Optical coherence tomography for the diagnosis and management of stent thrombosis

INTRODUCTION

In patients with ST-segment elevation myocardial infarction (STEMI), the treatment of choice is percutaneous coronary intervention (PCI) performed within the appropriate time window and by experienced operators. PCI has been shown to reduce mortality, reinfarction, and stroke compared with fibrinolysis.1 Among other factors, this benefit may be attributed to greater epicardial reperfusion and higher TIMI (Thrombolysis in Myocardial Infarction) flow grades and myocardial blush grades achieved with PCI in the culprit artery, all of which are known to influence survival.2,3

PCI have distinctive characteristics, including the presence of a high thrombus burden and performance in patients in a markedly thrombogenic state. To reduce the local thrombus burden in the culprit artery, manual thrombectomy (MT) has been widely used during PCI to reduce thrombus load, prevent distal embolization, and improve final myocardial perfusion.3,4 Despite its initial widespread adoption, routine use of MT in all patients undergoing PCI is no longer recommended, as randomized clinical trials have not demonstrated consistent clinical benefit.3-7

MT should be reserved for patients in whom it is most likely to provide benefit. To optimize the efficacy of PCI, the American8 and European9 clinical practice guidelines recommend MT in high-risk patients with a moderate-to-high thrombus burden who present with short ischemia times. Currently, however, there are no clearly defined criteria to precisely identify patients or thrombotic lesions that would derive the greatest benefit from MT.

The aim of our study was to evaluate the results of MT performed with the Hunter catheter (IHT–Iberhospitex SA, Barcelona, Spain), which has a high extraction capacity, in selected patients with STEMI undergoing PCI, and analyze the angiographic patterns associated with successful MT in our center.

METHODS

Study design and population

The Hunted registry is a single-center, observational, retrospective study. We included all patients diagnosed with STEMI who underwent PCI and in whom MT was performed using the Hunter thrombus aspiration catheter. This device was the first-choice catheter for MT in our center during the study period. The decision to perform MT was always left to the discretion of the operator performing the PCI, following homogeneous criteria among operators, subjectively based on angiographical evidence of a large angiographically visible thrombus. MT was not recommended in coronary vessels with a diameter < 2 mm or for the extraction of chronic thrombi or atherosclerotic plaques. The study period ranged from July 2020 through February 2022. Patients in whom MT was performed using a catheter other than the Hunter device were excluded.

The primary objective of the study was to evaluate the success of thrombus aspiration during PCI in patients with STEMI. Effective MT was defined as an angiographical reduction in thrombus burden, achieving a TIMI thrombus grade ≤ 2 (thrombus dimension < 50% of the vessel diameter). Moreover, angiographic factors associated with effective MT were assessed.

Secondary objectives included describing the clinical and angiographic characteristics of the patients and evaluating major adverse cardiovascular events (MACE) during hospitalization and follow-up.

MACE were defined as a composite endpoint of cardiovascular and noncardiovascular death, stroke, and acute myocardial infarction. Two follow-up time points were established at 1 and 2 years after PCI. Total ischemia time was defined as the interval, in minutes, between symptom onset and reperfusion, defined as passage of the intracoronary guidewire, in accordance with clinical practice guidelines.

Thrombus burden was graded according to the TIMI thrombus scale,10 which includes 5 grades: grade 1, possible thrombus; grade 2, thrombus dimension < 50% of the vessel diameter; grade 3, thrombus dimension 0.5 to 2.0 vessel diameters; grade 4, thrombus dimension > 2.0 vessel diameters; and grade 5, total vessel occlusion by thrombus. Grades 4 and 5 were considered high thrombus burden.

Successful PCI was defined as achievement of TIMI grade 3 flow with residual percent diameter stenosis < 20%, without device-related complications or intraoperative MACE. For study purposes, patients were categorized into 2 groups according to whether MT was effective or not. Furthermore, these groups were compared to identify angiographic parameters that could predict MT success.

Data for the variables included in the Hunted registry were collected using a dedicated electronic case report form. Retrospective angiographic analysis was performed exclusively by 3 operators. All measures were taken to ensure confidentiality and protection of patient health record information. The study protocol fully complied with international recommendations for clinical research outlined in the Declaration of Helsinki and was approved by the hospital Ethics and Research Committee.

Characteristics of the Hunter device and aspiration technique

The Hunter thrombus aspiration catheter is a 140 cm rapid-exchange aspiration catheter compatible with a 6-Fr guiding catheter. Its tip has a slightly conical, low-profile, atraumatic design. The effective distal aspiration area measures 0.95 mm², and it can aspirate up to 1.92 mL per second, one of the highest capacities available. The distal segment is coated with a hydrophilic surface to facilitate device navigability.

The standard thrombectomy technique used in our center consisted of advancing the Hunter catheter to a segment proximal to the culprit lesion. Aspiration was always initiated proximal to the culprit lesion. The catheter was then slowly advanced across the lesion under continuous aspiration to reach the distal segment, while continuous filling of the syringe was observed. If filling stopped, the catheter was withdrawn until aspiration resumed or removed completely if aspiration could not be reestablished. Continuous aspiration until removal from the guiding catheter was mandatory, as was thorough subsequent flushing of the guiding catheter, to prevent embolization of residual thrombotic material.

All retrieved material from the catheter and aspiration syringe was subsequently filtered using the filters provided with the device packaging. The procedure was repeated as many times as deemed necessary by the operator until the desired reduction in thrombus burden was achieved.

Statistical analysis

Qualitative variables are expressed as frequencies and percentages, and the quantitative ones as mean and standard deviation when normally distributed, and as median and interquartile range when distribution was nonnormal.

Qualitative variables were compared using the chi-square test, with odds ratios (OR) and 95% confidence intervals (95%CI) calculated. Quantitative ones were compared using the Student t test or nonparametric tests, as appropriate.

In univariate analysis, each variable was individually assessed for its association with effective MT. Variables showing a statistically significant association were included in the multivariate analysis. Multiple regression models were used to control for potential confounders and determine the independent effect of each variable on the outcome.

Survival curves were analyzed using the Kaplan–Meier method, and survival-related parameters were evaluated using Cox proportional hazards analysis. A 2-sided P value < .05 was considered statistically significant.

Statistical analyses were performed using STATA version 15 (StataCorp, United States).

RESULTS

During the study period, a total of 750 PCI were performed in patients diagnosed with STEMI. MT using the Hunter catheter was performed in 401 patients (53.47%). The clinical characteristics of the patients, STEMI features, and PCI are shown in table 1.

Table 1. Clinical characteristics of patients and procedural variables

Angiographic analysis of the culprit lesion and flow in the infarct-related artery is shown in table 2. Initial TIMI grade flow in the infarct-related artery, prior to intracoronary guidewire passage was 0 in 83% of cases. A high thrombus burden was observed in 87.5% of patients, corresponding to TIMI thrombus grades 4 or 5; 53.4% had grade 5 and 34.2% had grade 4. Only 50 patients (12.5%) had a TIMI grade < 4 flow, defined as a low thrombus burden.

Table 2. Angiographic analysis of patients treated with percutaneous coronary intervention and manual thrombectomy

According to the predefined criteria, effective MT was achieved in 327 of 401 patients (81.5%). Thrombectomy was considered ineffective in the 6 patients who had initial TIMI thrombus grades 1 and 2. Final post-PCI coronary flow was TIMI grade < 3 flow in 15 patients (3.74%). Overall, the PCI was successful in approximately 97% of cases. Device-related complications were recorded in 17 patients (4.24%): severe arrhythmias (ventricular fibrillation or ventricular tachycardia) occurring during reperfusion in 10 patients; severe no-reflow due to distal thrombus migration that could not be successfully treated in 4 patients; and coronary dissection after passage of the MT catheter, which was successfully treated with stenting in 3 patients. There were no cases of perioperative stroke due to migration of aspirated thrombus.

Comparative analysis of angiographic factors and ischemia time between patients with effective and noneffective MT is shown in table 3. MT was effective more frequently in patients with culprit coronary vessels > 3 mm in diameter (112 [34%] vs 12 mm [16%]; P < .001) and in those with a TIMI grade ≥ 4 thrombus burden (297 [91%] vs 54 [73%]; P < .001). Among the 61 patients with vessels < 2.5 mm, MT was ineffective in 36 (59.01%) vs 38 of 340 patients (11.2%) with vessels > 2.5 mm (P < .001). There were no statistically significant differences in ischemia time in relation to MT success.

Table 3. Comparison of angiographic and procedural characteristics between patients with effective and noneffective manual thrombectomy

The 1- and 2-year follow-up was completed in 100% of included patients. MACE occurred in 32 patients (7.98%) at 30 days, 36 patients (8.98%) at 1 year, and 40 patients (9.97%) at 2 years (table 4). Kaplan–Meier curves for event-free survival during follow-up and cardiovascular death according to effective vs noneffective MT are shown in figure 1 and figure 2, respectively. Individual components of MACE at 1 year are shown in figure 3.

Table 4. Incidence rate of the composite endpoint of major adverse cardiovascular events during follow-up

Figure 1. Kaplan–Meier curves for survival free from major adverse cardiovascular events (MACE). HR, hazard ratio.

Figure 2. Kaplan–Meier survival curves for cardiovascular death. HR, hazard ratio.

Figure 3. Individual components of major adverse cardiovascular events at the 1-year follow-up. AMI, acute myocardial infarction; CABG, coronary artery bypass grafting; CV, cardiovascular; TVR, target vessel revascularization.

DISCUSSION

In selected patients, MT using the Hunter catheter is a safe and effective strategy to reduce angiographically assessed thrombus burden during PCI.

Current clinical practice guidelines do not recommend routine MT during PCI but suggest considering it in patients with a high thrombus burden, based on individual assessment and operator experience.8,9 In our series, MT was performed in 53.5% of patients with STEMI treated with PCI, a higher proportion than reported in other countries.11-13 Only 12% of patients undergoing MT did not have a TIMI grade 4–5 thrombus burden on angiographic analysis.

In this selected population with high thrombus burden, 88% had TIMI grade ≥ 4 flow, which is similar to the 79% observed in the TOTAL trial,14 with favorable results in both cases. In contrast, only 33% of patients from the TASTE trial7 had a high thrombus burden.

A high thrombus burden appears to be a key determinant of achieving effective MT and may be associated with better clinical outcomes. In our cohort, although effective MT was achieved in 82% of cases, it was not significantly correlated with final procedural success (97% vs 95%; P = .67), likely due to the small sample size of the 2 groups. MT efficacy was higher in vessels with high thrombus burdens (TIMI grade ≥ 4 flow; P < .001), which is consistent with a meta-analysis showing less cardiovascular death in patients with STEMI undergoing PCI with MT in the high thrombus burden subgroup vs PCI alone (2.5% vs 3.2%; hazard ratio [HR], 0.81; 95%CI, 0.65–0.98; P = .03).15 These findings reinforce the concept that appropriate patient selection is crucial to benefit from MT. Moreover, the same meta-analysis reported a higher risk of stroke (0.9% vs 0.5% in the PCI-alone group),15 a complication that may be related to the TM technique used.

Strict adherence to proper technique is essential to minimize complications and maximize success. In our series, emphasis was placed on following a standardized and rigorous technique, as described in the Methods section, resulting in a low complication rate (4.3%). Many of these complications were not directly related to MT per se but rather to reperfusion, such as ventricular arrhythmias. MT-related stroke is a potential complication; in the TOTAL trial,16 the stroke rate was 0.7%, twice that observed in the PCI-alone group, whereas in the large real-world SCAAR registry, there was no increase in the incidence rate of stroke across the groups,17 which are findings more consistent with our results and possibly related to differences in MT technique.

To identify angiographic predictors of MT success, we compared the characteristics of patients with effective and noneffective MT. The former had significantly larger culprit vessel diameters; 48% of patients with noneffective MT had vessels measuring 2.5 mm. Although MT is generally discouraged in vessels < 2 mm, larger vessels may harbor greater thrombus burden and thus derive greater benefit from MT with the Hunter catheter, which has demonstrated higher in vitro aspiration capacity compared with other devices. Therefore, vessel size is a critical differentiating factor: in vessels < 2.5 mm, MT was ineffective in 59% of cases, a significantly higher proportion than the 11.2% of noneffective MT observed in larger vessels. The other major difference between groups was thrombus burden, as effective MT was achieved more frequently in patients with higher thrombus loads (TIMI grade ≥ 3 flow).15 Furthermore, this factor represents a major difference among randomized clinical trials, in which the proportion of patients with high thrombus burden varied substantially.15 More complex lesions, such as American Heart Association/American College of Cardiology type C lesions, those associated with calcification in addition to thrombus, long lesions, and left circumflex artery lesions were associated with higher rates of noneffective MT. These differences should be considered when selecting appropriate candidates for MT with the Hunter catheter, favoring patients with vessel diameters > 2.5 mm, abundant thrombus burden (TIMI grade ≥ 4 flow), and culprit arteries other than the left circumflex one.

Studies have shown that total ischemia time, which we believe may determine differences in thrombus composition,18 may influence the efficacy of thrombus aspiration.19 In our series, there were no differences between the effective and noneffective MT groups with respect to infarction duration.

When assessing whether effective MT impacted the outcome of the PCI, slightly different procedural success rates were observed: 97% for effective MT vs 95% for noneffective MT (P = .61). Statistical significance was not reached, possibly due to the small sample size. Only 4 patients experienced no-reflow that could not be resolved, with no differences across groups and without demonstrating that MT could prevent distal embolization, as suggested in former studies.17-19

Short- and long-term clinical outcomes in patients with STEMI who required MT were favorable, with a low 1-year cardiovascular death rate of 3.5%, comparable to that reported in randomized clinical trials. In the TASTE trial,7 the 30-day mortality rate was 2.4% in the MT group and 2.9% in the PCI-alone group (HR, 0.84; 95%CI, 0.70–1.01; P = .06). In the TAPAS trial,20 the 30-day all-cause mortality rate was 2.1% in the MT group vs 4.0% in the conventional PCI group, reaching statistical significance at the 1-year follow-up (P = .07). Large registries have reported mortality rates similar to those observed in our series, with 2.8% vs 3.0% in the Swedish registry21 and comparable findings in the Japanese registry.22 The overall mortality rate observed in a meta-analysis with aggregated data from published MT studies was 3.7%.15 When selecting patients with a high thrombus burden (TIMI grade ≥ 3 flow in the meta-analysis subgroup), the cardiovascular death rate was 2.5% (170 of 6872 patients) in the MT group vs 3.1% (205 of 6599 patients) in the PCI-alone group (HR, 0.8; 95%CI, 0.65–0.98; P = .03).15 Proper selection of this subgroup of patients with a high thrombus burden is, therefore, crucial to maximize the therapeutic benefit.

Limitations

As a single-center, observational, retrospective registry, this study has inherent limitations related to its design. First, because it reflects the experience of a single center—albeit with more than 15 years of experience in PCI—operator homogeneity may limit extrapolation of the results. The decision to perform MT was always left to the discretion of the operator, and no control group without MT was available for patient comparison. Because of the retrospective design of the analysis, we could not determine the cause of ineffective MT in all patients, which is why this variable could not be included in the analysis. This study should not be interpreted as an evaluation of thrombus aspiration in general, but rather as an assessment of outcomes in a selected population treated with the Hunter device; these selection criteria represent the primary contribution of this work to current scientific knowledge. The study did not incorporate systematic criteria to address sex- and gender-related variables during methodological development or result analysis. Although angiographic analysis was not performed by an independent core laboratory, it was conducted by 3 experienced analysts. In this analysis, MT success, procedural success, and baseline thrombus burden were defined using the TIMI scale.

CONCLUSIONS

In patients with STEMI undergoing PCI, selective use of MT with the Hunter catheter in cases with high thrombus burden (TIMI grade ≥ 4 flow), non-circumflex culprit vessels, and vessel diameters > 2.5 mm is a safe and effective strategy associated with a low complication rate. Further studies are needed to assess the impact of this strategy on PCI outcomes, MACE, and stroke.

FUNDING

This project was supported by IHT-Iberhospitex S.A. (Lliçà de Vall, Barcelona, Spain). As sponsor, the company collaborated in the study design but had no role in data collection, analysis, or interpretation. Manuscript preparation and the decision to submit for publication were entirely independent of the sponsor and performed by the research team.

ETHICAL CONSIDERATIONS

The study was approved by Hospital Universitari Germans Trias i Pujol Ethics Committee (Barcelona, Spain) (CEIC code: PI-22-281) and conducted in full compliance with the principles outlined in the Declaration of Helsinki. SAGER guidelines were not applied to address gender bias.

STATEMENT ON THE USE OF ARTIFICIAL INTELLIGENCE

No artificial intelligence tools were used in the preparation of this manuscript.

AUTHORS’ CONTRIBUTIONS

D.G. Borraz-Noriega: angiographic analysis, data review, and manuscript drafting. J.F. Andrés-Cordón: angiographic analysis, data review, and statistical analysis. E. Cañedo: data collection and clinical follow-up. F. Panchano-Castro: angiographic analysis and data review. M. Trichilo: data collection and clinical follow-up. V. Vilalta: data collection and critical manuscript review. O. Rodríguez-Leor: data collection and critical manuscript review. E. Fernández-Nofrerias: critical manuscript review. I. Santos-Pardo: critical manuscript review. X. Carrillo: study design, overall supervision, and manuscript drafting. All authors reviewed and approved the final version.

CONFLICTS OF INTEREST

None declared.

REFERENCES

INTRODUCTION

Acute myocardial infarction (AMI) remains a leading cause of morbidity and mortality worldwide, despite advances in prevention and treatment.1-3 Evidence suggests that patients with AMI admitted outside working hours–nights, weekends, and holidays–experience worse outcomes,4-6 likely due to treatment delays,7 limited specialist availability, and operational constraints.4,8,9 While this association is well-documented in high-income settings, its impact in low- and middle-income countries such as Mexico, where health care resources are often constrained, remains less understood.10-12

Among Organisation for Economic Co-operation and Development (OECD) countries, Mexico reports the highest AMI mortality rate (23.7%), far exceeding the OECD average of 7%.3,12 Throughout the past century, the Mexican public health system has evolved from fragmentation to institutional organization, with key institutions such as the Mexican Social Security Institute (IMSS) (est. 1943) and the Institute for Social Security and Services for State Workers or Civil Service Social Security and Services Institute (ISSSTE) serving formal and public-sector workers, respectively.13,14 To address coverage gaps for the uninsured, programs such as Seguro Popular (2003), Institute of Health for Welfare (INSABI) (2020), and IMSS-BIENESTAR (2023) were established. In recent years, more than 53 million uninsured individuals, about 42% of the population, receive care through the Mexican Ministry of Health (SSA) across nearly 12 000 health centers and more than 680 hospitals.13

Despite system reforms, the management of AMI in Mexico remains hindered by limited access to specialized services, fragmented insurance coverage, and an underperforming emergency response system. These issues, exacerbated by socioeconomic disparities and insufficient preventive care, contribute to delays and reduced quality of treatment.15,16 The main aim of the current study was to evaluate the association between out-of-hours presentation and 30-day AMI-related mortality in Mexican emergency departments.

METHODS

Study design, guidelines, and data source

We conducted a retrospective cohort study in full compliance with the Guidelines for Strengthening the Reporting of Observational Studies in Epidemiology (STROBE).17 The present study used information from the public database of emergency department admissions from December 2014 through August 2023 in Mexico, published and validated by the General Directorate of Health (DGIS).18 This database compiles anonymous information from the SSA hospital registry (Seguro Popular/INSABI/IMSS-BIENESTAR programs) nationwide. Additional information on individual treatments (eg, percutaneous coronary intervention [PCI], thrombolysis, or door-to-balloon time), patient comorbidities, and delays in patient arrival was not available in the database.

Participants and sample size

The study included cases diagnosed with AMI (International Classification of Diseases [ICD-10, I21.0 to I21.9]), with lengths of stay of ≤ 30 days, classified as a qualified emergency, aged ≥ 18 years, and treated in a SSA hospital. Cases without complete information or not treated in the emergency department (referred to another health care unit, voluntarily discharged, or referred to outpatient care) were excluded from the analysis.

Definition of predictors and outcome

Sociodemographic (sex and age), time-related variables (admission and discharge dates), health care-related variables (type of emergency, discharge status, bed type, and AMI type according to ICD-10), and catheterization laboratories (cath labs) characteristics were collected. Several variables were grouped for analysis, including region of residence, death (yes/no), and length of stay (admission-to-discharge interval). The main predictor, out-of-hours, was defined as admissions during overnight shifts (19:00-6:59 h), weekends, or official holidays (per Article 74 of the Mexican Federal Labor Law).19 Furthermore, the COVID-19 pandemic (from 23 March 2020 to 9 May 2023)20 was considered. Hospitals were classified as PCI-capable and non-PCI-capable centers based on the SSA equipment database; and accurate and up-to-date information on PCI-capable centers was obtained; detailed definitions are provided in table S1.

Statistical analysis

Quantitative variables were expressed as mean ± standard deviation (SD), and categorical variables as frequencies and percentages. The primary sample was divided into 2 groups to observe the differences between PCI-capable and non-PCI-capable centers, and all analyses were systematically performed in each group. The chi-square and Student t tests evaluated the difference between the out-hours and on-hours groups. Univariate and multivariate Cox regression analyses were performed with proportional hazards adjusted for age, sex, year of admission, type of emergency, type of bed, territorial regions, type of AMI, admission during the COVID-19 pandemic, and characteristics of the cath lab. Moreover, Kaplan-Meier survival curves were constructed to estimate and visualize the cumulative mortality rate according to out-of-hours admission components. Statistical significance was set at P < .05. Confidence intervals (CI) were set at 95% (95%CI). Descriptive and analytical methods were performed in SPSS, version 25.0 (IBM, United States) and R software version 4.2.0 (R Foundation for Statistical Computing, Austria).

Ethical considerations

The study was based on data from the public emergency income database published by the DGIS.4 Accordingly, the confidentiality of the subjects is governed by the Mexican standard NOM-012-SSA3-2012,21 and the study is classified as risk-free research based on the principles of the General Law of Research for Health.22 The research was approved by the Ethics Committee of the Women’s Hospital, SSA (No. 202403-47).

RESULTS

General characteristics

The final dataset included a total of 29 131 cases: 4515 from PCI-capable centers (group 1) and 24 616 from non-PCI-capable centers (group 2) (figure 1). Out-of-hours admissions accounted for 46.7% in group 1 and 53.6% in group 2, with night admissions being the most frequent and holiday admissions the least common (table S2). Among PCI-capable centers, 32.2% provided nightshift coverage, while 53.1% offered weekend availability. Out-of-hours admissions were associated with a higher proportion of male patients, longer lengths of stay, greater use of shock beds, higher mortality, more pronounced regional differences, and a greater impact of the COVID-19 pandemic (table S3). Moreover, in PCI-capable centers, out-of-hours admissions were associated with a lower daily average number of procedures and fewer professionals per shift (table S4).

Figure 1. Flowchart of filtering process. Patients with AMI (ICD-10 I21.0-I21.9), aged ≥ 18 years, hospitalized ≤ 30 days in Ministry of Health (SSA) emergency services were included. Cases with incomplete data or not treated in emergency care were excluded. The final cohort was categorized as PCI-capable and non-PCI-capable centers, identified through percutaneous coronary intervention or cardiac catheterization records. AMI, acute myocardial infarction; PCI, percutaneous coronary intervention.

Association between out-of-hours care and the risk of death

Univariate analysis showed that the presence of a cath lab was protective (HR, 0.22; 95%CI, 0.204-0.246). Out-of-hours, overnight shifts, weekend admissions, and the COVID-19 pandemic were associated with higher mortality in both groups. The risks from out-of-hours (HR, 1.46; 95%CI, 1.21-1.73 vs HR, 1.18; 95%CI, 1.11-1.23) and night-shift admissions (HR, 1.18; 95%CI, 1.12-1.23 vs HR, 1.13; 95%CI, 1.08-1.18) were more pronounced in PCI-capable centers, and holidays increased mortality only in non-PCI-capable centers (HR, 1.21; 95%CI, 1.03-1.41) (figure 2). Among cath lab characteristics, higher mean procedure volume (HR, 0.888; 95%CI, 0.840-0.938), greater staffing (HR, 0.800; 95%CI, 0.760-0.842), and availability during afternoon (HR, 0.515; 95%CI, 0.427-0.620), night (HR, 0.290; 95%CI, 0.233-0.361), and weekend shifts (HR, 0.317; 95%CI, 0.264-0.380) were associated with reduced mortality.

Figure 2. Mortality risk associated with out-of-hours admissions for acute myocardial infarction: analysis stratified by hospital type (PCI-capable and non-PCI- capable centers). The graph shows results from unadjusted and adjusted Cox regression models. Adjustments included age, sex, emergency type, bed type, region, year of admission, AMI type, and COVID-19 period. For PCI-capable centers, additional adjustments were made for cath lab availability by shift, angiography type, and number of specialists per shift. Points represent mortality HRs with horizontal lines for 95%CI. Results are shown separately for PCI-capable and non-PCI-capable centers. HR > 1 indicates increased mortality risk; HR < 1 indicates reduced risk. *P < .05; ** P < .01; *** P < .001. 95%CI, 95% confidence interval; COVID-19, coronavirus disease 2019; HR, hazard ratio; PCI, percutaneous coronary intervention.

After adjusting for potential confounders, the presence of a cath lab remained protective (HR, 0.25; 95%CI, 0.23-0.28), whereas out-of-hours, night-shift admissions, and the COVID-19 pandemic continued to be associated with increased mortality in both groups. Adjusted mortality risk from out-of-hours (HR, 1.25; 95%CI, 1.04-1.50 vs HR, 1.18; 95%CI, 1.12-1.23) and night-shift admissions (HR, 1.27; 95%CI, 1.06-1.53 vs HR, 1.11; 95%CI, 1.06-1.17) persisted, while weekend and holiday effects remained significant only in non-PCI-capable centers (figure 2). Regarding cath lab characteristics, only nightshift availability (HR, 0.149; 95%CI, 0.089-0.248) and mean health care staff per shift (HR, 0.770; 95%CI, 0.710-0.834) were significantly associated with reduced mortality. Kaplan-Meier survival curves are shown in figure 3.

Figure 3. Kaplan–Meier analysis of mortality of out-of-hours acute myocardial infarctions by hospital type. Each column represents cath lab availability, while each row represents the components of out-of-hours admission. Each panel (A-H) shows the unadjusted (uHR) and adjusted mortality risk (aHR) with corresponding 95% confidence intervals.

DISCUSSION

Major findings

This nationwide study examined AMI admissions across a large number of Mexican hospitals, comparing PCI-capable and non- PCI-capable centers. Out-of-hours admissions were common and associated with higher mortality, longer lengths of stay, greater shock bed use, and a higher proportion of male patients. While the presence of a cath lab appeared generally protective after adjustment, out-of-hours and nightshift admissions showed a modest increase in mortality in both groups. Although weekend and holiday effects were more evident in non-PCI-capable centers, these results should be interpreted with caution given the absence of detailed clinical data, including reperfusion strategies, ST-segment elevation myocardial infarction/non-ST-segment elevation myocardial infarction (STEMI/NSTEMI) classification, and comorbidities.

A key finding is that PCI-capable centers did not fully eliminate excess mortality during out-of-hours periods. Advanced interventional capabilities may mitigate some risk during weekends and holidays; however, challenges remain during overnight shifts, as only 32.2% of studied centers offered cath lab availability during overnight shifts. Out-of-hours admissions were associated with fewer procedures, lower staffing, and limited specialized personnel compared with regular hours, which is consistent with former studies4,8,9 (table S4). Additional factors likely contributing to poorer outcomes include reduced staffing, delays in care, limited access to specialized personnel, and the impact of human factors such as fatigue and sleep deprivation during nightshifts, despite the availability of advanced cardiac interventions.4,8,9 Longer door-to-balloon times, limited 24/7 cath lab coverage, and the relatively low number of cath labs per capita in Mexico further exacerbate these challenges. Moreover, the increased out-of-hours risk from PCI-capable centers may reflect their role as referral centers, where longer patient arrival times and procedural delays are more frequent.16

Pre-hospital and COVID-19 challenges in the management of AMI

Several institutions in Mexico have implemented programs to standardize the management of AMI, such as the IMSS and SOCIME national “Code Infarction” (2015) and ISSSTE “AsISSSTE Infarto” (2018).23,24 In contrast, SSA-affiliated institutions (Seguro Popular/INSABI/IMSS-BIENESTAR) follow more varied protocols, leading to disparities in care. Consequently, heterogeneity in the management of AMI management across institutions likely exacerbates challenges in pre-hospital care. Regulated, by NOM-034-SSA3-2013 and coordinated via Medical Emergency Regulatory Centers,25 the pre-hospital system faces challenges such as poor inter-institutional coordination, insufficient ambulance equipment, and personnel shortages, especially in rural areas. These factors increase response times and complicate AMI emergency management.26 Former studies report prolonged door-to-balloon times (up to 648 minutes) due to traffic, fragmented health care, and delayed diagnosis, with weekend and nighttime admissions independently predicting treatment delays over 12 hours. While this study does not examine treatment or pre-hospital delays, other research highlights these as major issues in Mexico. Araiza-Garaygordobil et al.16 reported a mean door-to-balloon time of 648 minutes in a Mexico City referral hospital, 3 times longer compared with developed countries, due to traffic, health care fragmentation and delayed primary diagnosis. Baños-González et al.15 found that patient delays, often from symptom unawareness or limited resources, led to late arrivals (> 12 hours), with 60% being transported by ambulance. Weekend and nighttime admissions independently predicted delays > 12 hours, with a mean 11-hour wait for first medical contact.

The COVID-19 pandemic was associated with an increased AMI mortality in both hospital groups, with a more pronounced effect in PCI-capable centers. Globally, the pandemic disrupted health care delivery, reducing hospitalizations and PCI procedures, partly due to lower referral rates and patients’ reluctance to seek care over concerns of viral exposure. Consequently, delays from symptom onset to first medical contact and prolonged door-to-balloon times occurred, both of which are known to adversely affect AMI outcomes.27 In Mexico, Rodríguez-González et al.28 reported a 51% decline in STEMI diagnoses in hospitals during early 2020, accompanied by longer arrival times and a 4.9%-6.8% increase in STEMI-related mortality. In other countries, such as Spain, PCI rates decreased by 10.1% during 2020.29 A recent meta-analysis reported a significant reduction in the number of PCI (IRR, 0.72; 95%CI, 0.67-0.77), I² = 92.5%) and an increase in time from symptom onset to first medical contact by a mean 69.4 minutes ([11-127], I² = 99.4%) during the COVID-19 pandemic. However, no significant change was observed in door-to-balloon times (3.33 minutes [0.32-6.98]; I² = 94.2%).30

Although detailed clinical data were not available in our study, the impact of treatment on AMI mortality has been extensively investigated in Mexico. The nationwide RENASCA cohort31 (21 826 patients, 2014–2017) reported high prevalence of diabetes (48%), hypertension (60.5%), smoking (46.8%), dyslipidemia (35.3%), and metabolic syndrome (39.1%) relative to multinational registries.31,32 Patients with STEMI (14.9%) experienced higher cardiovascular mortality vs patients with NSTEMI (7.6%), reflecting referral patterns and more severe in-hospital complications, including cardiogenic shock and arrhythmias. Implementation of the IMSS “Code Infarction” program reduced patients without reperfusion from 65.2% to 28.6%, increased fibrinolysis to 40.1%, and PCI to 31.3%.33 By comparison, Spain’s well-structured “Code Infarction” protocol achieved a primary PCI rate of 87%, median door-to-balloon time of 193 minutes, and minimal pre-hospital delays, with 55.3% of patients receiving timely care and most remaining delays attributable to initial diagnosis (18.5%).34

Global trends in out-of-hours mortality: a comparison across regions

Compared with previous meta-analyses, which often lack data from developing regions such as Latin America, our findings highlight important disparities. Sorita et al.4 reported higher short-term mortality rates and delayed PCI in patients with STEMI admitted outside regular hours, especially outside North America. Wang et al.5 found a slight increase in short-term mortality but no effect on long-term outcomes or PCI risk. Other analyses showed no significant differences in mortality or door-to-balloon times.35 Studies from developing countries such as Indonesia36,37 reported no mortality differences by admission time, while in Iraq,38 resource shortages and conflict led to higher out-of-hours mortality and limited reperfusion access. Latin American studies from Brazil39-41 and Argentina42 generally found no outcome differences by admission timing. Differences with our results are likely due to hospital infrastructure and staffing; Mexican public hospitals face substantial limitations after hours vs tertiary referral centers with 24/7 cath lab availability. Additionally, socioeconomic barriers and low awareness of AMI symptoms delay care, and the lack of standardized protocols further increases mortality risk.

Strategies to improve the out-of-hours management of AMI

The findings highlight several areas for improvement within the Mexican health care system, particularly regarding non-PCI-capable centers. All 3 components were associated with higher mortality, underscoring the need for enhanced pre-hospital and in-hospital management. Efforts should focus on early diagnosis, including increased availability of electrocardiography and the implementation of telemedicine platforms to allow cardiologists to promptly assess and guide treatment. Expanding access to thrombolytic therapy and ensuring experienced personnel are available during out-of-hours shifts can help reduce treatment delays. Furthermore, improving coordination across health centers through clear referral pathways, effective communication strategies, and standardized protocols, especially in hospitals serving uninsured patients, can streamline care and reduce variability in management. In PCI- capable centers, mortality was particularly elevated during overnight shifts, likely due to reduced cath lab availability within these hours. Despite existing protocols and optimized door-to-balloon times, strengthening nighttime staffing and training remains essential. This may include targeted training and incentives to encourage health care professionals to work during these shifts, thereby ensuring that high-quality care is maintained at all times.

Strengths and limitations

The present study is one of the largest cohorts to date examining the association between AMI-related mortality and time of presentation in Mexico. It provides valuable insights into the potential impact of out-of-hours admissions on outcomes, underscoring the importance of optimizing staffing, patient flow, and timely access to interventions such as PCI. Nevertheless, the findings should be interpreted with caution. The retrospective design, absence of detailed treatment information (eg, PCI, thrombolysis, door-to-balloon times), lack of comorbidity data, and inability to differentiate between STEMI and NSTEMI limit the depth and precision of the analyses. Furthermore, the study does not capture key factors such as patient-level delays, cath lab utilization patterns, or disparities in infrastructure and resource availability. Future research should integrate more comprehensive clinical and procedural data and explore the influence of provider fatigue, variability in care standards, and treatment delays to better delineate gaps in care.

CONCLUSIONS

Out-of-hours presentations were associated with high mortality rates, and contrary to findings in developed countries, the presence of a cath lab did not improve out-of-hours outcomes, which suggests that factors beyond cath lab availability, including systemic inefficiencies, resource limitations, and health infrastructure deficiencies, may profoundly affect patient outcomes in Mexico.

FUNDING

None declared.

ETHICAL CONSIDERATIONS

Ethical approval for this study was granted by the Ethics Committee of the Women’s Hospital, SSA (Approval No. 202403-47). The research used publicly available emergency income data published by the DGIS, for which subject confidentiality is governed by the Mexican standard NOM-012-SSA3-2012. In accordance with the General Health Research Law, the study is classified as risk-free research. Sex variable was defined as according to the definition of the Sex and Gender Equity in Research (SAGER) guidelines (table S1).

STATEMENT ON THE USE OF ARTIFICIAL INTELLIGENCE

No generative artificial intelligence (AI) tools were used in the conception, design, data collection and analysis of this manuscript. All content was produced entirely by the authors.

AUTHORS’ CONTRIBUTIONS

D. Arriaga-Izabal contributed to the conceptualization, methodology, data curation, data analysis, investigation, and writing of the original draft. F. Morales-Lazcano was responsible for investigation, data curation, and drafting the original manuscript. A. Canizalez-Román provided supervision, project administration, and validation, and participated in the review and editing of the manuscript. All authors read and approved the final version of the manuscript.

CONFLICTS OF INTEREST

None declared.

SUPPLEMENTARY DATA

REFERENCES

INTRODUCTION

Currently, transfemoral transcatheter aortic valve implantation (TF-TAVI) is the treatment of choice for most patients with severe aortic stenosis, particularly those with high surgical risk or advanced age.1 Several clinical trials have demonstrated comparable clinical outcomes between TF-TAVI and surgical aortic valve replacement.2-4 Major and minor bleeding remain one of the most frequent procedural complications and are associated with higher morbidity and mortality rates.5 Although, in recent years, improvements in materials (reduction in caliber required for valve implantation) and increasing operator experience have substantially reduced perioperative bleeding rates, such rates remain significantly high. One of the main risk factors for bleeding is the requirement for perioperative dual antiplatelet therapy,6 most widely necessary when coronary revascularization is performed along with TAVI.

In addition, the high prevalence of coronary artery disease in patients undergoing TF-TAVI, reported in up to 80% of cases in published series, along with current clinical practice guideline recommendations to revascularize all ≥ 70% proximal coronary stenoses, results in a high rate of revascularization.1

Our group recently published data showing that systematic, complete revascularization in patients undergoing TF-TAVI does not provide prognostic benefit in terms of mortality or major adverse cardiovascular events (MACE) (a composite of death, myocardial infarction, stroke, and heart failure-related hospitalization.)7 Given the high rate of bleeding events in these patients, it is of substantial clinical interest to evaluate whether revascularization may confer an increased bleeding risk and assess its clinical impact.6

METHODS

We conducted a retrospective study based on the historical cohort of patients who underwent TF-TAVI at our center from 2008 through 2018. Study information was drawn from the local database (Géminis) and supplemented by electronic health record review to document follow-up events. The study was approved by the Clinical Research Ethics Committee of A Coruña-Ferrol (Spain). The primary endpoint of the study was to compare the rate of major bleeding—defined according to the Bleeding Academic Research Consortium (BARC) criteria (types 3–5)—occurring after diagnostic coronary angiography, during the index hospitalization for TAVI, and during follow-up. Additional endpoints included the rate of MACE and the composite endpoint of MACE plus major bleeding over the same period of time, comparing patients who underwent percutaneous coronary revascularization with those managed conservatively.

Statistical analysis

Continuous variables are expressed as mean ± standard deviation, and qualitative variables as proportions. We used the Student t test and analysis of variance (ANOVA) with first-order polynomial contrast for continuous variables. For categorical variables, we used the chi-square test for linear trend or Fisher’s exact test as appropriate.

We conducted survival analyses using the Cox proportional hazards model to determine whether an association existed between coronary revascularization and patient prognosis in terms of mortality, MACE, major bleeding (BARC 3–5), and a composite of MACE and major bleeding. Results were expressed as age- and sex-adjusted survival curves.

Statistical analysis was performed with SPSS 26.0 (IBM, USA) and R version 4.1.3 (R Foundation for Statistical Computing, Austria). Statistical significance was set at P < .05 for all comparisons.

RESULTS

A total of 379 patients who underwent TF-TAVI between 2008 and 2018 were included. Four patients were lost to follow-up, leaving 375 patients for the statistical analysis. Table 1 illustrates the patients’ baseline characteristics, with a mean age of 83.1 years and predominance of the female sex and intermediate surgical risk (Society of Thoracic Surgeons score, 4.3%). Although most baseline characteristics were well balanced between patients with and without revascularization, the latter had a slightly higher surgical risk (4.5% vs 3.5%) and a higher proportion of women (61.3% vs 47.8%).

Table 1. Baseline characteristics of the patients

Regarding symptoms, most patients were in New York Heart Association functional class III (74.7%). Among revascularized patients, 25.0% (23 patients) reported angina symptoms vs only 15.7% (41 patients) from the nonrevascularized group. Regarding anticoagulant therapy, 29.2% (83 patients) from the nonrevascularized group were on vitamin K antagonists and 3.9% (11 patients) on direct oral anticoagulants. These rates were slightly lower among revascularized patients, with 23.1% (21 patients) on vitamin K antagonists and 3.3% (3 patients) on direct oral anticoagulants (table 1).

Percutaneous revascularization and bleeding

Of the patients undergoing TF-TAVI, 92 (24.3%) underwent coronary revascularization, which in our center was always performed before valve replacement, with a median interval of 88 days (46–162) between revascularization and TAVI. Consequently, most revascularized patients were on dual antiplatelet therapy when they underwent TAVI. The decision to revascularize was made in a multidisciplinary meeting according to contemporary clinical practice guidelines (2008–2018). The prevailing strategy was to revascularize most coronary lesions, and the clinical criterion remained unchanged throughout the study period.

Clinical outcomes during follow-up are shown in table 2. There were no statistically significant differences in the time elapsed between diagnostic catheterization and TAVI between the 2 groups (median of 118 days for revascularized patients and 123 days for the nonrevascularized ones; P = .835). Figure 1 illustrates that the overall rate of major bleeding was 21.4% (81/375), with a rate of 28.3% (26/92) for revascularized patients and 19.0% (55/283) for the nonrevascularized ones, without reaching statistical significance (P = .074).

Table 2. Clinical outcomes during follow-up according to revascularization status

Figure 1. Central illustration. Overall rate of major bleeding across follow-up periods. BARC, Bleeding Academic Research Consortium; TAVI, transcatheter aortic valve implantation.

Table S1 illustrates the bleeding events classified by BARC criteria according to revascularization status during follow-up. The overall rate of major bleeding (BARC 3–5) between coronary angiography and TF-TAVI was 5.5% (21/375) and was significantly higher among revascularized patients (12.0% vs 3.5%; P = .007). During the index hospitalization for TF-TAVI, the overall rate of major bleeding was 6.1% (23/375), with no statistically significant differences between the 2 groups (8.7% vs 5.3%; P = .31). During post-TAVI follow-up, the overall rate of major bleeding was 9.6% (36/375), with no significant differences between the groups either (7.6% vs 10.2%; P = .545) (figure 1).

After a mean follow-up of 60 months, 55.1% (209/379) of patients experienced MACE: 55.7% in the nonrevascularized group (160 patients), and 53.2% in the revascularized group (49 patients). There were no statistically significant differences (P = .082) (figure 2). The overall mortality rate at 60 months was 42.2% (161/378): 43.7% in nonrevascularized patients (125 patients), and 39.0% in revascularized ones (36 patients). Again, no significant differences were reported (P = .380) (figure 3).

Figure 2. Kaplan–Meier curve for MACE-free survival by revascularization status.

Figure 3. Kaplan–Meier curve for overall survival by revascularization status.

The revascularized group exhibited higher rates of myocardial infarction (11.5% vs 2.8%; P = .001) and repeat revascularization (6.5% vs 1.0%; P = .003).

After a mean 60-month follow-up, the composite endpoint (MACE or major bleeding) occurred in 62.0% (235 patients) of the total sample, without significant differences across the groups (66.3% nonrevascularized vs 60.6% revascularized; P = .139) (figure 4).

Figure 4. Kaplan–Meier curve for combined event–free survival (ischemic or hemorrhagic) by revascularization status.

DISCUSSION

There is currently no definitive clinical evidence on the optimal management of coronary artery disease in patients scheduled for TAVI. Clinical guidelines recommend percutaneous revascularization in all patients undergoing TF-TAVI with percent diameter stenoses ≥ 70% in the target vessel proximal segments, with a Class IIa recommendation. However, to this date, only 2 randomized clinical trials have assessed the benefit of pre-TAVI revascularization, with conflicting results.8,9

Major bleeding remains one of the most frequent and prognostically relevant complications after TF-TAVI. In fact, in the PARTNER 2 trial conducted with intermediate-risk patients, major bleeding was reported in up to 15.2% of patients 1 year after TAVI.³ Despite this, evidence on the impact of pre-TAVI coronary revascularization on bleeding risk is scarce.

In the ACTIVATION trial,8 a total of 235 patients scheduled for TF-TAVI who had significant coronary artery disease were randomized to undergo percutaneous revascularization (n = 119) or receive optimal medical therapy (n = 116). Outcomes in the 2 groups were evaluated according to a composite primary endpoint of all-cause mortality and hospitalization. At 1 year, noninferiority of the strategy of adding percutaneous revascularization to TF-TAVI could not be demonstrated in patients who did not undergo revascularization; however, higher bleeding rates were observed in the intervention group.

The results of the NOTION-3 trial9 have been recently published. In this study, 452 patients scheduled to undergo TF-TAVI with significant coronary artery disease were randomized to receive either an invasive or a conservative strategy. In this case, the decision to revascularize was guided by the severity of stenosis as assessed by fractional flow reserve. Outcomes were evaluated according to a composite primary endpoint of all-cause mortality, myocardial infarction, and emergency revascularization. At 2 years, patients who had been revascularized showed a significant reduction in the risk of MACE compared with the conservative strategy (26.0% vs 36.0%), driven primarily by a higher rate of unplanned revascularization and without an effect on mortality. However, this risk reduction was accompanied by a higher rate of bleeding events (28.0% vs 20.0%). A major limitation of this trial is its open-label design and the fact that it did not exclude patients with angina, which may have contributed to the increased rate of unplanned revascularization.

In our study, there were no statistically significant differences in MACE or major bleeding at 60 months between revascularized and nonrevascularized patients. However, a higher rate of major bleeding occurred among the former during the time elapsed between diagnostic coronary angiography and TAVI, which is consistent with former studies demonstrating higher bleeding rates in patients on dual antiplatelet therapy.8,9 Furthermore, this group exhibited higher rates of myocardial infarction and subsequent revascularization. Although these findings were expected, they should be interpreted with caution because, despite statistically significant differences, the small number of events limits statistical power to draw definitive conclusions. A reasonable strategy may be selective revascularization aimed at symptom control, particularly in patients with angina.

Our study has several important limitations. The most significant one is that it is a single-center, observational, nonrandomized, retrospective analysis, which results in multiple sources of selection bias. First, a biological selection bias exists because the study includes patients who self-selected by surviving to a mean age of 83 years with sufficient biological status to be considered eligible for TF-TAVI. Second, a clinical selection bias is present regarding which patients were selected for TAVI, as the retrospective design makes it impossible to standardize the criteria originally used to determine candidacy. Finally, our cohort only includes patients in whom the procedure was ultimately performed—not those who initially underwent evaluation for TAVI—which means that some patients who underwent diagnostic cardiac catheterization (with or without revascularization) may not have proceeded to TAVI and are therefore not included. The proportion of such patients and the reasons for not completing the procedure are unknown. Although the causes may be diverse, given the procedural risks of coronary interventions and the frequent presence of complex coronary artery disease in this population, it is plausible that some candidates did not undergo TAVI because of revascularization-related complications; however, this cannot be demonstrated with our data and remains speculative. Moreover, this is a single-center study with a limited sample size, which may restrict the external validity of the findings and the statistical power to detect inter-group differences.

CONCLUSIONS

In our cohort, pre-TF-TAVI systematic coronary revascularization was associated with an increased early risk of major bleeding, specifically between diagnostic catheterization and valve implantation. There were no statistically significant differences in long-term major bleeding, MACE, or the composite endpoint between revascularized and nonrevascularized patients. These findings, together with recent evidence indicating that revascularization of stable coronary disease does not clearly improve prognosis,10 reinforce the need for high-quality clinical evidence to define the clinical impact of pre-TAVI systematic coronary revascularization.

FUNDING

None declared.

ETHICAL CONSIDERATIONS

The study was approved by the Clinical Research Ethics Committee of A Coruña-Ferrol. Informed consent was not required due to the retrospective design of the study and use of a preexisting clinical database. SAGER guidelines were followed to minimize potential sex-related bias.

STATEMENT ON THE USE OF ARTIFICIAL INTELLIGENCE

No artificial intelligence tools were used in the preparation of this article.

AUTHORS’ CONTRIBUTIONS

C. Vidau Getán contributed to data collection and manuscript drafting. D. López Vázquez was the main reviewer and contributed to refinement of statistical analysis. X. Flores Ríos conceived the study and conducted the initial statistical analysis. M. González Montes and G. González Barbeito participated in data collection. The remaining coauthors reviewed the final version of the manuscript. All authors gave their final approval.

CONFLICTS OF INTEREST

None declared.

SUPPLEMENTARY DATA

REFERENCES

INTRODUCTION

Mitral regurgitation (MR) is a prevalent valvular heart disease associated with significant morbidity and mortality, particularly in patients with heart failure (HF).1 Traditional management strategies include optimal medical therapy to treat symptoms and surgery for definitive correction.2 However, current clinical practice guidelines only recommend mitral valve repair if the patient is undergoing another intervention such as coronary artery bypass graft or aortic valve replacement; nonetheless, many patients are at a high risk for surgery due to comorbid conditions, for this reason, alternative therapeutic options is required.3

Mitral transcatheter edge-to-edge repair (M-TEER) has emerged as a minimally invasive option for patients with symptomatic mitral regurgitation who are ineligible for surgery.4 While M-TEER has shown promising results, its comparative efficacy vs guideline- directed medical therapy (GDMT) is still a matter of discussion.5

This meta-analysis aims to assess the impact of M-TEER plus GDMT vs GDMT alone on major clinical outcomes in patients with secondary MR. The primary endpoint evaluated was HF-related hospitalization. Secondary endpoints included all-cause mortality, cardiac death, improvement in New York Heart Association (NYHA) functional class (FC), and the incidence rate of stroke and myocardial infarction. These endpoints were selected for their clinical relevance and potential to inform therapeutic decision-making in this high-risk patient population.

METHODS

We conducted this meta-analysis in full compliance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement guidelines.6 All the stages of this study were performed following the Cochrane Handbook for Systematic Reviews of Interventions, version 6.3.7 The protocol for this meta-analysis was registered prospectively in PROSPERO on 10 February 2025 with protocol ID CRD42025645047.

Search strategy

We conducted a systematic search across 3 electronic databases (PubMed, EMBASE, and COCHRANE) from inception until February 2025. Search terms included combinations of the following keywords: “transcatheter”, “secondary”, “mitral valve”, “replacement”, regurgitation”, and “insufficiency”. These keywords were combined using Boolean operators AND, OR. The reference lists of eligible studies and previous reviews were checked to identify additional valuable articles.

Criteria of the included studies

Studies were considered for inclusion in the meta-analysis based on the following criteria: a) randomized clinical trials (RCT) or observational cohort studies in Spanish or English that b) compared transcatheter mitral valve replacement vs optimal medical therapy in adult patients with mitral valve regurgitation due to secondary causes; and c) studies that reported outcomes of interest such as the index HF-related hospitalization, HF-related readmissions or cardiac death, and all-cause mortality.

Studies were excluded if they were non-original articles such as systematic reviews, letters, abstracts, meta-analyses, case reports, or case series. Furthermore, studies were excluded if mitral regurgitation was due to primary causes or if prior surgical repair had been performed.

Although observational cohort studies were eligible, we only identified 1 cohort study that did not meet the inclusion criteria during full-text screening. As a result, only RCTs were ultimately included in the meta-analysis.

Data extraction

To ensure accuracy, we used Zotero to eliminate duplicate references. We screened each paper based on title and abstract as a first step, followed by a full-text review as the second step. Two authors (D.A. Navarro Martínez, and D. Paulino-González) independently screened each paper, with any disagreements being resolved by a third author (A.L. García Loera). In addition, references of the included studies were reviewed and added if they met our eligibility criteria. Data extraction was conducted using Excel spreadsheets capturing the following information: a) baseline characteristics of the studied population, including baseline medication, echocardiographic baseline characteristics, and comorbidities; b) summary of the characteristics of the included studies; c) outcome measures; and d) quality assessment domains.

Primary and secondary endpoints

The primary endpoints established for this meta-analysis included: a) all-cause mortality (hazard ratio [HR]), defined as all-cause mortality, assessed using time-to-event analysis,8 and b) HF-related hospitalization, defined as hospital admission for worsening HF following the intervention (M-TEER or initiation of optimal medical therapy).9

Secondary endpoints included a) stroke, defined as an episode of neurological dysfunction caused by focal cerebral, spinal, or retinal infarction;10 and b) myocardial infarction, defined according to the Fourth Universal Definition.11

All endpoints were assessed as dichotomous categorical variables, which were reported in percentages. The risk ratio (RR) with its corresponding 95% confidence interval (95%CI) was calculated.

Assessment of heterogeneity

To assess heterogeneity, we used Cochrane’s Q-statistic with a significance level of P < .05. Additionally, the I²-statistic was used to quantify the proportion of variability due to heterogeneity, with values > 50% being considered indicative of high heterogeneity.12

Statistical analyses

To assess dichotomous data, we evaluated event frequencies and totals from each study group to calculate the RR and its 95%CI. Moreover, we obtained the HR and the 95%CI to calculate the standard error (SE).

The variables analyzed in this study were based on data reported in the original studies of the intention-to-treat principle. We implemented a random-effects model using the DerSimonian-Laird13 method to account for variability and allow comparison across studies. Given the limitations and strengths of this method, we conducted a sensitivity analysis using the Hartung-Knapp-Sidik-Jonkman method. Forest plots were utilized as a visual representation of the estimated outcomes.

Considering the differences in the populations included in the clinical trials, we conducted an additional exploratory analysis that focused on the RESHAPE-HF214 and COAPT15 trials for the primary endpoints by removing MITRA-FR16 from the analysis.

All statistical analyses, including the calculation of RR and SE, were conducted using RevMan V.5.4.1 software.

RESULTS

The initial database search yielded a total of 164 potentially relevant articles. After removing duplicates, a total of 129 articles finally remained for title and abstract screening. Following this initial screening, 8 articles were identified as potentially eligible for full-text review. Finally, after a detailed evaluation of the full texts under the predefined inclusion and exclusion criteria, 3 studies were included. A summary of the study selection process is shown in the PRISMA flowchart figure 1.

Figure 1. PRISMA 2020 flow diagram for new systematic reviews which included searches of databases and registers only.

Our analysis included 3 RCTs.14-16 We consulted the extended report of the 2-year outcomes of the MITRA-FR trial17 and 1 additional article on the RESAHPE-HF2 to obtain more data regarding hospitalization.18 All included studies compared the use of M-TEER plus GDMT vs GDMT alone. To standardize the outcomes, the median follow-up was set at 24 months; only the NYHA FC was evaluated with 1-year results. All 3 included studies used the MitraClip (Abbott, United States) device.

Clinical baseline characteristics of the patients

We obtained a total population of 1423 patients; among them 704 received M-TEER and 719, GDMT alone. In the RESHAPE-HF214 the device group showed the following characteristics: mean age was 70 years old; left ventricular ejection fraction (LVEF), 32%; median left ventricular end-diastolic volume (LVEDV), 200 mL; median effective regurgitation orifice (EROA), 0.23 cm²; median N-terminal pro-B-type natriuretic peptide (NT-proBNP), 2651; and brain natriuretic peptide (BNP), 556. In the COAPT15 trial, median age was 71 years old; LVEF, 31%; LVEDV, 194 mL; median EROA, 0.41 cm²; median NT-proBNP, 5174; and BNP, 1014. Finally in the MITRA-FR trial16, mean age was 71 years old; LVEF, 33%; LVEDV, 136.2 mL; median EROA, 0.31 cm²; median NT-proBNP, 3407; and BNP, 765. Regarding etiology, all trials enrolled mixed ischemic and non-ischemic etiology populations. Patient and study baseline characteristics, including comorbidities, drugs, and echocardiographic data, are shown in table 1, table 2, and table 3.

Table 1. Baseline characteristics of patients

Table 2. Summary of included studies

Table 3. Echocardiographic baseline characteristics

Risk of bias assessment

To evaluate the quality of the included randomized controlled trials, we used the Risk of Bias 2 (RoB2) tool from the Cochrane Handbook of Systematic Reviews of Interventions 6.3.7 This methodology allowed us to systematically assess the methodological quality of each study, thereby strengthening the validity of our findings.

Of the 3 studies, 1 had a low risk of bias,14 while the other 2 raised some concerns.15,16 The studies with some concerns were limited by factors in their methodologies, as they were open-label studies (figure 2).

Figure 2. Risk of bias for each included randomized clinical trial. The bibliographical references mentioned in this figure correspond to Stone et al.,14 Obadia et al.,15 and Anker et al.16.

Outcomes

HF-related hospitalization

The analysis of this outcome showed a significant reduction, with a RR of 0.71 (95%CI, 0.56-0.90; P = .004). However, substantial heterogeneity was observed, with an I² value of 79%. Considering the differences in the MITRA-FR study population, we conducted an exploratory analysis excluding this trial. Results still demonstrated a significant reduction in the risk of hospitalization, with a RR of 0.63 (95%CI, 0.55-0.72, P < .00001), along with a marked improvement in heterogeneity (I² = 0%). Of note, this finding is exploratory and should be interpreted with caution.

The analysis, including all studies, is shown in figure 3A, and the exploratory analyses are shown in figure 3B. Furthermore, we analyzed HR for this endpoint and, given the limited number of studies, we conducted a sensitivity analysis using the Hartung-Knapp-Sidik-Jonkman (HKSJ) method; all these analyses are provided in the supplementary data (figures S1-S4).

Figure 3. Forest plot of risk ratios (RR). Lines denote the 95% confidence intervals (95%CI) for each trial. A: forest plot of RR for HF related hospitalization. B: forest plot of exploratory RR for HF-related hospitalization. C: forest plot of RR for all-cause mortality. D: forest plot of exploratory RR for all-cause mortality. 95%CI, 95% confidence interval; GDMT, guideline directed medical therapy; M-TEER, mitral transcatheter edge-to-edge repair. The bibliographical references mentioned in this figure correspond to Stone et al.,14 Obadia et al.,15 and Anker et al.16.

All-cause mortality

The RR for all-cause mortality was 0.80 (95%CI, 0.63–1.02; P = .07). However, substantial heterogeneity was observed across the studies (I² = 56%). In the exploratory analysis, a significant reduction in mortality was found, with an RR of 0.71 (95%CI, 0.59–0.86; P = .0005) and no heterogeneity across the studies (I² = 0%). The analyses, including all studies, are shown in figure 3C, and the exploratory analysis in figure 3D.

Both the HR analysis and the sensitivity analysis are provided in the supplementary data (figures S5-S8).

Cardiac death

The RR for death from cardiovascular causes was 0.81 (95%CI, 0.62–1.06, P = .12), with low to moderate heterogeneity (I² = 48%). In the exploratory analysis, a significant reduction in cardiac death was observed, with an RR of 0.74 (95%CI, 0.55–1.00; P = .05) and low heterogeneity (I² = 41%). The analyses, including all studies, are shown in figure 4A, while the exploratory analysis, in figure 4B. The sensitivity analysis and HR results are provided in the supplementary data (figures S9-S12).

Figure 4. Forest plot of risk ratio (RR). A: forest plot of RR for cardiac death. B: forest plot of exploratory RR for cardiac death. C: forest plot of RR of NYHA FC I/II at 1 year. D: forest plot of RR for stroke. 95%CI, 95% confidence interval; GDMT, guideline directed medical therapy; M-TEER, mitral transcatheter edge-to-edge repair. The bibliographical references mentioned in this figure correspond to Stone et al.,14 Obadia et al.,15 and Anker et al16. Lines denote the 95%CI for each trial.

NYHA FC

In this analysis of the NYHA FC, we observed that patients undergoing M-TEER are more likely to be found in NYHA FC I/II at 12 months, with a RR 1.26 (95%CI, 1.06–1.50; P = .010), respectively. Of note, the high heterogeneity (I² = 72%). Furthermore, it is also essential to note that the NYHA FC is a subjective classification, which is why the findings of this outcome should be interpreted with caution. These results are shown in figure 4C.

Stroke

There were no statistical differences between M-TEER and optimal medical therapy regarding stroke, with a RR of 1.44 (95%CI, 0.76 - 2.73, P = .27) without any heterogeneity being reported across the studies (I² = 0%). These results are shown in figure 4D.

Myocardial Infarction

Regarding the risk of myocardial infarction, our analysis demonstrated no statistical difference between the M-TEER and optimal medical therapy groups (RR, 0.83; 95%CI, 0.43-1.61; P = .58). No heterogeneity among the studies was observed (I² = 0%); this finding is shown in supplementary data (figure S13).

Exploratory analysis by severity

We conducted an exploratory analysis stratified by the baseline grade of mitral regurgitation. Data on the composite endpoint of all-cause death or HF-related hospitalization were available according to MR grade. For patients with MR 3+, the RR was 0.75 (95%CI, 0.56-1.00; P = .05; I² = 12%). For those with severe MR 4+, data were available only from the RESHAPE-HF2 and COAPT trials, yielding a RR of 0.55 (95%CI, 0.42-0.73; P = .0001; I² = 0%). This finding is shown in the supplementary data (figure S14 to figure S15).

DISCUSSION

Our meta-analysis provides an evaluation of 3 major RCTs, COAPT15, MITRA-FR16, and RESHAPE-HF214, evaluating M-TEER plus GDMT vs GDMT alone in patients with secondary MR. In patients with persistent symptoms despite adequate GDMT and who do not meet the criteria for definitive surgical replacement, mitral M-TEER has emerged as a promising alternative. In this updated meta-analysis, no statistically significant effect on all-cause mortality was found; however, the marked between-trial heterogeneity suggests that M-TEER provides meaningful clinical benefit when used in appropriately selected patients.

Our study showed that heterogeneity of results was mainly driven by MITRA-FR. COAPT and RESHAPE-HF2 have multiple similar differences compared with MITRA-FR, as evidenced in trial cohort baseline characteristics, methods, and outcome directions. This is better portrayed by an exploratory analysis comparing COAPT and RESHAPE-HF2 results, showing a 37% reduction in the risk of HF-related hospitalization risk and a 32% reduction in cardiac death (figure 3). The overall neutral mortality rate resulting from our study may be driven by heterogeneity across trials rather than a lack of intrinsic therapeutic benefit.

Among the main differences across trials, the MITRA-FR trial had broader inclusion of patients with tricuspid regurgitation, advanced LV remodeling, greater dilation, and smaller EROA vs the other 2 trials. Another key source of heterogeneity was baseline MR severity. COAPT and RESHAPE-HF2 primarily enrolled patients with grade 3+ and 4+ MR, whereas MITRA-FR is only focused on grade 4+, which may contribute to the observed outcome differences. In our exploratory analysis by MR grade, data showed consistent trends favoring M-TEER; however, these findings should be interpreted with caution due to limited subgroup data.

As proposed by Paul et al.,19 one strategy to evaluate secondary MR is to consider the proportion between EROA and LVEDV. In the MITRA-FR trial, many patients had smaller EROA with markedly dilated ventricles, a phenotype in which GDMT may have more impact than valve proceduren. In the COAPT and RESHAPE-HF2 trials, a greater proportion of patients had large EROA relative to the LVEDV,20 making MR a primary driver of symptoms and outcomes, and, therefore, with potential for greater benefit from M-TEER.

In the trial-level subgroup analyses, the COAPT15 and RESHAPE-HF214 trials reported no significant interaction between treatment assignment and ischemic vs non-ischemic etiology, which indicates consistency of benefit across etiologies. Similarly, the MITRA-FR16 did not identify significant heterogeneity regarding etiology across prespecified subgroups. This suggests that ischemic vs non-ischemic alone should not be used to decide the treatment option in secondary MR. Instead, clinical decisions should prioritize other factors, such as the MR severity, the degree of LV dysfunction, symptom burden, and response to optimal medical therapy.21

Furthermore, differences between trials in procedural success are noteworthy, defined as achieving MR ≤ 2+. This was substantially lower in the MITRA-FR (75.6% MR ≤ 2+ at discharge) vs the COAPT (94.8% MR ≤ 2+ at 12 months) and the RESHAPE-HF2 (90.4% MR ≤ 2+ at 12 months). Multiple reasons could have influenced such differences; for instance, operator experience requirement in MITRA-FR (≥ 5 prior MitraClip cases) vs the high-volume and more experienced centers from the other trials. Similarly, post-approval data from the U.S. SSTS/ACC TVT Registry22 demonstrated > 90% MR reduction to ≤ 2+ and survival rates consistent with trial outcomes, underscoring the reproducibility of clinical benefit in experienced centers. Furthermore, device technology is a main contributor to these differences, from first-generation clips in MITRA-FR, to second-generation in COAPT, to the fourth-generation in RESHAPE-HF2 with independent leaflet grasping and wider arm options, potentially explaining discrepant clinical results regarding MR durability and outcomes. Lastly, GDMT implementation differed across trials. MITRA-FR was conducted before the widespread use of ARNI and SGLT2 inhibitors; COAPT was conducted before the adoption of SGLT2 inhibitors; and RESHAPE-HF2 reflects the contemporary use of quadruple therapy.

Despite the differences in patient phenotypes across the trials, in many developing countries, access to transcatheter valve procedures remains limited, primarily due to their high cost and the need for specialized infrastructure.23,24 The pronounced disparities in access to cutting-edge technologies, coupled with the centralization of these procedures in a few urban centers or high-specialty hospitals, a phenomenon observed even in high-income countries such as the United States, leave a substantial portion of the population without viable treatment options.25 Consequently, the most impactful and broadly deployable intervention in these regions remains rigorous GDMT optimization and provider education. Nevertheless, patient phenotyping should still be performed to identify those who may potentially have an outsized benefit from future M-TEER referrals.

By integrating trial-level population characteristics with clinical outcomes, our work bridges the gap between isolated trial findings and real-world patient selection, thus offering a potential framework for future prospective studies and for refining guideline criteria. While the results from our study must be interpreted with caution, they provide hypothesis-generating evidence supporting the concept that patient selection, particularly considering the balance between EROA and LVEDV and GDMT optimization, may be critical to optimizing the benefit of M-TEER in secondary MR. This approach moves beyond the broad application of current guideline recommendations and points toward a more individualized strategy in which anatomical and functional parameters guide intervention.

Limitations

The primary limitation of this study lies in the heterogeneity of the populations enrolled in the randomized controlled trials, as evidenced by the I2 observed in the analyses, which may influence the overall results. Second, differences in MR severity across trials represent a key limitation. While the COAPT and RESHAPE-HF2 trials mainly included patients with MR grade 3+ or 4, MITRA-FR enrolled a more severe grade, which may partly explain divergent results. Lastly, an individual patient-level meta-analysis could not be conducted due to lack of data; this level of analysis would further increase statistical power, especially in subgroup analyses, enhancing the robustness and generalizability of the findings.

CONCLUSIONS

In this meta-analysis, M-TEER plus GDMT shows a lower risk of HF-related hospitalization vs GDMT alone. We did not find any differences in the risk of all-cause mortality, cardiac death, stroke, or myocardial infarction.

FUNDING

None declared.

ETHICAL CONSIDERATIONS

This meta-analysis was performed using data from previously published studies. As it is based entirely on secondary data, no new data were collected from human or animal participants; on the other hand, the use of SAGER guidelines was not applicable in this study. All included studies were approved from the relevant center ethics committees. The authors confirm that all data utilized were publicly accessible, and no confidential information was used without proper authorization.

STATEMENT ON THE USE OF ARTIFICIAL INTELLIGENCE

During the preparation of this work, the authors used ChatGPT 4o to review the document’s syntax and grammar. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.

AUTHORS’ CONTRIBUTIONS

D. Paulino-González: conceptualization; formal analysis, writing, review, and editing. A.L. García-Loera: methodology, investigation, writing, review, and editing. D.A. Navarro-Martínez: methodology, formal analysis. M.A. Pardiño-Vega: writing, review, and editing, supervision. K.P. Zúñiga-Montaño: investigation, writing, review, and editing.

CONFLICTS OF INTEREST

None declared.

SUPPLEMENTARY DATA

REFERENCES