blood flow came out from thecoronary artery. The upstream LIMA-coronary anasto-mosis was retrogradely checked by vascular probe andshown to be fully patent. The LIMA graft was nonethe-less kept clamped during the accomplishment of thedistal anastomosis. At the completion of the distal veingraft anastomosis and before fashioning the proximalanastomosis, the LIMA was reopened and blood flowfrom within the coronary artery unexpectedly appeared,indicating a reestablished flow through the arterial con-duit. Proximal anastomosis of the vein graft was finallyperformed. The ECMO was maintained for 2 hours anddiscontinued upon observation of complete resolution ofischemic-like ECG changes, while maintaining the IABPsupport. Postoperative echocardiography performed inthe operating theatre and in the intensive care unitdetected a gradual recovery of myocardial contractilitywith nearly normal cardiac function at hospital dis-

charge, which occurred at postoperative day 9. i

In patient No.7, femoro-femoral venoarterial ECMOwas applied in the catheterization lab under externalcardiac massage, and after almost an hour the ischemic-like ECG changes slowly disappeared, with normalizedand stable hemodynamic conditions. Control coronaryangiography showed resolution of the global vascularspasm. The ECMO was gradually discontinued (6 hours)and IABP maintained for 48 hours, with complete reso-lution at echocardiographic assessment of the left ven-tricular dysfunction. The patients were eventually dis-charged on postoperative day 8.

In contrast to diffuse RVS patients, all patients whoexperienced single coronary or graft spasm had func-tional stenosis affecting a limited portion of the coronaryor grafted conduit (from 0.5 cm to 10 cm). In these cases,ECG-related changes also appeared during the first post-operative hours (from 2 to 7 hours postoperatively; mean3.4 hours). Four patients experienced hemodynamic de-terioration (2 transient cardiac arrest), whereas all theothers had no major cardiocirculatory compromise. Allpatients in the single-spasm group, except 1 (refractorycardiac arrest and electromechanical dissociation) weresuccessfully managed with IABP and selective vasodila-tor infusion. Eight patients of this group experiencedovert signs of acute myocardial infarction, with ECG-related changes and significant rise in blood cardiac-specific enzymes. All 17 survivors had a subsequentuneventful course with a mean length of stay of 5 to 9days (range 5 to 8 days).

Comment

Perioperative spasm after CABG procedure was firstdescribed by Pichard and colleagues [17]. After thatdescription, a few other publications have reported aboutthe occurrence of RVS [1–15]. The RVS after CABGremains mostly unexplained, and once developed repre-sents a significant determinant of morbidity and mortal-ity during the perioperative phase [1–15]. It may occurintraoperatively soon after CABG or during the firsthours in the intensive care unit, with an incidence rang-ing from 0.8% to 1.3%, according to different reports [2, 3].A continuous Holter monitoring during the first hoursafter surgery, however, has shown that transient coro-nary or graft spasm may reach up to 11.5% of patientssubmitted to CABG [8]. Late presentation of graft spasmalso has been described [13, 18]. These percentages

robably underestimate the real incidence of RVS. In-eed, if prompt cardiac catheterization is delayed or noterformed, diagnosis cannot be achieved and coronarypasm may remain unknown. Further studies based onarly angiographic control of perioperative ischemiaave highlighted this undiagnosed spasm rate [8, 19].

nterestingly, long-term follow-up of RVS patients seemso indicate that patients are no longer at risk for lateostoperative recurrence of RVS-related events [20].Although the pathophysiology of perioperative coro-

ary spasm has not been elucidated in detail, severalnsights on the mechanisms and predisposing factors,

ncluding smoking, diabetes, lipid metabolic disorders,

Page 4

548 LORUSSO ET AL Ann Thorac SurgCORONARY AND GRAFT SPASM AFTER CABG 2012;93:545–51

AD

ULT

CA

RD

IAC

Page 5

549Ann Thorac Surg LORUSSO ET AL2012;93:545–51 CORONARY AND GRAFT SPASM AFTER CABG

AD

ULT

CA

RD

IAC

and gene expression, have been proposed [2, 3, 21].History of variant angina does not appear to be a predis-posing factor [22]. The relationship between the severityof native-vessel stenosis and graft patency has beenpreviously reported for LIMA, gastroepiploic, radial ar-tery, and saphenous vein grafts [3, 23, 24], but no majorpredictors have been identified for such a frighteningevent. Adhesion molecules, such as intercellular adhe-sion molecule–1 and chemoattractant cytokines, seem toplay a role in the early reperfusion injury of the revascu-larized myocardium by regulating the adhesion of neu-trophils to vascular endothelial cells, and their subse-quent transmigration [25]. Such a process seems topredispose to increased vascular tone and, therefore, toenhance exacerbated response to any potential vasoac-tive stimulus (like endogenous catecholamines, ino-tropes, low blood flow, and low pressure values) [25]. Thetranscription-dependent synthesis of cytokines and ad-hesion molecules ensures that the inflammatory re-sponse, elicited by ischemia-reperfusion, occurs for sev-eral hours after the reperfusion [25]. Chemotactic signalsalso provide stimuli for smooth muscle cell proliferationand migration. These ongoing subclinical host responsesmay also lead to intimal swallowing and vessel narrow-ing, or, in turn, to a sort of proactive condition of coronaryor coronary-related (grafts) vessels [3].

Perioperative vascular spasm has been described afterCABG or even after valve surgery, and, in the former,both in on-pump and off-pump procedures [4–7]. It hasbeen suggested that constant administration of vasopres-sors needed to support low blood pressure during off-pump coronary surgery could represent an additionalvasoactive stimulus [4–7].

Spasm may involve an isolated conduit [1] or a graftedartery [4], an uninvolved native vessel [2], or be diffuse[5–13]. The presence of a common or different pathoge-netic mechanism in case of isolated coronary or graftspasm as opposed to diffuse RVS remains undetermined.The global and concomitant involvement of coronary-related vessels (either native or grafted) may indicate thatsuch a hypercontractile vasoresponse may be linked toeither circulating vascular-related activators of vascularmuscular tone or other systemic factors which are, atpresent, still undefined. Iatrogenic wall damage (vesselmanipulation) or other potential causes of a limitedincrease in vascular tone of the native coronary or graftedconduit (local platelet aggregation and release of vaso-constrictive substances like serotonin or thromboxane A2

by a ruptured or injured plaque) are more likely to berelevant in the single spastic event [3, 4]. Notably, theisolated spasm of native coronary arteries, as confirmed

Fig 1. Coronary angiograms of a representative patient (Patient No. 5angiogram showing (A) preoperative and (B) postoperative right coronpostoperative angiogram performed at 8 hours from surgery in Patien(D) postoperative angiograms of the left coronary artery. Diffuse spasnary artery. (E) Selective angiogram of the left internal mammary art

along the entire vessel course.

also in our experience of the patient group experiencingisolated spasm, affects more frequently the right artery[2]. Furthermore, the spastic right coronary artery isfrequently not directly involved in CABG procedure orby less severe atherosclerosis, suggesting, therefore, thatvessel manipulation during CABG or atherosclerosis-related complications are not the sole or predominantpredisposing factors for such events. Based on thesefacts, we may speculate that postoperative RVS mayaccount for several patterns and different pathophysio-logical processes, but, in the majority of the cases, re-mains unexplained. Unfortunately, the life-threateningscenarios of these patients do not usually allow perfor-mance of any kind of assessment of blood parameters orother peculiar diagnostic work-up. Further investiga-tions, particularly to elucidate whether cardiac-relatedvascular spasm, either isolated or diffuse, is associated toincreased vascular tones in other body districts, wouldprovide useful insights.

Intracoronary injection of vasodilators, has beenshown to be effective in the presence of perioperativeRVS [4–17, 26], but may be not sufficient in the presenceof profound cardiogenic shock or refractory cardiac arrestand to face subsequent marked myocardial stunning.Intravenous infusion of levosimendan, coupled with aleft ventricular assist device, has been reported to con-tribute to sustained control of vasospasm, providingadditional vasodilatation and inotropic support for sev-eral days [11].

Our case series of diffuse RVS included 7 patientsaffected by isolated coronary artery disease, developingsevere coronary artery and graft spasm after on-pumpcoronary artery surgery. In our experience, coronaryangiography was quickly performed in all patients toidentify the etiology of ischemic-like ECG changes. Fivepatients died in the catheterization room, despite useIABP, systemic inotropes, vasopressors, and selectivevasodilators, as well as prolonged resuscitation maneu-vers. In the last 2 patients, based on the negative expe-rience previously observed, a more aggressive approachwas instituted, with almost immediate beneficial effects,either in terms of ECG tracings or vascular response, and,more importantly, with successful final outcome.

In our experience, in the presence of diffuse and refrac-tory RVS, selective intracoronary infusion of vasodilatorsand the mechanical augmentation of vessel flow wereineffective, calling for more effective cardiocirculatory sup-port. The use of IABP, although usually effective in mostcases of reduced coronary or CABG conduit flows, isobviously insufficient or useless in the presence of severecardiogenic shock or cardiac arrest. The need for more

he diffuse refractory vascular spasm group. Right-oblique coronaryartery. Diffuse coronary spasm (arrowheads; B) is evident in the6. Right-oblique coronary angiogram showing (C) preoperative androwheads; D) is also evident in the left anterior descending coro-rrowheads) which shows diffuse narrowing of the arterial conduit

) of tary

t No.m (arery (a

Page 6

550 LORUSSO ET AL Ann Thorac SurgCORONARY AND GRAFT SPASM AFTER CABG 2012;93:545–51

AD

ULT

CA

RD

IAC

powerful left ventricular support to counteract RVS-relatedcardiocirculatory compromise has been recently reported[11]. The efficacy of a full heart and lung support withECMO in case of unresponsiveness to conventional thera-pies also has been recently shown [14, 27]. The advantage ofcomplete cardiopulmonary assistance may be multifacto-rial. The prompt institution of cardiopulmonary bypassmay decompress the suffering and distended heart. Fur-thermore, because ECMO provides immediate effectivesupport to the ischemic myocardium and increases bloodoxygenation, usually impaired because of acute lungedema, the catecholamine release may be reduced, myo-cardial oxygen consumption usually rapidly declines, andthe marked ongoing metabolic acidosis may be counter-acted, eventually blocking the vicious cycle installed by arefractory myocardial ischemia and contractile depression.Extracorporeal membrane oxygenation effectively supportsthe systemic circulation, while reducing left ventricularworkload, both of which are advantageous in compromisedpatients with ischemic left ventricular failure and cardio-genic shock, or obviously lifesaving in case of cardiac arrest.The ECMO may also provide blood depuration from cyto-kines and other molecules likely implied in vasospasm,through hemofiltration that has already been pointed out asa determinant strategy to restore hemodynamic stability[14, 27]. Our last 2 successful cases, characterized by expe-dient institution of ECMO, clearly showed that, soon afterestablishing full cardiocirculatory assistance, a progressiveresolution of the coronary and graft spasm may occur withresumption of grafted conduit and coronary flow. In case ofdiffuse RVS after CABG, concomitantly involving the coro-nary tree and the grafted conduits, the global myocardialischemia certainly poses additional risk and representsalmost a lethal condition. Therefore, the application of fullcardiac and lung assistance may enhance a prompt solu-tion, allowing at the same time a quicker and more effectivemeans to also treat marked ongoing metabolic acidosis andlung impairment, which usually exacerbates vasospasm.Furthermore, the latest portable ECMO devices will prob-ably represent an additional valuable tool for an expedientapplication of such a therapy in the intensive care units orin the catheterization lab; maybe preventing profound car-diogenic shock and refractory cardiac arrest [28].

In conclusion, the occurrence of diffuse RVS after CABGis a rare, yet potentially lethal condition. The detection andinterpretation of perioperative spasm occurring after CABGstill remain a challenging problem for the clinician. Promptdiagnosis and aggressive treatment are mandatory. Partialcardiocirculatory support, such as IABP, as well as selectiveinfusion of vasodilators in the coronary arteries and graftedconduits, are often successful to counteract coronary orgrafted conduits but may fail to face refractory low cardiacoutput syndrome or reverse cardiac arrest in case of diffuseRVS. The prompt application of ECMO may represent avaluable option, acting as a temporary cardio-respiratory-circulatory assistance aimed at prompt correction of ongo-ing left ventricular compromise, and metabolic derange-ment and acidosis due to reduced blood oxygenation andperipheral organ perfusion. The peripheral approach by

femoral vessel cannulation and the ECMO application

already in the catheterization lab during coronary angiog-raphy, or even in the intensive care unit in case of severepatient deterioration, may represent an additional strategythat may lead to a favorable patient outcome otherwisecharacterized by a dismal prognosis. Further studies arewarranted to elucidate which preoperative or surgical con-dition may generate or predispose to such a dramaticcomplication.

References

1. Paterson HS, Hones MW, Baird DK, Hughes CF. Lethalpostoperative coronary artery spasm. Ann Thorac Surg 1998;65:1571–3.

2. Buxton AE, Goldberg S, Harken A, Hirshfeld J Jr, Kastor JA.Coronary artery spasm after myocardial revascularization.N Engl J Med 1981;304:1249–53.

3. Lemmer JH Jr, Kirsh MM. Coronary artery spasm follow-ing coronary artery surgery. Ann Thorac Surg 1988;46:108 –15.

4. Trimboli S, Oppido G, Santini F, Mazzucco A. Coronaryartery spasm after off-pump coronary artery bypass grafting.Eur J Cardiothorac Surg 2003;24:830–3.

5. Bittner HB. Coronary artery spasm and ventricular fibrilla-tion after off-pump coronary surgery. Ann Thorac Surg2002;73:297–300.

6. Fukui T, Takanashi S, Hosoda Y. Coronary artery and graftspasm after off –pump coronary artery bypass grafting. JpnJ Thorac Cardiovasc Surg 2005;53:109–13.

7. Kitamura H, Nakayama K, Kitano T. A case of refractoryspasm of multiple coronary arteries and left internal thoracicartery graft following off pump CABG. [Article in Japanese]Kyobu Geka 2001;54:1103–6.

8. Hannes W, Fasol R, Zajonc H, et al. Diltiazem providesanti-ischemic and anti-arrhythmic protection in patientsundergoing coronary bypass grafting. Eur J Cardio ThoracSurg 1993;7:239–45.

9. He GW, Fan KY, Chiu SW, Chow WH. Injection of vasodi-lators into arterial grafts through cardiac catheter to relievespasm. Ann Thorac Surg 2000;69:625–8.

10. Zaiac M, Renzulli A, Hilton CJ. Coronary artery spasmfollowing coronary artery bypass grafting. Eur J Cardiotho-rac Surg 1990;4:109–11.

11. Dopfmer UR, Braun JP, Grosse J, Konertz W. Temporaryleft ventricular assist assist and levosimendan for coro-nary artery spasm. Interact J Cardiovasc Thorac Surg2005;4:316 – 8.

12. Ashar V, Niccoli G, Porto I. Diffuse saphenous vein graftspasm. J Invasive Cardiol 2004;16:667–8.

13. Sarandia D, Benassi F, Massarenti L, D’anniballe G, Ste-fanelli G. Intractable spasm of saphenous vein conduits aftercoronary artery bypass grafts. J Cardiovasc Med (Hager-stown) 2010 Feb 11 [Epub ahead of print].

14. Smith C, Akowuah E, Theodore S, Brown R. Diffuse coronaryartery spasm treated by extracorporeal membrane oxygen-ation. Eur J Cardiothorac Surg 2009;36:208–9.

15. Kimura N, Kawahito K, Adachi K, et al. Effects of intra-coronary and intra-graft adinistration of nicorandil for cor-onary spasm after coronary artery bypass grafting. [Articlein Japanese] Kyobu Geka 2006;59:71–7.

16. Tepliakop AT, Vecherskiı IuIu, Mamchur SE, et al. Postop-erative spasm of mammary-coronary grafts and possibilitiesof its correction by dihydropyridine calcium antagonistsnifedipine and amlodipine. [Article in Russian] Kardiologiia2002;42:9–14.

17. Pichard AD, Ambrose J, Mindisch B, et al. Coronary arteryspasm and perioperative cardiac arrest. J Thorac Cardiovasc

Surg 1980;80:249–54.

Page 7

2

2

2

2

2

lracsicarotdaio

gr

551Ann Thorac Surg LORUSSO ET AL2012;93:545–51 CORONARY AND GRAFT SPASM AFTER CABG

AD

ULT

CA

RD

IAC

18. Ichikawa Y, Kajiwara H, Hamada T, Yamazaki I, Jin Y, SatoS. Late vasospasm of the inferior epigastric artery graft. AnnThorac Surg 1999;67:1800–1.

19. Fabricius AM, Gerber W, Hanke M, Garbade J, AutschbachR, Mohr FW. Early angiographic control of perioperativeischemia after coronary artery bypass grafting. Eur J Cardio-thorac Surg 2001;19:853–8.

20. Buxton AE, Hirshfeld JW Jr, Untereker WJ, et al. Periopera-tive coronary arterial spasm: long term follow-up. Am JCardiol 1982;50:444–51.

21. Miwa K, Fujita M, Sasayama S. Recent insights into themechanisms of predisposing factors and racial differences ofcoronary vasospasm. Heart Vessels 2005;20:1–7.

22. Kawano H, Ogawa H. Endothelial function and coronaryspastic angina. Intern Med 2005;44:91–9.

23. Sabik JF III, Lytle BW, Blackstone EH, Khan M, Houghtaling PL,

Cosgrove DM. Does competitive flow reduce internal thoracicartery graft patency? Ann Thorac Surg 2003;76:1490–6.

elated catecholamine release, hypoxia, and oxidative

srpnb

asngbbe

H

DS3Se

© 2012 by The Society of Thoracic SurgeonsPublished by Elsevier Inc

4. Desai ND, Cohen EA, Naylor CD, et al. A Randomizedcomparison of radial artery and saphenous-vein coronarybypass grafts. N Engl J Med 2004;351:2302–9.

5. Carden DL, Granger DN. Pathophysiology of ischaemia-reperfusion injury. J Pathol 2000;190:255–66.

6. Uydes-Dogan BS, Nebigil M, Aslamaci S, Onuk E, Kanzik I,Akar F. The comparison of vascular reactivities of arterialand venous grafts to vasodilators: management of graftspasm. Int J Cardiol 1996;53:137–45.

7. Min HK, Lee YT, Kim WS, Kim SW. The role of percutane-ous support in the treatment of native coronary spasm aftercoronary artery bypass grafting. Heart Surg Forum 2010;13:E126–8.

8. Phillip A, Artl M, Amann M, et al. First experience with theuse of ultra compact mobile extracorporeal membrane oxy-

genation system Cardiohelp in interhospital transport. In-teract Cardiovasc Thorac Surg 2011;12:978–81.

INVITED COMMENTARY

Rare but life-threatening is diffuse refractory spasm ofthe coronary arteries and bypass conduits 3 to 8 hoursafter operation, whereas spasm of a single coronary orgraft is more prevalent and infrequently fatal. Extracor-poreal membrane oxygenation appears to be necessaryfor survival in the former, but intraaortic balloon coun-terpulsation and vasodilators are usually adequate ther-apy for the latter.

Lorusso and colleagues [1] report their experience in aarge number of patients, in contrast to mostly caseeports. I have queried two senior, invasive cardiologists,nd combined with my exposure in three academicenters (one going back to 1970), we could not recalleeing diffuse coronary spasm after bypass operations. Its possible that the authors have more vigorously usedoronary angiography to assess patients requiring vaso-ctive support, intraaortic balloon counterpulsation, oresuscitation after coronary operations to explain theirbservations. Although their patients left the operatingheater without inotropic support, it is possible thaturing resuscitation immediately before, during transfer,nd after arrival in the catheterization suite, there wasnfusion of large doses of inotropic agents or vasopressors,r both, which resulted in refractory vascular spasm.Etiology is multifactorial and includes coronary and

raft manipulation, exogenous vasoconstrictors, stress-

tress. The propensity for arterial conduit spasm is wellecognized and variably treated. We use intraluminalapaverine and, occasionally, nitroglycerin; others preferitroglycerin, nitroprusside, and calcium channellockers.It is imperative that those doing coronary operations

re mindful of the potential for graft as well as coronarypasm in the perioperative interval and the appropriate-ess of diagnosis by angiography. Intracoronary andraft infusion of vasodilators combined with intraaorticalloon counterpulsation are adequate in most instances,ut extracorporeal membrane oxygenation has been nec-ssary for more extensive or resistant coronary spasm.

endrick B. Barner, MD

ivision of Cardiothoracic Surgeryt. Louis University635 Vista Ave, DT 8th Flrt. Louis, MO 63110-mail: [email protected]

Reference

1. Lorusso R, Crudeli E, Lucà F, et al. Refractory spasm ofcoronary arteries and grafted conduits after isolated coronary

artery bypass surgery. Ann Thorac Surg 2012;93:545–51.

0003-4975/$36.00doi:10.1016/j.athoracsur.2011.10.014