Access to side branches with a sharply angulated origin: usefulness of a specific wire for chronic occlusions

INTRODUCTION

Nowadays the orthotopic cardiac transplantation is the endgame for many patients with end-stage heart failure in developed countries.1 After the third year of heart transplantation, chronic rejection is one of the leading causes of morbidity and mortality.2 One of the main manifestations of chronic rejection is cardiac allograft vasculopathy (CAV) that affects nearly 50% of transplanted hearts at 5 years.3 CAV is characterized by diffuse intimal thickening that leads to progressive coronary luminal narrowing, with similar consequences to native heart atherosclerotic disease.4 It typically shows diffuse lesions in the distal territories with more focal stenosis in the proximal segments.5 Although the main treatment of CAV is based on titrating immunosuppressive therapy, a PCI is usually conducted here.

The use of PCIs for the management of CAVs has been reported to have success rates above 90% but with long-term restenosis rates of up to 36%.6 On the other hand, as a result of denervation following transplantation and subsequent incomplete reinnervation, most patients are asymptomatic or show atypical symptoms, despite silent progression to advanced stages of the disease. Therefore, angiographic findings of a CTO in this population are not rare.

Due to the high rate of restenosis associated with these procedures7 and the lack of solid evidence of a clinical benefit, medical treatment is advised in these patients especially when it comes to CTOs. Nevertheless, CTO recanalization has experienced a significant boost due to new techniques and technological advances made over the last few years. Therefore, in highly experienced centres performing PCIs of CTOs, this kind of procedures can be an alternative.

We know from registries published in recent years that the success rates of PCIs on CTOs, in non-transplanted patients are between 60% and 80% in the United States, Canada, and Europe.8^,9 However, there are no studies of the success rates and results of PCIs on CTOs in patients in whom CAV can play an important role. We don’t have data on short and long-term clinical benefits either.

Consequently, the main goal of our study is to evaluate the characteristics of this population, the feasibility of PCI in these patients and its clinical results. In addition, we will make a comparison with heart transplant recipients (HTR) who underwent a PCI on lesions without CTO criteria.

METHODS

Definitions

For the purpose of this paper, the main conditions are defined as follows. According to the EuroCTO definition, chronic total coronary occlusion (CTO) is defined as the presence of Thrombolysis in Myocardial Infarction 0 flow within the occluded coronary segment with an estimated occlusion duration of > 3 months.10 Percutaneous coronary intervention (PCI)-target lesion procedure success is defined as the achievement of < 30% residual diameter stenosis of the target lesion as assessed by visual inspection or quantitative coronary angiography, without an in-hospital major adverse cardiac event (death, acute myocardial infarction, or repeated coronary revascularization of the target lesion).11 In-stent restenosis is the re-narrowing of a stent implanted at a lesion site to treat a prior stenosis, to an in-stent diameter stenosis of > 50%, including the original treated site plus the adjacent vascular segments 5 mm proximal and 5 mm distal to the stent.11 Regarding PCI-related myocardial infarction related (and according to the 4th universal definition of myocardial infarction), stand-alone post-procedural increases of cardiac troponin values are enough to establish a diagnosis of procedural myocardial injury but not for the diagnosis of a type 4a myocardial infarction. Type 4a myocardial infarction requires the elevation of cardiac troponin values greater than 5 times the 99th percentile URL in patients with normal baseline values or patients with elevated pre-procedural cardiac troponin in whom the cardiac troponin levels are stable (≤ 20% variation) or dropping. The post-procedural cardiac troponin needs to rise > 20% to an absolute value more than five times the 99th percentile upper reference limit. In addition, there should be evidence of new myocardial ischaemia, either from electrocardiogram changes, imaging evidence, or procedural-related complications associated with reduced coronary blood flow such as coronary dissections, occlusions of a major epicardial artery or side branch occlusion/thrombi, collateral flow disruptions, slow flows or no-reflow, or distal embolizations.12

Patients and data analysis

First, we performed an analysis of the incidence of CTO in the context of coronary angiography screening of CAV in one of our centres. We also show the proportion of patient who underwent PCIs. Secondly, we conducted a retrospective analysis of all HTRs who underwent PCIs on a coronary CTO between January 1, 2006 and December 31, 2016 in 2 centres with an ongoing CTO program. Data from both centres were used for the analysis of the CTO PCI procedure characteristics, immediate results, clinical events and coronary angiography during follow-up. Also, we analysed clinical events during follow-up of the same previous HTR with PCI on a CTO and compared it to HTRs with PCI on a non-CTO lesion. The clinical endpoints analysed were these: readmission for heart failure or acute myocardial infarction, sustained ventricular arrhythmias, cardiovascular death, and all-cause mortality. Patient demographics and PCI-related data have been extracted from hospital databases. Digital images have been mined from dedicated storage servers. Transthoracic echocardiography images were obtained before and after CTO recanalizations to assess variations in the left ventricular ejection fraction using the Simpson method. Diagnostic coronary angiographies were analysed to assess the CTO characteristics prior to the PCI. Short- and long-term complications were obtained from medical records. In addition, follow-up coronary angiographies were performed in most patients, according to the protocols of the different centres.

This study was approved by the clinical trials committee of the Hospital Universitario Puerta de Hierro de Majadahonda, Madrid, Spain, in full compliance with the principles of the Declaration of Helsinki.

Statistical methods

For the assessment of the differences in the baseline demographic characteristics between the CTO group and the non-CTO group, the qualitative variables were expressed in percentages and analysed using the chi-square test. The quantitative variables were expressed as mean ± standard deviation and analysed using the Student t test. The quantitative variables without normal distribution were expressed as median ± interquartile range and analysed using the Wilcoxon test.

When it comes to the size of the sample, a comparative analysis of the clinical outcomes between CTO and non-CTO group was conducted using Fisher’s exact statistical test. To compare mortality between the CTO with the non-CTO group we used the Kaplan-Meier survival analysis. For all the tests, a P values < .05 were considered statistically significant. The statistical analysis was performed using the software SPSS package (V 21.0.0.0).

RESULTS

Incidence of chronic total coronary occlusion in heart transplant recipients

On the incidence of CTOs, only patients from Hospital Universitario Puerta de Hierro were studied. During the analysed period, 605 coronary angiographies were performed in HTRs. Among these, 74 patients (12%) had a CTO according to the EuroCTO criteria. Of these patients, PCIs were performed in only 10 (13%), leaving the remaining 64 patients under medical treatment (figure 1).

Figure 1. Study flowchart. CTO, chronic total coronary occlusion; PCI, percutaneous coronary intervention.

Figure 2. Kaplan-Meier survival analysis comparing chronic total coronary occlusion (green) vs non-chronic total coronary occlusion (blue) percutaneous coronary intervention.

Clinical characteristics. Procedural outcomes and follow-up

Clinical characteristics

Ten patients patients from Hospital Universitario Puerta de Hierro and 4 patients from Hospital Clinic de Barcelona, Barcelona, Spain were included in the study.

The baseline clinical characteristics and follow-up of each patient are shown on table 1. Eighty-five percent were males with an average age of 57 years [46.2-66] at the time of the PCI.

Table 1. Clinical variables and follow-up

Patient	Age, y	HT	DM	DL	Clinical presentation prior to CTO diagnosis	Time from heart transplantation to PCI, ears	Time from CTOs PCI to FUCA, days	ISR in follow-up	LVEF prior PCI	LVEF post-PCI	Follow-up post-PCI
1	73	Yes	No	No	Asymptomatic	14	238	No	59%	52%	Alive at 4 and 1 months
2	64	No	No	No	HF	8	192	No	30%	45%	Alive at 4 and 4 months
3	25	No	No	No	HF	5	1122	Yes	35%	40%	Alive at 10 and 4 months
4	31	Yes	No	Yes	Asymptomatic	8	175	-	50%	55%	Alive at 9 and 6 months
5	60	Yes	No	Yes	Angina	16	210	No	Unknown	Unknown	Death at 1 and 8 months due to ruptured iliac artery aneurysm.
6	57	Yes	No	Yes	Asymptomatic	10	No FUCA	Unknown	40%	60%	Alive at 4 and 2 months
7	53	Yes	No	Yes	Asymptomatic	12	No FUCA	Unknown	60%	Unknown	Sudden death at 2 and 4 months
8	32	No	Yes	No	Asymptomatic	-	No FUCA	Unknown	35%	40%	Alive at 3 and 1 months
9	73	Yes	Yes	No	Asymptomatic	10	161	No	50%	65%	Death due to metastatic pancreatic cancer at 2 and 3 months
10	63	Yes	No	No	Angina	6	No FUCA	Unknown	60%	55%	Sudden death at 267 d
11	57	No	No	No	Dyspnea	14	134	No	60%	Unknown	Alive at 1 and 6 months
12	51	No	No	No	Asymptomatic	10	246	No	55%	Unknown	Alive at 2 and 6 months
13	53	Yes	No	Yes	Dyspnea	18	90	Yes	55%	Unknown	Alive at 1 y
14	72	No	No	Yes	Asymptomatic	18	1439	No	65%	63%	Alive at 4 y
CTO, chronic total coronary occlusion; DL, dyslipidemia; DM, diabetes mellitus; FUCA, follow-up coronary angiography; HF, heart failure; HT, hypertension; ISR, in-stent restenosis; LVEF, left ventricular ejection fraction; PCI, percutaneous coronary intervention.

On the clinical manifestations when the CTO was diagnosed, 4 patients had angina or angina-like symptoms (28%), 2 patients required hospitalization for decompensated heart failure (14%) and the remaining 8 (57%) were asymptomatic. In asymptomatic patients, the diagnosis of ischemia was achieved by studying regional wall motion abnormalities in the follow-up echocardiograms (62%) and electrocardiographic changes suggestive of ischemia (12%). In the remaining patients (26%), the diagnosis was achieved based on the CAV screening coronary angiography.

The time elapsed from cardiac transplantation to coronary CTO PCI procedure varies from a minimum of 5 years to a maximum of 18 years (median 10 years).

Baseline coronary angiography and procedure

The angiographic characteristics of CTOs and details of each patient’s procedure can be found on table 2 and table 3.

Table 2. Chronic total coronary occlusion angiographic characteristics

Patient	CTO location	J-CTO score	Blunt entry	Calcification	Occlusion length > 2 cm	Bending > 45º	Re-try lesion
1	Proximal RCA	0	-	-	-	-	-
2	Proximal LAD	0	-	-	-	-	-
3	First OM	2	+	-	-	+	-
4	Mid LAD	1	-	-	+	-	-
5	Mid RCA	2	-	-	+	+	-
6	Proximal LAD	1	-	-	+	-	-
7	Mid LAD	1	-	-	-	+	-
8	Distal RCA	1	-	-	-	+	-
9	Mid LAD	0	-	-	-	-	-
10	Second OM	1	+	-	-	-	-
11	Mid RCA	2	-	+	-	+	-
12	Mid LAD	0	-	-	-	-	-
13	Mid LAD	2	-	+	+	-	-
14	Mid LAD	1	+	-	-	-	-
CTO, chronic total coronary occlusion; LAD, left anterior descending artery; OM, obtuse marginal artery; RCA, right coronary artery.

Table 3. Variables related to the percutaneous coronary intervention

Patient	Access	Guide catheter	Guidewires	Successful guidewire	Microcatheter	Stent	Total length treated with stent	Successful intervention	Contrast
1	Femoral	AR 2-6 Fr	2	Miracle 3g	Yes	CYPHER	33 mm	Yes	241 mL
2	Femoral-radial	EBU 3.5-6 Fr	2	PT Graphics	No	CYPHER SELECT	23 mm	Yes	218 mL
3	Radial	AL 2-6 Fr	2	Miracle Bros 3	No	CYPHER	33 mm	Yes	117 mL
4	Femoral-radial	EBU 4-8 Fr	3	Miracle 6	Yes	TAXUS Liberté x3	96 mm	No	132 mL
5	Femoral-radial	JR 4-6 Fr	2	Miracle 3	No	TAXUS Liberté XIENCE V x2	83 mm	Yes	300 mL
6	Femoral	EBU 3.5-7 Fr	Unknown	Pilot 50	Yes	CYPHER x3 Vision x1	66 mm	Yes	468 mL
7	Femoral	JL 4-7 Fr	Unknown	Miracle 3	Yes	CYPHER	33 mm	Yes	158 mL
8	Femoral-radial	JR 4-6 Fr	Unknown	Fielder FC	Yes	CYPHER	18 mm	Yes	182 mL
9	Femoral	EBU 3.5-6 Fr	2	Fielder XT	Yes	XIENCE V Prime x2	43 mm	Yes	225 mL
10	Femoral-radial	Hockey S	Unknown	Fielder XT	Yes	Xience xpedition	22 mm	Yes	240 mL
11	Femoral	JR 4-6 Fr	1	Gaia third	Yes	Resolute Onyx	38 mm	Yes	224 mL
12	Radial	EBU 4-6 Fr	1	Fielder XT	No	Resolute Integrity	22 mm	Yes	154 mL
13	Radial	EBU 4-7 Fr	2	Fielder XT	Yes	BioMatrix x2	53 mm	Yes	129 mL
14	Femoral-femoral	EBU 4-8 Fr	2	Fielder XT	Yes	PROMUS Element	18 mm	Yes	200 mL

The occluded artery was the left anterior descending artery in 8 patients (57%), the right coronary artery in 4 patients (29%) and the left circumflex artery in 2 patients (14%). Only in one case (patient 8) a distal occlusion was treated, while the remaining patients showed proximal or mid segment occlusions. No patient had more than 1 CTO.

Fifty percent of the patients (7 patients) had 1 vessel disease, 28% (4 patients) 2 vessels disease and 21% (3 patients) 3 vessels disease. Other non-CTO severe lesions were treated before the CTO procedure in those patients with multivessel disease. The mean J-CTO13 score was 1 (± 0.78).

In 11 patients (79%), the femoral artery was used as the main access, in 5 of these patients the radial approach was used simultaneously. In the remaining 3 patients (21%), only radial approach was used. Contralateral injections were used in 6 interventions (3%).

In all cases, the anterograde strategy was used. An average of 1.9 guidewires were used per procedure and an intravascular ultrasound was performed in 3 of the procedures. In 10 patients (71%), at least one dedicated microcatheter was used. In every case, a guidewire escalation approach was performed, starting with guidewires with lower tip load and penetration capacity to guides with higher tip load and penetration capacity.

In every case, drug-eluting stents were deployed with an average of 1.57 stents per patient and an average stent length of 41.5 mm.

The amount of contrast used in the procedures went from 117 to 468 mL with a median of 209 mL. However, no events of contrast-induced nephropathy were reported.

The PCI was successful in 13 patients (92.8%). The only failed attempt (patient 4) was a mid-segment left anterior descending artery CTO with a J-CTO score of 1. Three drug-eluting stents were deployed but final Thrombolysis in Myocardial Infarction flow was 1. Follow-up coronary angiography showed no improvement in coronary flow. No further attempts were made to recanalize the vessel.

In-hospital results

Regarding cardiovascular events during hospitalisation after the PCI, only 1 patient had a procedural myocardial injury, with a significant increase of myocardial necrosis markers (troponin I peak of 9 µg/dL for a 99th percentile upper reference limit of 0.06 µg/dL) but without haemodynamic impairment or new regional wall motion abnormality. No radiodermatitis was reported.

Clinical and angiographic follow-up

During a median follow-up of 33.5 months [20-50] mortality rate was around 28.5% (4 out of 14 CTO patients). Among these, 2 deaths were due to sudden cardiac death, 1 to advanced stage pancreatic carcinoma and 1 death was due to suspected ruptured right iliac artery aneurysm (unrelated to the procedure).

The improvement of left ventricular ejection fraction measured using the Simpson method was confirmed in 6 of the 9 patients who underwent an echocardiogram both before and after the CTO procedure, with 5.8% (± 0.87) of global mean improvement (P = NS).

All patients from centre #1 had a coronary angiography during follow-up. The median time from the CTO PCI to the follow-up angiography was 201 days (161-246). Two patients had in-stent restenosis (20%) after 3 and 37.4 months, respectively (one first-generation drug-eluting stent and one second-generation stent). In both cases reintervention was successful, and new drug-eluting stents were deployed with no further events at follow-up.

CTO versus non-CTO PCI in heart transplant recipients

We compared the results of 14 HTRs in whom one CTO lesion PCI was performed and 36 HTRs with non-CTO lesion PCI over the same period of time.

The long-term follow-up of patients was 100% (CTO and non-CTO patients), with median follow-up of 27 months [14.7-50.2], a minimum of 3, and a maximum of 124 months.

No statistically significant differences were found when the clinical and demographic baseline variables were compared in both groups (table 4).

Table 4. Demographic characteristics of heart recipient patients with percutaneous coronary intervention. Period 2005-2016

Variable	CTO	Non-CTO stenosis	P
Patient	14	36	-
Mean age, y [IQR]	57 [46.2-66]	62 [47.5-68.7]	.552
Woman (%)	14.3	22.2	.538
HT (%)	57.1	63.9	.667
DM (%)	21.4	22.2	.953
DL (%)	50	52.8	.863
Actual smoker/former smoker (%)	50	23	.685
CKD (%)	90	74.3	.303
Statins (%)	90	82.4	.572
Median age at the time of heart transplant in years [IQR]	44 [37.5-58]	46 [35.7-54]	.709
Drug-eluting stent (%)	100	100	1
Acute rejection (%)	70	51.4	.259
Mean LVEF (%) (standard deviation)	50 (11.4)	52.7 (10.2)	.558
CKD, chronic kidney disease; CTO, chronic total coronary occlusion; DL, dyslipidemia; DM, diabetes mellitus; HT, hypertension; IQR, interquartile range; LVEF, left ventricular ejection fraction.

During follow-up, there were no statistically significant differences in the all-cause mortality rate and cardiovascular mortality rate between both groups (table 5). The mortality rate was 28% (4 out of 14) in the CTO group and 14% (5 out of 36) in the non-CTO group (P = .245). The cardiovascular mortality rate was 21% (3 out of 14) in the CTO group and 8% (3 out of 36) in the non-CTO group (P = .331).

Table 5. Major cardiac events. Period 2005-2016

Variable	CTO	Non-CTO	P Fisher´s exact test
HF or AMI hospital admissions- n (%)	4 (28.6%)	11 (31.4%)	.844
Death-n (%)	4 (28.6%)	5 (13.9%)	.245
Cardiovascular death-n (%)	3 (21.4%)	3 (8.3%)	.331
Sustained ventricular arrhythmias-n (%)	0 (0%)	0 (0%)	-
AMI, acute myocardial infarction; CTO, chronic total coronary occlusion; HF, heart failure.

Regarding the 3 patients with cardiovascular death in the non-CTO group, all had severe CAV and advanced chronic kidney disease: 2 of them died due to refractory heart failure and 1 due to humoral rejection. The causes of death in the CTO group are described in the dedicated section.

The rate of readmission for heart failure or acute myocardial infarction during follow-up was 28.6% (4 of 14) in the CTO group and 31.4% (11 of 35) in the non-CTO group (P = .844).

No ventricular tachycardias were reported at follow-up in any of the groups.

DISCUSSION

There is little experience in the percutaneous management of CAVs. Until recently, CAV was considered a progressive and irreversible process with few therapeutic options. Since treatment with m-TOR inhibitors to delay the progression of the disease is effective in many cases, the PCI has emerged as an adjunctive treatment in these patients with good results.14 However, the results of the CTO PCI in this context are largely unknown. Considering CAV as a diffuse vessel disease, the clinical and angiographic outcomes of CTO recanalization in transplanted hearts are difficult to extrapolate based on the results observed in atherosclerotic coronary occlusions.

As far as we know, this is the very first systematic study on the prevalence of CTOs in HTRs. Also, it is the very first analysis of PCI results of CTOs in this population.

We have found a PCI-target lesion procedure success rate very similar to that of studies published in CTO secondary to atherosclerotic coronary heart disease (92.8%). The rate of intrastent restenosis is similar to that found in large series of PCI in CTO in non-transplanted patients. Although the procedures were performed over an extended period of time, there were no differences in success rate between older and recent procedures in witch we cannot exclude an inclusion bias.

Regarding the safety of PCI, there has only been one procedural myocardial injury with elevation of cardiac necrosis markers, although there has been no haemodynamic involvement or impact on the ejection fraction of the left ventricle.

Despite the fact that our patients were treated with immunosuppressants, we found no contrastinduced nephropathy events, which may reflect that an optimization in contrast volume, adequate prophylaxis and the correct selection of cases can decrease the rate of renal impairment.

Unexpectedly, given the progressive nature of CAV, only 2 of the cases developed in-stent restenosis that was successfully treated in both cases with no other restenosis during long-term follow-up.

Consistent with the poor clinical prognosis of CAV in series already published, in our registry there is a high mortality rate during the long-term follow-up (28.5%). Of the 3 patients who died of cardiovascular death, 2 of them developed sudden death (271 and 856 days after PCI to CTO) and 1 patient died due to hypovolemic shock secondary to the suspected and unrelated to the procedure ruptured iliac artery aneurysm.

Although statistically irrelevant, there is an apparent higher mortality rate in HTRs with CTOs compared to HTRs with nonocclusive stenosis, which may have to do with a more advanced stage of CAV or with more associated comorbidities.

We should mention here that in our study, the CTO recanalization produces a statistically nonsignificant improvement of the ejection fraction.

Accordingly, our study suggests that following an adequate selection of cases, centres experienced in the management of CTOs can feasibly handle PCIs on CTOs in transplanted hearts with a high PCI-target lesion procedure success and low periprocedural complications. However, probably due to the underlying disease, medium and long-term results are still poor with a high mortality rate and a significant rate of restenosis.

Limitations

Although it should be noted here that this is the first study ever on this subject, there are several limitations. First of all, its retrospective nature. Secondly, the number of heart transplant patients is limited despite combining the experience of 2 high volume HTR centres. Nevertheless, we should keep in mind that the overall experience with these patients is very scarce considering their special characteristics.

In addition, the proportion of HTRs in our centres who undergo CTO PCI is low. Considering the CAV inner nature, it is not rare to find CTO of distal vessels or diffuse distal disease making any PCI attempts futile if not impossible. In fact, most of our patients who underwent PCIs showed CTOs in proximal segments with a good distal vessel. We could conclude that our procedures were performed in highly selected patients. This would be a bias that would be favoring a high success rate by selecting less complex cases (medium J-CTO score 1).

When it comes to the statistical analysis, we should mention here that it has not been possible to perform more powerful analyses such as propensity score analyses due to the low number of patients included in the registry.

However, despite these limitations, we believe that our study is relevant because it shows the experience accumulated over the years in 2 high heart transplant volume centres and because it is the very first study on this subject with a long-term follow-up of the patients.

CONCLUSIONS

Coronary CTO is a common condition in HTRs. PCI is feasible in centres with extensive experience conducting CTO procedures and in selected patients, with a PCI-target lesion procedure success and periprocedural complications rate similar to global CTO procedures and non-CTO PCIs. The rate of in-stent restenosis in HTRs is similar to that found in large series of CTO PCIs in non-transplanted patients. There is a non-significantly higher all-cause and cardiovascular mortality in HTRs undergoing a CTO PCIs compared to those with non-CTO PCIs. This may have to do with more advanced stages of the CAV.

CONFLICTS OF INTEREST

The authors have no conflicts of interest to declare whatsoever.

 

 

REFERENCES

INTRODUCTION

The number of patients who receive chronic oral anticoagulation (COAC) is huge, and is expected to increase in the future due to the overall aging of the population and the increased incidence of conditions that will require COAC.

The prevalence of COAC among patients with coronary disease who undergo percutaneous coronary interventions is between 6% and 8%.1 Most cases are due to the concomitant presence of atrial fibrillation with moderate-to-high embolic risk.

On the other hand, up to between 20% and 30% of the patients with atrial fibrillation and an indication of COAC present with coronary disease.2 Taking into account that the prevalence of atrial fibrillation in the population is between 1% and 2%, up to 1-2 million anticoagulated patients in Europe will end up needing one coronary angiography procedure.

There is, therefore, a significant number of patients receiving COAC with vitamin K antagonists (VKA) or one of the most recent direct-acting oral anticoagulants (DOACs, Apixaban, Rivaroxaban, Dabigatran or Edoxaban) that require coronary angiography. The routine practice with these patients is very variable, but traditionally patients with COAC have discontinued the drug and moved on to bridging therapy (BT) with low molecular weight heparin a few days before and a few days after the procedure.3 However, an increase of hemorrhagic events with this strategy in interventional procedures and higher morbidity and mortality in these patients due to bleeding or prothrombotic situations due to the discontinuation and reset of anticoagulants has been reported.2^,4^-7

The safety of diagnostic catheterization through radial access under treatment with acenocoumarol (VKA) has been demonstrated previously. In our group, the safety profile of VKA has already been evaluated in this type of patients in the past.8 Since there is less evidence in patients on DOACs, with this work we want to provide new evidence on this regard, given the increasing number of patients anticoagulated with these new drugs over the last few years.

In this study, we report our experience and we evaluate the safety profile of transradial diagnostic cardiac catheterizations in patients on DOACs who were discharged the same day they underwent the procedure. Also, we compared these patients with other treated with heparin during catheterization and patients on COAC with VKA.

METHODS

Study population

This is a prospective and observational study where 270 patients who underwent elective transradial cardiac catheterizations were included from January 2013 through November 2017, divided into 3 groups of 90 patients based on the intake of VKA, DOACs or without oral anticoagulant treatment, and then matched according to the date of completion. All patients who underwent diagnostic catheterization and were having DOACs during this period were recruited. As control groups, we decided to recruit the next patient who underwent a diagnostic catheterization without anticoagulant treatment and the next one that was receiving acenocoumarol without withdrawal.

In no case was oral anticoagulant therapy withdrawn prior the procedure.

In patients treated with DOACs, 20 were on dabigatran (22.22%), 38 patients on rivaroxaban (42.22%), 29 patients on apixaban (32.22%) and 3 patients on edoxaban (3.33%).

In patients undergoing VKA treatment with international normalized ratio (INR) values < 2 (underdosing) and in DOAC patients who missed their last dose by mistake, intraprocedural sodium heparin was prescribed at a dose of 2500-5000 international units in one intra-arterial bolus. In our series, 21.3% of the patients treated with VKA received underdosing (INR < 2) and, therefore, needed heparin. In this group of patients, the mean INR was 2.5 ± 0.06. The range was 1.3-4.3 (75% INR > 2.1).

In a former article of our group8 we described the methodology of outpatient catheterizations at our center that we detail here.

Patients without COAC received the standard anticoagulant therapy with one intraarterial bolus of 5000 IU of unfractionated heparin (the routine clinical practice at our center). Also, 2.5 mg of verapamil were administered intraarterially to all patients to prevent any radial spasms.

They underwent elective transradial cardiac catheterization with same-day discharge, after removing the compression bandage and achieving hemostasis. All cases were conducted through one hydrophilic 5-Fr sheath. Hospitalized patients who underwent diagnostic catheterization were excluded as well those in whom an angioplasty procedure was anticipated.

All procedures were conducted in a single interventional cardiology unit with huge experience using transradial access (over 90% of all cases annually) and with an active program of same-day discharge outpatient cardiac catheterizations.8^,9

Procedural characteristics

All patients are welcome at a room near the interventionist laboratory and are evaluated by a nurse specialized in their monitoring and follow-up. This nurse is in charge of informing the patients, collecting the patient’s background, verifying the doses and time of the last COAC intake, and estimating the INR in patients on acenocoumarol. After informing the patient and collecting the informed consent, the best vascular access is selected, assessing the quality of the pulse and performing the Barbeau test.

Once the procedure is completed, the radial compression is performed using the patent hemostasis technique for, at least, 2 h. Radial patency and possible complications are assessed in this room prior to the patient’s discharged. The complications and outcomes of the vascular approach we assessed were: acute bleeding (with need to extend the length of compression), hematoma 5-10 cm at discharge, radial patency at discharge, bleeding and/or hematoma at 24 h, radial patency at 1 month and the presence of other unusual complications (pseudo-aneurysms, fistulas, arterial perforations, or compartment syndrome). For hematoma classification, the EASY criteria were used.10 The compression of the vascular access was made with swab and elastic bands for 2 h using the patent hemostasis technique11 where distal permeability is verified through plethysmography. Additionally, a 30 min extra-compression was performed if the puncture bled when the bandage was removed. In order to assess radial patency after the removal of compression, the test described by Barbeau et al. was used.12 was used here. Artery occlusion was defined by type D response (no recovery of the curve of pulse in 2 min). All patients were contacted via telephone over the following 24 h after the procedure to determine delayed local complications, and they were all followed during 1 month to determine the access occlusion.

Statistical analysis

Data are expressed as absolute rate and percentage for qualitative variables. Quantitative variables are expressed as mean (standard deviation) or median 25–75 interquartile range depending on variable distribution. Group comparisons were analyzed using the Student t test or its non-parametric equivalent; the Man– Whitney U-test for continuous variables, and chi-square test or Fisher’s exact test were used for the categorical variables. Statistical significance was defined as P values < .05. The statistical analysis was conducted using the statistical package SPSS 19.0 (SPSS, Inc.; Chicago, Illinois, United States).

RESULTS

The patients included in the study were assigned to three groups of 90 individuals each and matched by date of procedure: one group without oral anticoagulation (group A), another group with anti-vitamin K treatment (group B) and another with direct-acting anticoagulants (group C).

The indication of oral anticoagulation in group B mostly corresponds to patients with atrial fibrillation (74.4%), 26.7% of patients had valvular disease and 5.6% of them were carriers of mechanical prostheses; the remaining patients received anticoagulation due to a past medical history of embolism (5.6%), dilated cardiomyopathy (1.1%) and other causes.

The baseline characteristics of our patients are shown on table 1. The group treated with DOACs had a higher proportion of men than the VKA group (71.1% vs 47.8%; P = .01) and patients were younger in the group without oral anticoagulation (63.4 ± 11.5 vs 70.2 ± 9.3; P = .03). Group B had a lower percentage of diabetic patients (22.2% vs 36.67% in group C, P = .03). Group A patients had a past medical history of ischemic heart disease more frequently than the groups of anticoagulated patients (27.84% vs 14.44% in group C, P = .028) and therefore, they had undergone previous catheterization using the same access in higher percentages (20% group A vs 5.5% in group C, P = .04).

Table 1. Demographic and procedural baseline characteristics

Variable	Total	Group A: no OAC (n = 90)	Group B: VKA (n = 90)	Group C: DOACs (n = 90)	P value
Age (years), median ± SD	68.59 ± 10.63	63.45 ± 11.47	72.09 ± 8.97	70.22 ± 9.35	.03a NSb NSc
Men (%)	62.6	68.9	47.78	71.1	NSa .001b 041c
Hypertension (%)	75.9	83.33	67.78	76.67	NSa NSb NSc
Diabetes mellitus (%)	32.2	37.78	22.22	36.67	NSa .03b .048c
Dyslipidemia (%)	48.5	53.33	52.22	40	NSa NSb NSc
Smoking (%)	10	10	10	10	NSa NSb NSc
BMI (kg/m2) , median ± SD	30.10 ± 4.62	30.11 ± 4.45	29.22 ± 4.77	30.98 ± 4.51	NSa NSb NSc
Prior isquemic heart disease	17	27.78	8.89	14.44	.028a NSb NSc
ASA treatment	79	65	3	11	< .0005a .048b < .0005c
Clopidogrel treatment	25	21	0	4	< .0005a NSb NSc
Prior catheterization same access (%)	11.5	20	8.89	5.55	04a NSb .031c
ASA, acetylsalicylic acid; BMI, body mass index; DOAC, direct-acting oral anticoagulants; NS, non-significant; OAC, oral anticoagulation; SD, standard deviation; VKA, vitamin K antagonists. aNo OAC (group A) vs DOAC (group C). bVKA (group B) vs DOAC (group C). cDOAC (group C) vs groups A+B.

When it comes to the concomitant treatment with antiplatelet agents, patients without COAC took acetylsalicylic acid more frequently (72.2% vs 12.2%; P < .0005) compared to the DOACs group, as well as clopidogrel (23.3% vs 4.4%; P < .0005). Acetylsalicylic acid was also more widely used in the DOACs group compared to the VKA group (12.2% vs 3.3%; P = .048). All this is probably related to a greater suspicion of ischemic heart disease in these patients. There were only 2 patients treated with prasugrel and one with ticagrelor in the group without COAC.

There were no other significant differences on the remaining baseline characteristics (high blood pressure, dyslipidemia, body mass index, smoking…).

Radial access was the access of choice in most patients (98.2%), and ulnar access in the remaining patients. Regarding complications (table 2) of vascular access, during the procedure and during the 24 h and 1 month follow-up, there were no significant differences, showing a rate of hematoma and/or bleeding at discharge of 1.1% in the DOACs group and arterial occlusion rates both at discharge and at 1 month between 0-2 an 2% in this group. Only one patient needed hospitalization due to prolonged radial bleeding.

Table 2. Complications (bleedings and occlusions) and vascular access events

Variable	Group A: no OAC (n = 90)	Group B: VKA (n = 90)	Group C: DOACs (n = 90)	P value
Early bleeding or hematoma (%)	2.2	1.1	1.1	NSa NSb
Late bleeding or hematoma (%)	0	0	1.1	NSa NSb
Early occluded access (%)	3.3	3.3	2.2	NSa NSb
Late occluded access (%)	3	2.9	0	NSa NSb
Other complication (dissection, fistula, perforation)	0	0	0
aDOAC, direct-acting oral anticoagulants; NS, non-significant; OAC, oral anticoagulation; VKA, vitamin K antagonists. bNo OAC (group A) vs DOAC (group C). cVKA (group B) vs DOAC (group C).

DISCUSSION

The performance of diagnostic cardiac catheterizations without withdrawing COAC is recommended in the guidelines13 and our standard routine here at our unit of hemodynamics and interventional cardiology is using mostly radial access (95%) excellent safety results.8^,9

The information comes basically from studies conducted in patients under treatment with VKA. Two meta-analyses that addressed this issue14^,15 conclude that performing catheterizations without withdrawing COAC is safe and effective. The study published by the Finnish group led by Karjalainen16 also assessed the safety profile comparing it to a group of patients on COAC and heparin BT. They found a rate of bleeding significantly higher in the latter group (1.7% vs 8.3%), being higher in the COAC-BT withdrawal group compared to the withdrawal group of COAC without BT (2.5% vs 8.3%). Also, if the procedure is done through radial access, the results in terms of bleeding are even better.17

However, when talking about patients on DOAC treatment, the evidence is scarce and there is no consensus. The current guidelines on revascularization18 do not include the recommendation of keeping oral anticoagulation during the procedure, but in recent expert consensus statements published by different international societies, there is controversy on this issue. Thus, due to the lack of existing evidence, the European document of antithrombotic consensus19 recommends to not stop anticoagulation in the case of VKA, but pre-withdraw DOACs between 12 and 24 h (24-48 h in the case of dabigatran) in patients who will undergo the intervention. On the other hand, in the document on preoperative and perioperative antithrombotic management20 they recommend to not withdraw antithrombotic treatment with DOACs in low-risk hemorrhagic procedures such as transradial diagnostic catheterizations because of its lower rate of vascular complications, especially when it comes to bleeding.

Despite all this, the standard practice with these patients is variable, but as a rule of thumb most hemodynamic laboratories worldwide withdraw COAC and move on to BT with low molecular weight heparin a few days before and after the procedure.3 However, there are more hemorrhagic events with this strategy when performing interventional procedures as well as more morbidity and mortality of these patients due to bleeding or prothrombotic situations that are created when withdrawing and reseting anticoagulant therapy.2^,4^-7

With the increasingly use of DOACs, new problems arise in our routine clinical practice when these drugs need to be withdrawn before performing procedures and it is common to see that in these patients BT is prescribed the same way as it is prescribed in patients undergoing VKA treatment from many medical and surgical specialties, although its use in recent consensus documents is not recommended.20

When it comes to cost-effectiveness, the BT strategy is a tremendous cost overrun due to several aspects: the higher incidence of hemorrhagic complications in patients who will need medical attention, the high cost of low molecular weight heparin, longer hospital stays, and the need to analyze the levels of anticoagulation in patients under treatment with VKA. In other areas of interventional cardiology, progress has been made on this regard and cost-effectiveness studies have been conducted on this question, such as the study conducted by Coyle et al.21 that showed that the non-prescription of BT saved US$ 1800 per patient following some the aforementioned aspects. Also, If we compare DOACs and VKA drugs, a recent study conducted by Shah et al.22 in patients with permanent atrial fibrillation showed that all DOACs proved superior in a cost-effectiveness model that included quality-of-life adjusted survival, with dabigatran being the most cost-effective drug in patients with the highest thrombo-embolic risk of all.

The main conclusion that we can draw from this study is that it is safe to perform diagnostic cardiac catheterizations through transradial access in patients on chronic anticoagulation with DOACs. These patients do not have more frequent vascular or bleeding complications compared to those under standard therapy.

When it comes to bleeding complications, there was a low incidence rate in all groups without any significant differences. Another thing to take into consideration in the case of DOACs is the possibility of anticoagulation reversal with the specific antidote (currently only available for dabigatran and about to be on the market for factor Xa inhibitors) in case of serious complications which gives us a greater safety profile when performing invasive procedures with these drugs.23

In our population, the incidence of radial occlusion was exceptionally low in all groups. Early (< 24 h) and late occlusions (1 month) occurred in 2.2% and 0% of the patients from the DOAC group; in 3.3% and 2.9% of the patients from the acenocoumarol group compared to 3.3%; and in 3% of the patients from the heparin group without any statistically significant differences. Previous studies have reported extremely low rate of occlusions, even lower than 1%.24

We have not found in the medical literature any series similar to ours, although there are numerous series in which coronary angiographies are performed without withdrawing COAC with VKA, with results that are consistent with ours.8^,25^,26

There have been trials with DOACs in percutaneous coronary interventions that have not found any differences in the clinical adverse events (bleeding, embolism, ischemia) in patients in whom the percutaneous coronary intervention was conducted under different anticoagulation strategies.27^,28

Study limitations and future directions

The low incidence of complications and the small size of the sample did not allow us to conclude any significant statistical differences among the groups. It would be necessary that the size of the sample was larger, which is difficult in a single center. Maybe a multicenter registry could shed some more light on this issue.

Nowadays, the number of patients under acenocoumarol or warfarin treatment is rapidly dropping due to the exponential use of new anticoagulant drugs as dabigatran, apixaban, rivaroxaban, and edoxaban. In sum, new and larger studies on direct-acting oral anticoagulants should be conducted on this regard.

CONCLUSIONS

The performance of outpatient diagnostic catheterizations using the radial access without withdrawing the DOAC treatment seems to be safe and does not bring a greater deal of complications compared to patients under treatment with acenocoumarol or without anticoagulant treatment. It would be advisable to conduct randomized studies to be able to confirm these data.

CONFLICTS OF INTEREST

We declare no conflicts of interest whatsoever.

 

REFERENCES

INTRODUCTION

In-stent restenosis (ISR) is still a common problem and a therapeutic challenge due to the high rates of culprit lesion revascularization. The treatment of choice for the management of ISR is still to be established.1 According to several randomized trials, drug-eluting balloon (DEB) angioplasty has better results for the management of ISR compared to conventional angioplasty and similar results compared to in-stent implantation of first-generation drug- eluting stents (DES),2^-13 although it is inferior to second-generation DESs (especially in the ISR of DESs).14^,15 This strategy has a class I indication (level of evidence A) for the management of ISR both in bare-metal stents (BMS) and DES.16

The effectiveness of this long-term strategy (over one year of follow-up) has already been established by the medical literature,17^-22 yet it is largely unknown in the very long run (over 3 years of follow-up).23 Similarly, the studies conducted so far claim that high comorbidity is an exclusion criterion because the effect it has on real-world unselected patients is still unknown.

Our aim was to assess the effectiveness of DEBs in a registry of real world patients in a long-term follow-up period (5 years).

METHODS

Retrospective registry of a cohort of patients with ISR treated with DEBs at a high-volume center experienced in performing these procedures (> 1500/year) and percutaneous coronary interventions (> 800/year). The ISR was defined as > 50% angiographic stenosis as seen in 2 radiographic in-stent orthogonal projections or less than 5 mm away from its edges accompanied by symptoms of angina or objective ischemia. All lesions were treated with the same DEB (SeQuent Please, B. Braun Surgical, Melsungen, Germany). No clinical exclusion or angiographic criteria were established in the registry.

Both the clinical and the procedural characteristics were gathered from the center and cath. lab databases. Quantitative coronary angiography of the lesions was carried out using the Philips Xcelera system. Mehran’s classification for coronary restenosis was used to characterize the lesions.24 The procedural strategy and predilation with cutting-balloon or non-compliant high-pressure balloon was left at the discretion of the operator. Dilation with DEB was attempted for at least, 60 seconds at nominal pressure.

A 5-year follow-up period was established. The study was approved by the Clinical Trials Committee. All follow-up periods occurred were done in accordance to clinical criteria consulting the regional healthcare system electronic database, which keeps a comprehensive record of all patient-system communications.

All events were defined in a standard way following the Academic Research Consortium-2 (ARC-2) consensus.25 The primary endpoint was the need for target lesion revascularization (TLR) with the DEB and the total number of lesions treated. Secondary endpoints were any revascularization of acute coronary syndromes/acute myocardial infarctions according to the universal definition established by the European Society of Cardiology (the same or any location),26 all-cause mortality and cardiovascular mortality, and hemorrhage according to the Bleeding Academic Research Consortium (BARC) ≥ 3.27 Also, the following device-oriented composite endpoints (DOCE): TLR + acute coronary syndrome/acute myocardial infarction of the culprit vessel + cardiovascular mortality or patient-oriented composite endpoints (POCE): any revascularization + acute coronary syndrome/acute myocardial infarction + stroke + overall mortality were estimated on the total number of patients. Stent thrombosis was also defined following ARC-2 criteria and was also estimated on the total number of lesions.

For data analysis, the IBM SPSS 19.0 statistics software package was used. Quantitative variables were expressed as mean ± standard deviation, and qualitative variables was relative percentage. The cumulative incidence of events at follow-up was measured too. A bivariate analysis was conducted using the chi-square test or Fisher’s exact test, and the multinomial logistics regression analysis was used to estimate primary endpoint predictors (statistically significant value: P < .05). The Kaplan-Meier method was used to build the cumulative incidence curve during the follow-up of the primary endpoint.

RESULTS

A total of 53 ISRs in 48 patients were efficiently treated with DEBs from January 2010 through December 2013. In one patient, the DEB did not cross the lesion, so the patient was not included in the study (98.2% success rate of the procedure using the DEB). The baseline characteristics of the lesions and the coronary interventions are shown on table 1 and table 2; 49.1% (n = 26) of the lesions showed a Mehran I pattern of ISR and 24.5% (n = 13) of the lesions were located at the edges of the stent. Good angiographic results were obtained in all the patients (residual stenosis < 30%) and Thrombolysis in Myocardial Infarction grade 3 flow. The mean follow-up was 5.6 years (range: 0.2-8.2 years). There were no losses during follow-up.

Table 1. Baseline characteristic of patients

Characteristics of patients	n = 48
Age (years)	69.3 ± 11.8 (44-93)
Male	77.1% (37)
AHT	75% (36)
Dyslipidemia	56,3% (27)
Smoking	35,4% (17)
Diabetes
DM + OAD	33.3% (16)
IDDM	12.5% (6)
Atrial fibrillation (OAC)	22.4% (11)
Prior AMI	43.8% (21)
Prior indication for PCI
Stable IHD	37.5% (18)
ACS	62.5% (30)
Prior CABG	6.3% (3)
CRF (GFR < 60 mL/min)	41.3% (19)
LVEF (%)	55 ± 10.2 (65-29)
Multivessel disease	77.1% (37)
Incomplete revascularization	18.9% (9)
P2Y12 inhibitors
Clopidogrel	95.8% (46)
Ticagrelor/prasugrel	4.2% (2)
Duration of DAPT
3 months	22.4% (11)
6 months	33.3% (16)
12 months	43.8% (21)
ACEI/ARA-II	87.5% (42)
Beta-blockers	91.7% (44)
Statins	100% (48)
ACEI, angiotensin converting enzyme inhibitors; ACS, acute coronary syndrome; AHT, arterial hypertension; AMI, acute myocardial infarction; ARA-II, angiotensin II receptor antagonists; CABG, coronary artery bypass grafting; CRF, chronic renal failure; DAPT, dual anti-platelet therapy; DM, diabetes mellitus; GFR, glomerular filtration rate; IDDM, insulin dependent diabetes mellitus; IHD, ischemic heart disease; LVEF, left ventricular ejection fraction; OAC, oral anticoagulants; OAD, oral antidiabetics; PCI, percutaneous coronary intervention.

Table 2. Characteristics of the lesions and the procedure

Characteristics of the lesion/PCI	n = 53
Location
Anterior descending artery	50.9% (27)
Circumflex artery	15.1% (8)
Right coronary artery	26.4% (14)
Left main stem	5.7% (3)
Aortocoronary vein graft	1.9% (1)
Bifurcation	26.4% (14)
Diffuse ISR (Mehran II, III, IV)	50.9% (27)
ISR in the edges of the stent	24.5% (13)
Type of stent with ISR
BMS	64.2% (34)
DES	22.6% (12)
DES over previous BMS	7.5% (4)
DES over previous DES	5.7% (3)
Second generation DES/total DES	84.2% (16)
Stent ≥ 3 mm	66% (35)
Diameter of the stent (mm)	2.9 ± 0.4 (2-4)
Length of the stent (mm)	22.8 ± 7.1 (8-38)
Diameter of the DEB (mm)	3.1 ± 0.3 (2-3.5)
Length of the DEB (mm)	20 ± 5.3 (15-30)
QCA prior to the PCI
Reference diameter (mm)	3.08 ± 0.39
Minimal lumen diameter (mm)	0.9 ± 0.43
Stenosis (% diameter)	62 ± 15
Length	13 ± 5.3
QCA post-PCI
Reference diameter (mm)	3.1 ± 0.36
Minimal lumen diameter (mm)	2.27 ± 0.35
Stenosis (% diameter)	15 ± 6
Cutting balloon	30.2% (16)
Non-compliant balloon	62.3% (33)
Radial access	55.1% (27)
Successful PCI (residual stenosis < 30%)	100% (53)
BMS, bare-metal stent; DEB, drug-eluting balloon; DES, drug-eluting stent; ISR, in-stent restenosis; PCI, percutaneous coronary intervention; QCA, quantitative coronary angiography.

Most patients (95.8%. n = 46) received clopidogrel as a P2Y₁₂ receptor inhibitor as part of their dual anti-platelet therapy (DAPT) and 22.4% (n = 11) oral anticoagulants and short courses of DAPT (3 months).

The rate of TLR at 1 year was 9.4%, and 20.8% at 3 and 5 years. The rate of late TLR (after the follow-up year) was 11.4%. The Kaplan-Meier analysis conducted (figure 1) shows that events accumulated during the first 3 years. The rate of TLR at 1 year was significantly slower in the ISR of BMSs (2.9% vs 21.1%; P = .05), and no late TLR was seen in the ISR of DESs. The rate of 5-year TLR was not associated with diabetes (22.7% vs 19.2%; P = .76) and it was not significantly lower with the use of the cutting balloon (12.5% vs 24.3%; P = .47) in ≥ 3 mm stents (25.7% vs 11.1%; P = .29) or in the ISR of BMSs (20.6% vs 21.1%; P = .96). Neither the bivariate analysis nor the logistics regression analysis identified any variables that would act as independent predictors of TLR.

Figure 1. Kaplan-Meier analysis of the cumulative incidence of target lesion revascularization (TLR) at a 5-year follow-up.

There were no cases of stent thrombosis in the target lesion. Two cases of acute coronary syndrome/acute myocardial infarction on the target vessel were reported: 1 case due to another lesion proximal to the lesion treated with DEB (with good results), and another case of acute coronary syndrome/non-ST-segment elevation acute myocardial infarction due to new onset ISR of the lesion treated with DEB.

At follow-up, a new coronary angiography was performed in 50% of the patients (n = 24). In 29.2% of the patients (n = 14) there was angiographic confirmation of lack of recurrent ISR in the stent treated with DEB. One patient underwent 2 TLRs in 2 different lesions at follow-up.

The remaining secondary and combined endpoints are shown on table 3. The overall mortality rate at 5 years was 33.3% (n = 16), and neoplasms were the most common cause (n = 5). According to the ARC-2 criteria, the cardiovascular mortality rate was up to 10.4% (n = 5): 2 sudden deaths at the 3-year follow-up (both with previous coronary angiography and good results of the lesions treated with DEB) and 3 strokes at the 2-year follow-up (one hemorrhagic stroke on oral anticoagulants).

Table 3. Secondary and composite outcomes at the 5-year follow-up

Secondary outcomes at 5 years	n = 48
Any revascularization	29.2% (14)
ACS-AMI in the target vessel	4.2% (2)
ACS-AMI in a different location	10.4% (5)
Overall mortality	33.3% (16)
Cardiovascular mortality	10.4% (5)
Stroke	12.5% (6)
DOCE	31.3% (15)
POCE	54.2% (26)
BARC ≥ 3 hemorrhages	6.3% (3)
ACS, acute coronary syndrome; AMI, acute myocardial infarction; BARC, Bleeding Academic Research Consortium; DOCE, device-oriented composite endpoints; POCE, patient-oriented composite endpoints.

A total of 13 (27.1%) patients suffered hemorrhages at follow-up, although only 6.3% (n = 3) were on DAPT (one of them on oral anticoagulant medication). Three patients had BARC ≥ 3 hemorrhages: one patient, a BARC 3a GI hemorrhage, and 2 patients had BARC 5b hemorrhages (one due to a GI hemorrhage on DAPT and the other one due to an intracranial hemorrhage on oral anticoagulant medication).

DISCUSSION

As far as we know, this study is the first of its kind to describe the long-term progression of ISR lesions treated with DEBs in unselected (outside clinical trials) real-world patients (old and with high cardiovascular risk). The conclusions we can draw from our series are: a) although the per-year rate of TLR is similar to that of other studies already published, in the long run, it seems to be higher than the one reported in selected patients; b) late TLR (after 1 year of follow-up) represents half of the cases that require new revascularizations at follow-up; c) in the long term, TLR is similar in the ISR of both BMSs and DESs, yet late TLR occurs only in the ISR of BMSs; d) the rate of TLR seems to stabilize after 3 years; and e) the use of DEBs for the management of ISRs is a safe strategy from the standpoint of stent thrombosis, even in patients who receive short courses of DAPT.

The mid-term results (6-12 months) of the use of DEBs for the management of ISR have been widely described in randomized studies comparing this strategy with simple angioplasty or DES implantation in populations with clinical and angiographic exclusion criteria. Scheller et al.3 initially reported angiographic 6-month TLR rates of 0% and clinical 12-month-TLR rates of 4% in the PACCOCATH - ISR study (Treatment of in-Stent Restenosis by Paclitaxel Coated PTCA Balloons). Further studies show data more adjusted to actual results, with mid-term TLR rates of 6.6% to 8.8%,2^,7^,8^,10^,13^-15 with differences based on whether we were dealing with the ISR of BMSs (6% to 8.7%)2^,5^,14 or DESs (4.3% to 22.1%).4^,6^-8^,13^,15^,28 In the RIBS studies (Restenosis Intra-stent of Bare Metal Stents: Paclitaxel-eluting Balloon vs Everolimus-eluting Stent) V and IV (populations with a geographic location similar to that of the sample), Alfonso et al.14^,15 reported a 1-year TLR rate of 6% and 13% in the ISR of BMSs and DESs, respectively. Also, these two studies showed 26 TLRs in 249 patients (95 with BMSs and 154 with DESs), that is, a 10.4% rate very similar to the 9.4% rate from our series.

Several studies have published their 3-year follow-up results: PEPCAD21 and RIBS V19 in BMSs, and RIBS IV,20 PEPCAD-DES,22 and ISAR-DESIRE 318 in DESs, with TLR rates of 6.2%, 8%, 15.6%, 19.4%, and 33,3%, respectively. These studies reported a total of 94 TLRs in 524 lesions amounting to a 3-year TLR rate of 17.9% lower than the 20.8% rate from our series. Although in our sample diabetes was not associated with TLR, it is a powerful predictor of ISR.29 This finding may be explained by a much greater prevalence of diabetes in our sample (55.5%) compared to the aforementioned studies (32% to 40%).

Late TLR is defined as a TLR occurring after one year of follow-up. The studies published indicate that TLR usually occurs over the first year of follow-up, not being very relevant thereafter, with rates between 0% and 4.1% (0%-25% of total cases). In our series, late TLR rate was 11.4% in over half of the cases (54% of all TLRs). Only the ISAR-DESIRE18 study and the study conducted by Habara et al.30 reported rates of late TLR close to the rates shown by our study (14.5% and 7.2%, respectively; 43% and 39% of all TLRs). The differences seen between those randomized studies and our series where TLR is clinical may be explained by the fact that most patients were angiographically followed during the first year, which in turn may have shown prematurely, in asymptomatic patients, a new angiographically significant ISR (> 50%), but not significant enough to cause any symptoms.

In our series there were only cases of late TLR in BMSs, since all TLRs in DESs occurred within the first year of follow-up. Initially, the effectiveness of DEBs was greater in the ISR of BMSs, yet the incidence of TLR went up at follow-up to the point of matching the incidence of TLR in the DES group (figure 2). These findings may be explained by the greater initial effect of the DEB over the characteristic pattern of ISR in BMSs (smooth muscle cells) vs the pattern of neoatherosclerosis, more common in the ISR of DESs.28 Thus, this neoatherosclerosis may be causing the late events reported in BMSs as described by Nakazawa et al.31 Our findings are different from those published by Habara et al.30 in a long series (550 lesions in 468 patients), where late TLR occurs predominantly in the ISR of DESs (odds ratio = 6; P = .002). Although the authors say that there were no differences in the rates of TLR reported between first- and second-generation DESs, 70% of these devices were first-generation DESs (14% were paclitaxel-eluting stents), vs 15% in our series (3 sirolimus-eluting stents and 0 paclitaxel-eluting stents). This study follow-up ends after 2 years, so the question of whether these findings stand in the long run still remains.

Figure 2. Kaplan-Meier analysis of target lesion revascularization (TLR) at 5 years after the use of drug-eluting balloon for the management of in-stent restenosis of bare-metal stents (BMS) vs drug-eluting stents (DES).

Scheller et al.23 published the 5-year follow-up of the PACCOCATH-ISR study, with a 5-year TLR rate of 9.3% and a prevalence of diabetes of 17%. Miura et al.32 reported 5-year TLR rates of 34% in 216 ISRs of DESs, a much higher rate compared to the rate from our sample, with a similar prevalence of diabetes. Both studies suggest that TLR rate increases in the long term. Curiously, in our series, TLR rate stabilized after the third year. In these studies, cases of 2-to-5-year TLR rates have been described, but the exact moment when the TLR occurred was never reported. It is possible that clinical TLR post-DEB occurs after one year follow-up (1-3 years) and then stabilizes over time.

In the ISAR-DESIRE 4 study, the use of cutting-balloon for the management of ISR with DEBs obtained better angiographic results (binary restenosis) and a significant reduction of the 1-year TLR (16.2% vs 21.8%; P = .26).33 In our series, patients treated with cutting-balloon showed a numerically lower rate of TLR (12.5% vs 24.3%), but with no statistical significance due to the size of the sample. Only 30% of the lesions were treated using this technique since when the study was conducted, the benefits of such technique had not yet been established.33

Our sample had 2 characteristics of which there is scarce information in the medical literature: old age and use of oral anticoagulants. The average age was almost 70 years with data from patients who were extremely old (up to 93 years of age) and over 20% were on oral anticoagulants. Even though this is not an exclusion criterion per se, the studies published so far do not provide information on the percentage of patients on oral anticoagulants or the rate of hemorrhages at follow-up.18^-22 In these studies, DAPT was extended between 3 and 6 months. In our series, over 50% of the patients received DAPT over a course of 3-6 months and those patients on oral anticoagulants (22.4%) followed a 3-month strategy. The use of DEB with these courses was safe. The overall rate of hemorrhages was moderate, which is consistent with patients’ age, and only one patient had a BARC ≥ 3 hemorrhage while on DAPT. Yet despite the high percentage of patients on short courses of DAPT, no cases of definitive/probable target lesion thrombosis were reported. Except for the PEPCAD study, that did not provide information either on stent thrombosis of the lesion treated with DEB, the remaining studies reported long-term rates of stent thrombosis between 0.8% (ISAR-DESIRE study) and 2.6% (RIBS IV study).18^-22

ARC-2 criteria are consensus criteria to make results uniform, although in elderly populations with high ischemic risk they can diminish the effectiveness of the procedure. The high 3-yar rate of DOCE reported (31.3%) was this big due to cardiovascular mortality (10.4%), that according to the ARC-2 criteria should also include stroke-induced mortality. As we have already mentioned, based on chronology and causality only, it is unlikely that the cardiovascular mortality rate seen in our sample can be attributed to DEBs. Even though they are combined outcomes with different definitions, DOCEs may look like traditional MACE (major adverse cardiac events), with a similar incidence to the one reported by the studies published, somewhere around 20%-38% at 3 years.18^-22 Our patients’ high ischemic risk is evident on the high rates of acute coronary syndrome/acute myocardial infarction of the non-target lesions reported (10.2%) and the rates of stroke (12.5%) at follow-up. Both events together with age-related non-cardiovascular mortality (neoplasms) penalized the POCEs by doubling the DOCEs from our series (54.2%). Thus, approximately half of the composite outcomes according to the ARC-2 criteria at 5-year follow-up would not be attributable to the use of DEBs.

Limitations

Our study has some limitations. This was a retrospective, single-center study with limited cases that did not allow to obtain solid scientific evidence. In line with the results published by former studies and reported in the medical literature, our results support the use of DEBs for the management of ISR. Consequently, in the long run, around 80% of patients from unselected populations could avoid having to undergo the implantation of an additional layer of BMS, thus saving this strategy for DEB failure.

CONCLUSIONS

In a clinical practice cohort, DEBs for the management of ISR have a 5-year TLR rate of 20.8%. Late TLR accounts for half the cases at follow-up and occurs in the ISR of BMSs. TLR rate seems to stabilize after three years. No stent thrombosis was reported in lesions treated with DEBs.

CONFLICTS OF INTEREST

None declared.

 

 

REFERENCES

INTRODUCTION

The fractional flow reserve (FFR) measurement has become established as a valuable tool for the functional assessment of intermediate coronary stenoses.1 Maximum hyperemia is needed to be able to assess the FFR. The most widely used pharmacological agent to induce vasodilation is adenosine through an intravenous infusion or intracoronary injection.2

Regadenoson, a selective agonist of adenosine A2a receptors has been proposed as an alternative given how easy it is to use since only a single peripheral intravenous bolus at a fixed dose is needed regardless of the patient’s weight and renal function.3^-7

Several studies have compared regadenoson to intravenous adenosine,8 but to our knowledge there are no studies comparing regadenoson to intracoronary adenosine for the assessment of the FFR.

The main objective of this study was to establish individual variability in the measurement of FFR using intracoronary adenosine and intravenous regadenoson administered sequentially so each patient is case and control at the same time.

Secondary objectives included analysis of the hemodynamic response, measurement of hyperemia times and assessment of adverse events.

METHODS

Forty-one (41) intermediate coronary lesions were studied both prospective and consecutively in 39 patients referred to undergo a coronary angiography and who had been prescribed functional assessment with pressure guidewire. Stenoses of 30% to 70% estimated visually or through automatic quantification during the angiography procedure were categorized as intermediate lesions. FFR cut-off values ≤ 0.80 were established to indicate revascularization. An informed written consent was obtained from all patients included in this study.

Procedure

Coronary angiography was conducted following routine clinical practice. The FFR was measured using the PressureWire guidewire (St Jude Medical, St Paul, Minnesota, United States) after administering of unfractionated heparin (50 IU/kg) and placing the sensor distal to the lesion, following the standards recommended for acquisition purposes, registry and interpretation of pressure tracings.9

Pharmacological protocol

Based on former dose-response studies, an initial dose of intracoronary adenosine of 100 µg for the right coronary artery and 200 µg for the left coronary artery was established.10

In an attempt to achieve a degree of optimal hyperemia, patients with values close to the cut-off value (FFR < 0.85) were eligible to receive, at the discretion of the operator, repeated doses of 60 µg in each bolus. The minimum value obtained was selected as the true FFR.

After the administration of adenosine, the registry was started, and the phase of hyperemia was considered over when the FFR returned to baseline values. Then, a peripheral intravenous bolus of 400 µg of regadenoson was injected and the measurements re-taken.

The FFR was obtained using an analysis performed on a beat-to-beat basis and when in doubt or in the presence of artifacts, the tracings stored in the console were reviewed.

Additionally, values such as heart rate and arterial blood pressure, both at baseline level and during the phase of hyperemia, were recorded and the possible side effects monitored.

Lastly, data such as the time required by each drug to achieve hyperemia and the duration of hyperemia were recorded for further analysis.

Statistical analysis

Continuous variables were expressed as mean ± standard deviation and categorical variables as absolute value or percentage. The Student t test for paired data was used to compare the different FFR values, hemodynamic values (arterial blood pressure and heart rate) and times of hyperemia observed after the administration of adenosine and regadenoson. Symptoms were studied using the chi-square test. Using the linear regression analysis, Pearson’s correlation coefficient and the Bland-Altman plot, the correlation between the different FFR values and both drugs was studied. The statistical analysis was conducted using the SPSS v20 statistical software package (IBM, Armonk, New York, United States) and results were considered significant with P values < .05.

RESULTS

Study patients

Table 1 shows patients baseline characteristics. Overall, 41 intermediate lesions (average stenosis: 52 ± 9%) were studied, 29 of them located in the anterior descending artery, 7 in the right coronary artery, and 5 in the circumflex artery. The average dose of intracoronary adenosine administered was 236 ± 60 µg. No postprocedural complications were observed.

Table 1. Characteristics of the study population

Sample (n = 41)
Age (years)	65 ± 14
Women	27%
Body mass index (kg/m2)	30 ± 3
Prior medical history
Hypertension	73%
Dyslipidemia	58%
Diabetes mellitus	24%
Smoking	36%
Prior infarction	33%
Prior revascularization	49%
Heart failure	3%
Diseased vessels
1	59%
2	25%
3	16%
Artery studied
Anterior descending artery	67%
Circumflex artery	15%
Right coronary artery	18%
Degree of stenosis
30%-50%	42%
50%-70%	58%
70%-90%	0%
Diameter of the vessel (mm)	3.3 ± 0.5
Dose of intracoronary adenosine (µg)
Left coronary artery	243 ± 55
Right coronary artery	206 ± 76

Measurement of FFR with intracoronary adenosine vs intravenous regadenoson

With 1 or both drugs, 13 lesions (33%) showed FFR values ≤ 0.80, indicating a significant functional stenosis. Also, a strong linear correlation with both hyperemic stimuli was seen (r = 0.925; P < .001) (figure 1 and figure 2). However, the FFR measured after the administration of regadenoson was lower compared to the one obtained after the administration of intracoronary adenosine (0.838 ± 0.072 vs 0.852 ± 0.073; P = .002) (table 2). Also, in four different cases (9.8%) there were relevant discrepancies when FFR values > 0.80 with adenosine and ≤ 0.80 with regadenoson were obtained, which led to reclassifying the lesion (table 3).

Table 2. Hemodynamic effects of adenosine and regadenoson

	Fractional flow reserve	Average arterial blood pressure (mmHg)	Heart rate (beats/min)
Baseline	0.94 ± 0.05	114 ± 22	69 ± 12
Adenosine	0.85 ± 0.07	92 ± 21	70 ± 14
Regadenoson	0.84 ± 0.07*	85 ± 18*	89 ± 18*
Data are expressed as mean ± standard deviation. *P < .05 with respect to baseline values and adenosine.

Table 3. Individual values of fractional flow reserve with intracoronary adenosine and intravenous regadenoson

Case	Age (years)	Sex	Coronary artery	Dose of adenosine (µg)	Adenosine administration in FFR	Regadenoson administration in FFR
1	49	Male	ADA	240	0.88	0.85
2	82	Male	Cx	160	0.73	0.75
3	60	Male	ADA	300	0.84	0.81
4	79	Male	ADA	240	0.79	0.78
5	54	Male	ADA	120	0.70	0.70
6	54	Male	RCA	120	0.98	0.91
7	53	Male	RCA	180	0.85	0.86
8	67	Male	ADA	180	0.84	0.83
9	56	Male	ADA	180	0.86	0.84
10	81	Male	ADA	180	0.85	0.91
11	60	Female	Cx	240	0.94	0.95
12	46	Male	RCA	120	0.84	0.79*
13	62	Male	ADA	240	0.92	0.89
14	80	Female	ADA	180	0.93	0.88
15	89	Female	ADA	240	0.86	0.85
16	79	Male	RCA	300	0.95	0.93
17	71	Female	ADA	240	0.85	0.86
18	76	Female	ADA	300	0.78	0.80
19	52	Male	Cx	180	0.86	0.85
20	52	Male	ADA	300	0.81	0.75*
21	67	Male	Cx	300	0.99	0.97
22	63	Male	ADA	240	0.87	0.88
23	49	Female	ADA	300	0.71	0.74
24	45	Female	ADA	180	0.83	0.81
25	58	Male	ADA	240	0.94	0.94
26	62	Male	ADA	300	0.93	0.85
27	64	Male	ADA	240	0.82	0.82
28	64	Male	RCA	240	0.88	0.85
29	57	Female	Cx	180	0.99	1.00
30	54	Male	ADA	300	0.82	0.81
31	81	Male	ADA	300	0.79	0.79
32	81	Male	ADA	300	0.88	0.89
33	77	Female	ADA	240	0.75	0.74
34	75	Male	ADA	240	0.82	0.81
35	62	Female	ADA	300	0.90	0.90
36	58	Male	ADA	120	0.75	0.67
37	65	Male	ADA	240	0.90	0.90
38	67	Male	ADA	300	0.79	0.78
39	61	Female	ADA	300	0.81	0.77*
40	83	Female	ADA	300	0.82	0.78*
41	69	Male	RCA	240	0.91	0.91
ADA, anterior descending artery; Cx, circumflex artery; FFR, fractional flow reserve; RCA, right coronary artery. *Cases with clinically relevant discrepancies.

Figure 1. Linear regression analysis. Correlation of fractional flow reserve values measured with intracoronary adenosine and IV regadenoson in each patient. FFR: fractional flow reserve.

Figure 2. Bland-Altman plot. Graphic representation of the differences seen in the fractional flow reserve measured using intracoronary adenosine and intravenous regadenoson. FFR: fractional flow reserve.

Hemodynamic parameters

With both drugs we observed a significant drop in baseline average arterial blood pressure levels, which was even more pronounced with regadenoson. However, only regadenoson significantly increased baseline heart rate (table 2).

The average time elapsed until reaching maximum hyperemia was significantly lower with adenosine (15 ± 6 vs 61 ± 49 s; P < .001), and this effect was even more prolonged with regadenoson (44 ± 29 vs 174 ± 72 s; P < .001) (figure 3).

Figure 3. Times of hyperemia. Differences seen in the time elapsed until reaching hyperemia and its average duration (in seconds) between intracoronary adenosine and intravenous regadenoson.

Side-effects profile

Side effects were mild with both drugs (table 4). After the administration of regadenoson, most patients experienced some kind of discomfort that they tolerated well and was not an obstacle to continue with the study.

Table 4. Symptoms and adverse events while measuring fractional flow reserve

Event	Adenosine	Regadenoson	P
Symptoms*	11 (27)	30 (73)	.29
Asystole > 3 seconds	2 (5)	0	< .001
Other complications	0	0	-
Data express n (%). *Dyspnea, chest pain, headache, myocardial blush or nausea.

Two cases of blockade with pauses longer than 3 seconds after the administration of adenosine were observed in the right coronary artery that resolved spontaneously. No conduction disorders or any other kind of complications were reported with regadenoson.

DISCUSSION

The reliability of FFR measurements depends on the capacity to induce maximum coronary hyperemia.11 The pharmacological agent most widely used is adenosine in intravenous infusions, although intracoronary adenosine is also used by many laboratories because it is faster and achieves similar results. The doses recommended are 100 µg and 200 µg for the right and left coronary arteries, respectively. However, we know that the response to these different doses varies depending on the patient. Some authors recommend doses of 300 µg or even higher, but these high doses have adverse effects, particularly unwanted conduction disorders in diagnostic testing.12

This study shows that the use of regadenoson in a single peripheral intravenous bolus at a fixed dose regardless of the patient’s weight and renal function could be an alternative to adenosine. A good linear correlation has been described between the FFR measured using intracoronary adenosine or intravenous regadenoson (r = 0.925; P < .001).

Also, the comparative analysis conducted showed that the bolus of intravenous regadenoson achieved FFR values that were significantly lower compared to the FFR values obtained using boli of intracoronary adenosine (difference of 0.014 ± 0.028; 95% confidence interval, 0.005-0.023; P = .002).

Maybe the higher hyperemia achieved with regadenoson was the reason why, in four cases with FFR values > 0.80 after the administration of adenosine, FFR values ≤ 0.80 were obtained after the administration of regadenoson, which led to reclassifying the lesion as hemodynamically significant.

On the other hand, the exact administration of intracoronary adenosine dose required the positioning of the guide catheter inside the coronary ostium, which was not always possible, and was the reason why the dose administered was not always the established one.

Thus, due to the administration of an insufficient dose or the use of an inaccurate technique, the fact of the matter is that in 9.8% of the cases maximum hyperemia was not achieved with adenosine. Our data are consistent with former studies that show that up to 10% of the cases may show suboptimal hyperemia with intracoronary adenosine.13^-14 In this sense, one recent meta-analysis revealed differences in the FFR similar to those observed in our study when comparing intracoronary adenosine administration and the intravenous infusion of adenosine.15 This may be relevant to indicate revascularization.

Another interesting aspect of regadenoson is that maximum hyperemia was maintained for longer periods of time. Even though our study only analyzed vessels with focal lesions, the longer average time of hyperemia observed with regadenoson (174 ± 72 s) may be useful to perform multiple FFR measurements in vessels with serial lesions or diffuse disease.

One potential limitation of regadenoson is its higher cost compared to adenosine. In this sense, and with no cost-effectiveness studies, a more efficient use of resources has been reported when regadenoson was administered compared to adenosine and dipiridamol in stress tests with isotopes.16

Finally, in our study adverse events were mild with both drugs. However, the negative arrhythmogenic and dromotropic effects of adenosine are well-known. The administration of regadenoson in a single bolus at a fixed dose through peripheral intravenous route showed a good safety profile and complications such as bronchospasm and high-grade atrioventricular block were rare. On rare occasions, the more prolonged vasodilation effect of regadenoson may be unwanted. A 50 mg bolus of intravenous aminophylline may be administered to reverse its effect.17

Limitations

The data presented should be interpreted taking into consideration a series of limitations. This was an observational, open, multicenter study with a small sample, meaning that results may be affected by confounding factors common to this type of studies.

Even though the half-life of intracoronary adenosine is short, we cannot rule out that its prior administration may alter the posterior effect of regadenoson due to preconditioning phenomena.

The maximum dose of adenosine was used at the discretion of the operator, which may have turned out insufficient to induce maximum hyperemia.

Finally, we should not forget that new non-hyperemic indices have stormed into the coronary physiology setting as an alternative to the functional assessment of stenosis without requiring vasodilating agents.

CONCLUSIONS

The administration of regadenoson in a single intravenous bolus has shown greater effectiveness in the measurement of FFR compared to the administration of boli of intracoronary adenosine, and the differences seen may be relevant for the clinical decision-making process. Because of how easy it is to use and because of its safety profile, regadenoson seems like a useful alternative for the hemodynamic assessment of intermediate coronary stenoses.

CONFLICTS OF INTEREST

The authors declared no conflicts of interest whatsoever.

 

 

REFERENCES