Stent Fracture: How Frequently Is It Recognized?

Abstract

In spite of there being several case reports, coronary stent fracture is not a well-recognized entity and incidence rates are likely to be underestimated. In this article, we review different aspects of stent fracture, including incidence, classification, predictors, outcome, diagnosis, and management.

INTRODUCTION

Adramatic revolution in the field of intervention cardiology has been heralded by the use drug-eluting stents (DES) replacing bare metal stents (BMS), and thereby significantly reducing the restenosis rates and the need for repeat revascularization.[1,2,3,4] However, the occurrence of late-onset complications, such as stent thrombosis, has raised concern over the use of DES. In addition, there has been an increasing awareness of stent fracture (SF) as a potential complication following DES implantation.

Review of the literature shows there is increasing concern about SF as a potential cause of stent restenosis and thrombosis, which can lead to adverse clinical outcomes such as recurrent angina, myocardial infarction (MI), and even sudden death.

The objective of this article is to analyze controversial issues about the incidence, diagnostic tools, and clinical implications of SF.

BACKGROUND

In 1964, Charles Theodore Dotter and Melvin P. Judkins described the first angioplasty.[5] In 1977, Andreas Grüntzig, a German radiologist, successfully performed the first balloon coronary angioplasty,[6] a revolutionary treatment that led to the birth of interventional cardiology.

Coronary stents were first developed in the mid-1980s,[7] and have ultimately replaced plain balloon angioplasty after the observed improvements in angiographic and clinical outcomes seen with their use.[8,9] Nowadays, different types of stents are used in daily practice, ranging from conventional BMS and old-generation DES to newer generations of DES: DES with novel coatings, dedicated bifurcation stents, self-expanding stents, and biodegradable stents.

Coronary SF was first reported in 2002 after a BMS implantation in a venous bypass graft.[10] The first case of coronary DES fracture appeared in 2004,[11] after which several cases of SFs were reported.

INCIDENCE

The reported incidence of SF varies widely between different studies. These variations are related to many different factors including definition of SF among these studies, the methods used to detect SF, type of stent used, and the population studied.

The majority of studies report the incidence of SF between 1 and 8% [Table 1].[12] The incidence also varies depending on the percentage of patients who undergo follow-up with the available imaging modalities with variable sensitivity of detection of SF.

Table 1.

The incidence of SF, adverse clinical outcome, and percentage of TLR in patients with SF

Generally speaking, the overall incidence of SF is most likely underestimated due to different reasons given below:

1. Patients with SF might be asymptomatic, particularly in case of minor fractures, and therefore, if the angiographic or any other imaging modality follow-up were clinically driven, then many cases of SF were not reported. On the other hand, the incidence of SF is expected to be higher in studies where repeat procedure and the use of imagining modalities were clinically driven secondary to selection potential bias

2. Not all SFs can be detected with conventional angiography; therefore, many patients might be treated as stent thrombosis or stenosis without the detection of SF, especially if other more sensitive diagnostic imaging modalities were not used

3. As SF might lead to stent thrombosis, several patients might present with sudden cardiac death before diagnosis.

Although most SFs occurred in sirolimus stents, several cases of SF were also reported in other types of stents, including BMS,[10,13,14,15,16] Taxus,[17,18,19] Xience (everolimus-eluting stent),[20] and zatrolimus.[21,22]

In a series of 530 patients (of a total 2728 patients treated with DES) who underwent repeat angiography during follow-up, SF was identified in 10 patients (a prevalence of 1.9%). All occurred with sirolimus-eluting stents.[23]

Aoki et al., reported an incidence of SF of 2.6% when they studied 256 patients who underwent angioplasty with SES.[24]

Umeda et al., studied 422 patients treated with Cypher stents and found a 7.7% incidence of SF during follow-up.[25]

In an autopsy study, the incidence of SF was 29%, which is much higher than clinically reported. A high rate of adverse pathologic findings was observed in lesions with grade V SF, whereas fracture with grade I-IV did not have a significant impact on the pathological outcome [Figure 1].[26] Autopsy studies could be expected to have a higher incidence of SF compared to living patients as the studied population might have had a higher incidence of stent thrombosis and restenosis in which SF might be the cause, leading to a higher reported incidence of SF.

Figure 1.

Shows (a) types of stent fracture,(b) classification of calcification. Nakazawa G, Finn AV, Vorpahl M, Ladich E, Kutys R, Balazs I, Kolodgie FD, Virmani R. Incidence and predictors of drugeluting stent fracture in human coronary artery a pathologic analysis. J Am Coll Cardiol. 2009 Nov 17;54(21):1924-31. PubMed PMID: 19909872. (Figure added with permission from author)

DEFINITION AND CLASSIFICATION

The definition of SF varies from study to study and various morphologic classification schemes have been used. Some studies discriminate between isolated strut fractures and SF.[27] Some include both complete and partial types of fractures,[28] while others only include severe fractures with complete separation of stent segments [Table 2, Figure 1].[1,24]

Table 2.

Classification of SF according to different studies

PREDICTORS OF SF

Stents are more likely to fracture in the presence of the following factors:

1. Technical factors:

• Balloon or stent overexpansion, as it may theoretically weaken the stent struts[32]

• Stent overlap, which results in localized rigidity creating hinge points that deform the stent leading to fracture[33]

• Stent length: Longer stents may be subjected to higher radial forces[32]

• Inappropriate handling of stent[34]

• Stenting technique: An example of stenting technique that might cause SF is crush technique. A case has been reported of stent strut fracture in a bifurcation lesion treated with crush stenting, resulting in restenosis.[35]

2. Stent type and stent conformability:

• Stent conformability is defined as the degree to which a stent can bend around its longitudinal axis after deployment.[36] Decreased stent conformability can lead to longitudinal straightening of the vessel after stent deployment, which subjects the stent to the countervailing force of the vessel wall. This tends to revert the vessel axis to its original shape, leading to SF.[36]

Most cases of SF were reported with the use of sirolimus-eluting stent (SES). Several theories have been proposed to explain this association.

The first was attributed to the design of the SES, with its rigid, closed-cell structure that results in greater straightening of the vessel, which subjects the stent to greater forces during the cardiac cycle.[1] Secondly, it is easier to detect SF in SES as it is a more radio-opaque structure.[37] On the other hand, Taxus and BMS have greater neointimal coverage, which could mask the fractured struts, as well as it may strengthen and stabilize the struts to withstand mechanical forces.[38]

Despite the higher incidence of SF in SES, it does not necessarily translate into a clinically significant difference, as it has been shown by different studies that SES are associated with lower rates of stent restenosis and late luminal loss compared to paclitaxel-eluting stents (PES).[36,39,40,41]

3. Anatomic and pathologic factors which include the following:

• Tortuous and highly angulated vessel[37]

• Long lesions

• Change in vessel angulation after stent implantation, which can create a significant distortion force[25]

• Complex lesions: In the ACROSS/TOSCA 4 study, SF was more frequent in the complex lesion subset of chronic total occlusion.[42]

• Stent location: [Table 3] SF is more common in right coronary artery (RCA) and saphenous graft locations as these vessels are dynamic during cardiac contractions. Stents in these locations may be subjected to repetitive distorting forces, as some segments of these vessels have more flexion points during the cardiac cycle.[1,11] Repetitive cardiac contraction exposes the stent to compression, torsion, kinking, elongation, bending, and shear stress,[11,43] which can cause fracture from mechanical fatigue.[44] The points of SFs are usually located at hinges[24] subjected to either medial or shear forces created by non-uniform vessel anatomy.

  Table 2 shows the incidence of SF according to the vessel stent location in a meta-analysis of eight studies with 108 SFs in 5321 patients, where the incidence of SF in RCA was the highest while left main (LM) stents were less likely to fracture.[39]

Table 3.

Incidence of SF in different coronary arteries

4. Other possible causes:

• There are other possible causes that may lead to SF. Sanchez et al.,[45] reported a case of biventricular pacing leading to LAD coronary SF where the chief risk for SF was the vascular angulation; however, this risk was increased by abnormal myocardial contraction patterns that were induced by the ventricular pacing. Hoshi et al., reported a case of fatal ostial right coronary artery SF, aneurysm formation, and coronary artery perforation caused by mechanical stress between the sternum and dilated aortic root.[46] SF was also reported following stenting of myocardial bridge.[47] In one study, chronic kidney disease was found as an independent predictor of DES fracture.[48]

It is important to mention that clinical characteristics such as CAD risk factors and previous MI or CABG do not determine the risk of SF.[39]

CLINICAL PRESENTATION AND OUTCOME

In spite of its low reported incidence, SF represents an important clinical entity as it may present with serious clinical sequelae [Table 3].

It is important to mention that not all SFs are associated with clinical sequelae as it can be an incidental finding in asymptomatic patients, particularly with mild forms of SF (isolated strut fractures).

SF can present as recurrent angina, MI,[49,50] and even sudden death.[51] So, it is possible that some cases of sudden cardiac death in patients with previous stent implantation might be secondary to SF, leading to stent thrombosis.

Usually severe forms of SF have more adverse clinical outcomes, as reported by Nakazawa et al.,[26]

The most likely explanation for the development of in-stent restenosis in case of SF is the poor distribution or interruption of drug delivery as a result of strut fracture, which will suppress the inhibition of neointimal formation, leading to neointimal overgrowth and stenosis.[48]

In a study by Lee et al.,[23] 530 patients with DES underwent repeat angiography. SF was identified in 10 patients. None of these fractures were detectable at the time of stent placement. The median time interval from stent implantation to detection of fracture at repeat angiography was 226 days (ranging from 7 to 620 days). Six patients had binary restenosis and one patient had stent thrombosis, all necessitating repeat intervention.

On routine follow-up coronary angiography 6-9 months after SES implantation, Ino et al.,[28] reported 33% in-stent binary restenosis, 28% target lesion revascularization (TLR), and 0% stent thrombosis rates in SF lesions. All patients with SF had an additional follow-up for 24 months, but no major adverse coronary events were observed.

In a literature review by Chhatriwalla et al.,[52] a total of 289 SFs, with available information about patient presentation, were found. 10.4% of cases presented with ST segment elevation myocardial infarction [STEMI] or stent thrombosis and 26.3% presented with Non ST segment elevation myocardial infarction [NSTEMI] or Unstable angina [UA]. The review also highlighted the difference in clinical presentation between patients with DES fracture and patients with BMS fractures, and it was noticed that STEMI occurred more often in patients in the first group, indicating that the latter has more benign presentation. This might be explained by the thicker neointimal layer that develops around the BMS, which limits the contact of stent material with the arterial lumen inhibiting thrombus formation. In case of DES, because of the absence of this thick neointimal layer, fractures allow the stent's metal structure to come into contact with the vessel lumen, leading to thrombus formation.

Several studies showed that SFs, particularly severe forms, were associated with high rate of TLRs [Table 3].[23,27]

In a study by Park et al.,[48] the clinical presentation of patients with SF did not differ from those without SF, including the severity of angina, the incidence of Acute coronary syndrome [ACS], and event silent ischemia. Despite that, it is important to mention that in this study, 50% of SFs were of type 1, and overall, types 1, 2, and 3 represented 96.1% of all cases, while type 4 represented only 3.9% and there were no type 5 fractures. It is expected that the majority of these patients will have similar outcomes compared to patients without fracture, as the adverse clinical outcome in SF is associated with severe forms. It is worth mentioning that the rate of binary stenosis, as well as target lesion revascularization, was significantly higher in the SF group compared to the control group (41.7% vs. 11.4% and 33.3% vs. 8.1%, respectively).

It is also important to recognize that while the probability of ISR and TLR is increased in patients with SF, the opposite is also true. That is to say the probability of SFs is increased in patients with ISR and in patients requiring TLR.[39]

It is worth mentioning that the time of occurrence of SF, the time between SF and the development of stent stenosis or thrombosis, and the occurrence of symptoms are still not well recognized. Some patients presented as early as 3 days after stent implantation (Xience)[20] and others presented after several years.[53,54]

In the study by Lee et al., mentioned previously, the median time interval from stent implantation to detection of fracture at repeat angiography was 226 days (ranging from 7 to 620 days).[23]

An interesting finding in a study by Ino et al.,[55] is that late stenosis was not observed in SF sites without early restenosis during the midterm follow-up after SES implantation.

SF AND CORONARY ANEURYSM

There are several reported cases of coronary artery aneurysms associated with DES SF.[56,57] Previous reports suggested several potential causes for aneurysmal formation:[11,58,59,60]

1. The use of oversized balloon and high-pressure balloon dilatation, which result in intima media tearing

2. SF which may also tear the arterial wall

3. The drug and/or polymer of the DES might cause some sort of inflammation or hypersensitivity reaction causing adventitial inflammation, weakening of the media, excessive dilatation and, consequently, aneurysmal formation and marked malapposition of the stent.[59] The FDA reports and autopsy findings suggest that DES may be a cause of systemic and intrastent hypersensitivity reactions that, in some cases, have been associated with late thrombosis and death.[60]

   Jindal et al.,[61] reported a case of giant coronary aneurysm after Cypher implantation in a patient who presented with fever of unknown origin. Histopathology revealed a predominantly lymphocytic and eosinophilic infiltrate with an absence of giant cells, suggesting that local hypersensitivity reactions caused the aneurysm.

4. It was also suggested that the coating drug of DES might inhibit the process of healing, leading to aneurysmal dilatation. Kim et al., reported a case of simultaneous occurrence of coronary aneurysm at both the DES-implanted sites without high-pressure ballooning, which suggests that an aneurysmal dilatation may have preceded SF, rather than the fracture of the stent leading to aneurysmal formation. Kim et al., attributed the unique property of the DES as the potential cause of coronary aneurysm.[58]

   There might also be association between SF, pseudoaneurysm, and infection. Kelvin S. H. Loke[62] reported a case of pseudoaneurysm and coronary abscess secondary to coronary SF identified with Tc-99m hexamethylpropyleneamine oxime-labeled white blood cell SPECT/CT scintigraphy.

   W. Kyle Stribling reported a case of giant aortocoronary saphenous vein graft pseudoaneurysm caused by SF in an 80-year-old patient who presented with fever and was found to have positive blood culture for methicillin-sensitive Staphylococcus aureus.[63]

   In an intravascular ultrasound (IVUS) study which was done to identify SF as a cause of stent failure, 17 patients were evaluated by IVUS where 20 SFs were found. Five SFs occurred in a coronary aneurysm (accompanied by malapposition in three patients) despite the absence of aneurysm at index stenting. Comparing the SFs associated with aneurysm (5/20) with those that occurred without association with aneurysm (15/20), complete SF was more frequent (100% vs. 27%). All fractures were after Cypher stent implantation and all presented more than 1 year after index stenting.[38]

   To the best of our knowledge, there are no cases of Taxus SF associated with coronary aneurysm, despite several reports of cases of acquired coronary aneurysm after Taxus implantation.[61,62,63,64,65,66] It is worth mentioning that it is possible that the association between coronary aneurysm and SF is a reciprocal one. In other words, as SF might cause aneurysm, aneurysmal formation and malapposition might precede SF. This would lead to excessive motion of the stent, leading to SF.[38]

DIAGNOSTIC MODALITIES

Conventional fluoroscopy

Stent visibility is limited on conventional fluoroscopy. There are several factors that contribute to stent visibility, including the patient's build, a stent platform, and stent thickness.[34] Earlier generation stainless steel stents are more visible than the newer cobalt-chromium stents (such as Xience), as the first have a strut thickness of 0.0055 inches compared to the latter that have a strut thickness 0.0032 inches.[34]

IVUS

In several studies, IVUS was used to confirm the diagnosis of SF that was suggested by angiography. In other studies, IVUS detected several cases of SF that were missed by angiography. So, the use of IVUS increased the rate of SF detection in multiple studies.

Yamada et al.,[67] in a prospective study of 102 Cypher stents with 100% angiographic and IVUS follow-up, observed three SFs (3%), all detected with IVUS but not observed on angiography.

Another advantage of IVUS over conventional angiogram is the ability of IVUS to identify mechanisms of stent failure by providing information regarding neointima formation, vessel remodeling, perivascular tissue, stent expansion, stent strut distribution, and malapposition.[38] A limitation of IVUS is that the resolution is only approximately 150 micrometer and the echoes frequently cause artifacts.[38]

Another limitation of IVUS is the occasional difficulty in passing the IVUS across the lesion when there is SF, especially with stent displacement.

Stent boost

This technology has improved the visibility of stent struts.[68] It involves the automated detection of proximal and distal markers of balloon catheters in each cine frame.[69] This automated marker detection is done through the identification of blob-like structures.[70] This technology can be also used to position a stent precisely over a previously stented segment.[34] Another advantage of stent boost over IVUS is that it does not add to the procedural costs. A limitation of stent boost technology is that a balloon catheter needs to be placed in the vicinity of a stent or stented segment in order to acquire images.[34]

Multi-detector computed tomography (MDCT)

Several cases of SF detected by MDCT have been reported [Figure 2].[71,72,73] In a retrospective study, 18 SFs were detected by MDCT in 371 patients with 545 stents. Six SFs were not detected by initial conventional angiograms in this study.[74] Pang et al.,[75] evaluated the ability of 64-slice computed tomography (CT), conventional cine-angiography, and IVUS to detect SFs under ideal conditions. They concluded that under ideal in vitro conditions, CT has a high accuracy when used to evaluate coronary SFs. The overall accuracy, sensitivity, and specificity of detecting SFs are lower with conventional cine-angiography. SFs were not detected using IVUS in this study, which was attributed to the limitations of acoustic window inherent to in vitro procedures and the longitudinal orientation of the fractures.

Figure 2.

A Fractured right coronary stent by MDCT. The RCA stents became two pieces. Li P, Gai L. Coronary stent fracture detected by multidetector computed tomography. Int J Cardiovasc Imaging. 2010 Oct;26(7):729-30. Epub 2010 Mar 28. PubMed PMID: 20349340 (Figure added with permission from author)

Hecht et al.,[76] evaluated stent gaps in 292 consecutive patients with 613 stents. Correlations with catheter coronary angiography (CCA) were available in 143 patients with 384 stents. The authors concluded that stent gap by CT angiography [CTA] is associated with 28% of ISR, and ISR is found in 46% of stent gaps. They also noted that stent gap is infrequently seen on catheter angiography, and most likely represents SF in the setting of a single stent and may represent SF or overlap failure in overlapping stents.

Optical coherence tomography (OCT)

This imaging modality can also be used to detect SF [Figure 3]. In a study by Kashiwagi et al.,[77] it was found that the absence of stent strut was the most common morphological feature of SF in OCT. It was also noticed that both the mean and maximal neointimal area were larger in the SF group and the distribution of neointimal area showed a peak at the fracture site in the fractured stent group.

Figure 3.

An Exactly Matching Set among Angiography, Fluoroscopy, and OCT A nonfractured stent shows (a) no restenosis in angiography, (b) no distorted stent in fluoroscopy, (c) preserved coronary lumen (white circle), and (d) well-covered stent struts (yellow circle) in optical coherence tomography (OCT). (c–e) Neointima grows evenly. A fractured stent shows (f) focal restenosis in angiography, (g) distorted and fractured stent in fluoroscopy, and (i) excessive neointimal hyperplasia with absence of stent struts (white line) at fractured site in OCT. (h, j)

The advantage of OCT over IVUS is that it has better resolution (10 times the resolution of IVUS) and fewer artifacts.[38]

MANAGEMENT

There is no consensus about the ideal management of SFs. The decision should depend on the type of fracture, presence of ischemia, and the presence of factors that predict possible recurrence. If the reason of SF was stent overexpansion, then restenting the lesion again is possible with avoidance of stent overexpansion. On the other hand, when SF is caused by a non-modifiable factor like excessive vessel tortuosity, then referring the patient for CABG is more reasonable when there is a clear need for revascularization.

Khanna et al., reported a case of acute STEMI 6 years after implantation of a sirolimus-eluting stent, secondary to complete SF, which was treated with CABG because of the expected recurrence of SF.[53]

In a study by Lee,[17] 1009 patients with DES underwent a follow-up coronary angiography irrespective of symptoms. Seventeen SFs were detected in 15 patients (1.5%). All SF patients were continued on medication with combination antiplatelet therapy, regardless of angina symptoms. If in-stent restenosis at the fractured site was significant, coronary interventions were performed even in patients without ischemic symptoms. Some patients were treated with heterogeneous DES for restenosis lesions (5/8 patients) and the rest were treated with either homogenous DES (2 patients), or plain balloon angioplasty (1 patient) or medical treatment only (7 patients). The authors concluded that if patients with SF were continued on combination antiplatelet therapy irrespective of ischemic symptoms, there would be a low rate of major adverse cardiac events, especially cardiac death associated with SF. In case of SES fracture, we are not sure whether restenting with other types of stents – like BMS or other type of DES – might prevent or decrease recurrence of SF, as the higher incidence of SF in SES might be due to higher visibility of SES, rather than the design of the stent. However, some cases of SES fractures have been treated with reimplantation of BMS.[54] In one study, among 24 patients with fractured stents, 8 patients required TLR, where 3 patients were treated with balloon angioplasty only, 3 with PES, and 2 with zotarolimus-eluting stent (ZES).[48]

An important tip in the management of SF by restenting, when there is difficulty in passing the wire or balloon across the lesion, is to use stent boost guidance to manipulate the wire and balloon across the lesion.[34]

CONCLUSION

SF represents an important clinical entity which is most likely underestimated. Clinical presentation ranges from an incidental finding in an asymptomatic patient to a presentation of recurrent angina, MI, and even sudden death. Factors such as right coronary artery and saphenous graft stenting, lesion angulations, long stents, and the use of DES are all associated with increased prevalence of SF.

There is no consensus on the best diagnostic imaging modality for detection of SFs, but it is well recognized that conventional angiography is not sufficiently sensitive for this purpose. Multiple imaging modalities can be used, including IVUS, stent boost, MDCT, and OCT.

Management of SF should be individualized depending on the presence of ischemic signs and symptoms, the presence of predictors of recurrence, the type of stent and the type of fracture, and the available techniques. Further studies with frequent clinical follow-up and using different imaging modalities are needed for a better understanding of this entity.

In the future, with better understanding of this condition, we might be able to create a scoring index depending on the presence of the predictors mentioned, in order to estimate the risk of SF before stenting or restenting. In addition, understanding this entity will also influence the future of stent design, intended to minimize the risk of SF in the future.