Percutaneous treatment of the left main coronary artery in older adults. Impact of frailty on mid-term results

INTRODUCTION

Degenerative aortic stenosis (AS) is the most common valvular heart disease in Western countries, largely due to the increase in life expectancy and the progressive aging of the population. Transcatheter aortic valve implantation (TAVI) has revolutionized the treatment of AS, with documented efficacy in inoperable patients,1 high surgical risk patients,2 and more recently, in patients with intermediate or low surgical risk.3,4 This growing body of evidence has led to a progressive increase in the number of TAVIs performed each year,5 creating a need within health care systems to adapt to this demand and optimize care pathways and hospital admissions.6

AS is an age-related condition. The presence of frailty, comorbidity, and other geriatric syndromes is closely associated with the rate of complications, need for hospital readmissions, and mortality in both conservatively treated patients7 and those undergoing surgery8 or TAVI.9 Both frailty and comorbidity burden are also associated with a higher rate of procedureal complications,10 despite the fact that technological advancements have made TAVI an increasingly less invasive procedure. Furthermore, prior data suggest that patients with a higher comorbidity burden face higher rates of readmissions and non-cardiac mortality, which may limit or even negate the benefits of TAVI in some individuals.11 Therefore, optimizing patient selection to avoid procedural futility remains one of the most pressing clinical challenges.

On the other hand, the frailty phenotype is defined as a potentially reversible state of vulnerability to external stressors.12,13 In patients with cardiovascular disease, a significant portion of this frailty is attributable to the heart condition per se and is, therefore, potentially reversible through specific treatment. However, in patients with a higher comorbidity burden, it may be difficult to determine what proportion of this frailty is associated with other underlying health conditions. It has been proposed that treating patients with incipient frailty requires a comprehensive approach to achieve reversal,14 including physical exercise, adequate nutrition, and tight control of comorbidities, which can accelerate the development of frailty. Some publications have shown that exercise programs can partially reverse frailty in various cardiovascular settings, including patients with AS.15-17 However, implementing these strategies in the routine clinical practice faces major obstacles, due to the cost associated with including this growing patient population in cardiac rehabilitation programs and the difficulty of achieving sustained adherence. Previous experiences show low adherence to hospital-based exercise programs (around 30%) even in controlled clinical trial settings.18

Telemedicine offers potentially significant advantages in the management of older patients with frailty and cardiovascular disease, as it may enable adequate follow-up by overcoming logistical barriers (need for transportation, suboptimal adherence), which are particularly relevant in this context. Various telemedicine tools have demonstrated their usefulness in improving prognosis in patients with heart failure,19 thus allowing for early detection of decompensation and prevention of hospitalizations and other complications. However, to this date, there is no experience on the application of telemedicine to frail patients with AS undergoing TAVI.

For these reasons, the main endpoint of the TELEFRAIL TAVI clinical trial is to analyze, in patients with AS undergoing TAVI, the effect of a comprehensive telematic intervention on frailty reversal 3 months after the intervention vs standard care.

METHODS

Study design

We conducted a prospective, multicenter, randomized (1:1) clinical trial to compare a comprehensive telematic intervention targeting frailty vs standard post-discharge care in patients ≥ 75 years with AS and frailty criteria undergoing TAVI. The study is promoted by the Section of Geriatric Cardiology of the Spanish Society of Cardiology and will be conducted in 20 Spanish hospitals, with participation from clinical and interventional cardiologists, geriatricians, and other specialists experienced in the management of these patients, as well as trained nursing staff. The trial is registered at ClinicalTrials.gov (NCT06742970).

Study population

Eligible patients must meet the following inclusion criteria: a) severe AS, defined by a mean aortic gradient > 40 mmHg or an aortic valve area < 0.8 cm2 on echocardiography; b) age ≥ 75 years; c) undergoing TAVI during hospitalization; and d) meeting baseline (preoperative) frailty criteria defined by a Short Physical Performance Battery (SPPB) score < 1020 and a (Fatigue, Resistance, Ambulation, Illnesses, Loss of weight) FRAIL score ≥ 3.21

Exclusion criteria include a) refusal to participate; b) inability to complete geriatric assessments or follow study procedures; c) inability to understand or sign informed consent forms; and d) life expectancy < 12 months.

Beyond defined criteria, patient inclusion and derived procedures must be deemed reasonable by the responsible medical team. If complications occurred during hospitalization, they must be reasonably resolved prior to inclusion.

Treatment protocol

Prior to performing TAVI, baseline geriatric assessment will be conducted via interview with the patient, family, or caregivers by multidisciplinary trained personnel at participant centers. We’ll assess functional capacity for basic activities of daily living using the Barthel Index,22 an ordinal scale ranging from 0 to 100. This scale categorizes dependence into total (0–20), severe (21–40), moderate (41–60), mild (61–90), and independent (> 90). Instrumental activities will be assessed using the Lawton and Brody scale.23 Moreover, we’ll evaluate cognitive status with the Pfeiffer Test24 and Mini Mental State Examination (MMSE),25 and frailty using the SPPB20 including a) balance in 3 positions (feet together, semi-tandem, tandem); b) gait speed (4 m); and c) 5 chair stands. The total SPPB score goes from 0 to 12, with scores < 10 indicating increased risk of disability and falls. Furthermore, we’ll assess frailty using the FRAIL scale,21 the Clinical Frailty Scale,26 and the Essential Frailty Toolset8 including a) 5 chair stands; b) MMSE;25 c) hemoglobin values; and d) albumin levels. Additionally, we’ll assess comorbidity using the Charlson Comorbidity Index27 (maximum score, 37) and record the number of chronic prescription drugs prior to admission. Finally, we’ll assess nutritional using the Mini Nutritional Assessment-Short Form (MNA-SF)28 (scores < 11 indicate malnutrition risk). Quality of life will be evaluated using the EQ-5D-5L questionnaire.29

Patients meeting frailty criteria and consenting to participate will be randomized before hospital discharge (post-TAVI) to 2 treatment arms: a) comprehensive telematic intervention (nutrition + supervised exercise + health education) within the first 90 days, or b) standard post-discharge care. Patients will be randomized using an online computer-generated 1:1 scheme, with concealed allocation. In-hospital medical therapy will follow clinical practice guidelines and the treating team’s judgment.

Frailty intervention

We will perform the intervention to allocated patients within the 90 days following discharge via a specialized central telematic platform. Afterwards, expert health care professionals in nutritional support and adapted physical exercise will conduct videocalls on week 1, day 15, and every 2 weeks within the first 3 months.

For physical exercise, we’ll use an adaptation of the VIVI FRAIL program.30 On week 1, we’ll be using protocol A, moving to protocol B up to day 30 after discharge depending on good clinical tolerance, eventually moving to protocol C during months 2 and 3. The entire process will be monitored through periodic videocalls by specialists in physical exercise for elderly patients. In addition, patients and their families will have access to a phone number to resolve doubts, which will be available during working hours from Monday through Friday throughout the entire process.

Regarding nutritional support, patients allocated to the frailty intervention group will receive nutritional information best suited to each individual profile after hospital discharge. The research team will provide nutritional supplements for 3 months after discharge. Patients will receive nutritional supplementation with a hypercaloric and hyperproteic formula. Nutritional supplements will be taken once a day after completing the corresponding exercise regimen during that period (3 months).

Health education will consist of complementing the information received on nutrition and exercise during successive videocalls within the first 3 months and resolving any related questions. Moreover, we’ll provide information to optimize treatment adherence and control cardiovascular risk factors.

Study endpoints

The primary endpoint will be the percentage of patients whose frailty is reversed, as measured by SPPB20 (meaning they achieve an SPPB score of ≥ 10) 3 months after discharge. Staff from participant centers blinded to the allocated treatment group will be conducting this assessment (figure 1).

Figure 1. TELE-FRAIL TAVI study design. AS, aortic stenosis; SPPB, Short Physical Performance Battery; TAVI, transcatheter aortic valve implantation.

The following will be analyzed as secondary endpoints:

• – Number of days alive and out of hospital 1 year31 after TAVI.
• – Need for readmission (cardiac or non-cardiac) at 3 months and 1 year.
• – All-cause and cardiovascular mortality rates at 3 months and 1 year.
• – Rate of cardiovascular events (MI, stroke, revascularization) at 1 year.
• – Proportion of robust patients (SPPB >10) at 1 year.
• – Disability (Barthel Index22) at 3 months and 1 year.
• – Nutritional risk (MNA-SF28) at 3 months and 1 year.
• – Quality of life (EQ-5D-5L29) at 3 months and 1 year.

Geriatric assessment during follow-up

A trained staff member from the participant centers will conduct the geriatric assessment in person. This staff will be blinded to the allocated treatment group. Three months and 1 year after TAVI, we’ll re-evaluate functional capacity (Barthel Index), instrumental activities of daily living (Lawton-Brody Index), nutritional risk (MNA-SF), cognitive ability (Pfeiffer test), quality of life (EQ-5D-5L), and frailty using the FRAIL, Clinical Frailty Scale, SPPB, and Essential Frailty Toolset scales. Clinical follow-up includes an in-person visit at 3 months and 1 year.

Study committees

This project is an independent clinical trial, with no industry funding. A steering committee will be responsible for overseeing the scientific and operational aspects of the study. While patients and researchers won’t be blinded to the treatment group allocation, a blinded event adjudication committee will evaluate clinical events to prevent bias. Similarly, a data safety monitoring board will be responsible for making relevant recommendations to the steering committee regarding the endpoints, as well as any potential observations related to patient safety.

Statistical analysis and sample size

Previous data from elderly patients with severe AS and frailty criteria undergoing TAVI show an approximate 50% proportion of frailty reversal after TAVI. Assuming an estimated 70% frailty reversal in the group allocated to the post-discharge telemedicine intervention, with 80% statistical power and a two-sided alpha error of 0.05, and accounting for a 10% loss to follow-up, the calculated sample size is 206 patients (103 in each group). A multicenter approach is necessary to achieve this planned sample size.

All statistical comparisons will follow the intention-to-treat principle. Results will be expressed as frequency and percentage or as median and standard deviation, as appropriate. Inter-group comparisons will be drawn using Fisher’s exact test. Patient follow-up will be censored at the time of death or at the end of the study. Primary endpoints will be compared between the 2 groups using a logistic regression model, considering frailty reversal as the dependent variable, the intervention as a fixed independent variable, and other covariates with a significant association with exposure in the final statistical model. In addition to secondary endpoints, the effect of the intervention on clinical events will be described using the Kaplan-Meier method, and Cox regression will be used for its evaluation. Hazard ratios and their 95% confidence intervals will be calculated. For all analyses, a two-sided P < .05 will be considered statistically significant. If significant differences are observed in the distribution of covariates between the 2 groups (control vs intervention), all variables with a significantly different distribution between the 2 groups, in addition to the allocated treatment group, will be included in the adjusted analysis. Prespecified subgroup analyses will be performed based on sex and comorbidity burden (Charlson Comorbidity Index).

Ethical aspects

Both participants and their families will receive detailed information about the potential risks and benefits of participating in the study. Both the study protocol and the informed consent form have been evaluated by the reference ethics committee. Participants will have the opportunity to carefully read the consent form and ask questions before signing it. All participants must sign the informed consent form prior to being included in the study. A copy of such form will be provided to participants. The rights and benefits of participants will be protected, with an emphasis that the quality of medical care will not be negatively impacted if they refuse to participate in the study.

DISCUSSION

Despite TAVI has become the treatment of choice in most elderly patients with AS, avoiding futility in patients with a higher burden of comorbidity, frailty, and disability remains clinically challenging. Therefore, a considerable percentage of elderly AS patients present with frailty, partly due to their heart disease (potentially reversible with specific treatment), and partly due to other existing comorbidities. In patients with a high comorbidity burden, isolated treatment of AS may not provide clinical benefit in the absence of other additional measures.11 A detailed understanding of each patient’s profile is essential for appropriate therapeutic planning both for AS and for the overall health status once the valvular heart disease has been corrected.

Frailty is considered an intermediate state in the transition toward disability that is potentially reversible, especially in its early stages.13 A holistic, multidisciplinary approach supported by physical exercise, proper nutrition, and tight control of underlying disease is essential.14 Some physical activity programs have partially reversed frailty in various cardiovascular contexts, including patients with AS.15-17 However, implementing these strategies in the routine clinical practice faces substantial challenges due to cost and the difficulty in ensuring sustained patient adherence.18 The TELE-FRAIL TAVI clinical trial aims to assess the potential impact of a comprehensive intervention on frailty reversal and prognosis in frail elderly patients undergoing TAVI. The telematic design of the intervention is expected to facilitate greater adherence to exercise and nutrition recommendations, thus contributing to frailty reversal along with AS correction via TAVI. The study design should allow determination of the percentage of patients in whom frailty is reversed with this combined strategy vs conventional TAVI follow-up. Furthermore, it should identify the clinical and geriatric profile of patients in whom the strategy fails, indicating a high likelihood of procedural futility.

CONCLUSIONS

The results of the TELE-FRAIL TAVI clinical trial are expected to shed light on the optimal management of frail elderly patients with AS undergoing TAVI. Optimizing treatment and prognosis for these complex patients could have significant clinical, financial, and social implications.

FUNDING

This project received funding from the Spanish Society of Cardiology (Clinical Research Projects SEC 2024, SEC/FEC-INVCLI 24/21) and the Catalan Society of Cardiology (2024 Grant).

ETHICAL CONSIDERATIONS

Both the study protocol and the informed consent form were reviewed and approved by the relevant ethics committee. All participants must sign the informed consent form prior to being included in the study.

STATEMENT ON THE USE OF ARTIFICIAL INTELLIGENCE

Not used.

AUTHORS’ CONTRIBUTIONS

E. Bernal-Labrador, R. Romaguera, F. Formiga, and A. Ariza-Solé contributed to the study conception and manuscript drafting. S. García-Blas, A. Regueiro, V. Serra, H. Tizón-Marcos, L. Asmarats, V. Agudelo, C. Scardino, J.M. Casanova-Sandoval, T. Rodríguez-Gabella, C. Jiménez-Méndez, A. Pérez-Rivera, C. Robles-Gamboa, A. Ayesta, P. Díez-Villanueva, S. Raposeiras-Roubín, I. Amat-Santos, A. Esteve-Pastor, G. Veiga-Fernández, M. Anguita, D. Martí-Sánchez, N. Martínez-Velilla, L. Cortés, E. Calvo-Barriuso, and S. Asimbaya contributed to critical manuscript revision.

CONFLICTS OF INTEREST

R. Romaguera is Associate Editor of REC: Interventional Cardiology; the journal’s editorial procedure to ensure impartial handling of the manuscript has been followed. The remaining authors declared no conflicts of interest whatsoever.

REFERENCES

INTRODUCTION

Over the past 5 years, the collaboration between the Interventional Cardiology Association of the Spanish Society of Cardiology (ACI-SEC) and the Interventional Working Group of the Spanish Society of Pediatric Cardiology (GTH-SECPCC) has led to the consolidation of the Spanish Cardiac Catheterization in Congenital Heart Diseases Registry, substantiated thus far by the publication of its first 3 reports on the activities conducted in 2020,1 2021,2 and 20223 (figure 1). The first 2 reports demonstrated that the number of centers participating in the registry, although highly representative of pediatric activity, did not accurately reflect the activity of adult congenital heart disease conducted in Spain.4,5,6 Therefore, the analysis of the current report presented in this article, on the activity conducted in 2023, has incorporated—same as in the previous report—the interventional activity in congenital heart disease from the Registry of Hemodynamics and Interventional Cardiology of the ACI-SEC for that same year.7 This methodology has proven to more reliably quantify hemodynamic procedures in congenital heart disease across all age groups. The findings from this edition were presented on June 14, 2024, at the ACI-SEC Congress held in Las Palmas de Gran Canaria (Canary Islands, Spain).

Figure 1. Comparison of the number of interventional procedures in 2020, 2021, 2022, and 2023.

METHODS

Data come from an annually updated, retrospective, voluntary, and non-audited registry. The inclusion of interventional data on congenital heart disease from the Registry of Hemodynamics and Interventional Cardiology of the ACI-SEC was possible due to the harmonization of questionnaires between the 2 registries conducted the previous year, which continued to undergo improvements.

All hospitals already participating in the Registry of Hemodynamics and Interventional Cardiology of the ACI-SEC, and all hospitals represented in the GTH-SECPCC, were asked to participate. Data was collected through an electronic database managed by an external company (pInvestiga España), which analyzed the registry results and compared them with those from previous years in collaboration with members of the GTH-SECPCC and the ACI-SEC board. Discrepant or particularly noteworthy data were cleaned and verified with the responsible investigator.

Due to the methodological characteristics of the study and the fact that it is only an activity registry, both the approval from the ethics committee and the processing of informed consent were deemed unnecessary.

RESULTS

Resources and infrastructure

A total of 19 hospitals participated (3 less than in 2021), of which 15 belong to the public health care system and 4 to the private sector (table 1 of the supplementary data). In addition, the analysis incorporated data on adult congenital heart disease interventions from another 114 hospitals (15 more than in 2022) included in the 2023 Registry of Hemodynamics and Interventional Cardiology of the ACI-SEC (table 2 of the supplementary data).

Table 1. Number of interventional procedures and distribution by age groups

Table 2. Number of interventional catheterizations in patients > 18 years and distribution according to origin registry

A total of 38 cath labs performing interventional procedures in congenital heart disease were registered, 7 of which (18.4%) only deal with pediatric procedures; 9 are equipped with biplane systems and 8 implement rotational angiography. The median number of monthly days dedicated to congenital heart disease interventions per hospital was 7 (range, 3–18) vs 6 days in 2022. Thirteen centers (68.4%) provide 24-hour emergency interventional care, including services for pediatric patients.

Regarding medical staff, a total of 68 interventional cardiologists were dedicated to this activity, including 36 (52.9%) specialized in adult interventions and 32 (47.1%) in pediatric interventions.

Diagnostic procedures

A total of 1127 diagnostic studies were recorded, representing a 1.2% decrease compared with the previous year. The age distribution was as follows: 27 (2.4%) were performed in infants younger than 1 month, 95 (8.4%) in children aged 1 month to 1 year, 548 (48.6%) in patients aged 1–18 years, and 457 (40.6%) in patients older than 18 years.

A total of 70 procedures (6.5%) were categorized as emergencies. Regarding morbidity, there were 5 cases (0.4%) of severe complications: 2 vascular events, 1 arrhythmia with severe hemodynamic instability and cardiorespiratory arrest, 1 anaphylactic reaction, and 1 neurological event. There were no procedural deaths.

Interventional procedures

Reported activity increased by 53.7% compared with the previous year for a total of 3856 therapeutic catheterizations registered and categorized into 15 different groups. Their case distribution and age breakdown are shown in table 1. Of the 2439 procedures (63.2%) performed in patients older than 18 years, 2016 (82.7%) came from data added from the Registry of Hemodynamics and Interventional Cardiology of the ACI-SEC (table 2).

A total of 195 cardiac catheterizations were categorized as emergencies (10.9% of all procedures for which this information was available). The volume of interventional procedures per center was as follows: 7 hospitals (36.8%) recorded > 150 cardiac catheterizations; 3 hospitals (15.7%) between 75 and 150 procedures, and 9 (47.3%) < 75 procedures. The overall effectiveness reported for the different interventional techniques was 97.9%, with most procedures showing success rates exceeding 95% (table 3).

Table 3. Summary of reported efficacy of interventional procedures

Transcatheter valvuloplasties

A total of 80 aortic valvuloplasties were reported for congenital aortic stenosis, reflecting a 19.4% increase from 2022. Two of these procedures were performed in fetuses. Sixty cases (75%) involved patients younger than 18 years, and in 77% of procedures, the dilation was performed on native, previously untreated valves.

A total of 166 pulmonary valvuloplasties were performed, marking a 20% increase from 2022. This included 2 fetal catheterizations. Technical details were available for 143 cases, of which 115 (80.4%) involved native valves—15 of them (10%) imperforate. In 8 cases (5.5%), the procedure was combined with ductal stent implantation.

Mitral valvuloplasty was reported in 19 cases, only 1 of which was performed in a patient younger than 18 years.

Transcatheter angioplasties

A total of 108 right ventricular outflow tract dilations were recorded, which is a 20% decrease compared with 2022. Technical and anatomical data were reported for 89 procedures (82.4%); 55% involved angioplasty of native tracts, and the remaining 45%, angioplasty of surgical conduits. Conventional balloon dilation was performed in 52% of cases, and stent implantation in 48%.

Pulmonary branch angioplasty accounted for 196 procedures (a 16% decrease compared with 2022). In 95% of cases, proximal branches were dilated, while peripheral arteries (lobar-segmental) accounted for the remaining 5%. Stent implantation was used in 52% of catheterizations, conventional balloon dilation in 46%, and cutting balloon dilation in 2%.

A total of 221 aortic coarctation procedures were reported (a 75% increase compared with 2022). Anatomical data were available for 139 procedures (62.8%), most of which (64%) were reinterventions. The site of dilation was the aortic arch/isthmus in all but 5 cases (4 abdominal aorta dilations and 1 ascending aorta dilation). The technique used included conventional balloon dilation in 28%, uncovered stent implantation in 26%, covered stent implantation in 32%, and repeat dilation of a previously implanted stent in 12%.

Finally, a total of 131 catheterizations were reported under the “other angioplasties” category, representing a 31% increase compared with the previous year, among them, 29 ductus arteriosus dilations and 7 surgical fistula dilations. Stent implantation was associated with 72% of these procedures.

Shunt closures and other occlusion procedures

A total of 1498 patent foramen ovale closures were reported (a 112% increase compared with 2022), of which 1337 (90%) came from the Registry of Hemodynamics and Interventional Cardiology of the ACI-SEC (table 2). Adults accounted for 99.1% of all patients undergoing this procedure.

Atrial septal defect (ASD) closures saw a substantial increase, reaching 700 procedures. This represents a 62% rise compared with 2022. Among the defects classified, 67% were complex, with the rest being simple. Guidance data were provided in 351 (50.1%) cases: transesophageal echocardiography was used in 82% procedures, angiographic balloon sizing in 15.9%, and intracardiac echocardiography in 1.9%.

A total of 344 ductus arteriosus closures were performed. Patients aged 1–18 years accounted for 65.6% of these procedures, while premature infants represented 4.7%, with 15 cases reported (compared with 9.2% in 2022). Anterograde venous access was used in 63% of closures. Occluder devices were used in 81% of cases and controlled-release coils in the remaining ones.

There were 107 ventricular septal defect (VSD) closures, a 181% increase compared with the previous year. Anatomical substrate data were provided in 73 cases (68.2%): 54 (73.9%) perimembranous, 14 (19.1%) muscular, and 5 (6.8%) postoperative. Occluder devices were used in 93% of cases, and coils in the remaining ones. Two devices were implanted via hybrid approach and the rest (97.2%) via transcatheter approach.

A total of 85 cardiac catheterizations were included under the “other occlusion procedures” category. Closure of systemic-to-pulmonary collaterals accounted for 70% of these cases, and venous collateral occlusion for 14%. The most widely used materials were occluder devices (46.3%), followed by coils (34.1%) and particles either alone or in combination with other materials (18.2%).

Atrial septostomy

A total of 64 procedures were reported (an 11.1% decrease compared with the previous year). Regarding imaging support, echocardiography was used in 16.5% of cases, fluoroscopy in 26.1%, and a combination of both imaging modalities in 57.4%. Fifty-two (81%) of these were balloon atrial septostomies (Rashkind). Additionally, 5 procedures were performed with radiofrequency septal perforation, 9 with needle perforation, and 8 with septal stent implantation.

Transcatheter valve implantation

A total of 93 procedures were reported, of which 51 (54.8%) were performed in adults. All procedures were performed via transcatheter approach, except for 1 case which required hybrid implantation. The pulmonary position was the most widely used (95.6%), with 4 valves being implanted in the tricuspid position. The anatomical substrates for pulmonary valve implantation were the native right ventricular outflow tract in 54 cases (58%), surgical conduits in 23 cases (24.7%), and valve-in-valve procedures in 16 cases (17.2%).

Complications

Morbidity and mortality data were available for 3738 interventional procedures. A total of 40 serious adverse events were reported (table 4), including 2 deaths, resulting in a major complication rate of 1.07% and a mortality rate of 0.05%. The categories with the highest morbidity were mitral and aortic valvuloplasties (5.2% and 4.8%, respectively), atrial septostomy (4.6%), and VSD closure (3.4%). The most frequent complications were device embolizations, reported in 12 cases (30% of all complications): 8 in ASD and foramen ovale closures, 2 in ductus arteriosus closures, 1 stent in vascular angioplasty, and 1 in pulmonary valve implantation. In 3 ASD closures and 1 valve implantation, surgical retrieval of the embolized prosthesis was required; the rest were resolved via transcatheter approach. A total of 8 serious arrhythmia events occurred, including 1 cardiac arrest which required extracorporeal membrane oxygenation. Furthermore, there were 7 cases of vascular complications.

Table 4. Distribution of major complications and reported deaths across different interventional procedures

DISCUSSION

The most significant finding of this report is the considerable increase in recorded interventional procedures (3856), representing a 53.7% rise compared with 2022. The most notable increases were observed in VSD closures (181%), patent foramen ovale closures (112%), and interventions for aortic coarctation (75%). Another remarkable observation is that 62% (55% in 2022 and 31% in 2021) of all reported procedures were performed in patients older than 18 years, confirming the growing volume of procedures in this age group. This trend likely reflects the broader expansion of structural heart procedures in Spain.7 Of note, as in the previous report, data from 114 hospitals (99 in 2022) reporting adult congenital heart disease procedures to the 2023 Registry of Hemodynamics and Interventional Cardiology of the ACI-SEC were included in the analysis. Comparisons with prior reports must be interpreted considering this methodology.

Nonetheless, most interventional categories recorded (10 out of 15) still concentrate the largest number of patients in the pediatric population (table 1). Additionally, fetal intervention activity remains minimal in Spain, with only 4 cases reported (2 aortic and 2 pulmonary valvuloplasties), despite evidence supporting their value and efficacy across multiple prenatal scenarios.8

Reported success rates for the different interventional techniques yield an overall success rate of 98.3% (97.6% in 2022) and a mortality rate of 0.05% (0.2% in 2021). Although the voluntary and unaudited design of the registry could limit the strength of these findings, they remain the best reported to date and are consistent with most international studies.9,10 Similarly, the 1% rate of serious adverse events is the lowest recorded so far (1.4% in 2022). Device embolization remains the most frequent complication, accounting for 30% of all cases by arrhythmias (20%) and vascular complications (17.5%) in 2023. Recent national studies have validated the usefulness of specific methodologies to assess expected risk of complications across different techniques and various clinical scenarios.11 Their application could enhance the quality of information produced by these results and is expected to be incorporated in coming years.

Regarding valvuloplasties, the most notable development was the reporting of 19 mitral valve dilations—a category that had been nearly absent from previous reports. This technique, in use for more than 40 years in adult interventional cardiology, has significantly evolved with the integration of 3D imaging modalities for patient selection and procedural guidance.12 Aortic and pulmonary valvuloplasties continue to rise (19% and 20%, respectively), mostly in the pediatric setting, reaffirming their role as first-line options for congenital valve stenosis in our environment. Of particular note within pulmonary valvuloplasties is the treatment of 15 cases of imperforate pulmonary valves. A transjugular approach might simplify such procedures, as recently demonstrated by a national group.13

In the angioplasty category, the most striking observation is the rise in reported cases of aortic coarctation (75% increase compared with 2022); 50.2% of the 221 cases were reported in patients older than 18 years. Dilations of the aortic arch and isthmus continue to account for most cases, with covered stent implantation increasingly established as the preferred approach in this anatomical setting. The use of this technique in pediatric patients is also on the rise, driven by the availability of covered stents with lower delivery profiles, now applied to other congenital anatomical contexts too.14

Interatrial septal defect closures (patent foramen ovale and ASD) are the highest-volume procedures in the registry, accounting for 57% of all cardiac catheterizations (38%, patent foramen ovale closure; 19%, ASD closure). Inclusion of patent foramen ovale closure in this registry, and its categorization as a congenital heart disease, is essential to maintain consistent criteria and comparability with previous reports, especially given its massive growth in adult interventional procedures.7 While double-disc devices remain the most widely used with the largest cumulative experience, the emergence of suture-based devices has expanded transcatheter treatment options, offering an attractive alternative in selected patients.15 Furthermore, ASD closure experienced significant growth this year (62% more compared with 2022). Up to 33% of cases with anatomical data were categorized as complex, which along with an excellent safety (1.5% complication rate) and efficacy (97.5%) profile confirm the maturity of this technique in Spain.

Ductus arteriosus closure remains a predominantly pediatric technique (87% of cases in patients younger than 18 years). However, a significant decline was reported in premature neonates, who represented only 4.7% of cases this year (9.4% in 2022). This slowdown in adopting transcatheter options for neonates contrasts with continued international studies validating it over surgery.16

VSD closure showed the greatest growth among procedures (181%), with notable progress across all age groups. Safety and efficacy results continue to improve, with a rate of major complications reduced to 3.4% (compared with 18% in 2021 and 5.2% in 2022) and a success rate of 96.2% (compared with 77.3% in 2021 and 96.7% in 2022). Occluder devices were used in 93% of cases. These results confirm a paradigm shift for the technique in Spain, driven by the introduction of new closure devices and technical modifications that facilitate the approach.17,18

Transcatheter aortic valve implantation increased slightly by 6.8%. Notably, 63.4% of these procedures were performed in patients older than 18 years, highlighting the growth of this technique in adult congenital heart diseases. As in previous reports, most were pulmonary valve implantations, with only 4 implantations being performed in the tricuspid position (2 in 2022). A major innovation in Spain is the availability of new self-expandable valve designs that broaden anatomical applicability—especially relevant for the 58% of patients with a native pulmonary tract.19,20

Limitations

The design of the registry (retrospective, voluntary, unaudited) may weaken the robustness of its findings. Expanding the information collected on certain key techniques would improve report quality and should be considered in future editions.

CONCLUSIONS

The significant increase in the volume of interventional procedures recorded compared with previous years, along with the rise in the number of participant centers, is the primary finding of this report. This growth is accompanied by continued improvements in the safety and efficacy data of most techniques. The most prominent increases were observed in VSD closures, interatrial shunt closures, and aortic coarctation procedures. The data obtained offer a realistic representation of interventional activity in congenital heart disease across all age groups in Spain.

An increase in participant centers and continued registry updates will enhance the quality and reliability of the information generated, reinforcing the relevance and usefulness of the registry.

FUNDING

None declared.

ETHICAL CONSIDERATIONS

Due to the methodological characteristics of the study and its design as a procedural activity registry, approval from the ethics committee or processing of informed consent were both deemed unnecessary. The nature of the work precludes the consideration of sex and gender variables, and therefore, SAGER guidelines were not followed.

STATEMENT ON THE USE OF ARTIFICIAL INTELLIGENCE

No artificial intelligence tools were used.

AUTHORS’ CONTRIBUTIONS

All authors made substantial contributions to data collection, and to the critical review and approval of the final version of the manuscript. F. Ballesteros Tejerizo and F. Coserría Sánchez drafted the manuscript.

CONFLICTS OF INTEREST

S. Ojeda Pineda is an Associate Editor of REC: Interventional Cardiology, and R. Sanz-Ruiz is a Section Editor of REC: Interventional Cardiology. In both cases, the journal’s editorial procedure to ensure impartial handling of the manuscript has been followed. The remaining authors declared no conflicts of interest whatsoever.

ACKNOWLEDGMENTS

To all the members of the Interventional Working Group of the SECPCC and the ACI-SEC who collaborated on this project from the beginning, and especially to the successive boards of directors of the ACI-SEC, which have consistently provided decisive support for the creation, growth, and consolidation of the Spanish Cardiac Catheterization in Congenital Heart Diseases Registry.

SUPPLEMENTARY DATA

REFERENCES

INTRODUCTION

Drug-eluting stents (DES) remain the standard of treatment for patients undergoing percutaneous coronary intervention (PCI).1,2 However, DES are associated with a gradually and permanent increased risk of adverse events, particularly due to late stent thrombosis and in-stent restenosis, with a 2% incidence rate per year with no plateau observed.1 This risk is even higher when complex and long lesions are treated.3 In recent years, drug-coated balloons (DCB) have emerged as a potential alternative treatment option to DES. Following adequate lesion preparation, unlike traditional stents, DCBs can release an antiproliferative drug into the vessel wall without leaving behind a permanent metal scaffold. Notably, permanent scaffolding can distort and constrain the coronary vessel, thus impairing vasomotion and adaptive remodelling, while also promoting chronic inflammation.4 DCB-PCI is a well-established treatment for in-stent restenosis and small-vessel coronary artery disease (CAD).5,6 However, its role in de novo large vessel CAD remains controversial. In a recent randomized clinical trial (RCT) with patients undergoing de novo CAD revascularization, a strategy of DCB-PCI did not achieve non-inferiority vs DES in terms of device-oriented composite endpoint driven by higher rates of target lesion revascularization (TLR).7 Contrary to prior published research, our findings did not support similar clinical outcomes for DCB vs DES in patients with de novo large vessel CAD.8,9 A recent meta-analysis of 15 studies compared DCB-PCI or hybrid angioplasty vs DES-PCI in patients with vessels > 2.75 mm in diameter showing no significant differences in the clinical endpoints of TLR, cardiac death, and MI.10 However, 14 of the 15 included studies were non-RCT, and the recent previously reported RCT was not included. Nevertheless, individual non-inferiority studies often lack the statistical power needed to definitively compare these technologies, underscoring the need for a systematic appraisal of treatment effects and evidence quality. Therefore, we conducted a systematic review and meta-analysis of available RCT to provide a comprehensive and quantitative assessment of evidence on the efficacy of DCB vs the current-generation DES in de novo large vessel CAD in terms of adverse events at longest available follow-up.

METHODS

Search strategy and selection criteria

We conducted a meta-analysis of RCT according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2009 guidelines.11 Two reviewers independently identified the relevant studies through an electronic search across the MEDLINE and Embase databases (from inception to October 2024). In addition, we employed backward snowballing (eg, reference review from identified articles and pertinent reviews). No language, publication date or publication status restrictions were imposed. This study is registered with PROSPERO and the search strategy is available in the supplementary data.

Study selection

Two reviewers independently assessed trial eligibility based on titles, abstracts, and full-text reports. Discrepancies in study selection were discussed and resolved with a third investigator. Eligible studies needed to meet the following pre-specified criteria: a) RCT comparing PCI with DCB and PCI with DES; b) study population including patients with de novo large vessel CAD (eg, defined as vessel diameter ≥ 2.5 mm);12 c) availability of clinical outcome data (without restriction as to follow-up time). Exclusion criteria were a) lack of a randomized design; b) studies including patients undergoing treatment for in-stent restenosis; c) studies including patients with de novo small vessel CAD; d) lack of any clinical outcome data.

A reference vessel diameter ≥ 2.5 mm was established as the cut-off value to define large vessel based on a recent proposed standardized definition.12

Data extraction

Three investigators (J. Llau García, S. Huélamo Montoro and J. A. Sorolla Romero) independently assessed studies for possible inclusion, with the senior investigator (J. Sanz-Sánchez) resolving discrepancies. Non-relevant articles were excluded based on title and abstract. The same investigators independently extracted data on study design, measurements, patient characteristics, and outcomes using a standardized data-extraction form. Data extraction conflicts were discussed and resolved with the senior investigator.

Data on authors, year of publication, inclusion and exclusion criteria, sample size, patients’ baseline patients, endpoint definitions, effect estimates, and follow-up time were collected.

Endpoints

The prespecified primary endpoint was TLR. Secondary clinical endpoints were all-cause mortality, cardiac death, myocardial infarction (MI), target vessel revascularization (TVR) and major adverse cardiovascular events (MACE). Secondary angiographic endpoints were minimum lumen diameter (MLD) and late lumen loss (LLL). Each endpoint was assessed according to the definitions reported in the original study protocols, as summarized in table 1 of the supplementary data. All the endpoints were assessed at the maximum follow-up available.

Table 1. Main features of included studies

Risk of bias

The risk of bias in each study was assessed using the revised Cochrane risk of bias tool (RoB 2.0).11 Three investigators (J. Llau García, S. Huélamo Montoro and J. A. Sorolla Romero) independently assessed 5 domains of bias in RCT: a) randomization process, b) deviations from intended interventions, c) missing outcome data, d) outcome measurement, and e) selection of reported results (table 2 of the supplementary data).

Table 2. Baseline clinical characteristics of included patients

Statistical analysis

Odds ratios (OR) and 95% confidence intervals (95%CI) were calculated using the DerSimonian and Laird random-effects model, with the estimate of heterogeneity being obtained from the Mantel-Haenszel method. The presence of heterogeneity among studies was evaluated with the Cochran Q chi-square test, with P ≤ .10 being considered of statistical significance, and using the I² test to evaluate inconsistency. A value of 0% indicates no observed heterogeneity, and values of ≤ 25%, ≤ 50%, > 50% indicate low, moderate, and high heterogeneity, respectively. Prediction intervals (95%) in addition to conventional 95%CI around ORs were calculated to assess residual uncertainty. Publication bias and the small study effect were assessed for all outcomes, using funnel plots. The presence of publication bias was investigated using Harbord and Egger tests and visual estimation with funnel plots. We performed a sensitivity analysis by removing one study at a time to confirm that the findings, when compared with DES, were not driven by any single study. To account for different lengths of follow-up across studies, another sensitivity analysis was performed using the Poisson regression model with random intervention effects to calculate inverse-variance weighted averages of study-specific log stratified incidence rate ratios (IRRs). Results were displayed as IRRs, which are exponential ratios of the regression model. Additionally, random-effect meta-regression analyses were performed to assess the impact of the following variables on treatment effect with respect to the primary endpoint: eg, percentage of patients with acute coronary syndrome (ACS), percentage of patients with diabetes mellitus, mean reference vessel diameter and follow-up duration. The statistical level of significance was 2-tailed P < .05. Stata version 18.0 (StataCorp LP, College Station, United States), was used for statistical analyses.

RESULTS

Search results

Figure 1 illustrates the PRISMA study search and selection process. A total of 7 RCT were identified and included in this analysis. The main features of included studies are shown in table 1.

Figure 1. Flow diagram of the search for studies included in the meta-analysis according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Statement.

All studies had a non-inferiority design. A clinical primary endpoint was selected in 1 study,7 and an invasive functional endpoint was selected in another trial,9 while angiographic primary endpoints were prespecified in the remaining studies.8,13-16 The mean clinical and angiographic follow-up were 21.5 months and 8.9 months respectively. A total of 4 studies were conducted in the context of ACS9,14-17 and 1 study in the context of chronic coronary syndrome (CCS).13 Finally, 2 studies enrolled both ACS and CCS patients.7,8 A total of 3 trials enrolled patients treated with second-generation DES (Firebird 2.0 [Microport, China], Xience Xpedition [Abbott Vascular, United States], Orsiro [Biotronik, Germany]),7,9,13 and 2 studies enrolled patients treated with third-generation DES (Biomine [Meril Life Sciences, India], Cordimax [Rientech, China]).14,15 One trial enrolled patients treated with second and third-generation DES (Xience Xpedition [Abbott Vascular, United States], Resolute Integrity, [Medtronic, United States], Firehawk, [MicroPort, China]).8 All studies included patients who underwent paclitaxel-DCB-PCI ([Pantera Lux, Biotronik, Germany],9,14 [SeQuent Please, B Braun, Germany],7,8,13,15 [Bingo DCB, Yinyi Biotech,China]),16 and none with sirolimus-DCB-PCI.

Baseline characteristics

A total of 2961 patients were included, 1476 of whom received DCB and 1485, DES for de novo large vessel CAD. The patients main baseline characteristics are shown in table 2.

Publication bias and asymmetry

Funnel-plot distributions of the pre-specified outcomes indicate absence of publication bias for all the outcomes (figures 1-8 of the supplementary data).

Risk of bias assessment

Table 2 of the supplementary data illustrates the results of the risk of bias assessment with the RoB 2.0 tool. One trial was considered at low overall risk of bias,7 5 raised some concerns8,9,13,14,16 and 1 presented a high overall risk of bias.15

Outcomes

Clinical outcomes

DCB use compared with DES was associated with a similar risk of TLR (OR, 1.21; 95%CI, 0.44-3.30; I² = 48%), all-cause mortality (OR, 1.56; 95%CI, 0.94- 2.57; I² = 0%), cardiac death (OR, 1.65; 95%CI, 0.90-3.05; I² = 0%), MI (OR, 0.97; 95%CI, 0.58-1.61; I² = 0%) and MACE (OR, 1.19; 95%CI, 0.74-1.90; I² = 13.5%). However, DCB was associated with a higher risk of TVR (OR, 2.47; 95%CI, 1.52- 4.03; I² = 0%) (figure 2, figure 3 and figures 9-10 of the supplementary data).

Figure 2. Forest plot reporting trial-specific and summary ORs with 95%CIs for the endpoint of A) target lesion revascularization; B) all-cause mortality; C) myocardial infarction; D) MACE. 95%CI, 95% confidence interval; DCB, drug-coated balloon; DES, drug-eluting stents; MACE, major adverse cardiovascular events; OR, odds ratio. References: REC-Cagefree I.,7 Nishiyama et al.,13 Xue Yu et al.,8 REVELATION,9 Hao et al.,16 Wang et al.,14 and Gobic et al.15

Figure 3. Central Illustration. DCB, drug-coated balloon; DES, drug-eluting stent; RCT, randomized clinical trial; TVR, target vessel revascularization. References: REC-Cagefree I.,7 Nishiyama et al.,13 Xue Yu et al.,8 REVELATION,9 Hao et al.,16 and Wang et al.14

Angiographic outcomes

Compared with DES, DCB use yielded significant smaller MLD (SMD, −0.36; 95%CI, −0.56 to −0.15; I² = 34.5%) and similar risk of LLL (SMD, −0.35; 95%CI, −0.74 to 0.04; I² = 81.4%) at follow-up (figure 4).

Figure 4. Forest plot reporting trial-specific and summary ORs with 95%CIs for the endpoint of A: minimum lumen diameter, and B: late-lumen loss. 95%CI, 95% confidence interval; DCB, drug-coated balloon; DES, drug-eluting stents; MACE, major adverse cardiovascular events; MLD, minimum lumen diameter; SMD, standardized mean difference; OR, odds ratio. References: Nishiyama et al.,13 Xue Yu et al.,8 REVELATION,9 Gobic et al.,15 Hao et al.,16 and Wang et al.14.

Prediction intervals were consistent with CI for all the outcomes except for TVR and MLD, which included the value of no difference.

Sensitivity analysis

A leave-one-out pooled analysis by iteratively removing one study at a time was performed for all endpoints. Treatment effects were consistent with the main analysis for TLR, all-cause mortality, cardiac death, MI and MLD. The risk of TVR was no longer significantly higher among patients undergoing DCB when removing the CAGEFREE I trial,7 and the risk of LLL was significantly lower among patients undergoing DCB-PCI when removing the REVELATION trial.9 However, an increased risk of MACE was observed among patients undergoing DCB-PCI when removing the study by Xue Yu et al.18 (tables 3-10 of the supplementary data). A sensitivity analysis using estimated IRRs was performed to account for varying follow-up lengths, confirming that our main analysis findings remained unchanged (table 11 of the supplementary data).

Random effect meta-regression analysis found no significant impact of the proportion of patients presenting with ACS (P = .882), diabetes mellitus (P = .641), mean reference vessel diameter (P = .985) and follow-up duration (P = .951) on treatment effect with respect to the primary endpoint.

DISCUSSION

This meta-analysis provides a comprehensive and updated quantitative analysis of available evidence on the comparison of DCB vs DES in de novo large vessel CAD, including data from 2961 patients enrolled in 7 RCT. The main findings of the study are:

a) The use of DCB was associated with a similar risk of clinical events vs DES except for TVR. However, data for this outcome was only available in 3 of the 7 included studies and the increased risk in patients undergoing DCB-PCI was not significant when the CAGEFREE I trial was removed. In addition, prediction intervals were not consistent with the CI. Therefore, the results of this outcome should be interpreted with caution.

b) The effect of DCB on the risk of TLR was not affected by the proportion of patients presenting with ACS or diabetes, as well as the mean reference vessel diameter or follow-up duration as assessed by meta-regression analysis.

c) DCB was associated with lower MLD at angiographic follow-up, but with similar LLL vs DES.

DES are the standard of treatment for patients undergoing PCI. However, complications such as stent thrombosis and in-stent restenosis still occur with rates estimated at 0.7-1% and 5-10% at the 10-year follow-up respectively.19,20 Therefore, in recent years there has been a growing concern for developing strategies to reduce stent-related adverse events. In this context, DCBs have emerged as a potential treatment alternative based on a “leaving nothing behind” strategy. Nevertheless, data of patients presenting with de novo large CAD is scarce and conflicting. The CAGEFREE I is the only available clinically powered RCT that included 2272 patients undergoing de novo non-complex CAD revascularization across 40 centers in China. A strategy of DCB-PCI did not achieve non-inferiority vs DES in terms of device-oriented composite endpoint driven by higher rates of TLR in the DCB-PCI group (3.1% vs 1.2%, P = .002). On the other hand, in single-center RCT conducted by Nishiyama et al. with 60 patients with CCS undergoing elective PCI a trend toward lower rates of TLR in the DCB-PCI group (0% vs 6.1%, P = .193) was shown at the 8-mont follow-up.13 Similarly, in a RCT including 170 patients undergoing PCI for de novo large CAD lower rates of TLR at the 12-month follow-up were found in patients undergoing DCB-PCI (1.6% vs 3.4%, P = .306).14 In our analysis when pooling data from all available RCT, the risk of TLR was similar among patients undergoing DCB-PCI or DES-PCI. Notably, since this result was obtained with a moderate heterogeneity (I² ≈ 50%), it should be interpreted with caution regarding its general applicability. These findings remained unvaried at the leave-one-out analysis. In addition, prediction intervals were consistent with CI around ORs showing lack of residual uncertainty. Previous studies have shown that in-stent restenosis after DES is not a benign phenomenon, presenting as an ACS in about 70% of the cases, with 5-10% of these resulting in MI.21 We could speculate that the lack of permanent scaffold with DCB vs DES may predispose to a less aggressive pattern of restenosis and not increase the risk of thrombotic vessel closure beyond 3 months when vessel healing after DCB-PCI has occurred.22

Notably, 5 of the 7 studies included in this meta-analysis enrolled patients presenting with ACS. A total of 34% of the patients included in the CAGEFREE study presented with ACS, with 16% being STEMI cases.7 Four other studies only included STEMI patients.7,9,14-16 Although the performance of DCB in the STEMI scenario is unknown, its use in clinical practice is increasing.23 Culprit lesion plaques in STEMI patients are usually soft and adequate plaque modification can be easily achieved through DCB-PCI (< 30% residual stenosis and low grade of dissection).23 Moreover, the ruptured lipid rich plaque can potentially be an ideal reservoir for effective paclitaxel uptake.24 On the other hand, DCBs carry specific risks for STEMI patients, such as acute recoil and culprit lesion closure, because they don’t provide vessel scaffolding.

In our study, the proportion of patients presenting with ACS had no impact on treatment effects on the meta-regression analysis. Nevertheless, further RCT with adequate sample size are needed to obtain more solid evidence in this field. Of note, complex lesions (eg, severe calcification and bifurcations with planned two-stent technique) were excluded from the studies that included patients presenting with CCS.7,8 Therefore, our findings might not be generalized to this population.

The better angiographic surrogate outcomes with DES-PCI vs DCB-PCI found in our meta-analysis after pooling data from 6 studies can be explained by the absence of a metal scaffold to expand the vessel lumen and the acute recoil following balloon angioplasty. This justifies the lower MLD achieved after DCB-PCI vs DES-PCI. While our analysis did not show significant differences regarding LLL during follow-up, the value of LLL was lower among patients undergoing DCB-PCI when excluding the REVELATION trial.9,17 This study showed extremely low LLL in both DCB and DES groups vs other available evidence from RCT.15,16 The presence of positive vessel remodeling with a late lumen enlargement after the use of DCB evaluated by intracoronary imaging modalities has been evidenced in multiple studies, and seems to be associated with small vessel disease, fibrous and layered plaques and a post-PCI medial dissection arc > 90°.25,26,27 However, evidence of this phenomenon in patients with large vessel CAD is less known.22 It should, therefore, be noted that all studies in this meta-analysis used paclitaxel-DCB. While the evidence comparing sirolimus and paclitaxel-DCB is scarce, 2 recent RCT have shown better angiographic results with the lipophilic component. In the first one, with 121 patients with the novo small vessel CAD, sirolimus-DCB failed to achieve non-inferiority for net-lumen gain at 6 months.28 In the second study, with 70 patients, the 2 devices showed similar results of LLL at 6 months, although patients treated with paclitaxel-DCB had more frequent late luminal enlargement.29 Due to the small sample size and although there is not enough evidence to evaluate differences across clinical endpoints, we cannot assume that there is a class effect across all DCBs. There are larger ongoing RCT to evaluate the outcomes of sirolimus DCB vs DES in large vessels that will provide evidence in this field.30,31

Limitations

The results of our investigation should be interpreted in light of some limitations. First, this is a study-level meta-analysis providing average treatment effects. The lack of patient-level data from the included studies prevents us from assessing the impact of baseline clinical, angiographic and procedural characteristics on treatment effects. Second, minor differences in definition were present for some endpoints (eg, MACE), limiting the reliability of effect estimates. Third, one study which accounted for approximately 75% of all patients included did not included angiographic follow-up,7 thus limiting the evaluation of DCB and DES on angiographic outcomes. Fourth, the clinical follow-up varied from 6 to 24 months. Ideally, outcomes such as TLR should be compared at uniform follow-up across studies (eg, at 1 year), which was not consistently possible in the current analysis. Nonetheless, these differences in follow-up duration were accounted with the IRRs, as detailed in the Methods section. However, longer follow-ups are needed to establish the safety and efficacy profile of DCB vs DES throughout time. Fifth, the definition of large vessel is inconsistent across trials, which might be a source of bias. Finally, the limited number of studies and patients, and the small event rate for some endpoints, such as all-cause mortality may reduce the power for detecting significant differences across groups.

CONCLUSIONS

This meta-analysis provides the most updated quantitative evidence on the use of DCB vs DES for the treatment of de novo large vessel CAD in both CCS and ACS. DCB-PCI is associated with similar TLR and LLL at mid-term follow-up representing an appealing treatment option for patients with large vessel CAD.

FUNDING

None declared.

ETHICAL CONSIDERATIONS

Ethics approval was deemed unnecesary for this meta-analysis as all data were collected and synthesized from previous studies. Additionally, no informed consent was required as there were no patients involved in our work. The meta-analysis of RCT was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2009 guidelines. We confirm that sex/gender biases have been taken into consideration.

STATEMENT ON THE USE OF ARTIFICIAL INTELLIGENCE

No artificial intelligence has been used in the preparation of this article.

AUTHORS’ CONTRIBUTIONS

J. Llau Garcia, S. Huelamo Montoro and J. A. Sorolla Romero participated in literature research and study selection. J. A. Sorolla Romero, L. Novelli and J. Sanz Sánchez contributed to the conception, design, drafting and revision of the article. P. Rubio, JL Luis Díez Gil, L. Martínez-Dolz, IJ. Amat Santos, B. Cortese, F Alfonso, and Hector M. Garcia-Garcia contributed to the critical revision of the intellectual content of the article.

CONFLICTS OF INTEREST

F. Alfonso is an associate editor of REC: Interventional Cardiology; the journal’s editorial procedure to ensure impartial handling of the manuscript has been followed. The authors declared no relevant relationships with the contents of this paper.

SUPPLEMENTARY DATA

REFERENCES

INTRODUCTION

Transcatheter closure of ventricular septal defects (VSD) offers numerous advantages, including less trauma, faster recovery, and a reduced length of stay.1 However, this technique has complications, such as device embolization, valve malfunction, and arrhythmias. One of the most concerning complications of transcatheter closure of perimembranous VSD is the development of complete atrioventricular block (CAVB).2 Although this complication is more likely to occur when an inappropriate device is selected, pinpointing the exact cause of the block can sometimes be challenging. Factors significantly contributing to CAVB include young age, low body weight, device malapposition due to septal aneurysm, selection of an excessively large device, and direct device compression. Despite its rarity, CAVB remains a severe complication associated with this procedure.3

The atrioventricular node is located at the posterior superior area of the membranous ventricular septum and branches into the left and right bundles at the lower posterior edge. This close anatomical relationship increases the risk of developing heart block during the transcatheter closure of perimembranous VSD.1,4,5 Left anterior fascicular block, a variant of left bundle branch block (LBBB), can result in ventricular asynchrony, which negatively impacts hemodynamics and left ventricular function.6

CAVB has been reported in 0-6.4% of cases after the transcatheter closure of VSD.7 Recent publications indicate that this rate is gradually declining. A systematic review by Yang et al. found that 107 of 4394 patients, 107 (2.4%) required permanent pacemaker implantation after the interventional closure of VSD, with a higher incidence rate being reported in young children.8 Additionally, Bergman et al. reported that CAVB was observed in 1 of 149 (0.7%) patients after the procedure involving various VSD devices at a 6-year follow-up.7

We evaluated a total of 180 patients, 42 of whom were under 10 kg, who underwent transcatheter closure of VSD in our center in the last 5 years, focusing on block development in young children. In this article we detail the treatment and follow-up of 1 patient who developed complete CAVB and 5 patients who developed LBBB.

METHODS

From January 2019 through December 2023, a total of 180 pediatric patients (42 of whom weighed less than 10 kg) underwent transcatheter closure of perimembranous ventricular septal defects (VSD) at our center.

The indications for closure included a left ventricular end-diastolic diameter Z score ≥ 2.0; Qp/Qs > 1.5, treatment-resistant heart failure, a cardiothoracic ratio ≥ 0.55 on chest radiography, and growth retardation unrelated to recurrent respiratory infections or malnutrition.

Patients with subaortic edge regurgitation, significant aortic regurgitation, ventricular outflow tract obstruction, mean pulmonary artery pressure > 20 mmHg, or associated surgical heart anomalies were excluded from the study.

The KONAR-MF VSD occluder (Lifetech, China) and Amplatzer Duct Occluder (ADO I and II, AGA Medical Corp., United States) devices were used in the procedures. The Konar MF was used more frequently due to its flexible design (Konar MF: 157, ADO I + ADO II: 23).

The device size was selected based on angiographic measurements, typically choosing a device 1–2 mm larger than the size of the left ventricular defect. In VSD with aneurysmal tissue, the left disc of the device was positioned inside the aneurysmal tissue.

All patients were continuously monitored with electrocardiography during the procedure and underwent serial electrocardiograms (ECG) and echocardiographic evaluations at the follow-up.

RESULTS

Heart block developed in 6 patients, all of whom weighed less than 10 kg: 1 CAVB and 5 LBBB.

Case 1

A 2-year-old female patient, weighing 9.9 kg (3^rd to 10^th percentile), was being followed by pediatric cardiology for a diagnosis of a VSD. She had a past medical history of failure to gain weight, growth retardation, and 2 hospitalizations due to lower respiratory tract infections. An echocardiogram revealed a perimembranous VSD, measuring 5 mm on the left ventricular (LV) side and 4 mm on the right ventricular (RV) one.

Due to the clinical and hemodynamic significance of the patient’s VSD, a decision was made to perform a transcatheter closure. Prior to the procedure, the patient was administered cefazolin (50 mg/kg) and heparin (100 U/kg). The VSD was successfully closed using a Lifetech Konar MFO 6-4 device via antegrade access while the patient remained under general anesthesia. There were no signs of conduction disturbances in the ECG performed intra- and postoperatively. An ECG performed on postoperative day 2 confirmed that the device was correctly positioned in the absence of residual shunt. The patient was prescribed a 6-month regimen of aspirin at a dosage of 3 mg/kg/day and was discharged without any complications. Three days after discharge, the patient exhibited cyanosis. An ECG revealed the presence of CAVB (figure 1).

Figure 1. Case 1: electrocardiography of complete atrioventricular block after transcatheter closure of ventricular septal defect.

Atropine was administered twice at a dose of 0.02 mg/kg. The intervention successfully raised the peak heart rate to 135 beats per minute, and the patient’s rhythm normalized to a junctional ectopic rhythm. However, as the CAVB persisted, a temporary transvenous pacemaker was implanted, and the patient was admitted to the pediatric intensive care unit under continuous follow-up. Dexamethasone was initiated at a dosage of 0.6 mg/kg per day.

On hospitalization day 3, the patient’s ECG showed a return to sinus rhythm. After the temporary pacemaker was turned off, the patient underwent 24-hour Holter ECG monitoring. The Holter ECG showed a consistent sinus rhythm, meaning there was no evidence of CAVB or advanced second-degree block. On hospitalization day 5, the patient, whose ECG was still showing a consistent sinus rhythm, was discharged with a plan to complete a 10-day regimen of dexamethasone.

During the 3- and 6-month follow-up visits, the patient’s ECG continued to show a normal sinus rhythm without the need for medication.

Case 2

A 15-month-old male patient, weighing 8 kg (which is below the 3^rd percentile), presented with a VSD and a large patent ductus arteriosus who underwent transcatheter closure at 3.5 months of age due to symptoms of heart failure that remained unresponsive to optimal medical therapy. During follow-up, the patient showed signs of inadequate weight gain and fatigue during feeding. Due to these clinical and hemodynamic indicators, a decision was made to close the VSD at 15 months of age. Echocardiography revealed a defect measuring 5 mm on the LV side and 4 mm on the RV side in the perimembranous region. The defect was closed using a transcatheter approach via retrograde access with a Lifetech Konar MFO 6-4 device.

Postoperative follow-up revealed the widening of the QRS complex. An ECG showed that the patient had developed a LBBB. As a result, the device was removed without being released. The patient then began dexamethasone at a dosage of 0.6 mg/kg per day.

By the end of week 1 of postoperative follow-up, the patient’s ECG showed a normal sinus rhythm with no evidence of LBBB.

Case 3

An 8-month-old patient, weighing 6.4 kg (below the 3^rd percentile), was monitored for a VSD measuring 5 mm on the LV side and 4.5 mm on the RV side in the perimembranous region. Due to poor weight gain and left ventricular enlargement on the echocardiography, transcatheter closure was performed.

A Lifetech Konar MFO 6-4 device was successfully implanted under general anesthesia without immediate complications. However, 3 hours later, the patient developed a LBBB on the ECG (figure 2). Although dexamethasone was started at 0.6 mg/kg/day, the LBBB persisted by day 4, and the patient was discharged.

Figure 2. Case 3: electrocardiography of left bundle branch block after transcatheter closure of ventricular septal defect closure.

During the 1-week follow-up, an incomplete LBBB was noted on the ECG. Dexamethasone treatment went on for another 2 weeks, and at the 1-month follow-up, the LBBB had resolved, indicating successful treatment.

Case 4

A 14-month-old female patient, weighing 8 kg (which falls within the 3^rd to 10^th percentile), was being monitored for a VSD. The ECG indicated a 6 mm perimembranous VSD. A decision was made to perform the transcatheter closure of the defect. The procedure was performed with a Lifetech Konar MFO 8-6 device via retrograde access in the absence of immediate complications.

However, after the procedure, an ECG showed the development of LBBB. The patient began dexamethasone at a dosage of 0.6 mg/kg/day. After discharge, she was closely monitored through frequent outpatient follow-up. Despite ongoing treatment, LBBB persisted, and echocardiography performed at the 1-week follow-up showed onset of aortic regurgitation. At the 3-week follow-up, the device was surgically removed and the VSD repaired. This decision was made because her echocardiography showed increased aortic regurgitation, and the lLBBB persisted on her ECG.

Case 5

A 12-month-old male patient, weighing 7 kg (below the 3^rd percentile), was admitted to the clinic with symptoms of growth retardation and evidence of left ventricular enlargement on echocardiography. The patient exhibited a perimembranous VSD measuring 6 mm on the LV side and 3.5 mm on the RV side. The defect was closed using a Lifetech Konar MFO 6-4 device, delivered through a transcatheter procedure, which went completed smoothly and without conduction disturbances being observed on the ECG at the follow-up. Echocardiography confirmed that the device had been implanted appropriately and in the absence of residual shunt. However, at the 4-year follow-up, LBBB was observed on the ECG. Since the left ventricular functions remained normal on echocardiography, the patient remained under close follow-up in the outpatient clinic without any additional treatment.

Case 6

A decision was made to perform a transcatheter closure of a VSD in an 11-month-old female patient who weighed 9 kg (falling within the 25^th to 50^th percentile). She had been on optimal medical therapy for heart failure and exhibited left ventricular dilatation on echocardiography. The defect measured 7 mm on the LV side and 4 mm on the RV side.

The procedure was performed via retrograde access using a Lifetech Konar MFO 7-5 device. After device implantation, QRS complex widening was observed on the monitor, and an ECG confirmed the development of LBBB. The device had to be removed without being released.

The patient began dexamethasone at a dose of 0.6 mg/kg/day. Four weeks after the procedure, the patient’s ECG showed a return to sinus rhythm in the absence of LBBB.

The patients’ demographic and clinical characteristics are shown in table 1.

Table 1. Demographic and clinical characteristics

DISCUSSION

Blocks that occur after transcatheter closure of perimembranous VSDs are primarily caused by the conduction bundle close proximity to the defect.9,10 The edge of the perimembranous VSD is located in an area of fibrous continuity between the atrioventricular valves, which forms the posteroinferior border. In this region, the atrioventricular conduction bundle leaves the central fibrous body and runs just subendocardial. This position makes it vulnerable to damage from devices used to close perimembranous VSDs.9

AVB due to direct mechanical compression of the atrioventricular node typically occur immediately after performing the procedure or 2 to 7 days after percutaneous closure. Later onset AVB may result from inflammation and fibrosis.2,9 CAVB are usually observed in the early postoperative period. In patients undergoing transcatheter closure, the timing of AVB formation can be unpredictable, with most cases being detected 2 to 7 days after the procedure.7,10 However, late-onset AVBs have been reported as late as 2 to 4 weeks or even 10 to 20 months after the procedure. The need for permanent pacemaker implantation is greater in younger patients.7 Although in our patient with complete AVB, symptoms developed 4 days after the procedure, there was no need for permanent pacemaker implantation.

After the perimembranous closure of VSD, bundle branch block is a more finding than CAVB. Right bundle branch block occurs more frequently than LBBB, likely because the right bundle branch is smaller and more prone to damage. While bundle branch blocks usually develop within 1 week after transcatheter closure, cases have been reported up to 3 years after the procedure.11 Most bundle branch blocks may resolve spontaneously or with steroid treatment, such as IV dexamethasone at 1 mg/kg/day or oral prednisone at 1-2 mg/kg/day.2,9 Close follow-up of patients is essential within the first 7 days after the procedure.10 LBBB has been reported to lead to abnormal left ventricular remodeling and heart failure.11

If CAVB occurs intraoperatively while crossing the defect, it is advisable to abandon the procedure. For postoperative CAVB, high doses of IV steroids followed by a 2-week regimen of oral steroids are recommended.9 The decision to remove the device is complex and depends on the patient’s symptoms, parental preference, and the experience of the clinic.9

If AVB resolves with steroid therapy, leaving the device in place is recommended. In symptomatic patients, a temporary pacemaker should be implanted, and response to steroid treatment should be monitored.9 In our patient with complete AVB, and in the 2 patients who developed postoperative LBBB, these blocks resolved after 2 weeks of steroid treatment, and sinus rhythm was restored. These patients have been closely monitored for any potential recurrence of the block.

For those patients who developed intraoperative bundle branch blocks, the devices were removed without release, as suggested in the literature.

In the patient who developed postoperative LBBB, which did not regress during follow-up, the device was surgically removed, and the VSD was repaired. The LBBB regressed with the removal of pressure on the left bundle branch.

Factors such as young age, low body weight, improper device positioning according to septal aneurysm and the choice of a large device have been identified as significant contributors to the development of conduction block.3 In our 5-year review, we observed that 5 of 180 cases of LBBB occurred in children weighing under 10 kg, which underscores the importance of age and body weight in the risk of developing LBBB.

To minimize the risk of a heart block, it is essential to prevent trauma and inflammation to the heart conduction tissue.4,7 This means an experienced operator should perform the procedure, using appropriately sized and flexible devices for the defect, and avoiding large carrier sheaths.7,9 The KONAR-MF VSD occluder, or KONAR-MFO, has become the primary choice in recent years for device selection due to its procedural flexibility, soft structure, and defect compatibility. We prefer to use KONAR-MFO in patients with low body weight and young age.3,12 While keeping septal aneurysmal tissue within the device during device implantation increases the risk of block, placing the left disc of the device inside the aneurysm may reduce the risk of block by removing it from the conduction system.13 Additionally, it is important to note that optimal medical therapy may be effective in cases without complete AVB basing the final treatment decision on the patient’s response.9

The reported rate of complete heart block after the surgical closure of VSD is < 2%. While the risk of CAVB (1-5%) in interventional closure of VSD raises concern, recent publications indicate a decreasing trend in the rates of CAVB.2,9,10 In our series, CAVB developed in only 1 patient (0.5%) and resolved with steroids after temporary transvenous pacing. Yang et al. (2012) reported that 8 of 228 patients (3.5%) developed postoperative LBBB.14 In a retrospective study of 2349 patients published in 2019, Wang et al. reported LBBB in 57 patients (2.4%) after the transcatheter closure of perimembranous VSD.11 In our center, LBBB developed in 5 of 180 transcatheter closures of VSD (2.7%), and the device was not implanted in 2 patients due to the development of intraoperative LBBB. Follow-up continues for our patient who developed late-onset LBBB.

Limitations

This study was conducted at a single center and retrospectively. Although patients were regularly monitored, longer follow-up periods are required, especially to detect conduction disturbance that may arise in the late period. Results may vary depending on the use of different devices or results obtained from different centers.

Considering these limitations, the findings should be interpreted with caution, particularly on the development of conduction block in low-birth-weight children. Further studies with larger sample groups, multicenter designs, and prospective follow-up data are required.

CONCLUSIONS

The risk of heart block in transcatheter procedures performed at experienced centers is lower than anticipated. Interventional closure of VSD has emerged as a viable alternative to surgery, providing benefits such as less trauma, faster recovery, and a reduced length of stay. With the arrival of newly developed devices, the risk of heart block in the transcatheter closure of VSD is steadily decreasing. Additionally, treatment often restores sinus rhythm in patients, and any heart block that may occur typically does not persist.

DATA AVAILABILITY

The raw data supporting the conclusions of this article will be made available by the authors upon request to any qualified researcher.

FUNDING

None declared.

ETHICAL CONSIDERATIONS

The study protocol was approved by the SBU Tepecik Training and Research Hospital ethics committee in full compliance with national rules and regulations and the ethical principles outlined in the revised Declaration of Helsinki (2008). Prior written informed consent and assent was obtained to participate in this study from each patient or caregiver. Furthermore, the authors confirm that sex and gender variables were considered in full compliance with the SAGER guidelines.

STATEMENT ON THE USE OF ARTIFICIAL INTELLIGENCE

No artificial intelligence technologies were utilized in the conception, data analysis, writing, or revision of this manuscript.

AUTHORS’ CONTRIBUTIONS

S. Oksuz and K. Yildiz designed the study protocol, analyzed the integrity of clinical data, and revised it. N. Narin and R. Aktas contributed to the conception and design, acquisition, and critically revised the manuscript, gave final approval, and agreed to be accountable for all aspects of work, ensuring integrity and accuracy. M.A. Atlan and S. Oksuz critically reviewed the article. R. Aktas and E. Gerceker contributed to the design, collected clinical data, and interpreted the results. C. Karadeniz provided editing and supervision. S. Oksuz took the lead in writing and reviewing the entire draft. All authors critically discussed the results and read and approved the final draft.

CONFLICTS OF INTEREST

None declared.

REFERENCES