Statistics from Altmetric.com
To gain familiarity with the commonly encountered complications associated with percutaneous coronary intervention (PCI).
To develop strategies to treat and overcome complications encountered in the cath lab.
To learn methods of avoidance of complications and improve the safety of PCI procedures.
In the 40 years since Gruntzig’s first angioplasty,1 percutaneous coronary intervention (PCI) has become one of the most frequently performed therapeutic interventions in medicine.2 Devices and techniques have evolved during this period and facilitated improved patient outcomes, so unlike Gruntzig it is no longer necessary to have a cardiac surgeon present for every procedure. However, these advancements have resulted in the treatment of increasingly complex patient populations, including acute coronary syndrome, chronic total occlusion (CTO), calcified coronary disease and cardiogenic shock. Consequently, despite improved device and procedural safety, complications associated with PCI continue to be encountered. It is of paramount importance that interventional cardiologists are equipped with the knowledge and skills required to rapidly recognise complications, and have strategies to overcome them, thereby minimising the risk of injury to their patients.
It is important to recognise that all invasive vascular interventions are associated with a risk of bleeding and vascular injury; however, for the purpose of this review, we will focus only on complications relating specifically to coronary intervention.
Traumatic coronary dissection
Coronary dissection is a pathological separation of the layers of the vessel. Traumatic coronary dissection can be induced by the guide catheter, wire manipulation, equipment in the coronary artery (imaging probes, ‘mother-and-child catheters’) or excessive balloon and stent expansion.
Guide catheter-induced dissection affects the ostioproximal segment of the vessel initially but may extend retrogradely into the aortic root or propagate down the coronary artery. Ostial disease and supportive guide catheters, particularly in patients with small aortic root dimensions, increase the risk of dissection. Careful attention should be paid to guide catheter engagement with coaxial alignment and monitoring of the arterial waveform for damping or ventricularisation (figure 1). Side-hole catheters are preferred by some operators if there is ostial disease, but there is a trade-off between gaining feedback from the pressure trace to acknowledge the presence of disease versus the risk of ischaemia. The use of side-hole catheters is generally discouraged as they provide a false sense of security, may increase the risk of dissection and are usually associated with an increased volume of contrast used.
Guide catheter dissection may occur without symptoms; however, patients can develop chest pain with associated ST segment change or, if dissection results in abrupt vessel closure, haemodynamic collapse may be observed. Urgent recanalisation is the priority; however, wiring the true lumen can be challenging and surgical bailout with haemodynamic support (extracorporeal membrane oxygenation or peripheral left ventricular support) may be considered.
Early recognition of guide catheter-induced dissection is important to avoid propagation of the false lumen. Further contrast injections should be minimised, and care should be taken to achieve true lumen wire placement. As the tip of the guide catheter may remain in the dissection plane, consideration should be given to cautious, partial withdrawal of the guide catheter, prior to wiring the vessel. Intravascular ultrasound can play an important role in confirming wire position and guidance of the subsequent intervention without the need for contrast injection (figure 1). The risk of dissection propagation by contrast injection precludes the use of optical coherence tomography.
More commonly, coronary dissection is caused by trauma from balloon angioplasty, but passage of intracoronary imaging probes and other equipment into the vessel also carries a small risk of dissection.3 4 If dissection occurs with a wire in the true distal lumen, it is critical not to lose wire position. Stents easily seal dissection flaps to prevent acute or abrupt vessel closure. If dissection involves the ostioproximal segment of a vessel close to the aortic root, adequate sealing of the ostium with an appropriately expanded stent is important to minimise the risk of retrograde propagation. Small asymptomatic dissections (types A and B) may settle conservatively (figure 1).
Air embolisation is a rare complication, reported in 0.1%–0.3% of cases, due to failure to clear air from the injection manifold system, and can be associated with significant haemodynamic compromise and death.5 Guide catheters should always be aspirated prior to connection to the pressure line. During the procedure, entrainment of air may occur when there is limited space within the internal lumen of the guide catheter, for example, when using more than one device in a 6Fr system, using 5Fr equipment or when aspirating blood through the manifold.
Numerous methods have been described to overcome the haemodynamic embarrassment of air embolism, including inotropic support to increase mean arterial pressure and disperse the air, microvascular vasodilation with adenosine or sodium nitroprusside to increase the capillary bed space, and wiring and aspiration of the embolised vessel to disperse and extract the air, respectively. Additionally, use of 100% oxygen, in ventilating the patient, produces a diffusion gradient favouring reabsorption of air.
Iatrogenic coronary thrombosis
Iatrogenic coronary thrombosis can arise from thrombus injected from the guiding catheter, thrombus formation in situ because of suboptimal antithrombotic therapy and disturbed haemorrheology from intracoronary instruments or accidental thrombus migration from one vessel to another (aspiration thrombectomy). Long procedural time, complex procedures and inadequate antithrombotic therapy are important predictors of periprocedural intracoronary thrombosis.6
All patients undergoing PCI should be pretreated with dual antiplatelet therapy and receive a loading dose of 70–100 IU/kg of unfractionated heparin. Activated clotting times (ACT) should be checked every 15–20 min and maintained above 250 s with further bolus doses during the procedure. For antegrade and retrograde CTO procedures, an ACT of >300 and >350 s, respectively, is recommended, although supporting evidence is lacking.7 Intracoronary instruments, such as balloon or imaging catheters, should be removed from the coronary artery and guide catheter immediately after use.
Embolisation of thrombus from the guide catheter can be avoided with meticulous aspiration or allowing the contents of the guide catheter to ‘bleed back’. In addition, close attention should be paid to the arterial waveform prior to contrast injection as a damped trace might imply thrombus in the catheter.
Measures to disperse intracoronary thrombus include administration of glycoprotein (GP) IIb/IIIa inhibitors and manual thrombus aspiration. Intracoronary bolus tends to be preferred. Intracoronary abciximab has been shown to reduce the risk of major cardiovascular events, reinfarction and congestive heart failure when compared with intravenous therapy.8 Gentle inflation or agitation with a balloon may be required to disperse small thrombi causing distal vessel occlusion. Figure 2 shows a case of thrombus embolisation from a guide catheter treated with manual aspiration.
Coronary perforation can be caused by disruption of the vessel wall secondary to instrumentation, for example, balloon angioplasty, stenting or atherectomy, or can occur distally secondary to coronary guidewire exit. The clinical consequences of a coronary perforation are clearly dependent on the location and extent of the disruption. The Ellis classification system (figure 3) differentiates the extent of extravasation and can be used to guide management and predict outcome.9 Types I, II and cavity spilling perforations are generally without clinical sequelae. Type I perforations may seal spontaneously and type II perforations can usually be managed with balloon inflation, rarely requiring additional pericardiocentesis. We will therefore focus on the management of type III coronary perforation and guidewire exit perforation, which is sometimes referred to as type V perforation.
Pericardial and intramyocardial haematomas can result from either type II or type III perforation.10 Large haematomas may create an inaccessible collection of blood compressing a vital structure or cause arrhythmias. Close patient monitoring is required as haematoma may develop hours post-PCI. Transthoracic echocardiography or CT may be used to delineate the extent of haematoma and guide the need for intervention. Surgical treatment of the haematoma may be required if there is ischaemia or infarction secondary to compression of an epicardial artery or haemodynamic instability.
Risk factors associated with perforation are female sex, increasing age, coronary calcification, use of a cutting balloon or atheroablation, balloon to artery ratio of >1.1, concomitant GPIIb/IIIa use and treatment of a CTO.11–13 The growing number of CTO procedures has resulted in improved methods of managing perforations.10
Management of large vessel perforation
Prompt intervention following a clear management algorithm is critical as patients often deteriorate rapidly, particularly in type III perforations. The first step is always balloon tamponade of the vessel at the perforation site to stop further extravasation of blood. Haemodynamic status should be assessed immediately. Unstable patients require urgent echocardiography followed by pericardiocentesis if tamponade is evident, fluid resuscitation, transfusion (autotransfusion could be considered) and cardiovascular support as necessary. Large bore venous access should be obtained and consideration given to connecting the pericardial drainage bag direct to the venous line or using a cell saver device where practicable.
Concurrently antithrombotic status requires evaluation. This can be difficult as the risk of ongoing bleeding has to be judged against the risk of vessel thrombosis. GPIIb/IIIa inhibitor infusions should stop and consideration given to partial reversal of heparin in some situations (very high ACT or distal guidewire exit perforation). Heparin should not be fully reversed with the wire and balloon in the vessel or if covered stents are required, as thrombosis of the whole vessel will lead to a higher mortality than the perforation.
If there is evidence of continuous contrast extravasation despite prolonged balloon inflation or it is not tolerated (due to ischaemia), covered stent implantation is required. The ‘ping-pong’ guide technique facilitates ongoing control of extravasation through the index guide system and delivery of the covered stent via the second guide. To achieve the ‘ping-pong’ guide technique, a second access site is obtained to allow delivery of a second guide catheter, backing the first guide out slightly to allow intubation. A second guidewire should be placed to the proximal edge of the inflated balloon and passed to the distal vessel with minimal balloon deflation time. Subsequently a covered stent can be placed over this second guidewire to the proximal edge of the inflated balloon. The balloon is deflated and withdrawn with the first guidewire as the covered stent is positioned and immediately deployed. The covered stent may require postdilatation to ensure good apposition if there is ongoing contrast extravasation. Covered stents may be placed using a single guide (quick exchange with the inflated balloon), but there is a substantial risk of rapid haemodynamic collapse in the time taken to do this. A case illustrating the management of a type III perforation is shown in figure 3.
These complications tend to occur in the most complex lesion subsets and consequently involve difficult stent delivery. As covered stents are bulky and the perforated vessel is likely to be calcified, 7Fr or 8Fr guides should be considered if the ‘ping-pong’ guide technique is used. Additionally, a mother-and-child guide catheter extension such as a GuideLiner (Vascular Solutions, Minneapolis, Minnesota, USA) may be required, but it is important to consider the impact of such devices on the internal dimension of the guide system and consequent ability to deliver bulky devices. Details of the device characteristics of covered stents available in Europe are shown in table 1.14
Management of guidewire exit perforation
Distal vessel perforations are caused by inadvertent coronary guidewire exit. Hydrophilic and polymer-jacketed guidewires are more likely to perforate distal vessels and should therefore be exchanged for a workhorse wire when crossing has been achieved.15 The risk of perforation is also increased when treating complex coronary disease (CTOs) and concomitant use of GPIIb/IIIa inhibitors.
Unlike large vessel perforations, the angiographic appearance caused by guidewire exit perforations is more subtle and may be more challenging to detect. In addition, haemodynamic compromise from cardiac tamponade may be delayed for several hours due to the slower rate of bleeding.
As with large vessel perforations, the first treatment step is balloon inflation proximal to the perforation site. Occasionally this is sufficient to seal the leak, but if contrast extravasation persists, definitive intervention is required with fat, coil, thrombin or autologous clotted blood embolisation.
Fat embolisation of persistent guidewire exit perforation is gaining popularity as it is universally available and can be delivered through any microcatheter. Fat can be harvested from local subcutaneous tissue using a scalpel and forceps and delivered through a microcatheter, selectively engaged just proximal to the perforation site.16 Figure 4 illustrates a case of delayed cardiac tamponade from guidewire exit perforation treated with fat embolisation. Fat particles can be dipped in iodinated contrast before embolisation to allow X-ray visualisation.16
Coils are made of steel or platinum and can be retrievable (detachable) or not retrievable (pushable).17 Some coils require a large microcatheter for delivery, but others (Cook microcoils, Cook Medical, Bloomington, Indiana, USA) can be delivered through standard coronary microcatheters, such as the Finecross microcatheter (Terumo, Japan) (see table 1 for device details). Interventional cardiologists should ensure familiarity with the coils available in their cardiac catheterisation laboratory.
Coronary no-reflow is defined as the inability to adequately perfuse myocardium after temporary occlusion of an epicardial coronary artery without evidence of a persistent mechanical obstruction.18 It occurs most commonly in the setting of acute myocardial infarction as prolonged ischaemia leads to endothelial damage, myocyte oedema and microvascular dysfunction. During primary PCI, embolisation of atherosclerotic debris, thrombus and platelet plugs may add further insult to the coronary microcirculation, expanding the no-reflow area. The incidence of no flow/slow flow in primary PCI was 1.5% (BCIS Audit Returns 2015) as compared with 0.7% in Non-ST elevation myocardial infarction (NSTEMI)/unstable angina cases and 0.3% in stable angina. No-reflow can also be a major problem in the treatment of saphenous vein grafts (SVG), thrombus-containing lesions and rotational atherectomy due to atheroembolism with subsequent release of vasoconstrictor substances.18
In the setting of PCI, no-reflow is characterised by chest pain, persistent or new ST segment change, and thrombolysis in myocardial infarction (TIMI) flow <3, or in the case of TIMI 3 flow when myocardial blush grade is 0 or 1.19 No-reflow following PCI is associated with poor functional recovery, increased recurrence of ischaemia, larger infarct size, worse left ventricular ejection fraction at 6 months and increased 1-year mortality (16.7% in patients with no-reflow vs 5.5% in patients with normal flow in one study).20–22
Several strategies can be used to reduce the risk of no-reflow (figure 5).23 Distal embolic protection devices are recommended for PCI of SVG lesions if technically feasible (class 1 recommendation, European Society of Cardiology (ESC) guidelines on myocardial revascularisation 2014).24–26 The role of GPIIb/IIIa antagonists and other medications (adenosine, verapamil) in preventing no-reflow is unclear. Some studies suggest encouraging results, but no specific recommendations can be made for the routine use of these drugs in preventing no-reflow.27–30 Direct stenting (vs predilatation) in acute myocardial infarction has been shown to be feasible and associated with lower microvascular dysfunction and no-reflow.31–33
Administration of intracoronary nitrates and an ACT check should be considered when faced with no-reflow, and the diagnosis can be reliably and quickly established by passing a double lumen catheter, microcatheter or an over-the-wire balloon distal to the occlusion and injecting contrast. If the contrast fills the vessel proximally and does not clear distally, a diagnosis of no-reflow is made, while normal distal flow of contrast suggests a mechanical obstruction with contrast filling back to the point of obstruction. Intravascular imaging may be needed to exclude dissection, thrombus or intense spasm, as an alternative, mechanical, cause.
Numerous drugs have been considered for the treatment of no-reflow, predominantly targeting enhancement of microvascular perfusion and thrombus dissolution. There is no gold-standard drug of choice. Adenosine (100 µg up to 4 mg over 1 min), verapamil (100–200 µg up to 1000 µg) and sodium nitroprusside (50–200 µg up to 1000 µg) have been shown to be effective and can be used alone or in combination (figure 5). Table 2 outlines the preparation and administration of commonly used drugs for the treatment of no-reflow and extensive thrombosis.34–36 Distal drug delivery is recommended to ensure the impaired microvascular bed receives the agent in the absence of antegrade flow from the guide catheter. GPIIb/IIIa inhibitors should be considered (class IIa recommendation, ESC guidelines on myocardial revascularisation 2014).26 In drug-resistant no-reflow, intra-aortic balloon pump insertion may improve myocardial perfusion.37 38
Side branch occlusion
Side branch (SB) occlusion is a potentially serious complication associated with PCI of bifurcation lesions. Factors associated with the risk of SB ostial impingement and occlusion during main vessel (MV) stenting include a high burden of plaque at the carina, reduced TIMI flow grade in the SB before stenting, bifurcation angle <70°, diameter stenosis of SB before MV stenting >50% and increasing diameter ratio between MV/SB.38 Proximal MV diameter stenosis >50% and acute coronary syndrome have also been shown to be independently associated with SB occlusion.39
Patients with SB occlusion during bifurcation PCI have worse clinical outcomes than patients without SB occlusion, with one study showing a higher rate of cardiac death or myocardial infarction and stent thrombosis associated with SB occlusion.40
The provisional SB stent implantation strategy should be considered the standard approach for treatment of bifurcation lesions. A ‘two-stent strategy’ should be considered upfront for lesions with difficult wiring or large SB with extensive disease extending >5–10 mm beyond the bifurcation.41 The risk of SB occlusion can be minimised by selecting an appropriate bifurcation stenting strategy, having a low threshold to protect the SB with a wire and consideration of intracoronary imaging to assess plaque burden and bifurcation anatomy. SB predilatation before MV provisional strategy stenting should generally be avoided as there is an increased risk of dissection and repeat revascularisation associated with balloon trauma, but may be necessary in complex lesions to maintain SB patency.42 A jailed wire in the SB may not prevent occlusion, but is associated with a much higher chance of recovery of the occluded vessel.40
Loss of stents and other equipment
The incidence of stent dislodgment and embolisation with current generation, premounted, stent systems is very low (0.32% in a large series).43 Stent loss occurs more frequently in lesions with calcification or significant proximal angulation. Stents are more commonly lost in the right coronary and left circumflex arteries.44 Calcified, tortuous vessels should be predilated and modified adequately with use of non-compliant or cutting balloons and atherectomy if appropriate. Supportive guide catheters and mother-in-child guide catheter extension systems (eg, GuideLiner, Vascular Solutions) will help facilitate difficult delivery of stents and other equipment. However, aggressive catheter engagement or failure to achieve coaxial alignment of the vessel and guide catheter can increase the risk of stent detachment. If resistance is felt on the withdrawal of a device, it is important to assess these factors before proceeding. Additionally, use of dedicated X-ray algorithms to enhance stent opacification can be used to confirm or exclude detachment.
Lost stents can be deployed or crushed in the coronary artery or retrieved through various techniques. Maintaining guidewire position is critical to keep all management options open. If a stent comes off its balloon in a suitable location for deployment and guidewire position is maintained, serial insertion and dilatation of small to larger balloons within the stent should facilitate satisfactory deployment. If guidewire access is lost, the stent is damaged or it is not possible to advance balloons into the stent, the stent could be crushed against the vessel wall with a balloon and a second stent deployed to cover the lesion.
Stent retrieval may be attempted through various techniques. The most common method is to advance a low profile 1.5–2.0 mm balloon through the stent, inflate it to 1–2 atmospheres and withdraw gently ensuring the stent is removed with the balloon. Alternatively, the wire braiding technique involves navigation of one or more guidewires through the dislodged stent to the distal vessel, applying torque to all wires to cause wrapping around the stent and then gentle withdrawal of the whole system.
Lost stents can also be retrieved using a loop snare (preferred in cases of inadvertent loss of the guidewire position) or other retrieval devices (biliary forceps, Cook retained fragment retriever or basket retrieval device).43 When removing defective material from the coronary system or ascending aorta, it is best to remove the entire system to the iliac artery to reduce the risk of cerebral embolisation. If the stent cannot be withdrawn into the guiding catheter or sheath, using a gooseneck snare system or upsizing the size of the vascular sheath (eg, to 9 Fr) may avoid the need for vascular cutdown.
Figure 6 illustrates a case of a dislodged stent, which was successfully retrieved from the radial access site.
Cath lab complications: human factors
Significant lessons have been learnt from the airline industry regarding the ‘human factors’ associated with accidents. It has been acknowledged that >75% of all air accidents relate to human error, and consequently particular attention is given to minimising the risk of errors through checklists, standardised practices and effective communication.45 Similar strategies are now being deployed in healthcare.
The catheterisation laboratory closely resembles the environment of an operating theatre, and numerous allied healthcare professional groups provide services within this high-throughput clinical area. The WHO’s surgical safety checklist was introduced in surgery to minimise errors and reduce complications and mortality.46 The checklist’s benefit extended beyond avoidance of ‘wrong procedure’ or ‘wrong site’ errors, by promoting team role recognition, improved communication and a clear understanding of the planned procedure and required equipment/expertise. The British Cardiovascular Society has developed a checklist customised to the unique characteristics of the cath lab.47
Simulation training provides trainees with an opportunity to develop technical and non-technical catheterisation/interventional skills without exposing patients to their learning curve.48 Recent evidence supports a role for simulation in accelerating the learning curve with individual/patient and institutional benefits, through reduced radiation exposure and procedural time, respectively.49 Additionally, fully simulated cath lab environments facilitate team-based training focusing on ‘human factors’ and crisis resource management. We anticipate that this novel method of training will enhance communication, teamwork, decision-making and leadership, with potential for improvements in patient outcomes.50
Teamworking is essential to the safe provision of healthcare for patients. Breaking down hierarchy has been highlighted as an important method to improve communication and facilitate junior members of the team raising concerns to avoid errors.51 The use of an emergency equipment trolley, containing the commonly required devices and drugs, ensures immediate treatment of complications (tables 1 and 2 provide guidance on what should be included). Standard operating protocols (SOPs) and defined treatment algorithms for the commonly encountered complications provide the team with guidance, and we have provided a simple tabulation of steps to consider if the complications outlined in this review are encountered (see table 3). More detailed SOPs should be available for reference within the cath lab.
Additionally, the role of engaging with colleagues, at the time of a complication, can be invaluable. Local policies should be adopted for the provision of interventional, surgical and anaesthetic support. A second, experienced, interventionist can provide invaluable assistance in an environment that is commonly stressful and emotionally charged. A surgical opinion, where available, provides an opportunity to consider all therapeutic strategies to overcome the complication, and early anaesthetic support ensures that the patient is comfortable, optimises oxygenation and haemodynamics, and minimises the time to ventilation or circulatory support in the event of deterioration.
Developments in the technologies and techniques available to coronary interventionists have resulted in increasingly complex PCI undertaken in patients with greater comorbidity. Consequently, procedural complications will continue to be encountered. Continued training, checklists, adoption of standardised operating protocols and effective teamworking will minimise future events. The anticipation of complications and consideration of strategies of avoidance and treatment, as detailed in this review, will improve outcomes.
Familiarity with potential complications may facilitate avoidance and improve strategies to overcome procedural problems.
The use of checklists and treatment algorithms limits procedural errors and assists the team during stressful situations.
An emergency equipment trolley ensures rapid access to devices at the time of a procedural complication.
Effective teamworking is essential to limit the risk of procedural complications and ensure good outcomes when they do occur.
CME credits for Education in Heart
Education in Heart articles are accredited for CME by various providers. To answer the accompanying multiple choice questions (MCQs) and obtain your credits, click on the ‘Take the Test’ link on the online version of the article. The MCQs are hosted on BMJ Learning. All users must complete a one-time registration on BMJ Learning and subsequently log in on every visit using their username and password to access modules and their CME record. Accreditation is only valid for 2 years from the date of publication. Printable CME certificates are available to users that achieve the minimum pass mark.
We thank Dr Asri Ranga, Hospital Raja Permaisuri Bainun, Malaysia, for the images in figure 2, and Dr Emmanouil Brilakis, Minneapolis Heart Institute, USA, for images 8a, 8b and 8c in figure 4.
Contributors Both authors have contributed to the preparation, refinement and review of the article and are in agreement with its content.
Funding TWJ is supported by the NIHR Biomedical Research Centre at the University Hospitals Bristol NHS Foundation Trust and the University of Bristol.
Disclaimer The views expressed in this publication are those of the author(s) and not necessarily those of the NHS, the National Institute for Health Research or the Department of Health.
Competing interests None declared.
Patient consent Not required.
Provenance and peer review Commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.