Highlight
– A first-in-human phase 1 study of AB-1002 (a cardiotropic AAV vector encoding constitutively active inhibitor-1) delivered by single antegrade intracoronary infusion was feasible and well tolerated in 11 patients with advanced nonischemic heart failure.
– No investigator-attributed serious adverse events occurred; higher dose cohort demonstrated transient, mild asymptomatic liver enzyme elevations.
– Preliminary efficacy signals included improvements in NYHA class and left ventricular ejection fraction in both dose cohorts and improved exercise metrics in the lower-dose cohort.
Background: unmet need and rationale for cardiotropic AAV gene therapy
Heart failure with reduced ejection fraction remains a major global cause of morbidity and mortality despite advances in medical therapy and device-based care. For many patients, guideline-directed therapies reduce symptoms and prolong life but do not reverse established myocardial dysfunction. Gene therapy offers a potential strategy to modify cardiomyocyte biology directly and durably, addressing intracellular pathways that underlie impaired contractile reserve, calcium handling, and maladaptive remodeling.
One attractive target is the cardiomyocyte calcium-handling apparatus. Protein phosphatase 1 (PP1) dephosphorylates phospholamban, reducing sarcoplasmic reticulum calcium uptake via SERCA2a and thereby impairing relaxation and contractility. Inhibition of PP1 through a constitutively active form of protein phosphatase 1 inhibitor-1 (I-1c) can enhance phospholamban phosphorylation, augment SERCA2a activity, and improve calcium cycling. Preclinical work shows that cardiac expression of I-1c can improve contractile function in animal heart failure models.
Adeno-associated virus (AAV) vectors can deliver transgenes to cardiomyocytes with relatively low immunogenicity and long-term expression. However, prior clinical gene therapy attempts in heart failure (for example, trials targeting SERCA2a) produced mixed results and highlighted the need for optimized vector tropism, delivery approaches, dosing strategies, and robust clinical evaluation. AB-1002 is a chimeric cardiotropic AAV designed to favor myocardial transduction and deliver I-1c to cardiomyocytes.
Study design
The phase 1, first-in-human study evaluated safety and feasibility of a single antegrade intracoronary infusion of AB-1002 in patients with nonischemic cardiomyopathy, New York Heart Association (NYHA) class III symptoms, and left ventricular ejection fraction (LVEF) of 15–35%.
Eleven participants (nine men and two women) were enrolled into two sequential dose cohorts. Cohort 1 (n = 6) received 3.25 × 10^13 viral genomes (vg); cohort 2 (n = 5) received 1.08 × 10^14 vg. The route of administration was a single antegrade infusion into the coronary circulation to achieve myocardial delivery. The primary objective was safety and feasibility; exploratory efficacy assessments included NYHA class, LVEF, peak oxygen consumption (peak VO2), and 6-minute walk distance (6MWD). The trial is registered as NCT04179643.
Key findings
Safety and tolerability
Investigators did not attribute any adverse events (AEs) or serious adverse events (SAEs) to the study treatment. The majority of reported AEs were mild or moderate. One death occurred during the study period, and adjudication determined it was not related to AB-1002 treatment. Liver function test abnormalities were the most notable laboratory signal: self-limiting, mild, asymptomatic elevations in alanine aminotransferase (ALT) and/or aspartate aminotransferase (AST) were observed, predominantly in the higher-dose cohort (cohort 2). No clinically meaningful acute hemodynamic deterioration, infusion-related severe events, or clear treatment-related arrhythmias were reported in this small cohort.
Preliminary efficacy and functional outcomes
Although the study was not powered for efficacy, exploratory outcome measures suggested potential clinical benefit. Both dose cohorts demonstrated improvements in NYHA functional class and LVEF compared with baseline. In cohort 1 (lower dose), there were reported improvements in exercise capacity metrics, including peak VO2 and 6-minute walk distance. Cohort 2 showed LVEF and NYHA class improvements but less consistent gains in exercise testing metrics, though numbers were small. The timing, magnitude, and durability of these changes were presented as preliminary signals that require confirmation in randomized cohorts.
Interpretation of effect sizes and clinical significance
Effect sizes were not reported as definitive measures with confidence intervals in the summary; because of the limited sample size, open-label design, and lack of a control group, reported improvements should be regarded as hypothesis-generating. Placebo effects and regression to the mean are important considerations for subjective measures such as NYHA class and effort-dependent metrics like 6MWD and peak VO2. Objective measures such as LVEF by imaging are less susceptible to placebo influences but remain dependent on imaging modality, interobserver variability, and small numbers.
Expert commentary and mechanistic insights
Biological plausibility for the observed signals is strong: inhibition of PP1 via I-1c can increase phospholamban phosphorylation and thereby enhance SERCA2a-mediated calcium uptake, improving systolic and diastolic performance at the cellular level. Use of a cardiotropic AAV backbone and intracoronary antegrade infusion is intended to maximize myocardial transduction and reduce systemic exposure. The transient hepatic transaminase elevations in the higher-dose cohort are biologically plausible: AAV vectors can transduce hepatocytes after coronary infusion, and vector-related immune activation or hepatocellular stress can cause modest, self-limited LFT rises.
Key unresolved mechanistic and translational questions include the degree and heterogeneity of myocardial transduction in human hearts (regional distribution, cell-type specificity), durability of transgene expression, host immune responses (including the impact of pre-existing or induced neutralizing antibodies), and potential off-target effects, including arrhythmogenesis. Long-term risks such as vector genome persistence, rare integration events, and oncogenic potential are theoretical concerns and require prolonged surveillance.
Limitations and generalizability
This phase 1 study provides initial human safety and feasibility data but has important limitations that constrain generalizability. The sample size is small (n = 11) and unblinded, with no randomized control arm. The cohort composition (nine men, two women) limits assessment of sex-specific effects. Short- to mid-term follow-up presented in phase 1 reports may not capture delayed adverse events, late loss of efficacy, or durable clinical outcomes such as heart failure hospitalization and mortality. The dose range explored is limited; the relationship between dose, transduction efficiency, and safety requires further study, and the apparent greater LFT signal at higher dose underscores the need for careful dose optimization.
Implications for future trials and clinical practice
AB-1002 demonstrates a favorable early safety profile and provides signals that justify controlled efficacy testing. Critical design elements for next-phase trials should include randomization with sham or placebo-controlled arms to mitigate placebo effects, sufficient sample size to detect clinically meaningful differences in patient-centered outcomes (NYHA class, HF hospitalizations, mortality), pre-specified imaging and biomarker endpoints (quantitative LVEF by core-lab echocardiography and/or cardiac MRI, NT-proBNP), systematic arrhythmia monitoring, standardized exercise testing, and protocolized safety surveillance for liver function and vector shedding.
Additional important considerations: screening and stratification by pre-existing anti-AAV neutralizing antibodies, planned duration of follow-up for safety (multi-year), analyses of myocardial transduction and transgene expression where feasible (endomyocardial biopsy or novel imaging tracers), inclusion of broader and more diverse patient populations (including more women and racial/ethnic diversity), and cost-effectiveness and manufacturing scalability planning. Because antibody responses typically preclude repeated systemic AAV dosing, durability of benefit from a single administration will be critical for clinical utility.
Conclusion
The phase 1 trial of AB-1002 represents an important translational step for cardiotropic AAV-mediated delivery of a constitutively active PP1 inhibitor in patients with advanced nonischemic heart failure. The intervention was feasible and generally well tolerated, with a manageable laboratory signal (transient liver enzyme elevations) at higher dose. Early signals of improved LVEF and functional status are encouraging but must be interpreted cautiously given the small, uncontrolled nature of the study. A randomized, adequately powered phase 2 trial with rigorous safety monitoring and clinically meaningful endpoints is the appropriate next step to determine whether AB-1002 can offer a durable, disease-modifying therapy for selected patients with heart failure.
Funding and clinicaltrials.gov
The trial is registered at ClinicalTrials.gov as NCT04179643. Detailed funding sources and conflicts of interest are reported in the primary publication (Henry TD et al., Nat Med 2025).
References
1. Henry TD, Chung ES, Alvisi M, et al. Cardiotropic AAV gene therapy for heart failure: a phase 1 trial. Nat Med. 2025 Oct 21. doi: 10.1038/s41591-025-04011-z. Epub ahead of print. PMID: 41120766.
Thumbnail prompt
A high-resolution, clinical-science–style image: a cardiac catheterization lab scene with an interventional cardiologist performing an intracoronary infusion, overlaid with a semi-transparent stylized AAV viral particle and a background schematic of a cardiomyocyte showing calcium cycling proteins (phospholamban, SERCA2a). The tone is optimistic and scientific, with cool blues and clean clinical lighting.