Long-Term Vascular Healing in Response to Sirolimus- and Paclitaxel-Eluting Stents: An Optical Coherence Tomography Study
Clinical Research
Objectives:
This study sought to assess stent strut coverage, malapposition, protrusion, and coronary evaginations as markers of healing 5 years after implantation of sirolimus-eluting stents (SES) and paclitaxel-eluting stents (PES), by optical coherence tomography (OCT).
Background:
Early-generation drug-eluting stents have been shown to delay vascular healing.
Methods:
A total of 88 event-free patients with 1 randomly selected lesion were suitable for final OCT analysis 5 years after drug-eluting stent implantation. The analytical approach was based on a hierarchical Bayesian random-effects model.
Results:
OCT analysis was performed at 5 years in 41 SES lesions with 6,380 struts, and in 47 PES lesions with 6,782 struts. A total of 196 struts were uncovered in SES (1.5%) compared with 185 struts in PES lesions (1.0%, 95% credibility interval [CrI]: 0.5 to 1.6; p = 0.32). Malapposed struts were present in 1.2% of SES compared with 0.7% of PES struts (0.7%, 95% CrI: 0.03 to 1.6; p = 0.23). Protruding struts were more frequent among SES (n = 114; 0.8%) than PES lesions (n = 24; 0.1%, 95% CrI: 0.3 to 1.3; p < 0.01). Coronary evaginations were more common among SES- than PES-treated lesions (17 vs. 7 per 100 cross sections, p = 0.003). During extended clinical follow-up, 2 patients suffered from very late stent thrombosis showing a high degree of malapposition, protrusion, and coronary evaginations at the time of OCT investigation.
Conclusions:
Early-generation drug-eluting stents show a similar degree of strut coverage and malapposition at 5 years of follow-up. Despite an overall low degree of uncovered and malapposed struts in event-free patients, some lesions show a clustering of these characteristics, indicating a heterogeneous healing response, which may be the source for very late adverse events.
Introduction
Although early-generation drug-eluting stents (DES) have a similar safety profile as bare-metal stents do, the phenomenon of late stent thrombosis (ST) emerged as a distinct entity complicating their use (1,2). Experimental studies and autopsy reports identified delayed endothelialization, chronic inflammation, and neoatherosclerosis as morphological features differentiating early generation DES from bare-metal stents (3–6). Incomplete endothelial coverage was identified as the most important predictor of late ST in an autopsy study with a risk continuum that increased with the numbers of uncovered struts (7). In addition, a high incidence of late acquired stent malapposition and positive vessel remodeling correlating with the extent of inflammatory cell infiltration was observed in intravascular ultrasound studies of patients suffering from late ST (8). Recently, differences in the vascular healing response as well as differential mechanisms leading to late ST have been reported for lesions treated with either sirolimus-eluting stents (SES) or paclitaxel-eluting stents (PES) (9).
Optical coherence tomography (OCT) allows for high-resolution intracoronary imaging and has been validated for assessment of stent strut coverage and apposition with an accuracy resembling that of histological examinations (10,11). Using OCT, early generation DES have been associated with a higher frequency of uncovered, malapposed, and protruding struts than bare-metal stents have (12,13). The use of this technology among event-free patients may contribute to the understanding of mechanisms underlying the continuous risk of late ST, may potentially identify patients at risk, and may offer guidance in the need for long-term dual antiplatelet therapy. Most OCT studies to date assessed strut coverage and apposition within the first year of DES implantation (12,14–17). However, autopsy studies indicate that arterial healing after DES implantation is delayed, warranting longer-term imaging follow-up. The present study provides quantitative OCT findings at 5 years after DES implantation complemented by geographic maps integrating the pattern of strut coverage, apposition, and protrusion and describes differences in the vascular healing response between SES and PES.
Methods
Patient population
Patient population
The design and results of SIRTAX (Sirolimus-Eluting Stent Compared With Paclitaxel-Eluting Stent for Coronary Revascularization) and SIRTAX LATE (Sirolimus-Eluting Stent Compared With Paclitaxel-Eluting Stent for Coronary Revascularization-Late) have been reported previously (18,19). For the purpose of the present study, all consecutive patients undergoing angiographic follow-up at 5 years during the period between December 2008 and July 2009 (n = 145) were eligible for OCT imaging (Fig. 1). The inclusion period commenced with the availability of the OCT console at Bern University Hospital in December 2008. The present study was limited to 1 lesion to ensure optimal image quality and minimize patient discomfort. Among patients scheduled for repeat angiography between December 2008 and July 2009, who had more than 1 study lesion (n = 19), all lesions were randomly allocated a numerical code of 1, 2, or 3 by an independent statistician. OCT was routinely performed in the lesion with the lowest number. In 4 patients with multiple lesions, OCT was technically not feasible. In none of these patients, the second or third lesion underwent OCT to respect the random selection. Thus, in 15 of 88 patients suitable for final analysis, lesion selection was random. The study complied with the Declaration of Helsinki regarding investigation in humans and was approved by the institutional ethics committees at Bern University Hospital, Switzerland. All patients provided written informed consent.
Flow Chart Showing Study Design and Patient Flow
OCT = optical coherence tomography; PES = paclitaxel-eluting stent(s); SES = sirolimus-eluting stent(s).
OCT imaging and analysis
OCT was performed with a time domain M2 system (Lightlab Imaging, Westford, Massachusetts) using a pullback speed of 2 mm/s and the nonocclusive flushing technique. After the diagnostic angiography and administration of 5,000 IU unfractionated heparin, the ImageWire (Lightlab Imaging) was carefully advanced distal to the study lesion. Following administration of 200 μg of nitroglycerin intracoronary, the target vessel was flushed via the guiding catheter with nonionic, isosmolar contrast liquid (Iodixanol 320, Visipaque, GE Healthcare, Cork, Ireland) using a power injector with flush rates between 3 and 4 ml/s. OCT pullbacks were assessed offline using a proprietary software (Lightlab Imaging). Lesions were analyzed performing OCT cross sections at 1-mm intervals and assessed for strut coverage, apposition, and protrusion by a single analyst blinded. All frames were reviewed by a second analyst who in case of disagreement consulted with a third referee, and final decision was based on consensus. Pullbacks were excluded in case >30% of the total stent length was not analyzable. Frames were considered not analyzable when more than one-quarter of the circumference was not visible due to insufficient flush or out of zoom. A strut was defined as a signal-intense bright spot with a typical dorsal shadowing. Thickness of strut coverage was measured as the distance between the endoluminal side of the strut in the midpoint of its long axis and the intersection of the lumen contour with the straight line between the endoluminal side of the strut and the gravitational center of the vessel. Struts were considered uncovered in case of a partial or complete absence of tissue coverage. Strut protrusion was defined as strut extension into the lumen for more than 160 μm but with no obvious separation from the vessel wall (Fig. 2). Apposition was assessed by measuring the distance between the center of the endoluminal strut surface and the intersection between lumen contour and the line connecting the center of the endoluminal strut side and the gravitational center of the vessel. Strut malapposition was defined as a distance ≥160 μm based on the consensus derived from the strut thickness of SES (153 μm) and PES (148 μm) plus the minimal axial resolution of OCT (10 μm). This consensus allowed a blinded assessment. Geographic maps were created displaying struts using color codes for strut characteristics, including strut coverage, apposition, and protrusion. The resultant map represented the stented vessel cut longitudinally along the reference angle 0° (corresponding to the 12 o'clock position in the respective OCT cross section) and spread out on an area (Fig. 3A).
Qualitative and Quantitative Assessment of Stent Strut Coverage, Protrusion, and Malapposition
A cross section illustrating all 3 categories is shown on the left. The strut at position 1 is apposed to the vessel wall and covered by a layer (130 μm), whereas the strut in position 2 is uncovered but apposed to the vessel. The strut at position 3 was classified as protruding because the measured protrusion into the lumen relative to an imaginary lumen line (yellow) was >160 μm. Two malapposed struts are shown at position 4 with a separation of >160 μm from the lumen and with absence of tissue between strut and lumen.
Geographic Stent Strut Maps
(A) The concept of the creation of geographic stent strut maps is illustrated. Lesions are presented as areas assuming a cylindrical geometry of the stent. Struts are color coded according to coverage, apposition, and protrusion. The x axis represents the length of the stent (mm), whereas the y axis indicates the position of the strut in the individual cross section ranging from 0° to 360°. Continued on the next page
(B) Strut coverage is presented using green for covered struts and red for uncovered struts. (C) Strut apposition is presented. Apposed struts are shown in green; protruding struts in yellow; and malapposed struts in red. The x-axis represents the length of the stent (mm), whereas the y-axis indicates the position of the strut in the individual cross section ranging from 0° to 360°. Zones of stent overlap are marked with blue lines. Abbreviations as in Figure 1.
Evaginations were suspected whenever the luminal vessel contour extended in a pouchlike fashion beyond the line connecting all stent struts (stent contour). Under these circumstances, the maximal radial distance between the circular line connecting all struts and the luminal vessel wall was evaluated using the thickness ruler function. When the maximal depth exceeded 160 μm (similar cutoff as for the presence of malapposition), we considered the outward bulging as evagination (Fig. 4). By definition, the evagination is limited laterally by stent struts. In addition to the maximal depth of the evagination, the interstrut evagination area was assessed. The interstrut evagination area was defined as the area limited by the stent contour luminally and the lumen contour abluminally.
Coronary Evaginations
(A) Illustration of a cross-section showing 3 coronary evaginations. (B) The quantitative assessment of coronary evagination depth and area is shown. A1 to A3 relates to evagination areas; d1 to d3 relates to evagination depth.
Statistical analysis
We used a Bayesian hierarchical random-effects model based on Markov chain Monte Carlo simulation methods with vague priors to estimate differences between SES and PES (18). For analyses at the cross-section and strut level, the model included random-effects at the level of patients, fully accounting for the correlation of characteristics of cross-sectional areas or struts within patients and implicitly assigning analytical weights to each lesion depending on the number of cross sections or on the number of struts observed per lesion. For continuous outcomes, we assumed a log normal distribution; for counts, we used a Poisson distribution; and we used appropriate transformations to derive arithmetic means and rates, respectively. Differences in the percentage of lesions with any struts with unfavorable outcome, with at least 5%, and with at least 10% of struts with unfavorable outcome were calculated using a Bayesian hierarchical random-effects model assuming a Bernoulli distribution. For all other analyses at the lesion level, we used conventional linear and Poisson regression models, depending on the nature of the outcome (continuous or counts). We derived 95% credibility intervals (CrI) from the 2.5th and 97.5th percentiles of the posterior distribution, also calculating 2-sided p values from the posterior distribution. Statistical analyses were performed using WinBUGS (version 1.4.3, Imperial College and Medical Research Council, London, United Kingdom) and Stata (version 11.0, StataCorp, College Station, Texas).
Results
Patients
Patients
The flow of patients included into the OCT study 5 years after DES implantation is shown in Figure 1. Of 95 patients undergoing OCT at 5.3 years (interquartile range: 5.1 to 5.5 years), 46 patients had been treated with SES and 49 patients with PES. Five SES patients and 2 PES patients were excluded (Fig. 1), resulting in 41 SES and 47 PES patients included into the final analysis. Baseline clinical and angiographic characteristics among patients undergoing OCT at 5 years were well balanced for both groups (Table 1). Baseline angiographic and procedural characteristics at the time of DES implantation were similar for both groups, including lesion length; vessel size; and the number, diameter, and length of implanted stents (Table 2). Angiographic follow-up at 5 years showed similar results in terms of minimal lumen diameter, percentage of diameter stenosis, late loss, and restenosis for both stent types (Table 3).
| SES (n = 41) | PES (n = 47) | p Value | |
|---|---|---|---|
| Age >60 yrs | 16(39) | 28(59.3) | 0.10 |
| Male | 34(82.9) | 36(76.6) | 0.46 |
| Diabetes mellitus | 8(19.5) | 8(17.0) | 0.76 |
| Insulin dependence | 2(4.9) | 3(6.4) | 0.76 |
| Hypertension | 21(51.2) | 30(63.8) | 0.23 |
| Hyperlipidemia | 25(61.0) | 25(53.2) | 0.46 |
| Current smoking | 19(46.3) | 16(34.0) | 0.24 |
| Previous myocardial infarction | 11(26.8) | 14(29.89) | 0.76 |
| Stable angina pectoris | 15(36.6) | 21(44.7) | 0.45 |
| Acute coronary syndromes | |||
| Unstable angina | 4(9.8) | 1(2.1) | 0.45 |
| Non–STEMI | 9(22.0) | 11(23.4) | |
| STEMI | 13(31.7) | 14(29.8) | |
| Multivessel disease | 30(73.8) | 30(63.8) | 0.35 |
| Lesion(s) per patient | |||
| 1 | 35(85.4) | 37(78.7) | 0.71 |
| 2 | 5(12.2) | 8(17.0) | |
| 3 | 1(2.4) | 2(4.3) | |
| Left ventricular ejection fraction, % | 56.7 | 58.6 | 0.40 |
| SES | PES | p Value | |
|---|---|---|---|
| Lesions, n | 41 | 47 | |
| Target lesion coronary artery | |||
| Left main | 1(2.4) | 1(2.1) | 0.31 |
| Left anterior descending | 17(41.5) | 25(53.2) | |
| Left circumflex | 13(31.7) | 7(14.9) | |
| Right | 10(24.4) | 14(29.8) | |
| ACC-AHA lesion class | |||
| A | 4(9.8) | 9(19.1) | 0.53 |
| B1 | 17(41.5) | 20(42.6) | |
| B2 | 13(31.7) | 10(21.3) | |
| C | 7(17.1) | 8(17.0) | |
| Angiographic measurements | |||
| Lesion length | 16.95±7.84 | 15.70±7.23 | 0.44 |
| Reference vessel diameter | 2.87±0.40 | 2.89±0.41 | 0.82 |
| Minimal lumen diameter | 0.40±0.37 | 0.48±0.38 | 0.36 |
| Stenosis, % lumen diameter | 85.95±12.6 | 83.38±13.0 | 0.35 |
| Pre-procedure TIMI flow grade | |||
| 0 | 13(31.7) | 8(17.0) | 0.44 |
| 1 | 2(4.9) | 2(4.3) | |
| 2 | 3(7.3) | 4(8.5) | |
| 3 | 23(56.1) | 33(70.2) | |
| Post-procedure TIMI flow grade | 1.0 | ||
| 0 | 0(0) | 0(0) | |
| 1 | 0(0) | 0(0) | |
| 2 | 1(2.4) | 1(2.1) | |
| 3 | 40(97.6) | 46(97.9) | |
| Thrombus present | 13(32.5) | 16(34.0) | 0.88 |
| Procedures | |||
| Study stents per lesion | 1.17±0.4 | 1.15±0.4 | 0.80 |
| Stent diameter, mm | 2.88±0.39 | 2.96±0.37 | 0.34 |
| Total stent length per lesion, mm | 20.59±9.05 | 18.47±7.67 | 0.24 |
| Maximal pressure, atm | 15.32±3.6 | 14.43±3.2 | 0.22 |
| Direct stenting | 0.22±0.4 | 0.30±0.5 | 0.41 |
| Angiographic results | |||
| Reference vessel diameter, mm | 2.89±0.50 | 2.89±0.44 | 1.00 |
| Final minimal lumen diameter, mm | |||
| In-stent | 2.68±0.41 | 2.75±0.41 | 0.42 |
| In-segment | 2.60±0.43 | 2.75±0.52 | 0.25 |
| Final stenosis, % of lumen diameter | |||
| In-stent | 7.73±4.5 | 5.55±4.3 | 0.03 |
| In-segment | 9.80±6.4 | 6.12±6.6 | 0.04 |
| Acute gain, mm | |||
| In-stent | 2.28±0.50 | 2.28±0.51 | 1.00 |
| In-segment | 2.22±0.49 | 2.32±0.66 | 0.54 |
| SES | PES | Difference (95% CrI) | p Value | |
|---|---|---|---|---|
| Lesions, n | 41 | 47 | ||
| Reference vessel diameter, mm | 2.83±0.44 | 2.86±0.38 | −0.03(−0.21to0.14) | 0.69 |
| Minimal lumen diameter, mm | ||||
| In-stent | 2.39±0.71 | 2.45±0.76 | −0.06(−0.28to0.16) | 0.57 |
| In-segment | 2.31±0.72 | 2.33±0.77 | −0.02(−0.24to0.20) | 0.86 |
| % diameter stenosis | ||||
| In-stent | 15.82±17.2 | 14.78±18.4 | 1.04(−4.22to6.31) | 0.70 |
| In-segment | 18.23±17.9 | 18.41±19.2 | −0.19(−5.67to5.30) | 0.95 |
| Late loss, mm | ||||
| In-stent | 0.28±0.43 | 0.28±0.46 | 0.00(−0.13to0.13) | 1.00 |
| In-segment | 0.26±0.39 | 0.28±0.42 | −0.02(−0.14to0.10) | 0.72 |
| Binary restenosis | ||||
| In-stent | 2(4.9) | 2(4.3) | 0.62(−8.52to9.76) | 0.89 |
| In-segment | 2(4.9) | 3(6.4) | −1.50(−12.6to9.56) | 0.79 |
OCT data
Quantitative analysis of luminal, stent, and neointimal volume and percentage of volume obstruction showed no differences between SES and PES at 5 years of follow-up (Table 4).
| SES (95% CrI) | PES (95% CrI) | Difference (95% CI) | p Value | |
|---|---|---|---|---|
| Analysis at lesion level | ||||
| Lesions analyzed, n | 41 | 47 | ||
| Cross sections analyzed per lesion | 19.1(17.9to20.5) | 16.8(15.7to18.0) | 1.13(1.03to1.26) | 0.01 |
| Struts analyzed per lesion | 155.6(151.8to159.5) | 144.3(140.9to147.8) | 1.08(1.04to1.11) | <0.001 |
| Minimal luminal area, mm2 | 4.66(3.95to5.38) | 5.08(4.49to5.67) | 0.42(−0.48to1.32) | 0.36 |
| Minimal stent area, mm2 | 5.37(4.81to5.94) | 6.22(5.62to6.81) | 0.84(0.02to1.66) | 0.04 |
| Percentage of volume obstruction | 12.2(8.58to14.9) | 13.7(11.6to15.8) | 1.53(−2.43to5.48) | 0.45 |
| Analysis at cross-section level | ||||
| Cross sections analyzed, n | 785 | 790 | ||
| Struts per cross section | 7.98(7.52to8.45) | 8.36(7.90to8.82) | −0.38(−1.00to0.28) | 0.25 |
| Luminal area, mm2 | 5.67(5.05to6.32) | 6.33(5.72to7.04) | −0.66(−1.62to0.19) | 0.15 |
| Stent area, mm2 | 6.50(3.62to7.17) | 7.35(6.72to12.3) | −0.85(−8.83to0.14) | 0.09 |
| Neointimal thickness, mm | 0.11(0.09to0.12) | 0.11(0.10to0.13) | −0.01(−0.03to0.02) | 0.64 |
| Neointimal area, mm2 | 0.97(0.86to1.10) | 1.03(0.91to1.16) | −0.06(0.24to0.12) | 0.46 |
| Mean area ISA, mm2 | 0.70(0.50to0.96) | 0.68(0.49to0.94) | 0.02(−0.31to0.34) | 0.88 |
| Mean malapposition distance, mm | 0.27(0.24to0.31) | 0.29(0.26to0.34) | −0.02(−0.07to0.03) | 0.47 |
| Number of evaginations | 0.17(0.10to0.26) | 0.07(0.04to0.10) | 0.10(0.03to0.20) | 0.003 |
| Mean evagination area, mm2 | 0.20(0.17to0.24) | 0.23(0.19to0.28) | −0.03(−0.08to0.02) | 0.24 |
| Mean evagination depth, mm | 0.25(0.24to0.28) | 0.24(0.22to0.27) | 0.01(−0.02to0.04) | 0.47 |
Results of strut-level and lesion-level OCT analyses stratified according to stent type are presented in Table 5. A total of 6,380 struts in 41 SES lesions and 6,782 struts in 46 PES lesions were analyzed. Uncovered struts were observed among 1.5% (95% CrI: 0.8% to 2.6%) of all SES struts compared with 1.0% (95% CrI: 0.5% to 1.7%) of all PES struts (weighted difference: 0.5%, 95% CrI: 0.5% to 1.6%; p = 0.32). Lesion-level analysis showed no difference in the proportion of lesions with ≥5% (10.7% vs. 7.2%, 95% CrI: 9.6% to 18.7%; p = 0.60) as well as ≥10% uncovered struts (2.4% vs. PES 2.0%, 95% CrI: 4.8% to 7.9%; p = 0.81) between SES and PES (Table 5, Fig. 5A). A geographic map with the spatial distribution of uncovered and covered struts is provided in Figure 3B. A high density of uncovered struts is noted in Lesions #10 and #13 of SES-treated patients and Lesion #35 of PES-treated patients.
| SES (95% CrI) | PES (95% CrI) | Difference (95% CrI) | p Value | |
|---|---|---|---|---|
| Analysis at strut level | ||||
| Struts analyzed, n | 6,380 | 6,782 | ||
| Uncovered struts, % | 1.5(0.8to2.6) | 1.0(0.5to1.7) | 0.5(−0.5to1.6) | 0.32 |
| Protruding struts, % | 0.8(0.4to1.4) | 0.1(0.0to0.3) | 0.7(0.3to1.3) | <0.01 |
| Malapposed struts, % | 1.2(0.6to2.2) | 0.7(0.3to1.3) | 0.5(−0.3to1.6) | 0.23 |
| Analysis at lesion level | ||||
| Uncovered struts, lesions with | ||||
| At least 10% uncovered struts | 2.4(0.3to10.8) | 2.0(0.2to7.6) | 0.4(−4.8to7.9) | 0.81 |
| At least 5% uncovered struts | 10.7(2.9to26.7) | 7.2(1.7to20.6) | 3.1(−9.6to18.7) | 0.60 |
| Protruding struts, lesions with | ||||
| At least 10% protruding struts | 0.6(0.0to5.0) | 0.0(0.0to0.8) | 0.5(−0.3to5.0) | 0.12 |
| At least 5% protruding struts | 3.7(0.6to13.8) | 0.3(0.0to3.6) | 3.1(−0.2to13.0) | 0.07 |
| Malapposed struts, lesions with | ||||
| At least 10% malapposed struts | 5.4(1.0to17.6) | 0.4(0.0to6.9) | 4.6(−0.0to16.3) | 0.05 |
| At least 5% malapposed struts | 24.0(8.9to45.2) | 5.7(1.3to15.9) | 17.5(1.9to39.3) | 0.03 |
Frequency Distribution of Uncovered and Malapposed Struts
Cumulative distribution of the proportion of uncovered struts (A) and strut malapposition (B). Abbreviations as in Figure 1.
Overall, malapposed struts were observed in 1.2% (95% CrI: 0.6% to 2.2%) of all SES struts compared with 0.7% (95% CrI: 0.3% to 1.3%) of all PES struts (weighted difference: 0.5%, 95% CrI: 0.03 to 1.6; p = 0.23). The mean area of stent malapposition showed no difference between SES and PES (SES: 0.70 mm2 [95% CrI: 0.5% to 0.96%] vs. PES: 0.68 mm2 [95% CrI: 0.49 to 0.94]; p = 0.88). Lesion-level analysis of malapposition showed more lesions with ≥5% (24.0% vs. 5.7%, weighted difference: 17.5%, 95% CrI: 1.9% to 39.3%; p = 0.03), as well as ≥10% malapposed struts among SES- than PES-treated patients (5.4% vs. 0.4%, weighted difference: 4.6, 95% CrI: 0.0% to 16.3%; p = 0.05) (Table 5, Fig. 5B), indicating an accumulation of malapposed struts in some SES lesions. Geographic stent strut maps are confirmatory in this regard and allow a visual assessment of the distribution of malapposed struts within a lesion.
Protruding struts were more frequent among SES (0.8%, 95% CrI: 0.4% to 1.4%) than PES in the strut-level analysis (0.1%, 95% CrI: 0.0% to 0.3%; weighted difference: 0.7%, 95% CrI: 0.3% to 1.3%; p < 0.01). Similarly, the number of lesions with ≥5% (3.7% vs. 0.3%, weighted difference: 3.1%, 95% CrI: 0.2% to 13%; p = 0.07), as well as ≥10% protruding struts (0.6% vs. 0.0%, weighted difference: 0.5%, 95% CrI: 0.0% to 16.3%; p = 0.12) tended to be higher among SES than PES (Table 5) in the lesion-level analysis.
In a total of 5 SES lesions and 2 PES lesions, overlapping stents were observed. The overlapping zones were delineated in the strut maps in Figures 3B and 3C. Visual inspection shows that neither uncovered nor malapposed struts were more frequent in overlapping zones.
The number of coronary evaginations was higher among SES- than PES-treated lesions (0.17 vs. 0.07 per cross sections; p = 0.003), with no difference in mean area and depth of individual evaginations. A geographic map showing the spatial distribution of apposed, malapposed, and protruding struts is provided in Figure 3C. A high density of strut malapposition or protrusion (>20%) is visible in Lesions #7, #13, #21, and #40 of SES-treated patients. Of these 4 SES patients, 2 (#7 and #40) suffered very late ST at 6 months and 1 year after acquisition of OCT imaging, respectively.
Discussion
The present OCT analysis performed among event-free patients 5 years after the intervention focused on the vascular healing response to early-generation SES and PES implanted in the framework of an all-comers randomized trial and has the following principal findings:
| 1 | Neointimal thickness and volume are low and of similar magnitude for SES and PES at 5 years. | ||||
| 2 | Strut-level analysis shows an overall low frequency of uncovered, malapposed, or protruding struts at 5 years. | ||||
| 3 | Geographic maps identified a few patients with a high degree of uncovered, malapposed, or protruding struts suggesting a heterogeneous healing pattern 5 years after early-generation DES implantation. | ||||
| 4 | Lesion-level analysis and geographic maps demonstrate a clustering of malapposition and protrusion in SES- versus PES-treated lesions, and coronary evaginations were more frequently observed in SES, suggesting a potential difference in the healing response of the 2 devices at 5 years of follow-up. | ||||
Neointimal thickness, neointimal volume, and percentage of volume obstruction were low and of similar magnitude for SES and PES at 5 years. Although these data were obtained in selected, nonrandomized patients, the OCT findings of the present study confirm similar observations in a recent autopsy study as well as an angiographic study (19) with late follow-up indicating the absence of significant differences in neointimal hyperplasia between SES and PES during long-term follow-up (9). Figure 6 illustrates the spectrum of neointimal phenotypes encountered 5 years after implantation of SES and PES.
Spectrum of Neointimal Phenotypes Observed at 5 Years After Implantation of SES and PES
(A1 to A3) SES. (B1 to B3) PES. Panels A1 and B1 show absence of coverage, whereas in A2 and B2, minimal coverage with a thickness <100 μm can be observed. In A3 and B3, a more pronounced coverage is present with a maximal thickness of 820 μm in SES (A3) and 600 μm in PES (B3). Abbreviations as in Figure 1.
Strut coverage
A number of OCT studies have investigated strut coverage among SES-treated lesions at various time points, but only very few are available for PES-treated lesions. The rate of uncovered struts in SES-treated lesions amounted to 15% at 3 months (20), 11% at 6 months (15), and 2.1% at 9 months (16). Although a direct comparison between the present study and previous reports is limited due to patient and lesion heterogeneity as well as differences in the analytical approach, there is a consistent increase in strut coverage, which is most pronounced during the first year but continues to accrue over time, resulting in a rate of uncovered struts of only 1% to 2% at 5 years. This observation is corroborated in a recent autopsy study reporting a decrease in the incidence of uncovered struts over time particularly among DES implanted in on-label indications (9). The same study also observed no difference in the proportion of uncovered struts between SES and PES in analogy to our OCT findings. To date, only 1 autopsy study showed a correlation between uncovered struts and the risk of stent thrombosis, suggesting delayed endothelialization and incomplete healing as potential mechanisms of late ST after DES implantation. Specifically, the odds for late ST were 9-fold increased among stents with more than 30% of uncovered struts per cross section compared with stents in control subjects (7). This observation has not been validated among living DES-treated patients using intracoronary imaging so far. Moreover, most available OCT studies only addressed overall strut coverage using strut-level (cross-sectional) analyses without accounting for a potential clustering of uncovered struts within lesions (patients). To address this limitation, we performed both a strut-level and lesion-level analysis and provide geographic maps of strut coverage for individual lesions. Whereas overall strut coverage was found to be nearly complete, lesion-level analysis indicated that 10.7% of SES- and 7.2% of PES-treated lesions had at least 5% uncovered struts. Accordingly, clustering (>10%) with a higher density of uncovered struts was limited to few lesion numbers (SES #10, SES #13, PES #35, and PES #40) (Fig. 5A), whereas most geographic maps revealed only isolated single uncovered struts, suggesting an individual healing response after DES implantation.
Strut apposition
Stent malapposition 5 years after DES implantation may be related to persistent or late acquired malapposition after resolution of thrombus or due to a dynamic process with positive vessel remodeling over time related to DES-induced inflammation and toxicity (8). The assessment of malapposition in this study has to be interpreted in the light of 1 important limitation: In the absence of a baseline investigation, it is not possible to differentiate whether malapposed struts at 5 years were present at the time of the index procedure (persistent) or whether malapposition developed during follow-up (late acquired).
Overall, malapposed struts were rare and occurred with similar frequency among SES- and PES-treated lesions in the strut-level analysis. However, lesion-level analysis of strut malapposition revealed clustering, with a higher density of malapposed struts among SES- than PES-treated lesions.
Protruding struts according to our definition did protrude at least 160 μm into the lumen and were always in contact with the vessel wall. Protruding struts may represent a stage of healed, formerly malapposed struts related to incomplete stent apposition at the time of DES implantation or may be the result of an outward remodeling of the vessel wall giving the appearance of coronary evaginations between the struts. Although protruding struts were rare overall, they occurred more frequently among SES- than PES-treated lesions. Taken together with the more pronounced clustering of malapposed struts in SES lesions, this observation suggests a differential healing response following implantation of SES and PES during long-term follow-up. A differential healing response of the 2 stent types has also been reported in a recent autopsy study. Histologically, an increased inflammatory response resulting in positive remodeling and malapposition has been associated with SES, whereas an excessive para-strut fibrin deposition was observed in PES-treated lesions (9).
Although this study did not intend to investigate the impact of stent strut-related findings on clinical outcome, it is noteworthy that 2 patients with a high density of both protruding and malapposed struts as documented 5 years after DES implantation developed late ST 6 months (SES #7) and 1 year (SES #40) (Fig. 3C) after completion of this OCT investigation. Table 6 summarizes the findings in both patients at baseline, follow-up, and at the time of the very late ST. An OCT cross section of the first patient (SES #40) obtained at 5 years is shown in Figure 4, where excessive coronary evaginations are noted (no OCT available at the time of very late ST), whereas serial OCT findings (at 5 years and the time of very late ST) are shown in Figure 7 for lesion SES #7. The 2 cases illustrate that OCT may play a role in identifying patients at risk for future adverse ischemic events. With respect to pathomechanisms leading to very late ST, the 2 cases provide evidence that very late ST beyond 5 years after DES implantation is not solely related to neoatheroslerosis and late restenosis, as both findings were not present.
| Case #1 | Case #2 | |
|---|---|---|
| Baseline findings | ||
| Age, yrs | 28 | 56 |
| cvRF |
|
|
| Indication for PCI at baseline | STEMI | STEMI |
| Target lesion | Proximal LAD | Proximal RCA |
| Treatment at baseline |
|
|
| TIMI flow grade before | 0 | 0 |
| TIMI flow grade after | 3 | 3 |
| LVEF, % | 50 | 50 |
| Maximum CK, U/l | 5,640 | 2,922 |
| DAPT duration, yrs | 1 | 1 |
| OCT findings at 5 yrs | ||
| Uncovered struts, % | 1.4 | 2.9 |
| Malapposed struts,% | 31 | 23 |
| Protruding struts, % | 35 | 8 |
| Coronary evagination | ||
| Depth, mm | 0.62 | 0.42 |
| Area, mm2 | 0.58 | 0.37 |
| Per cross section, n | 1.75 | 1.75 |
| Map # | #40 | #7 |
| Findings at time point of VLST | ||
| Time point of VLST, yrs after index procedure | 6 | 5.5 |
| Time between OCT and VLST, yrs | 1 | 0.5 |
| Antiplatelet therapy before VLST | Only aspirin | Only aspirin |
| Clinical presentation | Anterior STEMI | Inferior STEMI |
| OCT performed | No | Yes (Fig. 7) |
| Interventional treatment | Thrombus aspiration and stent implantation (zotarolimus-eluting stent, 3.5 × 30 mm) | Thrombus aspiration and balloon dilation without stent implantation |
| TIMI flow grade before | 0 | 2 |
| TIMI flow grade after | 3 | 3 |
| LVEF, % | 40 | 50 |
| Maximal CK, U/l | 900 | Troponin T 0.2 ng/ml |
Paired OCT Investigation at 5 Years (Routine per Protocol) and 5.5 Years (at the Time of Very Late ST)
The serial cross sections in A1 to A4 show protruding struts with coronary evaginations of the vessel wall (A1) and malapposition in subsequent frames (A2 to A4). The vicinity of protrusion and malapposition suggests that coronary evaginations with protruding struts may precede a detachment of the vessel wall from the stent struts, leaving behind the visual appearance of (late acquired) malapposed struts (A2 to A4). The potential clinical relevance of these findings by optical coherence tomography (OCT) are supported by the occurrence of a very late stent thrombosis (ST) 6 months after the 5-year OCT investigation (B1 to B4). In B1 to B4, the zones of malapposition are filled with material suggestive of thrombus.
Coronary evaginations
Coronary evaginations reflect a distinct vessel wall morphology, which is characterized by an outward bulging of the lumen between stent struts (pouches) (Fig. 4). In case of >1 evagination per cross section, the involved stent struts may appear as protruding. Whereas the phenomenon has been reported in case reports, no single study has described the incidence and underlying mechanism to date. Histological evaluations of coronary evaginations have not been reported so far, which may be related to the fact that the outward ballooning is more apparent in vivo in a pressurized vessel than after histological processing. We describe, for the first time, systematically the incidence and the extent of this OCT finding and show that coronary evaginations are more common in SES- than PES-treated lesions but with similar cross-sectional areas and depth. The clinical significance of coronary evaginations remains unclear. Hypothetically, coronary evaginations may represent an early stage of positive remodeling. Advanced coronary evaginations may appear angiographically as peristent contrast staining, an entity that has been recently correlated with late adverse clinical outcome (21).
Study limitations
The presented data have to be interpreted in light of several limitations. First, the data were obtained in a highly selected patient population of event-free individuals 5 years after DES implantation. Second, the present study provides OCT findings only at 5 years without a baseline examination. This has implications regarding the analysis of malapposed struts as it cannot be excluded that differences in malapposed struts were already present at baseline (persistent rather than late acquired). However, intravascular ultrasound studies have shown that an important proportion of malapposed struts at follow-up is related to acquired rather than persistent malapposed struts (22). The dynamic changes in the interaction of stent struts with the arterial wall remain hypothetical and will require confirmation in prospectively designed, serial OCT investigations. As it relates to both strut protrusion and coronary evaginations, matched histological evaluations are not available; therefore, careful interpretations of these OCT findings are required. The clinical impact of protruding struts and coronary evaginations is not known and requires further evaluation in prospective studies.
Conclusions
Early-generation DES show a similar degree of strut coverage and malapposition at 5-year follow-up. Despite overall low rates of uncovered, malapposed, and protruding struts, some lesions show a clustering of these characteristics, indicating a heterogeneous healing pattern among patients treated with early-generation DES. Two very late ST cases in patients with a high number of malapposed or protruding struts illustrate that OCT may play a role in identifying patients at risk for future adverse ischemic events.
1. : "Outcomes associated with drug-eluting and bare-metal stents: a collaborative network meta-analysis". Lancet 2007; 370: 937.
2. : "Early and late coronary stent thrombosis of sirolimus-eluting and paclitaxel-eluting stents in routine clinical practice: data from a large two-institutional cohort study". Lancet 2007; 369: 667.
3. : "Long-term effects of polymer-based, slow-release, sirolimus-eluting stents in a porcine coronary model". Cardiovasc Res 2004; 63: 617.
4. : "Neointimal thickening after stent delivery of paclitaxel: change in composition and arrest of growth over six months". J Am Coll Cardiol 2000; 36: 2325.
5. : "Comparison of inflammatory response after implantation of sirolimus- and paclitaxel-eluting stents in porcine coronary arteries". Circulation 2009; 120: 141. 1–2.
6. : "The pathology of neoatherosclerosis in human coronary implants bare-metal and drug-eluting stents". J Am Coll Cardiol 2011; 57: 1314.
7. : "Pathological correlates of late drug-eluting stent thrombosis: strut coverage as a marker of endothelialization". Circulation 2007; 115: 2435.
8. : "Correlation of intravascular ultrasound findings with histopathological analysis of thrombus aspirates in patients with very late drug-eluting stent thrombosis". Circulation 2009; 120: 391.
9. : "Coronary responses and differential mechanisms of late stent thrombosis attributed to first-generation sirolimus- and paclitaxel-eluting stents". J Am Coll Cardiol 2011; 57: 390.
10. : "Does optical coherence tomography identify arterial healing after stenting?: An in vivo comparison with histology, in a rabbit carotid model". Heart 2008; 94: 217.
11. : "Accuracy of optical coherence tomography in the evaluation of neointimal coverage after stent implantation". J Am Coll Cardiol Img 2010; 3: 76.
12. : "Strut coverage and late malapposition with paclitaxel-eluting stents compared with bare metal stents in acute myocardial infarction: optical coherence tomography substudy of the Harmonizing Outcomes with Revascularization and Stents in Acute Myocardial Infarction (HORIZONS-AMI) trial". Circulation 2011; 123: 274.
13. : "Neointimal coverage of bare-metal and sirolimus-eluting stents evaluated with optical coherence tomography". Heart 2008; 94: 566.
14. : "Evaluation in 3 months duration of neointimal coverage after zotarolimus-eluting stent implantation by optical coherence tomography: the ENDEAVOR OCT trial". J Am Coll Cardiol Intv 2009; 2: 1240.
15. : "Neointimal coverage of sirolimus-eluting stents at 6-month follow-up: evaluated by optical coherence tomography". Eur Heart J 2007; 28: 961.
16. : "An optical coherence tomography study of a biodegradable vs. durable polymer-coated limus-eluting stent: a LEADERS trial sub-study". Eur Heart J 2010; 31: 165.
17. : "Strut coverage and vessel wall response to zotarolimus-eluting and bare-metal stents implanted in patients with ST-segment elevation myocardial infarction: the OCTAMI (Optical Coherence Tomography in Acute Myocardial Infarction) study". J Am Coll Cardiol Intv 2010; 3: 680.
18. : Bayesian Approaches to Clinical Trials and Health Care Evaluation . Chichester, UK: John Wiley & Sons2004.
19. : "Five-year clinical and angiographic outcomes of a randomized comparison of sirolimus-eluting and paclitaxel-eluting stents: results of Sirolimus-Eluting Versus Paclitaxel-Eluting Stents for Coronary Revascularization LATE trial". Circulation 2011; 123: 2819.
20. : "Comparison of neointimal coverage by optical coherence tomography of a sirolimus-eluting stent versus a bare-metal stent three months after implantation". Am J Cardiol 2008; 102: 27.
21. : "Incidence, risk factors, and clinical sequelae of angiographic peri-stent contrast staining after sirolimus-eluting stent implantation". Circulation 2011; 123: 2382.
22. : "Incidence, mechanisms, predictors, and clinical impact of acute and late stent malapposition after primary intervention in patients with acute myocardial infarction: an intravascular ultrasound substudy of the Harmonizing Outcomes with Revascularization and Stents in Acute Myocardial Infarction (HORIZONS-AMI) trial". Circulation 2010; 122: 1077.
Abbreviations and Acronyms
| CrI | credibility interval |
| DES | drug-eluting stent(s) |
| OCT | optical coherence tomography |
| PES | paclitaxel-eluting stent(s) |
| SES | sirolimus-eluting stent(s) |
| ST | stent thrombosis |
Footnotes
This study was supported by research grants from
