Ten-year follow-up of the first left ventricular assist device implantation program in Chile: a case series Authors: Oneglio Pedemonte Villablanca, Paulina Berthelon Vega, Paulina Espinoza Reed, Lorenzo Merello Norero, Fernando Aranda Gutierrez, Julio Ibarra Castillo, Manuel Novajas Balboa, Nicolás Meza Concha, Ernesto Aranguiz Santander, Andrés Vera Peralta Published: 2026 ABSTRACT Heart failure (HF) affects over 64 million people worldwide, with prevalence projected to reach 3% by 2030. In advanced stages, when medical therapy fails and transplantation is limited by organ shortages, left ventricular assist devices (LVADs) represent a viable alternative. This retrospective, single-center study reports 10-year outcomes from the first Chilean LVAD program using the HeartWare (HVAD) device. Between 2013 and 2015, nine patients with advanced HF received HVAD implants. The mean age was 42.2 ± 15.8 years, with a left ventricular ejection fraction of 22% ± 4.1%. One-year survival was 77.8%, and 10-year survival for those using LVAD as destination therapy was 62.2%. All patients experienced at least one infection, with driveline infections being most common (55.5%). Bleeding events affected 77.7%, including gastrointestinal bleeding and severe epistaxis (both 22.2%). Neurological complications occurred in 33.3% of patients. Despite these events, four patients remained on LVAD support after 10 years, with no mechanical device failures requiring replacement. This case series highlights the feasibility, durability, and clinical relevance of long-term LVAD support in low-donation settings. Findings support LVADs as a destination therapy in selected patients and emphasize the importance of multidisciplinary care and structured follow-up.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s13019-026-04331-2. Introduction Heart failure (HF) affects between 1 and 2% of the global population. Currently, 64 million people worldwide suffer from HF [1, 2]. The prevalence of HF is estimated to increase to 3% by 2030, a trend that is observable in both international and local incidences [3]. Recently, new and innovative drugs have been recommended in clinical practice guidelines for the management of HF, as they have shown significant improvements in important outcomes such as mortality and hospitalizations [4]. However, 10 to 15% of patients with HF progress to severe presentations of the disease, with high rates of short-term mortality and limited therapeutic options [5]. Heart transplantation remains the preferred treatment for patients with advanced HF refractory to medical therapy [4]. However, donor shortages across several countries, including Chile, limit organ availability [6]. In Chile, the donation rate is below 10 donors per million inhabitants, resulting in approximately 40 to 45 heart transplants annually [6]. In parallel, the number of patients on the heart transplant waiting list typically ranges between 20 and 30 per month [6], reflecting a persistent gap between organ availability and demand. A proportion of patients with advanced HF requiring transplantation cannot be included in active waiting lists due to clinical deterioration, while others may experience unfavourable clinical results while waiting [7, 8]. Left ventricular assist devices (LVADs) are implantable mechanical pumps designed to replace the function of the left ventricle in patients with advanced HF refractory to medical therapy [9]. Originally developed as a bridge to transplantation, LVADs have since evolved into a destination therapy option for patients ineligible for heart transplantation [9, 10]. First-generation devices, such as the HeartMate I, used pulsatile-flow pumps based on pneumatic or electromechanical mechanisms. While they mimicked the physiology of the native heart, their large size and mechanical wear limited durability and safety [10]. Second-generation LVADs, such as the HeartMate II, introduced continuous-flow pumps with axial rotors. These offered a more compact design and improved mechanical durability but were associated with specific complications, including rotor thrombosis and hemolysis [9]. Third-generation devices, such as the HeartWare and the HeartMate III, incorporated continuous-flow centrifugal pumps with magnetic or hydrodynamic levitation technologies [10]. The HeartMate III, in particular, introduced advanced features such as active flow modulation to reduce the risk of thrombosis and cerebrovascular events. Its compact design enables fully intracorporeal implantation and has supported the development of minimally invasive surgical techniques [10]. These innovations have significantly expanded the indications for LVAD therapy and improved clinical outcomes, reducing device-related complications and enhancing patients’ quality of life [9, 10]. In June 2021, the production of the HeartWare device was discontinued, leaving the HeartMate III as the only third-generation LVAD available in Chile [11, 12]. This shift, combined with factors such as the global impact of the new organ allocation policy in the United States, the high cost of devices, health system restrictions during the SARS-CoV-2 pandemic, and the absence of new manufacturers, has posed additional challenges for the implementation of LVAD programs in the country [13]. It also raises concerns regarding the long-term follow-up of patients who continue to rely on the HeartWare device, a now-discontinued device model [9]. In the last few years, some studies have reported positive long-term outcomes for LVADs used as destination therapy. For instance, Mehra et al. [10] presented favorable 5-year follow-up data with the HeartMate III; Schmitto et al., in the ELEVATE study, evaluated clinical experience [14]; and in early 2024, 5-year results from the INTERMACS registry were published [15]. In all three studies, 5-year survival exceeded 60%. These findings have informed clinical practice guideline recommendations [16] and renewed the interest among manufacturers in developing new and improved devices [17]. We report a case series of patients with advanced HF included in a pilot LVAD program initiated in 2013, by presenting 10-years follow-up findings on clinical and device-related outcomes. Materials and methods Study design We conducted a retrospective, single-center case series of patients with advanced HF who have received a HeartWare LVAD since 2013 at the Dr. Gustavo Fricke Hospital (Viña del Mar, Chile). The protocol study was approved by the local Ethical-Scientific Committee (Resolution 05/2025). Participants We included all patients with advanced HF who underwent implantation of a HeartWare LVAD. Patients were eligible regardless of whether they were alive, deceased, or had undergone heart transplantation at the time of last follow-up. At the time of program initiation (2013–2015), the HeartWare device was the only LVAD available at our institution due to institutional, logistical, and economic constraints. Our program’s criteria for LVAD implantation included age between 15 and 65 years and a left ventricular ejection fraction ≤ 30%, in patients refractory to optimal medical therapy for advanced HF. Setting Patients were followed by a multidisciplinary team specialized in advanced HF and heart transplantation. The team included cardiologists, cardiac surgeons, cardiovascular anesthesiologists, specialized nurses, and rehabilitation staff. All care was delivered within the Heart Transplant Unit of the hospital. Data collection and follow-up Data were collected from both paper-based and electronic medical records, including: demographic, clinical and perioperative characteristics; laboratory results; and imaging studies. Additional data were gathered on clinical outcomes, such as mortality, hospitalizations, complications, functional status, among others, at 1-month, 1-year, 2-year, 3-year, 5-year, and 10-year follow-up timepoints, as follows: Clinical data: Demographics (sex, age, weight, etc.), comorbidities, transplantation status, survival status, pre- and postoperative clinical history, and functional capacity according to the NYHA classification.Hemodynamic and echocardiographic data: Left ventricular ejection fraction, left ventricular end-diastolic diameter, filling pressures, cardiac output, and cardiac index.Adverse events: LVAD-related complications, including infections, pump thrombosis, bleeding, right ventricular dysfunction, cerebrovascular events, and others. Clinical data: Demographics (sex, age, weight, etc.), comorbidities, transplantation status, survival status, pre- and postoperative clinical history, and functional capacity according to the NYHA classification. Clinical data: Demographics (sex, age, weight, etc.), comorbidities, transplantation status, survival status, pre- and postoperative clinical history, and functional capacity according to the NYHA classification. Hemodynamic and echocardiographic data: Left ventricular ejection fraction, left ventricular end-diastolic diameter, filling pressures, cardiac output, and cardiac index. Hemodynamic and echocardiographic data: Left ventricular ejection fraction, left ventricular end-diastolic diameter, filling pressures, cardiac output, and cardiac index. Adverse events: LVAD-related complications, including infections, pump thrombosis, bleeding, right ventricular dysfunction, cerebrovascular events, and others. Adverse events: LVAD-related complications, including infections, pump thrombosis, bleeding, right ventricular dysfunction, cerebrovascular events, and others. Statistical analysis Categorical variables were summarized as counts and percentages, and continuous variables as means ± standard deviations. Long-term survival was assessed using descriptive statistics and Kaplan–Meier curves, with right-censoring at 10 years post-implantation or heart transplantation. Analyses were performed using R (v4.3.2) via RStudio (v2024.06.0). Results Between August 2013 and December 2015, nine patients underwent implantation of left ventricular assist devices (LVADs) at Dr. Gustavo Fricke Hospital in Viña del Mar, Chile. All patients received a HeartWare device (HVAD, Medtronic Inc.). Baseline characteristics of the case series are detailed in Table 1. At the time of implantation, the mean age was 42.22 ± 15.75 years. All patients had a diagnosis of advanced heart failure with reduced ejection fraction, with a mean LVEF of 22% ± 4.1, and 77.8% were in NYHA functional class III. The most frequent comorbidities were hypertension and coronary artery disease, both present in 33.3% of the patients. Most patients were classified as INTERMACS level 2 (44.4%) or level 3 (44.4%). Preoperatively, only one patient required mechanical circulatory support with central extracorporeal membrane oxygenation (ECMO), and one patient required an intra-aortic balloon pump via femoral access. Levosimendan pulse therapy was administered prior to LVAD implantation in 6 of 9 patients (66.7%). The drug was administered as a continuous intravenous infusion of 12.5 mg diluted in 250 mL of 5% dextrose solution over 24 h, without a loading dose. In all cases, levosimendan was administered at least 24 h prior to LVAD implantation. The mean pulmonary vascular resistance was 3.75 ± 2.6 Wood units, with a range of 1.63–8.94 Wood units. Inhaled nitric oxide was used in 55.5% of patients and administered at the time of separation from cardiopulmonary bypass at an average dose of 20 ppm, and was maintained for approximately 24–36 h. Nebulized iloprost was used in 22.2% of patients to prevent early right ventricular dysfunction. Table 1Baseline characteristics of the case seriesCharacteristicPatients (N = 9)Age (years) Mean42.22 ± 15.75 Median40 (30–54)Gender, n (%) Male7 (77.8%) Female2 (22.2%)Comorbidities, n (%) Hypertension3 (33.3%) Type 2 diabetes mellitus0 (0%) Coronary artery disease3 (33.3%) Myopathies2 (22.2%) Obesity2 (22.2%) Dyslipidemia1 (11.1%) Stroke1 (11.1%)Pre-LVAD BMI (kg/m²)22.27 ± 3.92Etiology of HF, n (%) Ischemic3 (33.3%) Non-ischemic6 (66.7%)LVEF (%)22 ± 4.1 (17–29)LVEDD (mm)78.89 ± 13.25 (56–103)TAPSE (mm)18.5 ± 4.38Pre-LVAD cardiac index (L/min/m²)1.95 ± 0.84 (1–3)NYHA class, n (%) I0 II0 III7 (77.8%) IV2 (22.2%)INTERMACS level, n (%) 11 (11.1%) 24 (44.4%) 34 (44.4%) 40CRT device, n (%)2 (22.2%)ICD device, n (%)2 (22.2%)Pulmonary vascular resistance (Wood units)3.75 ± 2.6 (1.63–8.94)Pre-LVAD IABP, n (%)1 (11.1%)Pre-LVAD ECMO, n (%)1 (11.1%)Pre-LVAD levosimendan pulse, n (%)6 (66.7%)Pre-LVAD creatinine (mg/dL)0.99 ± 0.29Therapeutic strategy, n (%) Bridge to transplant2 (22.2%) Bridge to destination therapy, n (%)7 (77.8%) Bridge to decision4 (44.4%) Bridge to recovery0 Destination therapy4 (44.4%)BMI = body mass index; LVAD = left ventricular assist device; HF = heart failure; LVEF = left ventricular ejection fraction; LVEDD = left ventricular end-diastolic diameter; TAPSE = tricuspid annular plane systolic excursion; CI = cardiac index; NYHA = New York Heart Association; INTERMACS = Interagency Registry for Mechanically Assisted Circulatory Support; ICD = implantable cardioverter defibrillator; IABP = intra-aortic balloon pump; ECMO = extracorporeal membrane oxygenation; CPB = cardiopulmonary bypass Table 1Baseline characteristics of the case seriesCharacteristicPatients (N = 9)Age (years) Mean42.22 ± 15.75 Median40 (30–54)Gender, n (%) Male7 (77.8%) Female2 (22.2%)Comorbidities, n (%) Hypertension3 (33.3%) Type 2 diabetes mellitus0 (0%) Coronary artery disease3 (33.3%) Myopathies2 (22.2%) Obesity2 (22.2%) Dyslipidemia1 (11.1%) Stroke1 (11.1%)Pre-LVAD BMI (kg/m²)22.27 ± 3.92Etiology of HF, n (%) Ischemic3 (33.3%) Non-ischemic6 (66.7%)LVEF (%)22 ± 4.1 (17–29)LVEDD (mm)78.89 ± 13.25 (56–103)TAPSE (mm)18.5 ± 4.38Pre-LVAD cardiac index (L/min/m²)1.95 ± 0.84 (1–3)NYHA class, n (%) I0 II0 III7 (77.8%) IV2 (22.2%)INTERMACS level, n (%) 11 (11.1%) 24 (44.4%) 34 (44.4%) 40CRT device, n (%)2 (22.2%)ICD device, n (%)2 (22.2%)Pulmonary vascular resistance (Wood units)3.75 ± 2.6 (1.63–8.94)Pre-LVAD IABP, n (%)1 (11.1%)Pre-LVAD ECMO, n (%)1 (11.1%)Pre-LVAD levosimendan pulse, n (%)6 (66.7%)Pre-LVAD creatinine (mg/dL)0.99 ± 0.29Therapeutic strategy, n (%) Bridge to transplant2 (22.2%) Bridge to destination therapy, n (%)7 (77.8%) Bridge to decision4 (44.4%) Bridge to recovery0 Destination therapy4 (44.4%)BMI = body mass index; LVAD = left ventricular assist device; HF = heart failure; LVEF = left ventricular ejection fraction; LVEDD = left ventricular end-diastolic diameter; TAPSE = tricuspid annular plane systolic excursion; CI = cardiac index; NYHA = New York Heart Association; INTERMACS = Interagency Registry for Mechanically Assisted Circulatory Support; ICD = implantable cardioverter defibrillator; IABP = intra-aortic balloon pump; ECMO = extracorporeal membrane oxygenation; CPB = cardiopulmonary bypass Baseline characteristics of the case series BMI = body mass index; LVAD = left ventricular assist device; HF = heart failure; LVEF = left ventricular ejection fraction; LVEDD = left ventricular end-diastolic diameter; TAPSE = tricuspid annular plane systolic excursion; CI = cardiac index; NYHA = New York Heart Association; INTERMACS = Interagency Registry for Mechanically Assisted Circulatory Support; ICD = implantable cardioverter defibrillator; IABP = intra-aortic balloon pump; ECMO = extracorporeal membrane oxygenation; CPB = cardiopulmonary bypass Perioperative characteristics A less invasive surgical (LIS) approach was defined as any thoracotomy-based or limited-access technique avoiding full sternotomy, including left anterolateral thoracotomy, in accordance with previously published approaches. In eight patients (88.9%), a LIS approach was used for LVAD implantation, consisting of a left anterolateral thoracotomy combined with a superior hemisternotomy, both performed through minimal incisions. A full sternotomy was performed in only one patient. The mean cardiopulmonary bypass time was 65 ± 18.8 min. All devices were implanted intrapericardially. Two patients (22.2%) required surgical reexploration for perioperative diffuse bleeding within the first 24 h. No patient required ECMO or IABP support after implantation. One late postoperative death occurred due to ventilator-associated pneumonia and multiorgan failure. The average hospital stay until discharge was 39.89 ± 58.13 days. Table 2 summarizes the main pre-, intra-, and postoperative characteristics of the case series. Table 2Perioperative and postoperative characteristics of LVAD patientsCharacteristicValue/n (%)CPB time (min)65 ± 18.8 (45–108)Post-LVAD inhaled nitric oxide use, n (%)5 (55.5%)Post-LVAD iloprost use, n (%)2 (22.2%)Post-LVAD ECMO, n (%)0Surgical technique LIS: Left anterolateral thoracotomy and superior hemisternotomy8 (88.9%) Full sternotomy1 (11.1%)Postoperative complications, n (%) Surgical reintervention for diffuse bleeding2 (22.2%) Post-LVAD atrial fibrillation2 (22.2%) Renal failure0% Gastrointestinal bleeding0% Infection1 (11.1%) Stroke1 (11.1%) Death1 (11.1%) RVAD implantation for right ventricular failure0%Hospital length of stay until discharge (days)39.89 ± 58.13LVAD = left ventricular assist device; CPB = cardiopulmonary bypass; ECMO = extracorporeal membrane oxygenation; LIS = less invasive surgical; RVAD = right ventricular assist device; RV = right ventricle Table 2Perioperative and postoperative characteristics of LVAD patientsCharacteristicValue/n (%)CPB time (min)65 ± 18.8 (45–108)Post-LVAD inhaled nitric oxide use, n (%)5 (55.5%)Post-LVAD iloprost use, n (%)2 (22.2%)Post-LVAD ECMO, n (%)0Surgical technique LIS: Left anterolateral thoracotomy and superior hemisternotomy8 (88.9%) Full sternotomy1 (11.1%)Postoperative complications, n (%) Surgical reintervention for diffuse bleeding2 (22.2%) Post-LVAD atrial fibrillation2 (22.2%) Renal failure0% Gastrointestinal bleeding0% Infection1 (11.1%) Stroke1 (11.1%) Death1 (11.1%) RVAD implantation for right ventricular failure0%Hospital length of stay until discharge (days)39.89 ± 58.13LVAD = left ventricular assist device; CPB = cardiopulmonary bypass; ECMO = extracorporeal membrane oxygenation; LIS = less invasive surgical; RVAD = right ventricular assist device; RV = right ventricle Perioperative and postoperative characteristics of LVAD patients LVAD = left ventricular assist device; CPB = cardiopulmonary bypass; ECMO = extracorporeal membrane oxygenation; LIS = less invasive surgical; RVAD = right ventricular assist device; RV = right ventricle Survival outcomes Cumulative survival was 77.8% at both 1 and 2 years, and 62.2% at 5 and 10 years post-implantation, with heart transplantation treated as a censoring event. Figure 1 shows the Kaplan–Meier survival curve. Fig. 1Kaplan–Meier survival analysis by subgroup Kaplan–Meier survival analysis by subgroup Among the nine patients included in this case series, three deaths occurred during follow-up at 3.57, 45.5, and 230.29 weeks after implantation. Two patients underwent heart transplantation and were censored at the time of transplant, while the remaining four patients were alive with LVAD support at the end of the 10-year follow-up and were censored accordingly. The average duration of support among patients still living with the device was approximately 10.8 ± 0.8 years. Table 3 summarizes patient status over time, including survival and transplantation status throughout the follow-up period. Table 3Patient status over a 10-year follow-up periodStatus10-year follow-up of the case series (n = 9)1 Month1 Year2 Years5 Years10 YearsWith LVAD97644Deaths02233Transplanted00122Explanted00000Survival of LVAD Destination Therapy Group (%)100%77.8%75%62.2%62.2%LVAD = left ventricular assist device Table 3Patient status over a 10-year follow-up periodStatus10-year follow-up of the case series (n = 9)1 Month1 Year2 Years5 Years10 YearsWith LVAD97644Deaths02233Transplanted00122Explanted00000Survival of LVAD Destination Therapy Group (%)100%77.8%75%62.2%62.2%LVAD = left ventricular assist device Patient status over a 10-year follow-up period LVAD = left ventricular assist device Adverse events Three patients died during the 10-year follow-up period. The first died early (at 3.57 weeks post-implantation) from ventilator-associated pneumonia caused by Klebsiella pneumoniae and multiorgan failure, in the context of prolonged hospitalization and support with central ECMO. The second died at 45.5 weeks due to aspiration pneumonia following an extensive ischemic stroke in the territory of the right middle cerebral artery, diagnosed on day 4 after LVAD implantation. The third died at 230.2 weeks due to multiple cerebral infarctions with hemorrhagic transformation following pump thrombosis, having previously developed a bilateral non-aneurysmal subarachnoid hemorrhage that required anticoagulation reduction two months before hospitalization. The most frequent complications over the 10-year follow-up are summarized in Table 4. The detailed longitudinal distribution of complications over time is provided in Supplementary Information. Table 4Complications and adverse events during LVAD follow-upComplication categorySpecific complicationTotal patients affected (n = 9)Bleeding complicationsDiffuse perioperative bleeding2 (22.2%)Gastrointestinal bleeding2 (22.2%)Epistaxis2 (22.2%)Urinary bleeding1 (11.1%)Pump thrombosis2 (22.2%)Technical complicationsOutflow graft kinking1 (11.1%)Driveline granuloma2 (22.2%)Outflow graft stenosis1 (11.1%)Neurological eventsIschemic stroke2 (22.2%)Hemorrhagic stroke1 (11.1%)Infectious complications (IMACS classification) Early infections (≤ 3 months)Driveline infection1 (11.1%)Thoracotomy site infection1 (11.1%)Respiratory infection1 (11.1%)Infectious complications (IMACS classification)Late infections (> 3 months)Driveline infection4 (44.4%)Non-LVAD related infection5 (55.5%)Other complicationsSevere RV failure1 (11.1%)Aortic regurgitation > mild1 (11.1%)Cataracts1 (11.1%)Glaucoma1 (11.1%)Intercostal neuralgia1 (11.1%)Values represent the total number of patients who experienced each complication at any time during follow-upLVAD = left ventricular assist device; GI = gastrointestinal; RV = right ventricle; Aortic regurgitation > mild = more than mild aortic insufficiency; Stroke = cerebrovascular accident Table 4Complications and adverse events during LVAD follow-upComplication categorySpecific complicationTotal patients affected (n = 9)Bleeding complicationsDiffuse perioperative bleeding2 (22.2%)Gastrointestinal bleeding2 (22.2%)Epistaxis2 (22.2%)Urinary bleeding1 (11.1%)Pump thrombosis2 (22.2%)Technical complicationsOutflow graft kinking1 (11.1%)Driveline granuloma2 (22.2%)Outflow graft stenosis1 (11.1%)Neurological eventsIschemic stroke2 (22.2%)Hemorrhagic stroke1 (11.1%)Infectious complications (IMACS classification) Early infections (≤ 3 months)Driveline infection1 (11.1%)Thoracotomy site infection1 (11.1%)Respiratory infection1 (11.1%)Infectious complications (IMACS classification)Late infections (> 3 months)Driveline infection4 (44.4%)Non-LVAD related infection5 (55.5%)Other complicationsSevere RV failure1 (11.1%)Aortic regurgitation > mild1 (11.1%)Cataracts1 (11.1%)Glaucoma1 (11.1%)Intercostal neuralgia1 (11.1%)Values represent the total number of patients who experienced each complication at any time during follow-upLVAD = left ventricular assist device; GI = gastrointestinal; RV = right ventricle; Aortic regurgitation > mild = more than mild aortic insufficiency; Stroke = cerebrovascular accident Complications and adverse events during LVAD follow-up Infectious complications (IMACS classification) Late infections (> 3 months) Values represent the total number of patients who experienced each complication at any time during follow-up LVAD = left ventricular assist device; GI = gastrointestinal; RV = right ventricle; Aortic regurgitation > mild = more than mild aortic insufficiency; Stroke = cerebrovascular accident Infectious complications were the most common adverse events and were classified as early (≤ 3 months post-implant) or late (> 3 months), according to IMACS definitions. All patients experienced at least one infection that required hospitalization. Driveline infections occurred in 5 patients (55.5%) within the first 5 years; two cases required surgical debridement for deep tissue involvement, with no progression to systemic infection. Non-LVAD-related infections, such as respiratory and gastrointestinal infections, also affected 55.5% of patients, mostly during the later follow-up period. The most frequently identified microorganism in driveline infections was Staphylococcus epidermidis, while respiratory and gastrointestinal infections were mainly associated with common pathogens including influenza virus, Klebsiella pneumoniae, and Clostridium species. Bleeding complications affected 77.7% of patients. Two experienced postoperative bleeding requiring surgical reintervention. Gastrointestinal bleeding occurred in two patients at late follow-up, one at approximately 1 year and another at 3 years after LVAD implantation, including one case of colonic angiodysplasia managed endoscopically. Epistaxis was reported in two patients: one resolved after INR adjustment; the other required multiple cauterizations, arterial ligation, and ultimately sphenopalatine artery embolization due to severe recurrent bleeding at year 10 (Video 1). This case required careful anticoagulation adjustment, including continuous intravenous heparin infusion and temporary withdrawal of antiplatelet therapy to prevent pump thrombosis. One patient had hematuria associated with a supratherapeutic INR. Screening for acquired von Willebrand syndrome was not performed due to resource limitations. Neurological complications occurred in 3 patients (33.3%): two ischemic strokes (one at 4 days post-implantation and another at 1 year due to a thrombus in the apical septal region, without neurological sequelae), and one bilateral non-aneurysmal subarachnoid hemorrhage around year 4. Pump thrombosis occurred in 2 patients (22.2%), both with subtherapeutic INR levels. One event occurred during late follow-up at approximately 5 years after LVAD implantation. Both cases were successfully treated with tenecteplase administered as a single intravenous bolus over 5–10 s at a dose of 0.5 mg/kg (maximum 50 mg). In both cases thrombolysis was successful, and the device continued functioning without the need for pump exchange. One patient subsequently died due to multiple cerebral infarctions, likely related to thrombolysis, rather than device dysfunction. Other complications included one case of severe right ventricular failure and severe aortic insufficiency, leading to heart transplantation at 2.2 years post-implantation. Two rare complications were also documented: (1) one patient developed progressive kinking and stenosis of the outflow cannula, requiring sequential stent placement in years three and four (Video 2), and (2) another patient developed intercostal neuralgia due to chronic mechanical contact between the pump and the chest wall, successfully managed with intercostal nerve blockade (Fig. 2). Among long-term survivors, age-related complications such as cataracts also emerged, requiring ophthalmologic surgery with specialized perioperative planning by the LVAD team. Fig. 2Intercostal neuritis secondary to mechanical compression by HVAD after 10 years of support.Images 1 A and 1B show AP and lateral chest X-rays with appropriate device positioning. Image 2 is a bone scintigraphy scan highlighting a focal osteoblastic lesion in the left 8th rib (red circle), likely of traumatic origin. Images 3 A and 3B display mediastinal and lung windows of a chest CT, revealing direct contact between the device and the thoracic wall, consistent with periostitis Intercostal neuritis secondary to mechanical compression by HVAD after 10 years of support.Images 1 A and 1B show AP and lateral chest X-rays with appropriate device positioning. Image 2 is a bone scintigraphy scan highlighting a focal osteoblastic lesion in the left 8th rib (red circle), likely of traumatic origin. Images 3 A and 3B display mediastinal and lung windows of a chest CT, revealing direct contact between the device and the thoracic wall, consistent with periostitis No patient required LVAD replacement due to technical failure of the device. At the end of the 10-year follow-up, four patients remain on their original HVAD device; three are in NYHA functional class I, and none has opted to be listed for heart transplantation. Electromechanical complications are summarized in Table 5. One-third of patients experienced driveline cable jacket rupture. Temporary disconnection due to driveline contamination occurred in 44.4% of cases. Controller replacements were performed in 77% of patients due to software updates, and in 44% due to internal battery depletion. Table 5Electromechanical complications during 10-year LVAD follow-upComplicationNo. of events (N = 9), n (%)Driveline cable fracture3 (33.3%)Driveline cable contamination4 (44.4%)Controller software update7 (77.8%)Internal battery depletion4 (44.4%)LVAD = left ventricular assist device Table 5Electromechanical complications during 10-year LVAD follow-upComplicationNo. of events (N = 9), n (%)Driveline cable fracture3 (33.3%)Driveline cable contamination4 (44.4%)Controller software update7 (77.8%)Internal battery depletion4 (44.4%)LVAD = left ventricular assist device Electromechanical complications during 10-year LVAD follow-up LVAD = left ventricular assist device Device-related alarms documented during follow-up included both medium- and high-priority alerts. High-priority alarms were associated with pump stoppage, controller failure, or power loss. Medium-priority alarms included low flow, high power, suction, electrical faults, and early signs of controller malfunction. The distribution of these events is shown in Table 6. Table 6Device alarm events of medium and high priority over 10-year follow-upPatientSustained low flowHigh wattageElectrical failureController failureCritical batteryController exchangeAverage events per patientPatient 10211431.83 ± 1.47Patient 20000100.17 ± 0.41Patient 32200532.0 ± 1.90Patient 40001331.17 ± 1.47Patient 51110010.67 ± 0.52Patient 61301221.50 ± 1.05Patient 72000020.67 ± 1.03Patient 85011142.0 ± 2.0Patient 90000200.33 ± 0.82Total Events, N (%)11 (17.7%)8 (12.9%)3(4.8%)4(6.45%)18 (29%)18(29%) Table 6Device alarm events of medium and high priority over 10-year follow-upPatientSustained low flowHigh wattageElectrical failureController failureCritical batteryController exchangeAverage events per patientPatient 10211431.83 ± 1.47Patient 20000100.17 ± 0.41Patient 32200532.0 ± 1.90Patient 40001331.17 ± 1.47Patient 51110010.67 ± 0.52Patient 61301221.50 ± 1.05Patient 72000020.67 ± 1.03Patient 85011142.0 ± 2.0Patient 90000200.33 ± 0.82Total Events, N (%)11 (17.7%)8 (12.9%)3(4.8%)4(6.45%)18 (29%)18(29%) Device alarm events of medium and high priority over 10-year follow-up 3 (4.8%) 4 (6.45%) 18 (29%) Table 7Longitudinal distribution of complications and adverse events during LVAD follow-upFollow-up time1 Month(n = 9)1 Year(n = 7)2 Years(n = 7)3 Years(n = 7)5 Years(n = 5)10 Years(n = 4)Bleeding, n (%) Postoperative2 (22.2%)00000 Gastrointestinal01 (14.2%)01 (14.2%)00 Epistaxis00001 (20%)1 (25%) Urinary000001 (25%) Pump thrombosis, n (%)001 (14.2%)01 (20%)0Technical complications, n (%) Outflow cannula kinking0001 (14.2%)00 Driveline granuloma01 (14.2%)001 (20%)0 Outflow cannula stenosis00001 (20%)0LVAD-related infection, n (%) Driveline infection1 (11.1%)1 (14.2%)1 (14.2%)1 (14.2%)1 (20%)0 Thoracotomy site infection1 (11.1%)00000Non-LVAD related infection, n (%) Respiratory1 (11.1%)1 (14.2%)1 (14.2%)1 (14.2%)02 (50%) Digestive02 (28.5%)1 (14.2%)01 (20%)1 (25%) Severe RV failure001 (14.2%)000Stroke, n (%) Hemorrhagic00001 (20%)0 Ischemic1 (11.1%)1 (14.2%)0000 Aortic regurgitation > mild, n (%)001 (14.2%)000Ocular complications, n (%) Cataracts000001 (25%) Glaucoma000001 (25%)Other, n (%) Intercostal neuralgia000001 (25%)The number of patients (n) was adjusted throughout the follow-up period based on deaths or transplantsLVAD = left ventricular assist device; GI = gastrointestinal; RV = right ventricle; Aortic regurgitation > mild = more than mild aortic insufficiency; Stroke = cerebrovascular accident Table 7Longitudinal distribution of complications and adverse events during LVAD follow-upFollow-up time1 Month(n = 9)1 Year(n = 7)2 Years(n = 7)3 Years(n = 7)5 Years(n = 5)10 Years(n = 4)Bleeding, n (%) Postoperative2 (22.2%)00000 Gastrointestinal01 (14.2%)01 (14.2%)00 Epistaxis00001 (20%)1 (25%) Urinary000001 (25%) Pump thrombosis, n (%)001 (14.2%)01 (20%)0Technical complications, n (%) Outflow cannula kinking0001 (14.2%)00 Driveline granuloma01 (14.2%)001 (20%)0 Outflow cannula stenosis00001 (20%)0LVAD-related infection, n (%) Driveline infection1 (11.1%)1 (14.2%)1 (14.2%)1 (14.2%)1 (20%)0 Thoracotomy site infection1 (11.1%)00000Non-LVAD related infection, n (%) Respiratory1 (11.1%)1 (14.2%)1 (14.2%)1 (14.2%)02 (50%) Digestive02 (28.5%)1 (14.2%)01 (20%)1 (25%) Severe RV failure001 (14.2%)000Stroke, n (%) Hemorrhagic00001 (20%)0 Ischemic1 (11.1%)1 (14.2%)0000 Aortic regurgitation > mild, n (%)001 (14.2%)000Ocular complications, n (%) Cataracts000001 (25%) Glaucoma000001 (25%)Other, n (%) Intercostal neuralgia000001 (25%)The number of patients (n) was adjusted throughout the follow-up period based on deaths or transplantsLVAD = left ventricular assist device; GI = gastrointestinal; RV = right ventricle; Aortic regurgitation > mild = more than mild aortic insufficiency; Stroke = cerebrovascular accident Longitudinal distribution of complications and adverse events during LVAD follow-up The number of patients (n) was adjusted throughout the follow-up period based on deaths or transplants LVAD = left ventricular assist device; GI = gastrointestinal; RV = right ventricle; Aortic regurgitation > mild = more than mild aortic insufficiency; Stroke = cerebrovascular accident Discussion This retrospective case series describes the long-term clinical outcomes of the first nine patients with advanced heart failure who received a HeartWare left ventricular assist device in Chile. At 10 years of follow-up, we observed a cumulative survival rate of 62.2%, with four patients still alive and supported by their original HVAD device, without requiring replacement. Moreover, most surviving patients were in NYHA functional class I, reflecting a good quality of life. Compared to other studies, our results are consistent. For example, the ELEVATE study reported a 5-year survival rate of 63.3% with the HeartMate III [14]. Similarly, when comparing our survival data with heart transplantation, the ISHLT registry reports a post-transplant median survival of 12.2 years and a 5-year survival of 70% [15], suggesting that, in selected populations, LVAD therapy can achieve survival outcomes comparable to heart transplantation. In particular, among selected patients who are not transplant candidates, such as those older than 75 years, a 3-year survival rate close to 34% has been reported [18], reinforcing the value of LVADs as a viable destination therapy in this population [19]. In this context, our case series demonstrates that LVAD therapy is a feasible and durable strategy, even in health systems with limited organ availability for transplantation, such as in Chile [11, 12]. Beyond these comparisons, our findings should be interpreted in the context of contemporary durable LVAD data [14]. The MOMENTUM 3 randomized trial reported a 5-year Kaplan–Meier overall survival of 58.4% in patients supported with a centrifugal-flow device [10]. In the STS Intermacs 2023 annual report [15], patients supported with magnetically levitated devices showed a 5-year survival of 64%. More recently, Hanke et al. [20]. reported the longest single-center 10-year follow-up after HeartMate 3 implantation, in which, among the first 8 implanted patients, 3 had died and 2 remained on device support at 10 years. Although direct comparison is limited by differences in device generation, study design, implantation era, patient selection, and access to transplantation, these data suggest that our long-term outcomes fall within the range of contemporary durable LVAD experience. In our view, several factors may have contributed to the favorable long-term outcomes observed in our cohort, including careful patient selection, sustained multidisciplinary follow-up, strict anticoagulation surveillance, and timely management of complications. Our findings support the evolution of LVADs from a bridge to transplant toward a destination therapy, as proposed by the latest international guidelines [4, 16]. The observed survival, together with the achieved functionality and quality of life, reinforces the utility of LVADs in settings where access to transplantation is limited or when it is not indicated due to age or comorbidities. Regarding adverse events, infections were the most frequent complication including both LVAD-related and non–LVAD-related infectious episodes, affecting 100% of patients at some point during follow-up. This finding aligns with previous reports that highlight driveline-associated infections as one of the main challenges in long-term support [21]. In this context, the development of wireless or transcutaneous energy transmission technologies promises to revolutionize device management by reducing driveline-related infections [22]. Hemorrhagic complications, particularly gastrointestinal and nasal, were also frequent, as described in other long-term follow-up cohorts [23, 24]. One patient required embolization of the sphenopalatine artery for severe epistaxis at 10 years post-implant, a rarely reported situation. Neurological complications occurred in one-third of the patients, including cerebral infarctions and a non-aneurysmal bilateral subarachnoid hemorrhage. These findings underscore the need for fine and dynamic adjustment of anticoagulation therapy, also considering the possible appearance of acquired von Willebrand disease. Pump thrombosis was observed in two patients, both associated with subtherapeutic INR levels, reaffirming the importance of strict anticoagulation control. Both cases were successfully treated with tenecteplase, although one later resulted in death due to neurological complications. This conservative treatment approach, without device replacement, has been previously described as an effective alternative in selected scenarios [23]. An unusual complication was the progressive kinking of the outflow graft, which became clinically relevant at 3 years post-implant and required sequential endovascular treatment at years 3 and 4. Outflow graft obstruction is an uncommon but increasingly recognized LVAD complication, and kinking has been associated in the literature with excessive graft length, graft angulation, and anatomical disposition rather than with a single clearly established mechanism [25–28]. In our case, although a contribution from the surgical approach cannot be excluded, the event was likely multifactorial, related to graft length, angulation, adhesion formation, fibrin and local anatomical factors over time. This case also underscores the importance of multimodality imaging and longitudinal surveillance for early diagnosis of late outflow graft complications. Another notable case involved intercostal neuralgia due to chronic mechanical contact between the pump and the chest wall, successfully resolved with intercostal nerve blockade. This type of event illustrates how non-cardiovascular complications can significantly impact quality of life in patients with long-term support. The market withdrawal of the HeartWare device in 2021 represented a major challenge for ventricular assist programs worldwide, affecting more than 4,000 patients dependent on this device and limiting its use in specific populations such as pediatric patients or small-sized adults [29, 30]. Despite this situation, our results show that even with this device, prolonged survival was achievable, provided that close monitoring, individualized therapeutic adjustments, and a skilled multidisciplinary team were in place. Next-generation devices, such as the HeartMate III, have shown improved hemocompatibility, reducing bleeding and thrombotic events, according to recent studies including the ARIES trial [31–34]. Although there are no randomized studies directly comparing the HeartWare to the HeartMate III, observational studies have suggested a significant reduction in adverse events with the latter, as well as the potential to avoid antiplatelet use, further decreasing hemorrhagic complications [31–34]. Our experience suggests that less invasive surgical techniques, such as left anterolateral thoracotomy, may contribute to a lower incidence of right ventricular failure, as previously reported [35, 36]. In addition, Rojas et al. [37] compared this approach with conventional full sternotomy for third-generation LVAD implantation and observed less postoperative bleeding, a reduced need for inotropic support, and a trend toward better 2-year outcomes in patients undergoing less invasive implantation. None of the patients in our series developed right ventricular dysfunction beyond the initial implant phase, which may be related to the surgical technique described by Strueber et al. in 2011 [35], preserving right ventricular architecture and minimizing hemodynamic changes associated with left ventricular support. Although our sample size does not allow firm conclusions, these observations support the potential clinical value of minimally invasive implantation strategies in selected patients. In conclusion, this first LVAD program in Chile has shown exceptional results, exceeding initial survival expectations. In addition, the early use of durable mechanical support before transplantation has been proposed as a strategy that could lead to a net extension of life, reinforcing the value of these devices beyond their role as a bridge to transplant [38]. Together with technological advances and well-structured multidisciplinary programs, LVAD support stands out as a therapeutic tool capable of significantly extending longevity in patients with advanced heart failure [38–41]. These achievements have been possible thanks to rigorous patient education and a highly trained nursing team, ensuring timely care and efficient resolution of alerts and concerns. We believe that such programs cannot operate effectively without the support of a solid and well-trained multidisciplinary team. Limitations This retrospective, single-center study with a small sample size limits the generalizability of its findings. The exclusive use of the HVAD device and absence of a control group reduce comparative insight, while unperformed tests—such as those for acquired von Willebrand syndrome—further limit interpretation. The small number of patients supported beyond 10 years reduces statistical power. Moreover, the market withdrawal of the HVAD hinders extrapolation to current devices. Future multicenter, randomized trials are needed to evaluate the long-term outcomes of newer-generation LVADs. Strengths, clinical value, and future directions This is the first report of a 10-year LVAD follow-up in Chile, demonstrating the feasibility of long-term mechanical circulatory support in a resource-constrained setting. Its primary strength lies in the extended follow-up with detailed clinical outcomes, which adds valuable regional data to the global LVAD literature. Clinically, our findings support the use of LVADs as a destination therapy in selected non-transplant candidates and emphasize the importance of structured multidisciplinary care. Future directions include expanding the study through multicenter collaboration and integrating newer-generation devices to evaluate long-term safety, quality of life, and comparative effectiveness. Emerging evidence suggests that in younger patients, LVAD survival may approach that of heart transplantation, reinforcing its evolving role in advanced heart failure management. Conclusion Despite its limitations, this case series provides valuable evidence on the feasibility, safety, and efficacy of long-term LVAD support within a resource-limited healthcare system. The observed survival is comparable to international reports, and complications were successfully managed without the need for device replacement. These findings support the role of LVADs as a viable long-term durable left ventricular support for patients with advanced heart failure in Latin America, particularly in settings with limited access to heart transplantation. Supplementary Information Supplementary Material 1: Video 1. Management of severe epistaxis in a patient 10 years after LVAD implantation. Supplementary Material 1: Video 1. Management of severe epistaxis in a patient 10 years after LVAD implantation. Supplementary Material 2: Video 2. Management of outflow graft kinking and in-stent restenosis in a patient with LVAD. Supplementary Material 2: Video 2. Management of outflow graft kinking and in-stent restenosis in a patient with LVAD. Supplementary Table Supplementary Table