Indications and outcomes of ventricular assist devices in children and young adults with neuromuscular diseases: A systematic review

Abstract

Background

Neuromuscular diseases (NMDs) are frequently complicated by progressive cardiomyopathy leading to advanced heart failure. Ventricular assist device (VAD) implantation has emerged as a potential option in selected pediatric and young adult patients, but evidence on indications and outcomes is limited. We aimed to systematically review evidence on VAD implantation in children and young adults with NMDs, focusing on device strategies and clinical outcomes.

Methods

Systematic review following Cochrane methodology and reported per the PRISMA guidelines. Six major biomedical databases were searched from inception to October 1, 2025. Studies including patients aged ≤25 years with NMDs undergoing VAD implantation were eligible. The primary outcome was clinical status at last reported follow-up, classified as alive on device support, alive after heart transplantation (HT), or deceased.

Results

Six retrospective studies comprising 37 patients were included. Duchenne muscular dystrophy was the most common diagnosis (48.6%). Destination therapy was the most frequently reported strategy (45.9%), followed by bridge-to-transplant (37.8%) and bridge-to-decision (16.2%). At last follow-up, 75.7% were alive: 43.2% on device support and 32.4% after undergoing HT; 24.3% had died. Complications were common, most frequently bleeding and infection, whereas neurologic events were the leading cause of death. Reporting of functional outcomes and quality of life was sparse and heterogeneous.

Conclusions

Evidence on VAD use in this population is limited to small retrospective cohorts. Although most patients were alive at last reported follow-up, morbidity remains substantial and patient-centered outcomes are poorly characterized. Prospective multicenter studies with standardized outcome definitions are needed.

Introduction

Neuromuscular diseases (NMDs) encompass a heterogeneous group of progressive disorders primarily affecting skeletal muscle and frequently complicated by cardiac involvement. Dilated cardiomyopathy (DCM) and conduction abnormalities are common and contribute substantially to morbidity and mortality, with heart failure (HF) emerging as a leading cause of death.1, 2 Advances in respiratory support, corticosteroid therapy, and multidisciplinary care have improved overall survival in several NMDs, particularly Duchenne muscular dystrophy (DMD), extending life expectancy into adulthood and shifting the clinical burden toward the management of advanced cardiac disease.2, 3 Nearly all patients with DMD develop DCM and arrhythmias over the disease course, and cardiac complications increasingly represent a major determinant of prognosis.4, 5

Despite optimized medical therapy, disease progression may culminate in refractory HF requiring advanced cardiac support. Heart transplantation (HT) is considered the definitive treatment for end-stage HF;⁶ however, many patients with NMD are deemed ineligible due to progressive neuromuscular decline, respiratory involvement, comorbidities, and concerns regarding post-transplant outcomes and organ allocation.6, 7 This therapeutic gap underscores the growing need to evaluate alternative advanced HF strategies in this population.

Ventricular Assist Devices (VADs) have emerged as a viable option for pediatric and young adult patients with NMD and severe HF. VADs may be implanted as bridge-to-transplantation (BTT) in patients considered eligible for HT, as destination therapy (DT) in those not expected to undergo transplantation, or as bridge-to-decision (BTD) when transplant candidacy or long-term goals of care remain uncertain.⁸

While DT is an established strategy in selected adult populations,⁹ its role in children and young adults with progressive NMD remains incompletely defined. The decision to pursue VAD implantation in patients with NMD is complex and raises unresolved questions regarding long-term survival, quality of life (QoL), progression of non-cardiac disease, and ethical considerations surrounding resource allocation.⁶ Existing evidence is limited and is largely derived from case reports and small retrospective series. Nevertheless, advances in VAD technology, including smaller and more durable devices with improved hemocompatibility, may have expanded the feasibility of VAD support in carefully selected patients.¹⁰

We, therefore, conducted a systematic review to characterize patients with NMDs receiving VAD support, their indications and device strategies, their survival and complication profiles, and the gaps in the current evidence base.

Materials and methods

Design

We conducted a systematic review on pediatric and young adults with NMDs who underwent VAD implantation, following Cochrane methodology¹¹ and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines¹² (Table S1). The protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO; www.crd.york.ac.uk; No. CRD42024512521). Institutional Review Board approval was not required for this systematic review.

Inclusion and exclusion criteria

The inclusion criteria were as follows: (a) study population including patients aged 25 years or younger with NMDs who underwent VAD implantation for refractory HF, and (b) study design, including randomized controlled trials, non-randomized trials, and observational cohort studies (both prospective and retrospective), as well as conference abstracts and proceedings (Table S2).

The exclusion criteria were as follows: (a) meta-analyses and reviews, editorials, commentaries, book chapters, and letters, (b) studies for which the primary outcome could not be extracted, and (c) studies involving only adults older than 25 years.

When studies included mixed pediatric and adult populations (>25 years), we included them only if an age-stratified estimate of the primary outcome was available. Other outcomes were reported for the overall cohort and interpreted descriptively when age stratification was not possible.

Search strategy

Two key concepts informed our search strategy: (i) patients with NMDs and HF and (ii) patients who underwent VAD implantation. We performed comprehensive searches of the following databases from inception through October 1st, 2025: Ovid MEDLINE (and Epub ahead of Print, In-Data-Review, In-Process), Embase, Cumulative Index to Nursing and Allied Health Literature (CINAHL), Web of Science, Scopus, and the Cochrane Central Register of Controlled Trials. We included all studies regardless of publication date. Only studies published in English were included. Full search strategies are reported in Table S3.

Data management and study selection

Citations were exported to Covidence (www.covidence.org) for title and abstract screening, full-text review, and data extraction. Three authors independently performed title and abstract screening to identify potentially relevant studies, followed by full-text review of selected studies. A fourth author resolved differences with reference to the eligibility criteria.

When multiple publications from the same institution or registry reported potentially overlapping patient cohorts, studies were cross-checked by comparing study period, center, and patient characteristics. To avoid double counting, we retained for data analysis the most comprehensive report on VAD-related outcomes. The other studies were reported for descriptive purposes (Table S7).

Risk of bias

We independently assessed the Risk of Bias (RoB) of the included studies using the Joanna Briggs Institute (JBI) Critical Appraisal Checklists for Case Series and Cohort Studies (Tables S4.1 and S4.2).13, 14 Two investigators independently evaluated each study across the respective methodological domains, according to the design-appropriate checklist. Disagreements were resolved by consensus with a third investigator. Study quality informed the interpretation of findings and was not used as an exclusion criterion. Given the rarity of NMDs requiring VAD implantation, we anticipated small observational studies and considered this design methodologically appropriate.

Data collection

Extracted variables included the study design, inclusion criteria, sample size, type of NMD, patient age and sex, type of VAD used, and support intent (DT, BTT, or BTD).

When relevant data were unavailable or unclear, the corresponding authors of the included studies were contacted by email for clarification. If requested data were not obtained, the study was either excluded from analysis or included only for outcomes that were explicitly reported.

Outcomes

The primary outcome was clinical status at last follow-up after VAD implantation, defined as alive on device support, alive after HT or deceased. For descriptive purposes, patients alive on device support and those alive after HT were combined as alive at last reported follow-up. Secondary outcomes, if available, included: duration of VAD support (“Time on VAD”), length of Intensive Care Unit (ICU) and total hospitalization, functional status pre- and post-implantation, need for mechanical ventilation pre- and post-VAD, QoL measures, adverse events and long-term complications (e.g., infections, thrombosis, stroke, right heart failure, and bleeding), and cost-effectiveness analyses, when reported.

Statistical analysis

We used a descriptive statistical approach to summarize the findings from the included studies. Categorical variables were summarized as frequencies and percentages, while continuous variables were summarized as means with standard deviations or medians with interquartile ranges, according to data distribution. Clinical status at last follow-up was described according to support intent (DT, BTT, or BTD).

Due to small sample sizes, clinical heterogeneity, and overlap of patient cohorts, we did not perform a quantitative meta-analysis. Instead, a structured narrative synthesis was conducted, focusing on feasibility, intent of support, survival patterns, and complication profiles. Case reports were retained for qualitative description but excluded from the quantitative synthesis due to susceptibility to publication bias.

Results

Study selection

The literature search identified 2999 records. After removal of 998 duplicates, 2001 records underwent title and abstract screening, resulting in the exclusion of 1915 records. Full text was sought for 86 studies but was not retrievable for one; of the 85 studies reviewed in full, 67 did not meet inclusion criteria (Figure 1). Following detailed assessment of 18 studies, five studies were excluded because of potential cohort overlap (Table S7), and seven case reports were retained for descriptive purposes only (Table S8),15, 16, 17, 18, 19, 20, 21 yielding six studies22, 23, 24, 25, 26, 27 for the quantitative data synthesis (Table 1).

Figure 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram.

Table 1.

Characteristics of Studies Reporting Ventricular Assist Device (VAD) Implantation in Children and Young Adults with Neuromuscular Disease (NMD)

First author, year	Study design	Single vs multi-center,Study period, country	Population	No. patients with NMD implanted with VAD	Type of NMD	Sex(M, %)	Age at implant (median, IQR or range)	Type of VAD	Device strategy
Hollander et al., 2016	Retrospective	Single-center, 2011-2012, US	BMD patients with DCM implanted with LVAD	2	BMD: 2 (100%)	2 (100%)	15.0 yrs (range 14.0–16.0 yrs)	HM II: 2 (100%)	DT: 2 (100%)
Seguchi et al., 2016	Retrospective	Single-center, 2004–2015, Japan	Muscular dystrophy patients implanted with VAD	6	BMD: 3 (50%) u-MD: 2 (33.3%) α-dystroglycanopathy: 1 (16.7%)	6 (100%)	21.0 yrs (IQR 17.3–21.8 yrs)	NIPRO: 5 (83.3%) HM II: 1 (16.7%)	BTT: 6 (100%)
Perri et al.,2017	Retrospective	Single-center, 2011–2016, Italy	Dystrophinopathies with DCM implanted with LVAD	7	DMD: 6 (85.7%) β-sarcoglycanopathy (LGMD): 1 (14.3%)	NR	15.6 yrs (IQR 15.3–17.4 yrs)	Jarvik 2000: 7 (100%)	DT: 7 (100%)
Wittlieb-Weber et al.,2019	Retrospective	Multi-center, 2005–2018, North America	DMD patients implanted with VAD	3	DMD: 3 (100%)	3 (100%)	17.0 yrs (IQR 16.0–17.0 yrs)	HVAD: 2 (66.7%) HM II: 1 (33.3%)	DT: 1 (33.3%) BTT: 1 (33.3%) BTD: 1 (33.3%)
Aydin et al., 2021	Retrospective	Single-center, NR, Turkey	BMD patients with DCM implanted with LVAD	2	BMD: 2 (100%)	2 (100%)	14.5 yrs (range 14.0–15.0 yrs)	NR	BTT: 2 (100%)
Nandi et al., 2023	Retrospective (registry-based)	Multi-center, 2012–2020, US	Muscular dystrophy patients with DCM implanted with VAD	17	DMD: 9 (52.9%) EDMD: 3 (17.6%) BMD: 2 (11.8%) LGMD: 2 (11.8%) LMNA-CMD: 1 (5.9%)	15 (88.2%)	17.2 yrs (IQR 15.7–19.1 yrs)	HM 3: 12 (70.6%) HVAD: 5 (29.4%)	DT: 7 (41.2%) BTT: 5 (29.4%) BTD: 5 (29.4%)
Total	6 (100%) Retrospective	4 (66.7%) Single-center 2 (33.3%) Multi-center Range 2004-2020	-	37	Dystrophinopathies: 27 (73.0%) DMD: 18 (48.6%) BMD: 9 (24.3%) Non-dystrophinopathies: 8 (21.6%) LGMD: 3 (8.1%) EDMD: 3 (8.1%) α-dystroglycanopathy: 1 (2.7%) LMNA-CMD: 1 (2.7%) Undetermined etiology: 2 (5.4%)	28/30 (93.3%)	17.0 yrs (IQR 15.1–19.1; range 11.7–25.0)	HM 3: 12 (32.4%) Jarvik 2000: 7 (18.9%) HVAD: 7 (18.9%) NIPRO: 5 (13.5%) HM II: 4 (10.8%) NR: 2 (5.4%)	DT: 17 (45.9%) BTT: 14 (37.8%) BTD: 6 (16.2%)

Denominators. Percentages within each study row are calculated on the number of patients in that study; percentages in the Total row are calculated on the 37 patients in the pooled cohort, except for sex, where the denominator is restricted to the 30 patients in the five studies that report it.

Abbreviations: BMD, Becker Muscular Dystrophy; BTD, Bridge-to-Decision; BTT, Bridge-to-Transplant; DCM, Dilated Cardiomyopathy; DMD, Duchenne Muscular Dystrophy; DT, Destination Therapy; EDMD, Emery-Dreifuss Muscular Dystrophy; HM II, HeartMate II; HM 3, HeartMate 3; HVAD, HeartWare HVAD; IQR, Interquartile Range; LGMD, Limb Girdle Muscular Dystrophy; LMNA-CMD, LMNA-related Congenital Muscular Dystrophy; LVAD, Left Ventricular Assist Device; NMD, Neuromuscular Disease; NR, Not Reported; u-MD, Undetermined Muscular Dystrophy; VAD, Ventricular Assist Device; yrs, years.

Characteristics of included studies

All studies were retrospective. Four (66.7%) were single center,22, 23, 24, 26 while two (33.3%) were multi-center studies,25, 27 including one registry-based analysis.²⁷ Study periods ranged from 2004 to 2020, with publications between 2016 and 2023 (Table 1).

Characteristics of included patients

Across the six studies, 37 patients with NMD underwent VAD implantation. Median age at implantation was 17.0 years (IQR 15.1–19.1 yrs; range 11.7–25.0 yrs). Sex was reported in five studies, comprising 30 patients, of whom 28 (93.3%) were male. Where the device was specified, all patients received a left VAD (LVAD). The most frequent diagnosis was DMD (48.6%), followed by Becker Muscular Dystrophy (BMD) (24.3%). Dystrophinopathies therefore accounted for 27 patients (73.0%). Non-dystrophinopathy muscular dystrophies accounted for 8 patients (21.6%) and comprised limb-girdle muscular dystrophy (3, 8.1%), Emery-Dreifuss muscular dystrophy (3, 8.1%), α-dystroglycanopathy (1, 2.7%), and LMNA-related congenital muscular dystrophy (1, 2.7%); the etiology of muscular dystrophy was undetermined in the remaining 2 patients (5.4%).

Device strategy was DT in 17 patients (45.9%), BTT in 14 (37.8%), and BTD in 6 (16.2%). The most frequently reported device was HeartMate 3 (HM 3) (12 patients, 32.4%), followed by Jarvik 2000 (7, 18.9%), HeartWare HVAD (HVAD) (7, 18.9%), the NIPRO device (5, 13.5%), and HeartMate II (HM II) (4, 10.8%). Device type was not reported in 2 patients (5.4%).

Pre-implantation status

Data relevant to candidacy were unevenly reported: ambulatory status in four studies and pre-implantation respiratory support in four (Tables S5.1–2). Among DT patients, 13 (76.5%) were non-ambulatory; the four ambulatory exceptions were not listed for HT because of patient preference, pulmonary hypertension, or adverse social circumstances. Ambulatory status was reported for only 6 (42.9%) BTT patients, 3 of whom were non-ambulatory. Before implantation, non-invasive ventilation (NIV) was used in 10 patients and invasive ventilation in 1, with the mode unspecified in a further patient. In the BTD group, the two patients who remained on VAD were non-ambulatory and on NIV, the two transplanted were ambulatory without respiratory support, and two died on device.

Primary outcome

Status at last follow-up was reported as alive on either mechanical support or after HT in 28 patients (75.7%) (Table 2). More specifically, 16 (57.1%) were alive on VAD and 12 (42.9%) had undergone HT, corresponding to 43.2% and 32.4% of the total cohort, respectively. The remaining 9 patients (24.3%) had died.

Table 2.

Studies reporting outcomes of Ventricular Assist Device (VAD) Implantation in Children and Young Adults with Neuromuscular Disease (NMD)

First author, Year	No. of patients with NMD implanted with VAD	Primary outcome (clinical status at last follow-up)			Time on VAD before HT (months, IQR)	Follow-up time (months, IQR)	Complications on VAD (n)	Causes of death (n, %)
		Alive at last follow-up	Alive on VAD	Alive after HT
Hollander et al., 2016	2	2 (100%)	2 (100%)	0 (0%)	-	28.5 and 25.4 mo	Major: 2 Bleeding (non-CNS): pump pocket hematoma requiring mediastinal exploration: 1 Device malfunction: driveline fracture requiring pump exchange: 1 Other reported events: 3 Fall with patellar fracture / hemarthrosis: 1 Obesity with loss of ambulation: 2	-
Seguchi et al., 2016	6	5 (83.3%)	1 (16.7%)	4 (66.7%)	NR	NR	NRa	Sepsis: 1
Perri et al., 2017	7	4 (57.1%)	4 (57.1%)	0 (0%)	-	15.0 mo (IQR 7.5–40.9 mo)	Major: 5 Bleeding (non-CNS): retropharyngeal 1; abdominal with splenectomy 1; thoracic after chest drain 1 Infection: pedestal osteolysis / infection: 1 Neurologic: ischemic stroke (non-fatal): 1 Other reported events: 2 Gastrostomy for poor feeding: 1 Cholecystectomy for gallstones: 1	ICH: 1 Severe ENT bleeding: 1 Lung infection / sepsis: 1
Wittlieb-Weber et al., 2019	3	2 (66.7%)	1 (33.3%)	1 (33.3%)	10 mo	10.0 mo (IQR 7.5–14.0)	Major: 2 Neurologic: stroke: 1 Device malfunction: pump thrombosis: 1 Other reported events: 0	Stroke (hemorrhagic conversion): 1
Aydin et al., 2021	2	2 (100%)	1 (50%)	1 (50%)	30 mo (n=1)	24 mo (n=1) NR after HT (n=1)	None reported	-
Nandi et al., 2023	17	13 (76.5%)	7 (41.2%)	6 (35.3%)	1.12 mo (IQR 0.68–5.12)	6.7 mo (IQR 1.1–22.3)	Major: 22 Bleeding (non-CNS): 5 Infection: 7 Neurologic (stroke): 2 Right heart failure: 4 Arrhythmia: 4 Other reported events: 8 Hemolysis 2; pericardial effusion with tamponade 1; respiratory failure 1; renal dysfunction 1; wound dehiscence 1; psychiatric episode 1; hypertension 1	Stroke: 2 Right heart failure: 1 Unknown (found unresponsive): 1
Total	37	28 (75.7%)	16 (43.2%)	12 (32.4%)	-	-	Major complication events: 31 Bleeding (non-CNS): 9 (29.0%) Infection: 8 (25.8%) Neurologic (stroke / ICH): 4 (12.9%) Right heart failure: 4 (12.9%) Arrhythmia: 4 (12.9%) Device malfunction / thrombosis: 2 (6.5%) Other reported events: 13 All reported adverse events: 44	Neurologic (stroke / ICH): 4 (44.4%) Infection / sepsis: 2 (22.2%) Major bleeding (non-CNS): 1 (11.1%) Right heart failure: 1 (11.1%) Unknown / NR: 1 (11.1%) Total deaths: 9 (24.3%)

Abbreviations: CNS, Central Nervous System; ENT, Ear Nose Throat; HT, Heart Transplantation; ICH, Intracranial Hemorrhage; IQR, Interquartile Range; mo, months; NMD, Neuromuscular Disease; NR, Not Reported; VAD, Ventricular Assist Device.

Extraction rule. Complications were extracted from every adverse event reported in each source article and assigned to one of six pre-specified major categories — bleeding (non-central nervous system), infection, neurologic events, right heart failure, arrhythmia, and device malfunction or thrombosis. Events falling outside these categories are reported separately as “other reported events” so that no reported event is omitted. Fatal events are counted once, in the Causes of Death column, and are not duplicated among complications.

Denominators. Percentages in the three clinical-status columns are calculated on the number of patients with NMD implanted with VAD in each study, and on 37 in the Total row. "Alive at last follow-up" is the overall survival figure for each cohort and is subdivided into the two mutually exclusive columns "Alive on VAD" and "Alive after HT", which sum to it. The table has no separate column for deceased patients: deaths are the difference between the number implanted and the number alive at last follow-up (9 of 37 overall), and their causes are itemized in the Causes of Death column. In the Total row, percentages in the complications column are calculated on the 31 major complication events and those in the Causes of Death column on the 9 deaths. Because multiple events could occur in the same patient and patient-level linkage was not reported, these values describe the relative distribution of event types and must not be interpreted as incidence, cumulative incidence, or event rates. A hyphen (-) denotes a category that is not applicable; NR denotes a reported outcome for which the value was not stated.

^aSeguchi et al. report complications only for the whole muscular dystrophy group of ten patients (LVAD-related infection 9 [90%], LVAD-related cerebrovascular disease 4 [40%], post-LVAD right heart failure 2 [20%]) and do not stratify them by age. They cannot be attributed to the six age-eligible patients and are therefore not included in the pooled counts, which consequently underestimate the true complication burden of the cohort.

Stratified by device strategy, 12 DT patients (70.6%), 12 BTT patients (85.7%), and 4 BTD patients (66.7%) were alive at last follow-up. Among DT patients, 12 (70.6%) were alive on VAD and 5 (29.4%) died. Among BTT patients, 10 (71.4%) underwent HT, 2 (14.3%) survived on device, and 2 (14.3%) died. In the BTD subgroup, 2 patients (33.3%) were successfully transplanted, 2 (33.3%) transitioned to long-term VAD support as DT, and 2 (33.3%) died. Patient flow from device strategy to clinical status at last follow-up is depicted in Figure 2.

Figure 2.

Patient flow from device strategy to clinical status at last reported follow-up after Ventricular Assist Device (VAD) implantation. Band width is proportional to the number of patients. Abbreviations: DT, destination therapy; BTT, bridge-to-transplant; BTD, bridge-to-decision; HT, Heart Transplantation; VAD, ventricular assist device.

Time on VAD support was inconsistently reported across studies. Among BTT patients, median time on VAD before transplant was 1.12 months (IQR 0.68–5.12) in the Nandi et al. cohort, ²⁷ while single patients were transplanted after 10 months²⁵ and 30 months²⁶ of support. In DT patients, support duration was not systematically reported, precluding meaningful comparison. Median follow-up varied substantially, from 6.7 months (IQR 1.1–22.3) in the largest cohort,²⁷ followed by 10.0 months (IQR 7.5–14.0) in Wittlieb-Weber et al.²⁵ and 15.0 months (IQR 7.5–40.9) in Perri et al.,²⁴ whereas individual follow-up extended to 28.5 and 25.4 months in the Hollander case series.²² Among the eight deaths for which timing was reported, the median interval from implantation to death was 11.7 months (IQR 5.5–36.2 months; range 0.4–55.2 months); timing was not available for the remaining death.

Secondary outcomes

Secondary outcomes were heterogeneously reported across studies (Table S5.1), with inconsistent data on respiratory support, advanced heart failure therapies, and length of stay, limiting cross-study comparisons. Functional outcomes were rarely assessed. One small case series reported loss of ambulation after VAD implantation despite survival due to obesity,²² while another described qualitative improvement in a functional scale (MMS).²³

A total of 31 major VAD-related complication events were reported. Bleeding events (non-Central Nervous System [CNS]) were the most frequent (9 events, 29.0% of all reported major complication events), followed by infection (8, 25.8%), whereas neurologic events (ischemic or hemorrhagic stroke), right heart failure, and arrhythmia each accounted for 4 events (12.9%) and device malfunction or thrombosis for 2 events (6.5%). A further 13 events fell outside these pre-specified categories and are reported separately, for a total of 44 reported adverse events. Complications reported by Seguchi et al.²³ were available only for the whole muscular dystrophy cohort of ten patients and could not be attributed to the six age-eligible patients; they were therefore excluded from the pooled counts, which consequently underestimate the true complication burden. Proportions refer to the 31 reported events rather than to patients and describe the relative distribution of complication types (Table 2).

Among the nine documented deaths across studies, neurologic events (stroke or intracranial hemorrhage) represented the most frequent cause (44.4%), followed by infection or sepsis (22.2%), major non-CNS bleeding (11.1%), right heart failure (11.1%), and unknown or unreported causes (11.1%). No single dominant cause of death emerged across studies (Table 2).

No studies reported standardized QoL measures or cost-effectiveness analyses. Functional outcomes were described only qualitatively in a limited number of case series.

Risk of bias

RoB was assessed with design-appropriate JBI checklists.13, 14 Two studies24, 27 were rated at low RoB and two at moderate RoB.23, 25 Hollander et al.,²² and Aydin et al.²⁶ were rated at high RoB because inclusion criteria were unclear and consecutive, complete inclusion could not be verified in these two-patient series (Table S6).

Case reports

Seven case reports15, 16, 17, 18, 19, 20, 21 were excluded from the cohort., presenting three patients bridged to transplant15, 19, 20 and 4 to DT.16, 17, 18, 21 All patients were alive at the end of the follow-up: two underwent HT,15, 19 while five remained alive on VAD support. (Table S8)16, 17, 18, 21

Discussion

This systematic review synthesizes the available evidence on VAD implantation in children and young adults with NMDs and advanced heart failure. Drawing on six retrospective studies encompassing 37 patients, three principal findings emerge: the evidence base is narrow and concentrated in a small number of sources; survival on device support or after transplantation is achievable in selected patients, but morbidity remains substantial; patient-centered outcomes were rarely and inconsistently reported.

The largest and most methodologically structured cohort derives from the ACTION registry analysis,²⁷ accounting for nearly half of included patients. A substantial portion of the remaining experience originates from a single Italian center across multiple publications²⁴. However, partial cohort overlap limits the effective expansion of independent patient-level data: five publications were excluded because reported overlapping study periods (Table S7). The number of truly independent patients remains modest, reinforcing the need for coordinated, multicenter data collection.

Beyond its limited volume, the evidence is heterogeneous across patient selection, device strategy, and outcome reporting. DT was the most frequently reported approach, while BTT and BTD were applied less consistently. Support intent appeared to reflect institutional policies, transplant eligibility criteria, and era of implantation rather than standardized selection frameworks. Ambulatory status and rehabilitation potential were the most commonly-cited determinants of strategy, yet only one study quantitatively assessed these variables.

Offering BTT or long-term DT support to patients with a progressive and currently incurable disease appears paradoxical. The literature increasingly resolves this tension not through blanket inclusion or exclusion by diagnosis, but through individualized assessment of disease trajectory and extracardiac burden. Candidacy is now rarely withheld solely because the diagnosis is categorically prohibitive; rather, the coexistence of respiratory insufficiency, generalized weakness and cardiac failure drives the decision. Our data suggest that functional status, rather than genotype alone, drives the decision: non-ambulatory status predominated among DT patients, whereas retained ambulation and freedom from invasive ventilation characterized those bridged to transplantation. The BTD subgroup demonstrates that these characteristics were not applied as fixed thresholds (Tables S5.1–2).

Despite these limitations, the clinical status at last reported follow-up was notable: 75.7% of patients were alive on VAD support or after HT. However, these figures must be interpreted alongside a considerable complication burden: bleeding and infectious events were the most frequently reported complications, while neurologic events, though fewer in absolute number, accounted for the largest proportion of deaths. VAD implantation is technically feasible in selected NMD patients, but associated morbidity is substantial and survival alone does not fully characterize the clinical trajectory.

When stratified by strategy, BTT patients have the highest proportion alive at last follow-up (85.7%), compared with 70.6% for DT and 66.7% for BTD. This gradient most plausibly reflects selection bias: by definition, transplant eligibility requires preserved respiratory function, adequate nutritional status, and favorable psychosocial circumstances, rather than intrinsic superiority of the BTT approach. These estimates are exploratory and preclude formal between-group comparison.

The overall survival proportion is broadly consistent with one-year survival rates reported in broader advanced heart failure VAD cohorts, exceeding 70%²⁷, in selected advanced heart failure cohorts receiving durable mechanical support, though direct comparison is limited by differences in patient selection, follow-up duration, outcome definitions, and likely publication bias favoring centers with favorable experiences.

One of the most consistent findings of this review is the absence of systematically reported patient-centered outcomes. Functional status, respiratory trajectory, and QoL were rarely assessed and, when described, reported only qualitatively. No study employed standardized neuromuscular functional scales or validated QoL instruments. In a population characterized by progressive multisystem disease, survival alone is insufficient to define treatment benefit.²⁸ The lack of structured functional and QoL reporting limits the ability to evaluate the true impact of VAD therapy genuinely improves lived experience or merely extends device dependency.

Several limitations warrant consideration. All included studies were retrospective with small sample sizes. Selection bias is likely, as only patients deemed appropriate candidates at individual centers underwent implantation. The predominance of dystrophinopathies (73.0%) limits generalizability to other NMD with distinct cardiac phenotypes. Device heterogeneity is also relevant: only 12 patients (32.4%) received HM 3, so these findings may not reflect contemporary HM 3 outcomes. Finally, complication data were denominated on reported events rather than patients and cannot be interpreted as incidence rates.

The clinical decision to implant a VAD in a patient with NMD demands individualized multidisciplinary evaluation integrating cardiology, neurology, pulmonology, rehabilitation, palliative care, and ethics expertise within shared decision-making frameworks. The current evidence does not support rigid selection criteria but underscores the complexity of these decisions. Prospective multicenter studies and shared registries with standardized definitions of indications, complications, functional outcomes, and QoL measures²⁷ are urgently needed to enable reliable comparisons and develop evidence-informed frameworks for this vulnerable population.

Conclusions

Available evidence on VAD implantation in children and young adults with NMD is limited to small retrospective cohorts. In carefully selected patients, survival on device support or after HT is achievable, and NMD per se should not be considered an absolute contraindication; however, complication burden remains substantial, and functional outcomes are rarely reported. Decisions should be individualized within multidisciplinary teams, incorporating anticipated disease trajectory, respiratory status, functional capacity, and patient and family preferences. Prospective, multicenter studies with standardized definitions of indications, complications, and outcomes are needed to identify which patients derive meaningful benefit from VAD therapy.

Financial disclosure statement

No funding was required for this project.

Consent for publication

All listed authors have approved the manuscript before submission, including the names and order of authors.

Author contributions

V.Sa., S.I., F.B., D.P., V.St., L.M. made fundamental contributions to the conception and design of this study. V.Sa., S.I., F.B. conducted the study selection and the data extraction. S.I. and D.P. conducted quality assessment of the included studies. V.Sa., S.I., and L.M. drafted the initial manuscript. V.Sa., S.I., F.B., D.P., M.D., A.Te., A.To., V.V., A.A., V.St., L.M. revised and approved the final manuscript.

Declaration of Generative AI and AI-assisted technologies in the writing process

During the preparation of this work the authors used Claude (Anthropic) in order to assist with proofreading, and editorial refinement. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

Declaration of Competing Interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Valeria Salme’ reports article publishing charges was provided by Padua University Hospital. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplemental material

Supplementary material