Single Center 7 Year Experience of Coronary Artery Perforation: Angiographic and Procedural Characteristics, Management and Outcome.

CONTEXT: Coronary artery perforation is a rare but potentially catastrophic complication of percutaneous coronary intervention (PCI). It is infrequent complication of PCI. AIMS: The objective of the study is to report the 7-year experience of coronary artery perforation with respect to incidence, clinical and angiographic characteristics, management and outcomes. SETTINGS AND
DESIGN: The study involved retrospective analysis of single centre 7 years of percutaneous coronary intervention data. Patients who had complication of coronary artery perforation during PCI were identified and included in the study. SUBJECTS AND METHODS: Retrospective analysis of clinical, angiographic and procedural characteristics as well as management and outcome of coronary artery perforation was done. STATISTICAL ANALYSIS USED: The whole data were tabulated, variables were presented as mean and percentages and comparison was done within them.
RESULTS: A total of 37 cases of coronary artery perforation were identified from 4532 PCI performed. Most of the coronary artery perforation belonged to Ellis Type II and Type III (both n = 15) followed by Type III CS and Type I. Lesions belonged to AHC/AHA Type C in 31 cases. Most frequent mechanism of coronary artery perforation was related to the use of guidewire and balloon (both n = 17). The total of 8 cases presented with cardiac tamponade requiring pericardiocentesis. Eleven cases required emergency covered stent implantation. In two cases microcoil was used while one case required polyvinyl alcohol particles to seal the perforation site. There was no in-hospital mortality while 30-day mortality occurred in one patient. One case was referred for emergency surgery.
CONCLUSIONS: Coronary artery perforation is rare but potentially fatal complication of percutaneous coronary intervention. Complication of coronary artery perforation can be managed effectively in the catheterization laboratory without the need of emergency of bailout surgery and in-hospital outcomes remain good in the majority of cases. Copyright:

INTRODUCTION

Coronary artery perforation (CAP) is a rare but potentially catastrophic complication of percutaneous coronary intervention (PCI). CAP can be caused by various percutaneous devices used during PCI such as wire, balloon, stent or atheroablative devices. Historically emergency surgery used to be standard treatment. However the development of covered stents and coil has enabled many cases of CAP to be treated in the catheterization laboratory. Despite improvements in interventional devices and skills, the incidence of CAP remains fairly constant.[1234567] This is probably due to increasing complexities of PCI performed nowadays. Even with increase in the number of PCI performed around the world, CAP remains an infrequent complication of PCI which is reflected in smaller sizes of the series published so far.[1234567]

The objective of the study is to report a single center experience of CAP with respect to incidence, clinical and angiographic characteristics and outcomes.

SUBJECTS AND METHODS

Study designs and participants

The study involved a single center, retrospective analysis of 7 year PCI data of patients treated from June 2011 to May 2018. During this period, patients who had complication of CAP during PCI were identified. Retrospective analysis of clinical, angiographic and procedural characteristics as well as the management of CAP and outcome was done.

Coronary artery perforation classification and definitions

CAPs were categorized using Ellis classification into Types I, II, III, III cavity spilling (III CS).[1] Cardiac tamponade was defined by the presence of pericardial effusion by fluoroscopy or echocardiography, systemic hypotension with systolic blood pressure <90 mmHg, presence of pulsus paradoxus and echocardiographic evidence of the right ventricle free wall diastolic collapse.

Periprocedural myocardial infarction was defined by the elevation of biomarker troponin T or creatine kinase-muscle/brain more than five times the upper limit of normal value within 24 h of PCI.[8]

Procedure

All the cases with CAP were analyzed with respect to clinical, angiographic and procedural characteristics and their management and outcome was registered.

Clinical analysis – baseline clinical characteristics such as different risk factors for coronary artery disease, antiplatelet therapy and indication of PCI were registered.

Angiographic analysis – detailed angiographic analysis was done for every case including the vessel involved and lesion location. Lesion type was decided as per the ACC/AHA classification taking into consideration different angiographic characteristics of lesions such as calcification, tortuosity, chronic total occlusion and lesion length[9]

Procedural analysis – mechanisms of perforation like predilation or postdilation with balloon, stent implantation, wire exit perforations, type of coronary guidewire used (hydrophilic/stiff), any rotablator or cutting balloon related perforations were registered in relation to the type of CAP.

RESULTS

A total number of 4532 PCI were performed from June 2011 to May 2018 [Table 1]. PCI included interventional procedures such as balloon angioplasty, stenting, fractional flow reserve, optical coherence tomography, and rotablation. Complication of CAP occurred in total of 37 cases which were included in the study for detailed analysis. Most of the CAP belonged to Ellis Type II (n = 15) and Type III (n = 15) followed by Type III CS (n = 4) and Type I (n = 3).

Table 1

Total percutaneous coronary intervention and coronary artery perforation

PCI and CAP	n
Total number of PCI	4532
Total number of CAP (%)	37 (0.79)

PCI: Percutaneous coronary intervention, CAP: Coronary artery perforation

Baseline clinical variables

Of 37 cases of CAP, indication for PCI was ST-elevation myocardial infarction in two cases, non-ST elevation acute coronary syndrome in six cases while stable angina with 24 cases formed the majority group for PCI indication [Table 2]. Mean age of the patients was 66.7 years and 22 were male.

Table 2

Indication for procedure

Indication of PCI	n
ST elevation myocardial infarction	2
Non-ST elevation- acute coronary syndrome	6
Stable angina	24
Silent ischemia	4
Ischemic left ventricle dysfunction	1

Of 37 cases, hypertension was present in 19 while diabetes mellitus was present in 15. Risk factor such as smoking was present in 8 and dyslipidemia was found in 17. Renal dysfunction was present in two cases. Majority of the patients were on dual antiplatelet aspirin and clopidogrel. In addition, 32 patients received glycoprotein IIb IIIa inhibitor tirofiban [Table 3].

Table 3

Baseline clinical variables

Clinical variables	n (%)
Mean age (years)	66.7
Sex
Male	22
Female	15
Hypertension	19 (51)
Diabetes mellitus	15 (40)
Smoking	8 (21)
Dyslipidemia	17 (46)
Family history of coronary artery disease	9 (24)
Chronic renal disease/renal dysfunction	2 (5)
Previous myocardial infarction	17 (45)
Previous PCI	3 (8)
Antiplatelet therapy
Aspirin	37
Clopidogrel	34
Prasugrel	2
Ticagrelor	1
Glycoprotein IIb IIIa inhibitor tirofiban	32

PCI: Percutaneous coronary intervention

Angiographic characteristics

Average number of lesions per patient was 2.13 indicating predominantly multivessel subset of patients [Table 4]. Moderate to severe calcification was present in the involved vessel in 20 cases. Twelve cases of CAP involved chronic total occlusion. Diffuse disease with lesion length >20 mm in culprit vessel was present in 31. Significant tortuosity was present in 12 cases. Lesions belonged to AHC/AHA Type A in 1, Type B in 5, and Type C in 31 cases. Two cases involved in-stent restenosis. The average number of stents implanted per patient with CAP was 2.43. Coronary arteries involved in perforation were left anterior descending in 15, left circumflex in 10, and right coronary in 12 cases. Distal CAP occurred in 14 cases and was predominantly due to wire exit while CAP in the proximal and middle part of the vessel occurred in 23 cases which were due to a variety of reasons including balloon dilation, stent implantation and stiff guidewire.

Table 4

Angiographic characteristics

Characteristics	n
Number of lesions per case
1	7
2	18
3	12
Target vessel
Left anterior descending	15
Left circumflex	10
Right coronary	12
CAP location
Ostial	0
Proximal	4
Mid	19
Distal	14
ACC/AHA lesion type
A	1
B	5
C	31
Moderate/severe calcification	20
Significant tortuosity	12
Chronic total occlusion	12
Lesion length >20 mm	31
Vessel size <2.5 mm	8
In-stent restenosis	2
Stents implanted per patient	2.43

CAP: Coronary artery perforation, ACC/AHA: American College of Cardiology/American Heart Association

Procedure characteristics

The most frequent mechanism of CAP was related to the use of guidewire (n = 17) and balloon (n = 17) [Table 5]. Guidewire exit perforations predominantly belonged to Type II with 11 of 17 cases accounting for it. Guidewire related CAP was predominantly due to the use of hydrophilic guidewires causing perforation in 13 cases while stiff guidewire use was responsible for CAP in 4 cases. One case of Type III CAP occurred during retrograde approach to chronic total occlusion. Of 17 cases of guidewire related CAP, 14 CAP occurred due to wire exit through the distal end of vessel while 3 cases occurred during lesion crossing with stiff guidewire. Balloon dilation as cause of CAP was predominantly due to the use of noncompliant balloons for postdilation responsible in 16 of 17 cases while 1 case involved semi-compliant balloon. Stent implantation was cause of CAP in 3 cases mostly due to oversized stents. All three cases due to stent implantation presented as Type III. In addition, two cases of CAP occurred during postdilation with a noncompliant balloon in cases of in-stent restenosis due to high pressure inflation. No case of CAP could be attributed to the use of rotablation or cutting balloon in our series. 32 cases received glycoprotein IIb IIIa inhibitor tirofiban during PCI.

Table 5

Procedure characteristics

	I	II	III	III CS	n
Predilation semi-compliant balloon	0	0	1	0	1
Predilation noncompliant balloon	0	0	0	0	0
Hydrophilic guidewire	2	11	0	0	13
Stiff guidewire	1	0	3	0	4
Stent implantation	0	0	3	0	3
Postdilation noncompliant balloon	0	4	8	4	16
Cutting balloon	0	0	0	0	0
Rotablation	0	0	0	0	0
Glycoprotein IIb IIIa inhibitor	3	12	14	3	32

CS: Cavity spilling

Clinical presentation

27 cases of CAP were asymptomatic and detected due to angiographic abnormalities [Table 6]. 6 cases presented with features suggestive of cardiac tamponade with hemodynamic instability in catheterization laboratory. Two cases of CAP were initially unrecognized and were detected later after few hours due to cardiac tamponade. Cause in both these cases was exit perforation due to hydrophilic guidewire. 6 cases of CAP were complicated by periprocedural myocardial perforation.

Table 6

Clinical presentation

	I	II	III	III CS	n
Initially unrecognized	0	2	0	0	2
Pericardial tamponade/effusion	0	2/4	6/11 (stiff wire - 3, balloon - 3)	0	8/15
Periprocedural myocardial infarction	0	2	4	0	6
Asymptomatic	3	13	7	4	27
Total	3	15	15	4	37

CS: Cavity spilling

Management

CAP related to Types I and III CS was asymptomatic and did not require any active intervention. A total of 20 cases of CAP were managed conservatively in view of hemodynamic stability [Table 7]. Of 17 cases of CAP due to guidewire exit, 12 could be managed conservatively as they were associated with hemodynamic stability and little symptoms. Rest of the cases of CAP required at least some form of intervention.

Table 7

Management

Treatment strategy	n
Conservative	20 (Type I-3, Type III CS-4, Type II-13)
Prolonged balloon inflation	9
Covered stent	11
Microcoil	2
Polyvinyl alcohol particles	1
Emergency surgery	1
Pericardiocentesis	8
Blood transfusion	8
Reversal of heparin	1

CS: Cavity spilling

Emergency pericardiocentesis was required in 8 cases with two cases among them requiring pericardiocentesis few hours after PCI procedure due to the delayed presentation with cardiac tamponade. Initial prolonged balloon inflation was resorted to in 9 cases. In two such cases with prolonged balloon inflation for 10–15 min, leak from perforation site was stopped. 11 cases required emergency covered stent implantation. In one case, the covered stent was implanted in the main artery to exclude the culprit branch vessel with perforation. In two cases, where CAP involved small caliber branches, microcoil was used to seal the perforation site while one case required polyvinyl alcohol particles to seal the perforation site. In one case, heparin reversal was done with the administration of protamine sulfate, and the patient was sent for emergency surgery to seal the site of perforation.

Outcome

There was no in-hospital mortality while 30-day mortality occurred in one patient [Table 8]. One case was referred for emergency surgery while the incidence of reinfarction within 30 days was 1%. The incidence of side branch occlusion was 5%.

Table 8

Outcome

	I	II	III	III CS
Death	0	0	0	0
30-day mortality	0	0	1	0
Emergency surgery	0	1	0	0
Reinfarction	0	0	1	0
Side branch occlusion	0	1	4	0

CS: Cavity spilling

DISCUSSION

The incidence of CAP in our study was found to be 0.79% indicating CAP to be a rare complication of PCI. The incidence of CAP in the literature across different studies ranges from 0.1% to 1.5% and our study result corresponds to the same.[1234567] The slightly higher CAP incidence in our study may be due to the more complex disease including more cases of multivessel disease, chronic total occlusion and diffuse disease.

Of 37 cases of CAP, 31 cases had ACC/AHA lesion Type C indicating predominantly more complex lesion subset of patients had CAP.

One important variable predicting CAP has been lesion calcification and acute myocardial infarction.[1011] The treatment of chronic total occlusion and use of stiff guidewire has been found as one of the strongest predictors of CAP.[1011] Coronary artery caliber <2.5 mm is another predictor of CAP.[10] The literature describes the use of hydrophilic wire as one cause of CAP.

In our series, we found hydrophilic wire use as a predominant cause of wire exit perforation.[12] Only one case of CAP in our study was associated with predilation. It involved fibrotic lesion where higher inflation pressure was required. The rest cases of balloon related CAP were associated with postdilation with noncompliant balloon. It is generally recommended that balloon artery ratio should be between 1 and 1.1 for optimal stent implantation.

However due to technical issues of calcified lesions, suboptimal results lead to the use of higher caliber balloons which when inflated to very high pressures for stent optimization may lead to perforation.[12] The risk involved in the use of balloon for aggressive predilation as well as postdilation can be decreased with optimal bed preparation by the use of atheroablative methods and cutting or scoring balloon. Rotablation and cutting balloon use in lesion preparation helps in less aggressive post dilation. One study found the cutting balloon to be protective against CAP.[13] Further study is required to test this novel finding.

In addition, the use of invasive imaging such as optical coherence tomography and intravascular ultrasound can help in stent optimization.

We think that CAP in our study attributed to oversized stent could be reduced with more use of these modalities, especially in complex PCI. Studies have found the use of glycoprotein IIb IIIa inhibitor as predictor of mortality in CAP.[13] However we could not attribute glycoprotein IIb IIIa inhibitor use with mortality in our study similar to one study of 39 cases of CAP reported in literature.[7]

Clinical presentations and outcomes in our study can be explained to a large extent by the severity of perforation as given by Ellis classification.[1] As per the classification, predominant cases of CAP belonged to Types II and III. This is in line with studies described in literature.[1014] Pericardial tamponade, periprocedural myocardial infarction and need for emergency surgery were the outcomes in Types II and III CAP while Type I and Type III CS remained asymptomatic and required no special treatment and managed conservatively [Figure 1]. Two cases due to guidewire use in our study became manifest only several hours following PCI.

Figure 1

Ellis Type III perforation of distal left circumflex artery (arrow) caused by guidewire exit causing cardiac tamponade and requiring emergency pericardiocentesis (a); asymptomatic Ellis Type III cavity spilling perforation of the left anterior descending artery (arrow) caused by postdilation with non-compliant balloon (b)

The management of CAP is dictated by the severity of CAP and hence type of CAP as per the Ellis classification. Type I and Type III CS usually require only observation and no active treatment as they are usually well tolerated. Majority of Type II CAP also are usually well tolerated and hence usually require no active treatment other than cessation of glycoprotein IIb/IIIa inhibitor and repeated echocardiogram to detect any pericardial effusion. Most guidewire exit perforations belong to Types I and II although they can be Type III as well and usually involve distal coronary bed. However they can be initially missed on fluoroscopy and may later present with cardiac tamponade due to persistent leak from perforation site.

The management in Type III CAP is simultaneously directed at two specific things, one is stabilizing the patient's hemodynamics by urgent pericardiocentesis and the second is efforts directed at sealing the site of perforation. The first most appropriate and feasible option is attempted balloon occlusion at the site of perforation. The selected coronary balloon should be of 1:1 caliber in comparison to vessel.[15] It should be inflated at lower pressures for prolonged period. Intermittent release of balloon may be required to restore blood flow due to ischemia induced downstream. The use of protamine sulfate to reverse the effect of heparin can help in sealing the site of perforation faster.

Platelet transfusion may be useful in case of abciximab use although they are not useful in case of other glycoprotein IIb IIIa inhibitors.[16] These actions may seal the site of perforation sometimes precluding the requirement of further interventions. These actions also give time to stabilize the patient and decide about further course of action.

Covered stents have revolutionized the treatment of CAP. Expanded polytetrafluoroethylene (PTFE) covered stents have enabled treatment of coronary perforations in catheterization laboratory without the need of emergency surgery, especially perforations involving large epicardial arteries in proximal and middle segments [Figures 2 and 3].[171819] All covered stents in our study could be delivered successfully with complete sealing of perforations [Table 9].

Figure 2

Ellis Type III perforation of middle left anterior descending artery caused by postdilation of stent with noncompliant balloon (a) and perforation sealed by placement of covered stent (b)

Figure 3

Ellis Type III perforation of obtuse marginal artery caused by distal edge of stent (a) and perforation sealed by placement of covered stent (b)

Table 9

Covered stents (in authors experience)

	Graftmaster	Prograft	Pk papyrus	Aneugraft dx
Manufacturer	Abbott vascular	Vascular concepts	Biotronik	ITGI medical
Graft material	Expandable PTFE	Expandable PTFE	Electrospun polyurethane	Equine pericardium
Stent material	Stainless steel	Stainless steel	Cobalt chromium alloy	Stainless steel
Remark	High crossing profile and low flexibility	High crossing profile and low flexibility	Thinner polymer, highly flexible and deliverable	Thin profile, highly trackable, requires flushing of balloon lumen before use

PTFE: Polytetrafluoroethylene

However, in one case, perforation could not be sealed with single covered stent and required additional covered stent at higher pressure to achieve complete seal of leaking site [Figure 4]. Studies evaluating safety and efficacy of covered stent implantation have shown high success rate in implantation with angiographic success rate of around 92%–96%.[1820]

Figure 4

Ellis Type III perforation of middle right coronary artery caused by stiff guidewire exit (a) causing cardiac tamponade and requiring emergency pericardiocentesis with multiple covered stents required for continuing residual leak (arrow) (b)

There are however some drawbacks of the covered stents. One of the important technical limitations is high profile and limited flexibility which may hinder their use in extremely tortuous vessel. The use of covered stent has risk of side branch occlusion. Hence they cannot be used in presence of major side branch to avoid major periprocedural myocardial infarction. There was side branch occlusion in 4 cases in our study although in one case covered stent was specifically used to exclude culprit leaking vessel.

Another concern regarding covered stent has been the issue of subacute thrombosis and late restenosis. In our study, there was death in one case after 4 weeks and it was probable case of stent thrombosis.

In a study of covered stents in various clinical settings, subacute stent thrombosis occurred in 5.7% of cases 7–70 days after stent implantation and overall restenosis rate was 31.6%.[20]

In RECOVERS trial, evaluating PTFE covered stents in saphenous vein grafts, higher incidence of 30-day myocardial infarctions as compared to the bare metal stent group.[21] Increased incidence of subacute thrombosis and late restenosis can be associated with delayed endothelialization.

Newer generation covered stents have been made from electrospun polyurethane and pericardium to improve stent flexibility and deliverability and reduce thrombogenicity.[222324] CAP related to coronary guidewire exit predominantly involve distal coronary bed and smaller caliber vessel. Variety of devices have been described in the literature for therapeutic embolization such as microcoil, polyvinyl alcohol particles, gel foam, thrombin fibrin glue, cyanoacrylate glue, newer agent called onyx, autologous blood clots, and fat.[142526] Microcoils are one of the most feasible and commonly used methods for therapeutic embolization. Appropriate size microcoils which are slightly oversized up to 1.5 times the target vessel caliber can be delivered using microcatheters and help in achieving sealing of perforation site due to their thrombogenic properties.

Emergency surgery was required in one case of CAP where there was persistent leak despite prolonged balloon inflation and reversal of heparin effect the appropriate size covered stent was not available. We did not find a need for reversal of heparin in our cases except one due to prompt sealing of leaking site with the use of covered stent or microcoil.

In our series, only one patient required emergency surgery and there was no in-hospital mortality. The course was complicated by periprocedural myocardial infarction in 6 and side branch occlusion in 5. Thirty day mortality in our study was a single case related to probable stent thrombosis. Mortality in various studies of CAP ranges from 5% to 11%.[2345] However, a study of 62 cases and another study of 18 cases did not have a mortality while other studies of 41 cases and 39 cases of CAP had reported a single in-hospital mortality.[171218] Far less mortality in these series is similar to our study and indicate the effectiveness of covered stents and other devices in the management of CAP in the catheterization laboratory without the need of emergency surgery in the majority of cases.

Our study does have some limitations. We did not have any experience of CAP related to the use of atheroablative devices. Another limitation may be less use of invasive imaging to guide the complex PCI. Optical coherence tomography became available for use at the center only in the past 2 years of the study.

Invasive imaging such as optical coherence tomography and intravascular ultrasound can influence the incidence of CAP. With all these limitations, the experience of the authors with the fairly large number of cases of CAP in the presented study indicates that this catastrophic complication can be managed the catheterization laboratory successfully without the need of referral to emergency surgery in the majority of cases and reinforces the confidence to carry out more complex PCI in hospitals without on-site cardiac surgery backup.

CONCLUSIONS

CAP is a rare but potentially fatal complication of PCI. Availability of covered stents and various means for therapeutic embolization have allowed many cases of CAP to be treated in the catheterization laboratory without the need of emergency bailout surgery. With the increasing number of complex PCI being done, the incidence of CAP will increase.

Increasing the use of advanced techniques such as rotablation and cutting balloon and adjunctive invasive imaging will influence the incidence of CAP and requires further studies. The lower mortality in our study suggests that the complication of CAP can be managed in the catheterization laboratory and hence more complex PCI could be done off-site cardiac surgical backup.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.