Early stent thrombosis. Aetiology, treatment, and prognosis.

Stent thrombosis (ST) is an uncommon but life-threatening complication after percutaneous coronary intervention (PCI), frequently manifesting as acute coronary syndrome (ACS) or even cardiac death.

According to the academic research consortium (ARC), the definition includes definite, probable, or possible ST and is described in detail elsewhere [1].

Traditional classification categorises this complication into early (including acute and subacute ST, within 24 h and from 24 h to 30 days, respectively), late (from 30 days to 1 year), and very late (after 1 year). However, this classification does not include intraprocedural coronary thrombosis, which occurs in nearly 1% of patients [2] and is more common in the setting of ACS [3].

The majority of these events seem to occur within the first month after PCI. Among 21,009 patients treated with bare metal stents or drug eluting stents (DES) from the Dutch Stent Thrombosis Registry, 437 patients experienced ST and only 27% occurred late or very late [4]. Similar results were observed with bioresorbable vessel scaffolding (BVS) within large multicentre GOUST-EU registry (1189 patients included), where ST mostly clustered within 30 days [5]. A shift toward later ST occurrence was observed within the Japanese ST RESTART registry. This included patients treated with sirolimus eluting stents and comprised 611 patients with definite ST. Among them 47% occurred after 1 year [6]. The higher rate of late and very late ST in the Japanese registry may be associated with prolonged healing of the vessel after implantation of DES with potent antiproliferative sirolimus drug.

Finally, a completely different pattern of ST timing was observed within the impressive number of 401,662 ACS patients from the CathPCI registry [7]. Among them, definite ST events were identified in 7315. Very late ST constituted as much as 61%, and only 19% of patients presented as early ST.

The broad spectrum of risk factor categories is related to the patient (incl. clinical presentation), lesion, stent, and antiplatelet therapy (Table I). Among them, premature cessation of dual antiplatelet therapy (DAPT) seems to be the strongest single risk factor for ST. However, this seems only partially true for early ST, as the majority of patients experiencing ST within the first month remain on DAPT (88% in the Dutch ST Registry) [4]. Furthermore, as shown in the ST-elevation myocardial infarction (STEMI) patient population from the HORIZONS-AMI study, there are differences between ST risk factors for acute, subacute, late, and very late ST [8]. Acute ST was more common in ulcerated lesions and in coronaries with impaired flow at baseline (TIMI 0/1), in younger patients, in patients randomised to bivalirudin rather than unfractionated heparin (UFH) plus glycoprotein IIb/IIIa inhibitors (GPI), and in those not receiving pre-randomisation UFH. The multivariable predictors of subacute ST included insulin-treated diabetes mellitus, history of congestive heart failure, baseline platelet count, baseline and final TIMI 0/1 flow, and non-use of a loading dose of clopidogrel (600 mg). Finally, cigarette smoking and prior ACS were the only independent predictors of late ST, whereas insulin-treated diabetes mellitus, prior PCI, baseline platelet count, and use of UFH plus GPI rather than bivalirudin were the only predictors of very late ST.

Table I

Potential risk factors for early stent thrombosis from the Dutch Stent Thrombosis Registry

Risk factors for early ST	Odds ratio
Premature clopidogrel discontinuation	36.5
Stent under-sizing	13.4
Coronary dissection	6.1
Postprocedural TIMI flow	5.2
≥ 50% lesion proximal to the culprit lesion	4.1
Malignancy	3.0
No aspirin at PCI	2.8
Impaired left ventricle ejection fraction (< 30%)	2.7

Resistance to antiplatelet therapy as a cause for ST is controversial. In a case-controlled, multicentre study, three genes (CYP2C19, ABCB1, and ITGB3) and two clopidogrel-related factors (loading dose and proton pump inhibitors) were identified to be independently associated with early stent thrombosis [9]. However, recently published European Society of Cardiology guidelines on myocardial revascularisation recommend platelet function testing or genetic testing only in specific high-risk situations. These include history of stent thrombosis, suspicion of drug resistance, compliance issue or high bleeding risk (class of recommendation IIb, level of evidence C) [10].

Other lesion/stent-related risk factors for this complication have also been identified, such as bifurcation treatment, stent length, etc.

Several trials have reported ca. 50% reduction of ST with new-generation DES, compared to first generation DES [11]. However, this difference was primarily driven by a significant risk reduction in terms of very late ST, and not early ST.

As the first-generation DES has become practically out of use, the question of the safest profile among second-generation DES is much more intriguing.

Data from a meta-analysis [12] including 85,490 patients suggest that biolimus-eluting stents covered with bioresorbable polymer were characterised by higher rates of 1-year and long-term definite ST, compared to cobalt-chromium everolimus-eluting stents with durable polymer. As stated before [11], this study confirmed the highest risk for very late ST for paclitaxel- and sirolimus-eluting stents.

Thrombosis with the use of bioresorbable vessel scaffoldings. The single centre real-world prospective registry included 135 patients [13]. As much as 3% of ST were observed within 6 months. Further, higher incidence of ST was also observed in the European multicentre GOUST-EU registry [5]. A total of 1189 patients were included, and the cumulative 6-month ST was 2.1%. Finally, among STEMI patients from the prospective RAI registry, treated with BVS implantation, the 6-month ST rate was as high as 2.7% [14]. On the other hand, the group of ACS patients from another registry also experienced relatively high rate of ST at 1 month (2.0%). However, this was comparable to the control DES group (1.9%) [15]. The worrying results from these registries exceeds the ST incidence typically reported in contemporary all-comers registries of second-generation DES. Some authors suggest that, in contrast to the current recommendations for the 6-month DAPT after elective PCI with DES, in cases of BVS 12-month DAPT should be considered [13]. On the other hand, the group of ACS patients from single-centre registry also experienced the relatively high rate of ST at 1 month (2.0%). However, this was comparable to the control DES group (1.9%) [15].

Only a small percentage of STs present as a benign event. In the Japanese RESTART registry as many as 89% of STs presented as ASC [16]. Among the American ST patients from the CathPCI registry, ca 60% presented with STEMI, 23% with non-STEMI, and 17% with unstable angina [17].

Primary PCI is the treatment of choice. Despite a lack of convincing proof, the use of thromboaspiration and IIb/IIIa platelet inhibitors are frequently advocated.

Aggressive post-dilatation with high-pressure balloons seems reasonable to correct stent underexpansion and malapposition. Intravascular imaging tools like optical coherence tomography or intravascular ultrasonography (IVUS) may be of value to guide coronary re-intervention

Both the short- and long-term outcomes are unfavourable; however, some differences were observed. Short-term outcome: in-hospital mortality from the CathPCI registry was significantly higher in early ST (7.9%) compared with late (3.8%) and very late thrombotic event (3.6%) [17]. The 1-year mortality rate from the RESTART registry was significantly lower in patients with very late ST (10.5%) compared with those with early (22.4%) or late occurrence of this event (23.5%) [16]. Long-term outcome: some studies indicate that long-term outcome may be worse after ST within 1 year compared to very late ST. In a study by Kubo et al., which included 152 ST patients, five-year follow-up was reported. Significantly lower cumulative incidence of MACE and all-cause mortality and recurrent ST were observed after very late ST than in early ST and late ST groups. Additionally, significantly lower target lesion revascularisation rate was seen in the very late ST group. However, no statistical difference was seen for cardiac death between these groups [6].

The incidence of recurrent ST is not negligible. Based on the Dutch stent thrombosis registry, recurrent ST is time-dependant. The cumulative incidence of definite or probable recurrent stent thrombosis from this registry was 11.6% during hospitalisation, 14.4% at 1 month, 18.2% at 1 year, 19.6% at 2 years, and 20.1% at 3 years [18]. Patients with subacute ST may be more prone to develop second ST than those with acute ST [19].

The case below illustrates the possible procedural factors that can be associated with this complication.

A 60-year-old male patient with a history of ACS and PCI of diagonal branch in 1994 was admitted for control angiography due to recurrent angina CCS class III and positive exercise treadmill test (ETT), with ST-segment depression in II, II, aVF, V5, and V6 leads and chest pain. Occluded right coronary artery, borderline lesion in the left circumflex coronary artery, and tight lesion in the left anterior descending coronary artery (LAD) were visualised (Figures 1A –C). Percutaneous treatment was chosen and accepted by the patient. Thus, IVUS-guided PCI was performed. Pre-interventional IVUS showed calcified, annular lesion in proximal LAD with minimal lumen area (MLA) of 1.7 mm2. After predilatation with a non-compliant balloon, two biolimus eluting stents (3.5 × 19 mm and 2.5 × 18 mm) were implanted. Despite post-dilation with non-compliant 3.5 mm balloon at 16 atm, the proximal stent edge remained elliptical in shape and underexpanded, with MLA 3.9 mm2 (as compared to reference stent lumen of 7.9 mm2) (Figures 1b and c, respectively). The lesion was left without further post-dilatations due to extensive, annular calcifications. Spot stent strut malapposition was visualised in distal stented segment (Figure 1d).

Figure 1

Images A, B – significant stenosis in proximal segment of the left anterior descending coronary artery. C – Magnification of image B. Intravascular ultrasonography cross-sections after stent implantation: a – excessive calcification, minimal lumen area (MLA) 8.0 mm2; b – ellipsoidal, underexpanded stent, MLA 3.9 mm2; c – well-expanded stent, MLA 7.9 mm2; d – malapposed struts, MLA 4.5 mm2; e – calcified, unstented lesion, MLA 3.2 mm2; f – distal reference, MLA 7.1 mm2; g – longitudinal view

Three days later the patient was re-admitted with STEMI. Urgent angiography revealed totally occluded LAD in the proximal segment due to ST. Several attempts to restore reflow using non-compliant balloons (NC Quantum Apex 3.5 mm) were done, but no more than TIMI I flow was achieved. Thus, manual thrombectomy was performed with removal of thrombi. Next, a biolimus eluting stent 2.5 × 14 mm at 16 atm was implanted distally to previously implanted stents and complex post-dilatation with a non-compliant 3.5 mm balloon was done. TIMI III flow was then restored. Maximal Troponin T rise was 1741 ng/l (UNL < 14). No ECG Q waves were observed after the event. The patient was discharged with no angina or signs of heart failure on the fourth day post-LAD occlusion.

During repeated ETT at 5-month follow-up, 1.2 mm ST-segment depression in only the V5 lead was seen at a workload of 150 Watts (100% of predicted value for age and sex), without any chest discomfort.

Procedural factors leading to acute ST are discussed above and include stent underexpansion and malapposition (especially proximal in location), smaller stent diameter, and coronary dissection [20, 21]. Adequate preparation of calcified lesion before stent deployment is critical for good stent expansion. Interventional options to break annular calcifications and avoid stent underexpansion include rotational atherectomy and cutting balloon or hugging balloon technique [22].

Pharmacological prevention options for high-risk lesions include antiplatelet drugs more potent than clopidogrel.

Primary PCI for STEMI resulting from ST, even if successful, is associated with larger infarct territory and poorer outcome as compared to native STEMI [23]. The role of thrombectomy as an adjunctive tool for primary PCI after positive TAPAS trial and negative INFUSE-AMI and TASTE trials is uncertain. Surprisingly, no differences were observed in the TASTE trial subgroups, defined according to thrombus burden and coronary flow before PCI [24]. However, these studies were addressed to STEMI patients in native coronary arteries and not STEMI secondary to ST. Thus, the impact of thrombectomy in the ST subset of STEMI patients remains unclear. Waldo et al. observed coronary flow improvement and procedural success after use of thrombectomy in ST patients. However, this was not associated with improved long-term outcomes [25]. Conversely, in the study by Lemesle et al., the incidence of the clinical endpoint (death, recurrent MI, and recurrent ST) at 30 days was significantly lower in the thrombectomy group [26]. Similar results were observed in a retrospective study by Mahmoud et al. Among 113 patients with identified ST, manual thrombus aspiration was used in 51 patients. The use of manual thrombectomy was associated with greater epicardial and microvascular myocardial reperfusion. Additionally, mortality was lower in patients treated with thrombus aspiration, although it was not statistically significant [27].

In our case, complete flow restoration was possible only after several passages of thrombectomy. New stent implantation should be avoided unless there is a clear indication such as a flow-limiting dissection or significant lesion, left before without-stent coverage. As observed in the multicentre Spanish ESTROFA registry, implantation of a new stent was associated with five-fold increased risk of ST recurrence [28]. Increased risk of recurrent ST after new stent was also observed in the OPTIMIST study [29].

Single reports of successful systematic fibrinolysis included subacute or late ST and not acute ST [30]. A small series of subacute ST cases showed that systematic thrombolytic therapy may be very effective [31]. In some acute ST cases this treatment may also be the initial option.

Finally, as the patients with ST constitute the group of increased risk of ST recurrence, alternative regimens with more potent antiplatelet drugs may be considered. However, no trials, such as switching to prasugrel or ticagrelor or higher-dose clopidogrel, have been performed in this particular group. Some authors suggest that the benefit from the use of new antiplatelet therapy or prolongation of DAPT over 1 year outweighs the increase in risk of bleeding in most of these patients.