Introduction
In-stent restenosis (ISR) is well recognized as a clinically significant complication after coronary stent implantation, observed in 20–30% of patients undergoing bare-metal stent implantation and 5–15% undergoing drug-eluting stent (DES) implantation [1,2]. It may be due to biologic, mechanic or technical factors [3]. Of these, stent under-expansion plays an important role and coronary artery calcium (CAC) has been recognized as a major determinant of stent underexpansion during the percutaneous coronary intervention (PCI). Thus, an accurate evaluation of CAC is of clinical significance in planning a PCI strategy.
According to the results from studies of intravascular imaging, stent expansion is mainly associated with the angle, length and thickness of calcium deposition. Using intravascular ultrasound (IVUS), stent expansion was found to be inversely correlated with the arc of calcium [4]. An optical coherence tomography (OCT)-based calcium scoring system further demonstrated that lesions with a maximum angle greater than 180°, maximum thickness more than 0.5 mm and length more than 5 mm would be at risk of stent underexpansion [5]. Although it was generally agreed that coronary angiography was less sensitive to detect CAC compared to intravascular imaging, Wang et al., [6] recently reported that angiographically invisible calcium (only detectable by IVUS or OCT) did not appear to inhibit stent expansion and angiographically visible
calcium seemed to be a good marker for predicting stent underexpansion [6]. It might be because some of the angiographically invisible calcium were thinner (<0.50 mm) in thickness, which is associated with a greater stent expansion because of calcium fracture during PCI [7,8]. Therefore, the evaluation of calcium density and total volume is imperative for predicting stent expansion.
Computed tomography (CT) is the only noninvasive modality with high sensitivity and specificity for calcium detection. Usually, the severity of CAC assessed by CT images is quantified with a method introduced by Agatston et al., [9] called the Agatston score. It takes into account both the area and peak density of calcified lesions. Previous studies have suggested that the Agatston score correlates to late lumen loss [10] and might be helpful in determining the treatment strategy for complex coronary artery lesions with severe calcification [11]. However, the exact relationship between the calcium score and ISR is not well understood.
Traditionally, Agatston-score assessment is based on ECG-gated coronary CT angiography (CCTA) [9]. However, nongated noncontrast chest CT (NCCT) is more technically simple and widely used in routine practice compared with CCTA. Previous studies have illustrated a high correlation between NCCT and CCTA for CAC quantification [12]. However, the utility of CAC quantification using NCCT for predicting ISR after coronary stent implantation has not been well described. Moreover, a clinically useful cutoff value of CAC severity for optimizing interventional strategy to reduce ISR risk has not been well determined.
Accordingly, we performed this study to explore the associations between CCTA- or NCCT-derived CAC score and ISR.
Methods
Study design
This single-center, retrospective study was conducted in Shanghai Chest Hospital, Shanghai Jiao Tong University, following the principles of the Declaration of Helsinki and local regulations. This study was approved by the ethics committee of Shanghai Chest Hospital, Shanghai Jiao Tong University. Written informed consents were waived.
Study population
We retrospectively screened patients (≥18 years) undergoing DES implantation for de novo coronary lesions in the Department of Cardiology, Shanghai Chest Hospital, Shanghai Jiao Tong University from January 2010 to January 2020. We enrolled patients who had documented coronary angiographic images at both index procedure and follow-up after 6–18 months and received a CT (i.e. NCCT or CCTA) scan within 3 months before index procedure. Exclusion criteria were ST-segment-elevation myocardial infarction, prior coronary artery bypass grafting, chronic renal failure, hemodialysis, or metal implants affecting CAC quantification. Angiographic exclusion criteria were stent fracture or reference vessel diameter of the target lesion below 2.5 mm assessed by quantitative coronary angiography (QCA). We assumed that ISR in the vessel with reference diameter below 2.5 mm may be mainly caused by a small in-stent lumen diameter but not CAC.
Study endpoints
The primary endpoint of this study was ISR at angiographic follow-up, defined as lumen diameter stenosis over 50% at the stent segment or its proximal or distal edges (5-mm segments adjacent to the stent) [13]. We assessed lumen diameter stenosis and late lumen loss by QCA, that is, diameter stenosis = (reference vessel diameter − minimal lumen diameter)/reference vessel diameter × 100%), and late lumen loss = postprocedural minimal lumen diameter − minimal lumen diameter at follow-up. Target lesion revascularization (TLR) was also recorded with medical chart review.
Quantitative coronary angiography analysis
QCA was performed following the standard procedure of analysis using a validated software (QAngio XA 7.3, Medis Medical Imaging Systems, Leiden, the Netherlands). QCA was analyzed by two independent, well-trained technicians who had performed QCA in at least 50 patients, who were blinded with data on CAC score. Case examples were present in Supplementary Figure S1–S6, Supplemental digital content 1, https://links.lww.com/MCA/A484.
Computed tomography and coronary artery calcium quantification
Images of CCTA or NCCT were acquired using single-source ≥64-row CT scanners. Using reproducible landmarks such as the ostium of main vessel and side branches, the target segment undergoing PCI was determined on images of CCTA or NCCT with the use of angiographic images as a reference. Scans were interpreted by two investigators independently, who were blinded with data on QCA analyses. CAC was quantified using the Agatston method [9] with a commercially available software (OsiriX, Pixmeo SARL, Bernex, Switzerland). Case examples were present in Supplementary Figure S1–S6, Supplemental digital content 1, https://links.lww.com/MCA/A484.
Statistical analyses
Descriptive statistics are presented as mean ± SD, median (interquartile range) or number (percentage) as appropriate. Data were presented on a per-patient basis for clinical characteristics and per-vessel basis for the remaining analyses. We compared target lesion and procedural characteristics and angiographic parameters before and after the index procedure and at the angiographic follow-up between groups with the use of unpaired Student’s t-test, Mann–Whitney U tests or chi-square tests as appropriate. The lesion CAC score was correlated with in-stent, proximal-edge and distal-edge diameter stenosis and lumen late loss. CAC score in lesions with ISR was compared with that in the lesion with no ISR with the use of Mann–Whitney U tests. Considering that (1) CCTA with 2.5-mm detector row width (CCTA/2.5 mm) and NCCT with 5-mm detector row width (NCCT/5 mm) account for the highest percentage of included lesions and patients as shown in Table 1 (indicating that these two types of CT scan may be the mostly used in clinical practice) and (2) we recognized that the detector row width of CT scan may affect the cutoff value, we chose to perform subgroup analyses in lesions receiving CCTA/2.5 mm (168 lesions in 90 patients) and NCCT/5 mm scans (138 lesions in 71 patients) and generated specific cutoff values for each type of the CT scan. We set the lesions with CAC score = 0 as the reference group, and divided the lesions with CAC score > 0 into quartiles. We compared the in-stent diameter stenosis and late lumen loss among groups with different CAC scores with the use of one-way analysis of variance followed by Dunnett’s tests using the group with CAC score = 0 as the reference group. Divided by the cutoff point of the CAC score, target lesion and procedural characteristics and angiographic parameters before and after the index procedure and at the angiographic follow-up were compared as described above in the CCTA/2.5 mm group or the NCCT/5 mm group. A receiver-operating characteristic (ROC) curve analysis was used to determine the ability of the lesion CAC score to distinguish between lesions with and without ISR at angiographic follow-up and identify the optimal cutoff point of the lesion CAC score that provided the greatest sum of sensitivity and specificity in the CCTA/2.5 mm group or the NCCT/5 mm group, respectively. Univariable and multivariable logistic regression models were performed to determine the association between CAC score and ISR at angiographic follow-up. Age greater than 65 years, male sex, hyperlipidemia, smoking, stent diameter and stent length were included in multivariable analyses.
Table 1. -
Baseline clinical characteristics of all patients included in analyses (
n = 194).
| Characteristics |
|
| Age, years |
67.2 ± 8.5 |
| Male sex |
136 (70.1) |
| Diagnosis |
| Stable angina or silent ischemia |
104 (53.6) |
| Unstable angina |
53 (27.3) |
| Non-ST-segment elevation myocardial infarction |
37 (19.1) |
| Hypertension |
142 (73.2) |
| Diabetes mellitus |
77 (39.7) |
| Hyperlipidaemia |
138 (71.1) |
| Smoking |
111 (57.2) |
| Multivessel disease |
| 1 vessel |
61 (31.4) |
| 2 vessels |
70 (35.6) |
| 3 vessels |
63 (33.0) |
| Type of CT/detector row width |
| CCTA/≤1.5 mm |
9 (4.6) |
| CCTA/2.5 mm |
90 (46.4) |
| NCCT/≤3 mm |
24 (12.4) |
| NCCT/5 mm |
71 (36.6) |
Values are mean ± SD or n (%).
CCTA, coronary computed tomography angiography; CT, computed tomography; NCCT, noncontrast chest computed tomography.
All analyses were conducted with the STATA software version 14.0 (Stata Corp, College Station, Texas, USA) and GraphPad Prism version 5.0.1 (GraphPad Software, San Diego, California, USA). A two-sided P value of <0.05 was considered statistically significant.
Results
Baseline clinical, lesion and procedural characteristics
We screened 5158 patients receiving DES implantation. Finally, 368 lesions in 194 patients which met the inclusion criteria and had none of the exclusion criteria were included in the analyses (Fig. 1). The demographic and clinical characteristics of the patients are summarized in Table 1. The mean age of patients was 67.2 ± 8.5 years, and 70.1% were male. Ninety-five patients received the NCCT scan. Of these, 71 had NCCT with 5-mm detector row width and 24 with ≤3-mm detector row width. Ninety-nine patients received the CCTA scan. Of these, 90 had CCTA with 2.5-mm detector row width and 9 with ≤1.5-mm detector row width. The demographic and clinical characteristics of the patients in the CCTA/2.5 mm group and NCCT/5 mm group are summarized in Supplementary Table S1, Supplemental digital content 1, https://links.lww.com/MCA/A484. The percentage of patients with the acute coronary syndrome (i.e. unstable angina or non-ST-segment elevation myocardial infarction) in the NCCT/5 mm group was higher than that in the CCTA/2.5 mm group. Other baseline characteristics were comparable. The baseline lesion and procedural characteristics are presented in Table 2. Half (49.5%) of target lesions were located at the left anterior descending artery, 20.6% at the left circumflex artery and 29.9% at the right coronary artery. The median CAC score was 32, and the minimal and maximal scores were 0 and 1085, respectively. Two patients received advanced techniques for severe CAC. One with a CAC score of 307 received cutting balloon inflation, and the other with a CAC score of 272 received rotational atherectomy. Most lesions (78.3%) were treated by implanting a sirolimus-eluting stent. The mean stent diameter and length were 3.0 ± 0.4 mm and 27.5 ± 7.3 mm, respectively. Table 2 presents the angiographic characteristics of target lesions at index procedure and follow-up. The mean duration between the index procedure and angiographic follow-up was 12.8 ± 5.4 months. Preprocedural diameter stenosis of the target lesion was (72.8 ± 14.5)%, and in-stent diameter stenosis was (9.9 ± 4.2)% after the procedure and (22.8 ± 17.4)% at angiographic follow-up. Thirty-two (8.7%) lesions received TLR at the angiographic follow-up. The procedural and angiographic characteristics of the CCTA/2.5-mm and NCCT/5-mm groups were present in Supplementary Table S2–S5, Supplemental digital content 1, https://links.lww.com/MCA/A484.
Table 2. -
Baseline lesion, procedural, angiographic characteristics of target lesions at index procedure and follow-up, stratified by coronary artery calcium score
|
Overall (n = 368) |
CAC score = 0 (n = 94) |
CAC score > 0 (n = 274) |
P value |
| Baseline lesion and procedural characteristics |
| Location of target lesions |
|
|
|
0.010 |
| LAD |
182 (49.5) |
34 (36.2) |
148 (54.0) |
|
| LCx |
76 (20.6) |
23 (24.5) |
53 (19.3) |
|
| RCA |
110 (29.9) |
37 (39.4) |
73 (26.6) |
|
| CAC score |
32 (0, 153.8) |
0 |
67 (22, 192.5) |
<0.001 |
| ACC/AHA classification |
|
|
|
0.001 |
| A |
29 (7.9) |
10 (10.6) |
19 (6.9) |
|
| B |
120 (32.6) |
43 (45.8) |
77 (28.1) |
|
| C |
219 (59.5) |
41 (43.6) |
178 (65.0) |
|
| TIMI flow at baseline |
|
|
|
0.41 |
| 0 |
25 (6.8) |
3 (3.2) |
22 (8.0) |
|
| 1 |
27 (7.3) |
6 (6.4) |
21 (7.7) |
|
| 2 |
113 (30.7) |
30 (31.9) |
83 (30.3) |
|
| 3 |
203 (55.2) |
55 (58.5) |
148 (54.0) |
|
| Stent type |
|
|
|
0.24 |
| Sirolimus-eluting stent |
288 (78.3) |
69 (73.4) |
219 (79.9) |
|
| Zotarolimus-eluting stent |
48 (13.0) |
13 (13.8) |
35 (12.8) |
|
| Everolimus-eluting stent |
32 (8.7) |
12 (12.8) |
20 (7.3) |
|
| Stent diameter, mm |
3.0 ± 0.4 |
2.9 ± 0.4 |
3.0 ± 0.4 |
0.26 |
| Stent length, mm |
27.5 ± 7.3 |
27.3 ± 7.5 |
27.6 ± 7.2 |
0.75 |
| Maximal balloon pressure, atm |
18.5 ± 3.8 |
16.9 ± 3.1 |
19.1 ± 3.9 |
<0.0001 |
| Angiographic characteristics of target lesions at index procedure and follow-up |
| Before procedure |
| Minimal lumen diameter, mm |
0.8 ± 0.4 |
0.8 ± 0.4 |
0.8 ± 0.5 |
0.22 |
| Reference vessel diameter, mm |
2.9 ± 0.4 |
2.9 ± 0.4 |
2.9 ± 0.4 |
0.78 |
| Diameter stenosis, % |
72.8 ± 14.5 |
70.8 ± 13.2 |
73.4 ± 14.9 |
0.14 |
| Length of target lesions, mm |
22.4 ± 7.2 |
21.5 ± 7.2 |
22.7 ± 7.2 |
0.16 |
| After procedure |
| Minimal lumen diameter, mm |
| Proximal |
3.0 ± 0.5 |
3.0 ± 0.5 |
3.1 ± 0.6 |
0.27 |
| In-stent |
2.8 ± 0.4 |
2.8 ± 0.4 |
2.8 ± 0.4 |
0.20 |
| Distal |
2.5 ± 0.5 |
2.5 ± 0.5 |
2.5 ± 0.5 |
0.58 |
| Residual diameter stenosis, % |
| Proximal |
7.1 ± 7.4 |
7.2 ± 6.7 |
7.1 ± 7.7 |
0.91 |
| In-stent |
9.9 ± 4.2 |
7.4 ± 2.4 |
10.8 ± 4.4 |
<0.001 |
| Distal |
8.7 ± 8.0 |
9.1 ± 7.9 |
8.6 ± 8.0 |
0.56 |
| Follow-up |
| Minimal lumen diameter, mm |
| Proximal |
2.9 ± 1.1 |
3.1 ± 1.8 |
2.8 ± 0.7 |
0.012 |
| In-stent |
2.2 ± 0.6 |
2.6 ± 0.5 |
2.1 ± 0.6 |
<0.001 |
| Distal |
2.3 ± 0.6 |
2.4 ± 0.5 |
2.3 ± 0.6 |
0.038 |
| Diameter stenosis, % |
| Proximal |
10.5 ± 15.5 |
9.2 ± 15.5 |
11.0 ± 15.5 |
0.33 |
| In-stent |
22.8 ± 17.4 |
12.7 ± 11.3 |
26.3 ± 17.8 |
<0.001 |
| Distal |
12.0 ± 13.1 |
9.7 ± 10.2 |
12.8 ± 13.8 |
0.052 |
| Late lumen loss, mm |
| Proximal |
0.2 ± 1.0 |
−0.1 ± 1.8 |
0.3 ± 0.5 |
0.001 |
| In-stent |
0.5 ± 0.5 |
0.2 ± 0.4 |
0.6 ± 0.5 |
<0.001 |
| Distal |
0.2 ± 0.5 |
0.1 ± 0.4 |
0.3 ± 0.5 |
0.060 |
| Target lesion revascularization |
32 (8.7) |
4 (4.3) |
28 (10.2) |
0.077 |
Values are mean ± SD, median (interquartile range), or n (%). CAC score was calculated using the Agatston score. P values were calculated for the comparison between groups with CAC score = 0 and CAC score >0 with the use of unpaired Student t-tests, Mann–Whitney U tests, or chi-square tests as appropriate.
ACC/AHA, American College of Cardiology/American Heart Association; CAC, coronary artery calcium; LAD, left anterior descending artery; LCx, left circumflex artery; RCA, right coronary artery; TIMI, Thrombolysis In Myocardial Infarction.
;)
Fig. 1.: Study flow diagram. CABG, coronary artery bypass grafting; CAC, coronary artery calcium; CCTA, coronary computed tomographic angiography; DES, drug-eluting stent; NCCT, noncontrast chest computed tomography; PCI, percutaneous coronary intervention; QCA, quantitative coronary angiography; STEMI, ST-segment elevation myocardial infarction.
Coronary artery calcium score, diameter stenosis, late lumen loss and in-stent restenosis
In Fig. 2, the lesion CAC score was significantly correlated with in-stent diameter stenosis (Spearman r = 0.7357; P < 0.0001) but not proximal-edge (Spearman r = 0.0724; P = 0.17) or distal-edge diameter stenosis (Spearman r = 0.0762; P = 0.14) assessed at angiographic follow-up. The lesion CAC score was significantly correlated with in-stent (Spearman r = 0.7306; P < 0.0001), proximal-edge (Spearman r = 0.1931; P = 0.0002) and distal-edge late lumen loss (Spearman r = 0.2136; P < 0.0001). Among the 368 lesions, 32 (8.7%) had ISR. Lesions with ISR had significantly higher CAC scores than those with no ISR [189 (28, 512) vs. 29.5 (0, 131.5); P = 0.0002]. No ISRs were observed in patients who received advanced techniques.
Fig. 2: Correlations between the coronary artery calcium (CAC) score and in-stent diameter stenosis (a), proximal-edge diameter stenosis (b), distal-edge diameter stenosis (c), in-stent late lumen loss (d), proximal-edge late lumen loss (e), or distal-edge late lumen loss (f) assessed at the angiographic follow-up in all lesions.
Coronary artery calcium score, in-stent diameter stenosis and in-stent restenosis in lesions receiving the coronary computed tomography angiography/2.5 mm scan
In the CCTA/2.5 mm group, the lesion CAC score was significantly correlated with in-stent diameter stenosis (Spearman r = 0.7702; P < 0.0001; Fig. 3). In-stent diameter stenosis and late lumen loss were significantly higher in lesions with CAC scores of 26–87, 88–220 or >220 than those with a CAC score of 0 (all P < 0.05; Table 3; Fig. 4). Among the 168 lesions in the CCTA/2.5 mm group, 15 (8.9%) had ISR. Lesions with ISR had significantly higher CAC scores than those with no ISR [251 (43, 620) vs. 47 (5, 176.5); P = 0.0019).
Table 3. -
In-stent diameter stenosis and late lumen loss assessed at the angiographic follow-up in lesions with a coronary artery calcium score of 0, 1–25, 26–87, 88–220 and >220 in the ECG-gated coronary computed tomography angiography with 2.5-mm detector row width group and 0, 1–22, 23–67, 68–192, and >192 in the nongated noncontrast chest computed tomography with 5-mm detector row width group
|
CAC score |
0 (n = 25) |
1–25 (n = 36) |
26–87 (n = 36) |
88–220 (n = 36) |
>220 (n = 35) |
P value |
| CCTA/2.5 mm |
Diameter stenosis |
10.9 ± 5.8 |
16.8 ± 17.3 |
23.1 ± 16.7a |
28.4 ± 6.5b |
42.1 ± 17.3b |
<0.001 |
| Late lumen loss |
0.2 ± 0.1 |
0.4 ± 0.4 |
0.6 ± 0.4b |
0.7 ± 0.3b |
1.0 ± 0.5b |
<0.001 |
|
CAC score |
0 (n = 51) |
1–22 (n = 21) |
23–67 (n = 22) |
68–192 (n = 22) |
>192 (n = 22) |
P value |
| NCCT/5 mm |
Diameter stenosis |
13.3 ± 14.4 |
12.2 ± 6.3 |
23.4 ± 23.1c |
26.5 ± 9.2a |
44.6 ± 17.2b |
<0.001 |
| Late lumen loss |
0.3 ± 0.5 |
0.2 ± 0.2 |
0.6 ± 0.6c |
0.7 ± 0.2a |
1.1 ± 0.5b |
<0.001 |
Values are mean ± SD or n (%). P values were calculated using one-way ANOVA followed by Dunnett’s tests using the group with CAC score = 0 as the reference group.
ANOVA, analysis of variance; CAC, coronary artery calcium; CCTA, coronary computed tomographic angiography; NCCT, noncontrast chest computed tomography.
aP<0.01.
bP<0.001.
cP<0.05.
Fig. 3.: Correlations between in-stent diameter stenosis and the coronary artery calcium (CAC) score assessed by ECG-gated coronary computed tomography angiography with 2.5-mm detector row width (CCTA/2.5 mm) and nongated noncontrast chest computed tomography with 5-mm detector row width (NCCT/5 mm).
Fig. 4.: In-stent diameter stenosis and late lumen loss assessed at the angiographic follow-up in the ECG-gated coronary computed tomography angiography with 2.5-mm detector row width (CCTA/2.5 mm) group (a and c) and nongated noncontrast chest computed tomography with 5-mm detector row width (NCCT/5 mm) group (b and d), stratified by the coronary artery calcium (CAC) score.
Coronary artery calcium score, in-stent diameter stenosis and in-stent restenosis in lesions receiving the noncontrast chest computed tomography/5 mm scan
In the NCCT/5 mm group, the lesion CAC score was significantly correlated with in-stent diameter stenosis (Spearman r = 0.7105; P < 0.0001; Fig. 3). In-stent diameter stenosis and late lumen loss were significantly higher in lesions with CAC scores of 23–67, 68–192, or >192 than those with a CAC score of 0 (all P < 0.05; Table 3; Fig. 4). Among the 138 lesions in the NCCT/5 mm group, 15 (10.9%) had ISR. Lesions with ISR had significantly higher CAC scores than those with no ISR [130 (28, 325) vs. 10 (0, 96); P = 0.0082).
Receiver-operating characteristic analyses
In the CCTA/2.5 mm group, the ROC curve analysis demonstrated that a lesion CAC score distinguishes between ISR and non-ISR [area under the curve, 0.744; 95% confidence interval (CI), 0.602–0.886; P = 0.002; Fig. 5a]. The lesion CAC score greater than 245 was identified as the optimal cutoff value providing the greatest sum of sensitivity (60.0%; 95% CI, 32.4–83.7%) and specificity (83.7%; 95% CI, 76.8–89.1%). Lesions with a CAC score greater than 245 had higher residual diameter stenosis after the procedure and in-stent minimal lumen diameter, in-stent diameter stenosis and in-stent late lumen loss at angiographic follow-up as compared with those with CAC score ≤245. In the 34 lesions with a CAC score greater than 245, 9 (26.5%) had ISR, and in the 134 lesions with a CAC score ≤245, 6 (4.5%) had ISR. The risk of ISR was significantly higher in lesions with a CAC score >245 than in those with a CAC score ≤245 (P < 0.001). More lesions received TLR at angiographic follow-up in the group with a CAC score >245 than those in the group with a CAC score ≤245 (26.5 vs. 4.5%; P < 0.001). Other demographic, clinical and procedural parameters were comparable between lesions with a CAC score ≤245 and >245 (Supplementary Table S1, and S2, Supplemental digital content 1, https://links.lww.com/MCA/A484). In the univariable and multivariable logistic regression analyses, the CAC score was significantly associated with ISR (all P < 0.01; Table 4). A CAC score >245 in lesions in the CCTA/2.5 mm group was identified to be independently associated with a higher risk in ISR at the angiographic follow-up. Lesions with a CAC score >245 had an 8.46-fold increased risk in ISR as compared with those with a CAC score ≤245 after adjustment [26.5 vs. 4.5%; adjusted odds ratio (OR), 8.46; 95% CI, 2.23–32.13; P = 0.002).
Table 4. -
Univariable and multivariable logistic regression analyses for in-stent restenosis assessed at the angiographic follow-up in the ECG-gated coronary computed tomography angiography with 2.5-mm detector row width group and nongated noncontrast chest computed tomography with 5-mm detector row width group.
|
Univariable model |
Multivariable model |
|
OR (95% CI) |
P value |
OR (95% CI) |
P value |
| CCTA/2.5 mm group |
| CAC score >245 |
7.68 (2.51, 23.50) |
<0.001 |
8.46 (2.23, 32.13) |
0.002 |
| Age >65 years |
0.89 (0.31, 2.58) |
0.83 |
1.21 (0.35, 4.12) |
0.77 |
| Male sex |
1.79 (0.38, 8.32) |
0.46 |
1.41 (0.20, 9.91) |
0.73 |
| Hyperlipidaemia |
5.47 (0.70, 42.90) |
0.11 |
5.95 (0.70, 50.78) |
0.10 |
| Smoking |
1.01 (0.32, 3.16) |
0.99 |
0.61 (0.14, 2.73) |
0.52 |
| Stent diameter |
1.04 (0.29, 3.68) |
0.95 |
0.67 (0.12, 3.71) |
0.65 |
| Stent length |
1.04 (0.96, 1.12) |
0.36 |
1.04 (0.95, 1.13) |
0.44 |
| NCCT/5 mm group |
| CAC score >209 |
9.89 (2.97, 32.92) |
<0.001 |
21.89 (4.19, 114.35) |
<0.001 |
| Age >65 years |
1.38 (0.46, 4.12) |
0.56 |
0.63 (0.14, 2.74) |
0.54 |
| Male sex |
1.23 (0.37, 4.11) |
0.74 |
3.63 (0.52, 25.46) |
0.20 |
| Hyperlipidaemia |
5.13 (0.65, 40.58) |
0.12 |
9.74 (0.99, 95.99) |
0.051 |
| Smoking |
0.78 (0.27, 2.30) |
0.66 |
0.31 (0.05, 1.72) |
0.18 |
| Stent diameter |
0.60 (0.14, 2.52) |
0.49 |
0.32 (0.05, 2.01) |
0.23 |
| Stent length |
1.08 (0.99, 1.18) |
0.095 |
1.07 (0.97, 1.19) |
0.19 |
CAC, coronary artery calcium; CCTA, coronary computed tomographic angiography; CI, confidence interval; NCCT, non-contrast chest computed tomography; OR, odds ratio.
Fig. 5.: Receiver-operating characteristic curve for the coronary artery calcium (CAC) score assessed by ECG-gated coronary computed tomography angiography with 2.5-mm detector row width (CCTA/2.5 mm) (a) or nongated noncontrast chest computed tomography with 5-mm detector row width (NCCT/5 mm) (b) to distinguish between lesions with and without in-stent restenosis. The optimal cutoff value for in-stent restenosis is CAC score > 245 (sensitivity: 60.0%; specificity: 83.7%; area under the curve: 0.744; 95% CI, 0.602–0.886; P = 0.002) for ECG-gated CCTA/2.5 mm and > 209 for NCCT/5 mm (sensitivity: 46.7%; specificity: 91.9%; area under the curve: 0.704; 95% CI, 0.549–0.860; P = 0.010).
In the NCCT/5 mm group, the ROC curve analysis demonstrated that the lesion CAC score distinguishes between ISR and non-ISR (area under the curve, 0.704; 95% CI, 0.549–0.860; P = 0.010; Fig. 5b). The lesion CAC score >209 was identified as the optimal cutoff value providing the greatest sum of sensitivity (46.7%; 95% CI, 21.3–73.4%) and specificity (91.9%; 95% CI, 85.6–96.0%). Lesions with a CAC score >209 had higher residual diameter stenosis after the procedure and in-stent minimal lumen diameter, in-stent stenosis and in-stent late lumen loss at angiographic follow-up in comparison with those with a CAC score ≤209. In the 17 lesions with a CAC score >209, 7 (41.2%) had ISR, and in the 121 lesions with a CAC score ≤209, 8 (6.6%) had ISR. The risk of ISR was significantly higher in lesions with a CAC score >209 than in those with a CAC score ≤209 (P < 0.001). More lesions received TLR at angiographic follow-up in the group with a CAC score >209 than those in the group with a CAC score ≤209 (41.2 vs. 6.6%; P < 0.001). Other demographic, clinical and procedural parameters were comparable between lesions with CAC score ≤209 and >209 (Supplementary Table S3 and S4, Supplemental digital content 1, https://links.lww.com/MCA/A484). In the univariable and multivariable logistic regression analyses, the CAC score was significantly associated with ISR (all P < 0.05; Table 4). CAC score >209 in lesions in the NCCT/5 mm group was identified to be independently associated with a higher risk in ISR at the angiographic follow-up. Lesions with a CAC score >209 had a 21.89-fold increased risk in ISR in comparison with those with CAC score ≤209 after adjustment (41.2 vs. 6.6%; adjusted OR, 21.89; 95% CI, 4.19–114.35; P < 0.001).
Discussion
In this study, we demonstrated that the CAC score was significantly associated with increased risk in ISR, irrespectively of clinical or procedural characteristics. In the CCTA/2.5 mm group, a lesion CAC score >245 was significantly associated with an 8.46-fold increase in ISR. Similarly, in the NCCT/5 mm group, a lesion CAC score >209 was significantly associated with a 21.89-fold increase in ISR.
ISR was recognized as a clinically important vessel-oriented complication, which may lead to recurrent angina or even myocardial infarction [2,3]. Thanks to the evolution of the coronary stent (e.g. from bare-metal stent to DES, and from the first-generation DES to second-generation DES) and improvement of the stent deployment technique (to achieve no residual narrowing, no presence of dissection and complete stent expansion and apposition), the burden of ISR had been reduced dramatically [1]. However, approximately 5–10% of patients receiving stent implantation may still experience ISR [14–16]. Intravascular imaging studies have demonstrated that stent under-expansion was observed in about half of ISR lesions [17]. It is thus essential for interventionists to identify lesions with a higher risk in stent under-expansion and then to optimize interventional strategy for eliminating the occurrence of ISR. However, a clinically useful tool for preoperative assessment is lacking.
CAC has been well recognized as an important contributor of stent underexpansion. Previous studies have shown a higher risk in ISR in calcified lesions. In a subgroup analysis of the Cypher Post-Marketing Surveillance Registry study, a significantly higher rate of restenosis was observed in calcified lesions than that in noncalcified lesions (39.5 vs. 17.0%; P = 0.029) [18]. Angiographically visual evaluation of CAC is commonly used in routine clinical practice. But this assessment is qualitative and subjective, limiting its standardization and extrapolation. Intracoronary imaging modalities, including IVUS and OCT, could detect the CAC and provide detailed information, but the devices may be hard to cross some lesions with severe stenosis or calcium. As compared with angiographically visual evaluation and intracoronary imaging modalities, a CT scan is a quantitative, objective and noninvasive approach to detect CAC. Also, it can provide additional information for optimizing pretreatment strategy for target lesions even before interventional procedure. However, the quantitative analyses on the association between the magnitude of CT-detected CAC and ISR risk were less investigated. In a study of 69 lesions, Tanabe et al. [10] found higher preprocedural Agatston calcium scores in lesions with ISR than those without [(629 ± 718) vs. (153 ± 245); P = 0.08]. In this study, we found that a CAC score >245 in the CCTA/2.5 mm group or a CAC score >209 in the NCCT/5 mm group was significantly associated with increased risk in ISR, irrespectively of clinical or procedural characteristics. Our study provided a clinically feasible tool of preprocedural assessment of CAC for predicting long-term ISR. Its utility in identifying lesions with high ISR risk is warranted for further validation.
Compared with CCTA, NCCT is more widely used in routine clinical practice, with a simpler procedure and lower cost. The usefulness of NCCT to evaluate the severity of CAC has been explored in previous studies. In a meta-analysis of 661 patients, the correlation coefficient for the agreement of CAC scoring between nongated and ECG-gated CT examinations was 0.94 (95% CI, 0.89–0.97) [19]. However, its association between CAC severity and ISR has not been compared with CCTA. In our study, we found that a CAC score >245 in the CCTA/2.5 group and a CAC score >209 in the NCCT/5 mm group were both independently associated with a higher risk in ISR, which suggests that NCCT scan has the potential for clinical utility in CAC assessment for ISR prediction in comparison with CCTA scan. Further, it is interesting to observe that in this study, the median value of the CAC score in the NCCT/5 mm group was lower than that in the CCTA/2.5 mm group (17 vs. 56.5; P < 0.001). Similarly, in a previous study, the false-negative NCCT for CAC was 8.8% when noted on the ECG-gated scans and 19.1% of high CAC scores were underestimated [19]. It might be explained by the lower radiation dose and wider detector slice in NCCT/5 mm as compared with CCTA/2.5 mm. Thus, a single cutoff value of the CAC score might be not useful for different types of CT scans and different slice thickness. In this study, we determined the individual cutoff value for each of the two CT scan types separately. In both types of CT scans, lesions with a CAC score higher than the cutoff value had significantly increased risk in ISR, and more importantly, more events on TLRs. This suggests that although the absolute values of the CAC score are different, CAC scoring by the NCCT/5 mm scan might have a similar prognostic value as compared with CCTA, and thus, might be as clinically useful as CCTA in predicting ISR. The utilities of CAC assessment by the NCCT/5 mm scan and CCTA/2.5 mm scan are required for further validation.
Previous studies have described multiple contributors of ISR, including patient characteristics (e.g. age, female sex, diabetes, multivessel coronary artery disease and genetic variation), lesion characteristics (e.g. prior ISR, bypass graft, chronic total occlusion, small vessel lesion, calcified lesion and ostial lesion) and procedural characteristics (e.g. type of DES and postprocedural minimal lumen diameter) [3,20]. Among these contributors, most are unlikely to be modified. In addition, although the selection of a newer generation of DES and the optimization of postprocedural minimal lumen diameter to obtain optimal acute angiographic results could improve long-term outcomes, these approaches are required for all target lesions. Unlike other contributors, CAC could be identified and quantified before the procedure and be theoretically ‘modifiable’ for better stent deployment and expansion during the procedure with the use of more aggressive interventional approaches, such as cutting ballooning [21], rotational atherectomy [21,22], orbital atherectomy [23], excimer laser coronary atherectomy [24] and coronary intravascular lithotripsy [25]. These approaches could be planned even before conducting the invasive procedure. Our study provided a clinically useful tool of preprocedural assessment on CAC severity for tailoring interventional strategy.
There are some limitations in this study. First, as a retrospective study, there is selection bias. For example, there is a selected group of patients who underwent coronary CT. Second, the sample size of this study is limited. However, the findings were statistically significant and consistent among different subgroup analyses. Larger-scale studies are warranted for further confirmation.
In conclusion, our study found that either a CAC score >245 in CCTA/2.5 mm or a CAC score >209 in NCCT/5 mm was independently associated with a higher risk in ISR. Further investigations are warranted to prospectively validate these findings and test the clinical usefulness of the CAC score assessed by CCTA or NCCT in optimizing interventional strategy for severely calcified lesions.
Acknowledgements
The authors thank Dr. Weituo Zhang from the School of Medicine, Shanghai Jiao Tong University, Shanghai, China for the statistical consultation, Dr. Yifeng Jiang from the Department of Radiology, Shanghai Chest Hospital, Shanghai Jiao Tong University, Shanghai, China, for technical assistance with CT images and Mr. Hao Zhang from Catheter Laboratory, Department of Cardiology, Shanghai Chest Hospital, Shanghai Jiao Tong University, Shanghai, China, for technical assistance with angiographic images.
This work was supported by the National Natural Science Foundation of China (81830010, 81770428); Shanghai Science and Technology Committee (18411950400); Emerging and Advanced Technology Programs of Hospital Development Center of Shanghai (SHDC12018129); and Shanghai Sailing Program (20YF1444200, 19YF1444600).
Conflicts of interest
There are no conflicts of interest.
References
1. Torrado J, Buckley L, Durán A, Trujillo P, Toldo S, Valle Raleigh J, et al. Restenosis, stent thrombosis, and bleeding complications: navigating between scylla and charybdis. J Am Coll Cardiol 2018; 71:1676–1695.
2. Moussa ID, Mohananey D, Saucedo J, Stone GW, Yeh RW, Kennedy KF, et al. Trends and outcomes of restenosis after coronary stent implantation in the United States. J Am Coll Cardiol 2020; 76:1521–1531.
3. Dangas GD, Claessen BE, Caixeta A, Sanidas EA, Mintz GS, Mehran R. In-stent restenosis in the drug-eluting stent era. J Am Coll Cardiol 2010; 56:1897–1907.
4. Henneke KH, Regar E, König A, Werner F, Klauss V, Metz J, et al. Impact of target lesion calcification on coronary stent expansion after rotational atherectomy. Am Heart J 1999; 137:93–99.
5. Fujino A, Mintz GS, Matsumura M, Lee T, Kim SY, Hoshino M, et al. A new optical coherence tomography-based calcium scoring system to predict stent underexpansion. EuroIntervention 2018; 13:e2182–e2189.
6. Wang X, Matsumura M, Mintz GS, Lee T, Zhang W, Cao Y, et al.
In vivo calcium detection by comparing optical coherence tomography, intravascular ultrasound, and angiography. JACC Cardiovasc Imaging 2017; 10:869–879.
7. Maejima N, Hibi K, Saka K, Akiyama E, Konishi M, Endo M, et al. Relationship between thickness of calcium on optical coherence tomography and crack formation after balloon dilatation in calcified plaque requiring rotational atherectomy. Circ J 2016; 80:1413–1419.
8. Kubo T, Shimamura K, Ino Y, Yamaguchi T, Matsuo Y, Shiono Y, et al. Superficial calcium fracture after PCI as assessed by OCT. JACC Cardiovasc Imaging 2015; 8:1228–1229.
9. Agatston AS, Janowitz WR, Hildner FJ, Zusmer NR, Viamonte M Jr, Detrano R. Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol 1990; 15:827–832.
10. Tanabe K, Kishi S, Aoki J, Tanimoto S, Onuma Y, Yachi S, et al. Impact of coronary calcium on outcome following sirolimus-eluting stent implantation. Am J Cardiol 2011; 108:514–517.
11. Sekimoto T, Akutsu Y, Hamazaki Y, Sakai K, Kosaki R, Yokota H, et al. Regional calcified plaque score evaluated by multidetector computed tomography for predicting the addition of rotational atherectomy during percutaneous coronary intervention. J Cardiovasc Comput Tomogr 2016; 10:221–228.
12. Shin JM, Kim TH, Kim JY, Park CH. Coronary artery calcium scoring on non-gated, non-contrast chest computed tomography (CT) using wide-detector, high-pitch and fast gantry rotation: comparison with dedicated calcium scoring CT. J Thorac Dis 2020; 12:5783–5793.
13. Mehran R, Dangas G, Abizaid AS, Mintz GS, Lansky AJ, Satler LF, et al. Angiographic patterns of in-stent restenosis: classification and implications for long-term outcome. Circulation 1999; 100:1872–1878.
14. Windecker S, Serruys PW, Wandel S, Buszman P, Trznadel S, Linke A, et al. Biolimus-eluting stent with biodegradable polymer versus sirolimus-eluting stent with durable polymer for coronary revascularisation (LEADERS): a randomised non-inferiority trial. Lancet 2008; 372:1163–1173.
15. Stone GW, Ellis SG, Cox DA, Hermiller J, O’Shaughnessy C, Mann JT, et al.; TAXUS-IV Investigators. A polymer-based, paclitaxel-eluting stent in patients with coronary artery disease. N Engl J Med 2004; 350:221–231.
16. Morice MC, Colombo A, Meier B, Serruys P, Tamburino C, Guagliumi G, et al.; REALITY Trial Investigators. Sirolimus- vs paclitaxel-eluting stents in de novo coronary artery lesions: the REALITY trial: a randomized controlled trial. JAMA 2006; 295:895–904.
17. Jensen LO, Vikman S, Antonsen L, Kosonen P, Niemelä M, Christiansen EH, et al. Intravascular ultrasound assessment of minimum lumen area and intimal hyperplasia in in-stent restenosis after drug-eluting or bare-metal stent implantation. The Nordic Intravascular Ultrasound Study (NIVUS). Cardiovasc Revasc Med 2017; 18:577–582.
18. Fujimoto H, Nakamura M, Yokoi H. Impact of calcification on the long-term outcomes of sirolimus-eluting stent implantation: subanalysis of the Cypher Post-Marketing Surveillance Registry. Circ J 2012; 76:57–64.
19. Xie X, Zhao Y, de Bock GH, de Jong PA, Mali WP, Oudkerk M, Vliegenthart R. Validation and prognosis of coronary artery calcium scoring in nontriggered thoracic computed tomography: systematic review and meta-analysis. Circ Cardiovasc Imaging 2013; 6:514–521.
20. Jørgensen E, Kelbaek H, Helqvist S, Jensen GV, Saunamäki K, Kastrup J, et al. Predictors of coronary in-stent restenosis: importance of angiotensin-converting enzyme gene polymorphism and treatment with angiotensin-converting enzyme inhibitors. J Am Coll Cardiol 2001; 38:1434–1439.
21. Amemiya K, Yamamoto MH, Maehara A, Oyama Y, Igawa W, Ono M, et al. Effect of cutting balloon after rotational atherectomy in severely calcified coronary artery lesions as assessed by optical coherence tomography. Catheter Cardiovasc Interv 2019; 94:936–944.
22. Abdel-Wahab M, Toelg R, Byrne RA, Geist V, El-Mawardy M, Allali A, et al. High-speed rotational atherectomy versus modified balloons prior to drug-eluting stent implantation in severely calcified coronary lesions. Circ Cardiovasc Interv 2018; 11:e007415.
23. Redfors B, Sharma SK, Saito S, Kini AS, Lee AC, Moses JW, et al. Novel micro crown orbital atherectomy for severe lesion calcification: Coronary Orbital Atherectomy System Study (COAST). Circ Cardiovasc Interv 2020; 13:e008993.
24. Ojeda S, Azzalini L, Suárez de Lezo J, Johal GS, González R, Barman N, et al. Excimer laser coronary atherectomy for uncrossable coronary lesions. A multicenter registry. Catheter Cardiovasc Interv 2021; 98:1241–1249.
25. Hill JM, Kereiakes DJ, Shlofmitz RA, Klein AJ, Riley RF, Price MJ, et al.; Disrupt CAD III Investigators. Intravascular lithotripsy for treatment of severely calcified coronary artery disease. J Am Coll Cardiol 2020; 76:2635–2646.
