Intracoronary β-irradiation prevents excessive in-stent neointimal proliferation in de novo lesions of patients with high plasma ACE levels. The BetAce randomized trial

1. Introduction 

The recurrence of stenosis after percutaneous coronary interventions (PCI) with bare metal stent implantation still affects nearly 20% of patients [1]. The advent of drug eluting stents (DES) technology has significantly reduced restenosis below 10% [2], [3], [4], [5], but at least in Europe, new-generation bare metal stents are still the first-treatment choice in over 60% of patients undergoing PCI. In-stent restenosis (ISR) is essentially due to the exuberant proliferation of neointimal tissue that ensues vessel wall injury caused by the implantation of the metallic prosthesis [6]. This event occurs predominantly in diabetic patients and in subjects bearing an individual predisposition to increased vascular wound-healing response. In patients treated with DES, restenosis presents a different angiographic distribution compared with bare metal stents that suggests a biological all-or-none response of restenosis [7]. Vascular brachytherapy (VBT) was developed as a very effective treatment to prevent restenosis by means of its potent antiproliferative action [8]. The initial experience with the application of VBT after balloon dilatation in de novo lesions was very promising [9], [10], but with the combined use of elective stent implantation and VBT, the limitations of this first-intention antirestenotic approach emerged unexpectedly [9], [11], [12]. This combination aimed to reduce the restenosis rates by two different theoretical mechanisms: avoiding vessel shrinkage by stent implantation and preventing excessive neointimal growth within the stent by VBT. However, it was associated with a high incidence of stent thrombosis (and mortality) and the occurrence of late need for repeat revascularization due to a delayed restenotic process induced by low-radiation doses at sites of geographical miss [13].

Two recently published randomized trials have tested the combination of stent implantation and VBT in de novo lesions according to updated recommendations, aimed at reducing adverse events; however, stent thrombosis and late restenosis remained higher in the VBT+stent group [14], [15].

In the BetAce trial, patients suitable for entering a randomized study and bearing an augmented risk for ISR have been selected to test the effects of VBT. The deletion allele of the insertion (I), deletion (D) polymorphism of the angiotensin 1-converting enzyme (ACE) gene correlates with the plasma level of the enzyme [16], [17], and this has been suggested to increase the risk for ISR [17], [18], [19], [20].

2. Methods 

The criteria for inclusion in the study were age 40 years or older with symptoms of stable angina, silent ischemia, or unstable angina (class ≤II); stenosis ≤15 mm long, located in the native coronary artery, suitable for stenting; ACE genotype D/D or I/D, with plasma ACE level >34 UI/l; and feasibility of angiographic follow-up. The threshold value of plasma ACE concentration 34 UI/l was identified previously by maximizing the specificity for predicting the occurrence of ISR [17] and subsequently confirmed in a large series of cases [20]. The biochemical method applied in the BetAce study to measure plasma ACE activity is the same as used previously (quantitative kinetic determination at 340 nm, with the use of FAPGG substrate, by SIGMA Diagnostics). Such technique has been shown to offer reproducible results in different laboratories and in populations of different genetic background [21].

The exclusion criteria were recent myocardial infarction with abnormal CK/CK-MB values, malignant disease within previous 5 years, previous thoracic irradiation, significant left main disease, abrupt closure or severe dissection following balloon angioplasty, and contraindication to the use of aspirin or ticlopidine.

Patients were randomized to bare metal stent implantation alone or in association with VBT. The final angiographic result was assessed by quantitative coronary analysis (QCA) on-line and intravascular ultrasound. Additional in-stent dilatations were performed when needed in both groups. In patients assigned to VBT, β-irradiation was delivered after balloon dilatation and prior to coronary stenting using a sealed 90Yttrium (90Y) source. The Schneider–Sauerwein–Boston Scientific afterloader system and centering catheter were used in 64% of cases, and the Beta-Cath delivery system (Novoste), in 36% of cases, depending on source availability. The radiation dose as recommended for each device [9], [12] was prescribed based on the average diameter of the reference vessel segment as measured by QCA. The procedure was performed in collaboration between the cardiologist, the radiation oncologist, and the medical physicist according to local regulations. The standard radiation protection measures normally used in the interventional suite were applied.

QCA was performed according to described methods [22] at baseline, immediately after the stented dilatation, and at 6 months. Angiograms obtained in at least two orthogonal projections after intracoronary nitrates were forwarded to the InterCorNet Core Angiography Laboratory, which was blinded to the patient's genotype, plasma ACE level, and randomization arm. The analysis was extended to a long vessel segment (30 mm) in all patients as to include the stent edges and the irradiated segments when applicable.

The primary endpoint of the study was the minimal lumen diameter (MLD) at 6 months, assessed by QCA. In previous studies, QCA performed 6 months after stent implantation in patients carrying the DD genotype of the ACE gene polymorphism showed restenosis rates that ranged from 29% to 34% and MLD that ranged from 1.65±0.71 to 1.77±0.80 mm [17], [18], [19]. We assumed that, in this prospective trial, a similar MLD would be observed in the control group. Compared with the control group, a clinically meaningful 30% improvement in MLD was the expected treatment effect with the combined use of stents and VBT (expected MLD of 2.2 vs. 1.65 mm in the control group). To verify this hypothesis, 29 patients in each arm are necessary to obtain a β=.1 (90%). The sample size was increased by an additional 10% to account for possible technical failures and dropouts.

Secondary objectives were additional angiographic results (restenosis, mean in-stent diameter, in-stent lumen volume, loss in MLD, and loss in in-stent lumen volume between postprocedure and follow-up and positive remodeling defined as reported below) and the clinical outcome at 6 and 12 months.

Other definitions were as follows—restenosis: percent diameter stenosis (%DS) ≥50 at follow-up; mean in-stent diameter: QCA of the luminogram; in-stent lumen volume: volume of contrast within the stented lumen; loss in MLD: difference between MLD postprocedure and follow-up; loss in in-stent lumen volume: between postprocedure and follow-up QCA; positive remodeling: an increment in lumen (MLD) and vessel (D-ref) dimensions between postprocedure and follow-up; major adverse cardiac events: ranked as death, myocardial infarction (both Q and non-Q-wave, defined as CK-MB >3 times the upper normal limit), repeat procedures, albeit bypass surgery or target vessel revascularization (TVR) by percutaneous technique.

Statistical analysis was performed according to the intention to treat principle. The study was designed as a prospective, single-center, single-blinded, randomized (1:1) investigation. Treatment allocation to regular stent implantation or to VBT prior to stent implantation was obtained after written informed consent. Only patients who had been genetically screened prior to the planned procedure were considered. The trial was approved by the Ethics Committee of the OLV hospital. Continuous data are expressed as means and standard deviations; discrete variables are given as absolute values and percentages. Two-tailed Student's t test was used for comparison of parametric variables and the chi-square or exact test for discrete variables. Event-free survival analysis was performed using Kaplan–Meier curves (log-rank test). Calculations were made using the SPSS software (7.5 release for Windows).

3. Results 

Sixty-six patients were allocated either to regular stent implantation [32] or to stent implantation following VBT [34]. Five patients did not undergo angioplasty and were excluded (Fig. 1). According to the intention to treat analysis, 33 lesions in 31 patients were treated with regular stent implantation (control group), and 31 lesions in 30 patients were treated with adjunctive VBT. Angiographic follow-up is available in all but one stented patient who declined repeat catheterization and is not available in three patients in the VBT+stent group (one patient who died before 6 months and two asymptomatic patients who refused to undergo repeat catheterization). Clinical outcome data are available in all patients in both treatment arms.

View Large Image Download to PowerPoint

Fig. 1. Study flow chart. Out of 66 patients initially allocated, 5 did not undergo stent implantation because angiographic exclusion criteria became apparent during the intervention. One patient received balloon angioplasty only and stent implantation was not attempted. Among patients initially randomized to receive stent and VBT, four were deregistered. In two cases, surgical revascularization was preferred because of clinical instability and multivessel disease in one case and because of an ostial stenosis of the left anterior descending coronary artery in the other. One patient received balloon angioplasty without stent, and another presented with total coronary occlusion that could not be reopened.

Baseline clinical and laboratory data were similar in the two groups, except for a higher number of patients with previous myocardial infarction in the VBT+stent group (Table 1). Angiographic and procedural data were similar between groups (Table 2, Table 3). The medical treatment at hospital discharge was not different between the two groups. Twenty percent of patients in both groups were treated by ACE-inhibitor drugs. After VBT, patients were advised to continue aspirin (160 mg daily) during lifetime, and ticlopidine (at least 250 mg daily) for 7 months.

Table 1. Patients baseline characteristics

		Stent (31)	VBT+stent (30)	P value
	Clinical data
	Age	61.6±9	60.6±10	.7
	Female gender	7 (22.6%)	3 (10%)	.1
	Diabetes	3 (9.7%)	4 (13.3%)	.4
	Smokers	19 (61.3%)	17 (56.7%)	.5
	Family history of CAD	9 (29%)	13 (43.3%)	.2
	Hypertension	13 (41.9%)	14 (46.7%)	.5
	Hypercholesterolemia	11 (35.5%)	11 (36.7%)	.9
	Previous MI	1 (3.2%)	6 (20%)	.05
	Stable angina or silent ischemia	28 (90.3%)	28 (93.3%)	.5
	Laboratory data
	Glycemia (mg/dl)	109.9±30	113.4±34	.7
	Total cholesterol (mg/dl)	215±44	219±40	.7
	HDL cholesterol (mg/dl)	45.9±8	45.3±10	.9
	Fibrinogen (mg/dl)	383±90	376±74	.9
	ACE U/L	40.2±13	45.3±13	.1
	DD polymorphism	25 (78.1%)	20 (69%)	.4
	ID polymorphism	7 (21.9%)	9 (31%)	.4

ACE: angiotensin-converting enzyme; CAD: coronary artery disease; D: deletion; I: insertion; LVEF: left ventricular ejection fraction; MI: myocardial infarction; VBT: vascular brachytherapy.

Table 2. Angiographic baseline characteristics of 64 lesions

	Angiographic data	Stent (33)	VBT+stent (31)	P value
	Single-vessel disease	19 (61.3%)	18 (54.5%)	.3
	LVEF (%)	73.7±10	68.9±16	.1
	Target vessel
	LAD	16 (48.5%)	12 (38.7%)
	RCA	13 (39.4%)	14 (45.2%)	.7
	LCx	4 (12.1%)	5 (16.1%)
	Lesion type
	A+B1	26 (75.8%)	25 (80.6%)	.3
	B2+C	8 (24.2%)	6 (19.4%)
	Thrombus	0	0	1

LAD: left anterior descendent; LCx: left circumflex; LVEF: left ventricular ejection fraction; RCA: right coronary artery.

Table 3. Procedural data

	PCA data	Stent	VBT+stent	P value
	TIMI flow pre-PCA
	Grades 1–2	3 (9.1%)	1 (3.2%)	.3
	Grade 3	30 (90.9%)	30 (96.8%)
	TIMI flow post-PCA
	Grade 3	33 (100%)	31 (100%)	1
	Stent type
	Self-expanding	19 (57.6%)	13 (41.9%)	.2
	Balloon-expandable	14 (42.4%)	18 (58.1%)
	Nominal stent diameter	3.61±0.7	3.7±0.8	.7
	Mean stent length	19.3±7.8	16.6±4.3	.2
	Max balloon diameter	3.5±0.5	3.5±0.5	.9
	Max balloon pressure	12.4±3.5	10.9±3.4	.1
	Dwelling time (s)		220.8±100.7	–

D-ref: reference diameter; MLD: minimal lumen diameter; %DS: percent diameter stenosis; QCA: quantitative coronary analysis; TIMI: thrombolysis in myocardial infarction.

By QCA, there was no difference in stenosis severity both prior to and immediately following stented angioplasty. At 6 months, the MLD in the control group was 1.74±0.8 mm. Out of 32 lesions, 12 had a diameter stenosis ≥50%, 2 of which were total occlusions. Accordingly, the restenosis rate was 37.5%. In the VBT+stent group, the reference vessel diameter was larger (3.42±0.6 vs. 3.07±0.42 mm, P=.04) and the MLD was greater than in the stent group by 35% (2.25±1.05 mm, P=.04). The late loss in MLD was 0.70±.9 compared with 1.08±0.8 mm in controls (P=.1). Out of 28 stenoses, 5 had a diameter stenosis ≥50%, of which 2 were total occlusions and 1 was a stenosis at the distal edge of the stent. Accordingly, the restenosis rate was 17.9% (Table 4). Four edge stenosis were observed, one corresponded to a focal restenosis partially inside the stent and three edge effects were nonrestenotic (DS% <50), one placed distally and two at both edges of the stent. Comparing poststenting to follow-up measurements, similar reference vessel diameters (3.47±.6 vs. 3.42±.6, P=.2) and similar in-stent lumen volume values (124±62.9 vs. 101.2±57.6, P=.08) were observed in the VBT+stent group. In the stent-only group, a significant reduction of both the reference vessel diameter and the in-stent lumen volume was observed (respectively, 3.39±.7 vs. 3.07±.42, P=.0001, and 160.2±91.8 vs. 92.8±78.8, P=.0001). No coronary aneurysms or unhealed dissections were seen. The cumulative distribution frequency of the MLD (Fig. 2) shows that coronary lumen at 6 months was larger in the VBT+stent group over the entire range of vessel sizes. The cumulative distribution frequency of the loss in in-stent lumen volume (Fig. 3) illustrates the treatment effect, i.e., the antiproliferative effect of β-irradiation.

Table 4. Angiographic data (quantitative coronary analysis)

	QCA data	Stent (33)	VBT+stent (31)	P value
	QCA pre-PCI	33 lesions	31 lesions
	Lesion length (mm)	11.7±7	9.6±5.3	.3
	D-ref (mm)	3.3±0.7	3.3±0.6	.7
	MLD (mm)	1.06±0.50	1.17±0.51	.4
	%DS	68.4±15	66.2±15	.5
	QCA poststent	33 lesions	31 lesions
	D-ref (mm)	3.39±0.7	3.47±0.6	.6
	MLD (mm)	2.8±0.4	2.9±0.4	.1
	%DS	15.45±9.3	14.0±8.6	.5
	Mean in-stent diameter (mm)	3.2±0.4	3.4±0.4	.1
	In-stent luminal volume (ml)	160.2±91.8	124±62.9	.07
	QCA follow-up	32 lesions	28 lesions
	D-ref (mm)	3.07±0.42	3.42±0.6	.04
	MLD (mm)	1.74±0.8	2.25±1.05	.04
	%DS	42.7±23.6	34.8±27.8	.2
	Mean in-stent diameter (mm)	2.3±0.8	2.9±1.05	.02
	In-stent luminal volume (ml)	92.8±78.8	101.2±57.6	.6
	Acute gain in MLD (mm)	1.76±0.6	1.81±0.5	.7
	Late loss in MLD (mm)	1.08±0.8	0.70±0.9	.1
	Net gain in MLD (mm)	0.68±0.87	1.06±1.08	.1
	Loss mean stent diameter (mm)	0.92±0.8	0.51±0.9	.07
	Loss stent volume (ml)	69.12±67.95	24.19±70.67	.01
	Restenosis rate (DS ≥50)	12/32 (37.5%)	5/28 (17.9%)	.08

D-ref: reference diameter; MLD: minimal luminal diameter; %DS: percent diameter stenosis; QCA: quantitative coronary angiography.

View Large Image Download to PowerPoint

Fig. 2. Cumulative frequency distribution curves of the MLD (mm) at 6-month follow-up angiography. Time to follow-up was 199.5±118 days in the control group and 194±54.4 days in the VBT+stent group (NS). The MLD values are larger in the VBT+stent group, and this effect is maintained across the entire range of vessel size.

View Large Image Download to PowerPoint

Fig. 3. Cumulative frequency distribution curves of the loss in in-stent lumen volume (ml) at 6-month follow-up angiography. In-stent lumen volume loss was smaller in the VBT+stent group across the entire range of vessel size. Furthermore, 29% of the patients treated with VBT and stent show negative values (enlargement in in-stent lumen volume or positive “remodeling”), a phenomenon only observed in 9% of patients in the control group (P=.04).

The clinical outcome at 6 and at 12 months was not different between the two groups in terms of death and incidence of myocardial infarction. Within 6 months, one patient in the VBT+stent group died following acute infarction 4 months after stenting (ticlopidine treatment had been interrupted 2 months earlier). Nonfatal infarction occurred in three patients in each group, which was due to occlusion of the target vessel in four out of six. Of these, a non-Q wave myocardial infarction was observed in two patients in the stent group and in one in the VBT+stent group. A significantly higher need for TVR was observed in the stent group (35% vs. 13%, P=.04). Between 6 and 12 months, two additional events occurred in the VBT+stent group: One patient underwent TVR and non-TVR at 229 days from the index procedure; another patient died postoperatively following peripheral vascular surgery 1 year after the index procedure. All TVR procedures were performed in the presence of recurrent angina or inducible ischemia. No patient underwent bypass surgery within 12 months following the index procedure.

4. Discussion 

The main findings of this investigation can be summarized as follows: Patients carrying the D allele of the ACE gene polymorphism in whom high plasma ACE level is expressed are at a high risk for restenosis after elective bare metal stent implantation and the preventive use of VBT reduces significantly in-stent neointimal proliferation.