CircRNAs are endogenous single-stranded RNAs formed from precursor mRNAs through a non-canonical process known as back-splicing, in which a downstream 5’splice site is covalently linked to an upstream 3’splice site to generate a stable, closed circular structure^[10,11]. Lacking both a 5’cap and a 3’poly(A) tail, this covalent loop renders circRNAs highly resistant to exonuclease-mediated degradation, giving them stability that far exceeds their linear RNA counterparts^[12]. The biogenesis of circRNAs is facilitated by two primary drivers: cis-acting elements, such as inverted repeat sequences (e.g., Alu repeats) within flanking introns that bring splice sites into proximity, and trans-acting RNA-binding proteins (RBPs) that “bridge” distal exons to promote circularization.

This process yields a diverse array of circRNA species. The most common are exonic circRNAs (ecRNAs), composed entirely of one or more exons. Other forms include exon-intron circRNAs (EIciRNAs), which retain intronic sequence, and circular intronic RNAs (ciRNAs), derived solely from introns. Functionally, circRNAs exhibit pronounced tissue-specific and developmental stage-specific expression patterns. While generated in the nucleus, many are exported to the cytoplasm to execute their functions. There, their unique structure enables multi-layered interactions with miRNAs, RBPs, and even DNA, establishing them as critical components of gene regulatory networks^[11].

Persistent pathological myocardial hypertrophy is a key early event in the development of HF and a primary therapeutic target. While initially an adaptive response, sustained stress drives the heart from compensation to decompensation. Numerous studies^[9] have now confirmed the crucial role of circRNAs within the multi-layered regulatory network of hypertrophy, providing new directions for understanding myocardial remodeling.

Cardiomyocyte death, encompassing apoptosis, necroptosis, ferroptosis, and dysregulated autophagy, is a core driver of adverse remodeling and the progression to HF. Recent studies reveal that circRNAs are deeply integrated into these cell death pathways, employing a diverse toolkit of molecular mechanisms that extend far beyond miRNA sponging to include RBP interactions, epigenetic control, and the maintenance of mitochondrial homeostasis^[65].

In the regulation of apoptosis, the miRNA sponge mechanism remains a prominent theme. circRNAs such as circ_0040414^[30], hsa_circ_0097435^[21], and circ23679^[51] have all been shown to modulate canonical life/death signaling pathways [e.g., phosphatase and tensin homolog (PTEN)/protein kinase B (AKT) signaling pathway, mitogen-activated protein kinase (MAPK) signaling] by sequestering specific miRNAs under various stress conditions^[66]. This mechanism also extends to inflammatory cell death such as pyroptosis, where specific circRNAs have been shown to exacerbate myocardial damage by modulating inflammatory pathways^[31,32]. In necroptosis, however, protein interactions take center stage. For instance, cardiac- necroptosis-associated circRNA (CNEACR)^[33] directly binds to histone deacetylase 7 (HDAC7), preventing its inhibition of a key transcription factor and thereby blocking the RIPK3-mediated (receptor-interacting serine/threonine-protein kinase 3-mediated) necroptotic pathway, showcasing a precise protein-localization-based control mechanism.

circRNAs also play a crucial role in regulating metabolic forms of cell death. In ferroptosis, circRNAs such as ferroptosis‐associated circRNA (FEACR)^[43] can act as metabolic regulators, binding to and stabilizing key enzymes to maintain redox balance and prevent iron-dependent cell death. In the context of autophagy and mitochondrial homeostasis, circRNAs demonstrate remarkable versatility. Some, including the nuclear circRNA ACR^[47], epigenetically regulate the transcription of key mitochondrial protective genes. Others, such as circSamd4 [a circular RNA derived from the SAMD4 (sterile alpha motif domain containing 4) gene]^[28], function in the cytoplasm to directly maintain mitochondrial protein localization and function, thereby preventing mitochondrial instability and cell death.

In summary, circRNAs orchestrate the complex landscape of cardiomyocyte death through a multi-dimensional toolkit, including miRNA sponging, RBP scaffolding, and epigenetic modulation. They are closely linked to classical signaling pathways such as PTEN/AKT, RIPK3, and the NAMPT/Sirt1 axis (nicotinamide phosphoribosyltransferase/sirtuin 1), underscoring their central role in cardiac injury and repair. Further research, particularly in large animal models, is needed to translate these mechanistic insights into precise, circRNA-based therapeutic interventions for HF.

Enhancing the extremely limited regenerative capacity of the adult heart is a core goal of HF research. While strategies such as cell transplantation exist, their clinical efficacy is limited. Consequently, identifying novel molecular targets to boost endogenous regeneration is a critical research direction. Against this backdrop, circRNAs have emerged as key regulatory molecules in cardiac proliferation and repair.

The regulatory landscape of circRNAs in regeneration is characterized by a fascinating duality, with distinct species acting as either potent inhibitors or promoters of cardiomyocyte proliferation. circNfix^[25] is the quintessential “inhibitory” circRNA. It is highly expressed in the adult heart and restricts proliferation via a dual mechanism: it facilitates the ubiquitination and degradation of key cell-cycle proteins, and it sponges miR-214 to inhibit pro-regenerative β-catenin signaling. Consequently, inhibiting circNfix has been shown to enhance cardiomyocyte proliferation and improve cardiac function post-myocardial infarction (post-MI), highlighting its significant therapeutic potential. Conversely, circHipk3^[25,67] is a classic “pro-regenerative” circRNA, highly expressed during early cardiac development. It promotes proliferation and angiogenesis^[68] through a multi-pronged approach, including stabilizing the neurogenic locus notch homolog protein 1 (Notch1) signaling pathway and sponging miR-133a to de-repress downstream growth factors. More recently, circCDYL has also been identified as a pro-proliferative molecule that functions primarily by sponging miR-4793-5p to promote myocardial repair^[69].

In summary, molecules such as circNfix, circHipk3, and circCDYL represent distinct but archetypal modes of circRNA action in cardiac regeneration, from dual-mechanism inhibition to multi-target activation. While this field is still nascent, these studies reveal the crucial role of circRNAs in governing the heart's regenerative potential. They provide a strong theoretical and practical foundation for the future development of RNA-targeted therapies aimed at repairing the injured myocardium and ultimately treating HF.

Myocardial fibrosis, characterized by excessive extracellular matrix deposition, is a core pathological driver of cardiac stiffness and the progression to HF. In this complex process, circRNAs have been shown to play a key role, modulating fibrosis through multiple signaling pathways^[70].

circRNAs regulate fibrosis by targeting a web of crucial signaling cascades. Within the pivotal transforming growth factor-beta (TGF-β) signaling pathway, for instance, some circRNAs such as circPVT1 [a circular RNA originating from back-splicing of the PVT1 (plasmacytoma variant translocation 1) gene]^[35] are pro-fibrotic, while others such as circSMAD3 [circSMAD3, a circular RNA derived from the SMAD3 (SMAD family member 3) gene]^[71] can be inhibitory. This pathway is also targeted by molecules such as circSMAD4^[72]. The complexity is further underscored by different circRNAs from the same gene locus exhibiting opposing functions; for instance, the myosin IXA (MYO9A) gene produces both pro- and anti-fibrotic circRNAs^[36]. Beyond TGF-β, circRNAs also target other key cascades, such as the Wnt/β-catenin pathway, which is activated by circNSD1^[37], and inflammatory feedback loops driven by peptides such as Nlgn173^[53]. The regulatory landscape is further enriched by numerous other circRNAs that modulate fibrosis by targeting distinct molecular nodes. These include those that modulate the crucial Yes-associated protein 1 (YAP) signaling pathway (e.g., circYap^[39] and circHelz^[40]), and others that act through different axes to exert either anti-fibrotic (e.g., circNFIB^[38], circ-sh3rf3^[73]) or pro-fibrotic (e.g., circ_0002295^[41]) effects.

In summary, circRNAs act as key upstream regulators in myocardial fibrosis by modulating a dense network of signaling pathways including TGF-β/Smad, Wnt/β-catenin, and YAP. Crucially, these studies reveal the pronounced bidirectional nature of circRNA regulation. From pro-fibrotic players (e.g., circPVT1 and circHelz) to anti-fibrotic molecules (e.g., circSMAD3 and circYap), this duality is a recurring theme. This provides a strong foundation for developing highly specific, RNA-targeted therapeutic strategies that either inhibit pro-fibrotic circRNAs or augment protective circRNAs to combat HF.

Beyond structural changes, HF is driven by a complex interplay of metabolic imbalance, mitochondrial dysfunction, oxidative stress, ferroptosis, and inflammation. Recent studies have positioned circRNAs as crucial molecular hubs that connect these metabolic and stress pathways to myocardial pathology, employing a range of mechanisms from miRNA sponging to direct protein interaction.

A key theme emerging is the role of specific circRNAs as guardians of cellular homeostasis, particularly against metabolic and oxidative stress. circSamd4^[42], for example, is a mitochondrial-localized circRNA that protects cardiomyocytes by reducing mitochondrial reactive oxygen species (ROS), inhibiting apoptosis, and promoting repair. Its mechanism highlights a direct link between circRNA function and the mitigation of oxidative damage. Similarly, FEACR^[43] acts as a crucial regulator of ferroptosis, a form of iron-dependent cell death critical in ischemia/reperfusion injury. It achieves this by stabilizing a key enzyme (NAMPT) to ultimately upregulate the ferroptosis inhibitor ferritin heavy chain 1 (FTH1), representing a novel protective axis against ischemia/reperfusion injury (I/R injury). Furthermore, circRNAs such as circ-SNRK^[74] contribute to energy metabolism by modulating their host gene’s expression via a miRNA sponge mechanism, thereby protecting myocardial energy balance. These studies collectively demonstrate how circRNAs can maintain metabolic homeostasis and protect the heart from various forms of stress-induced cell death.

In summary, circRNAs are deeply integrated into the networks governing myocardial metabolic homeostasis, mitochondrial function, oxidative stress, and ferroptosis. They can act as potent protective factors, as exemplified by circSamd4, FEACR, and circ-SNRK, helping the myocardium resist injury. However, their dysregulation can also exacerbate energy depletion and cell death, promoting cardiac dysfunction. Incorporating circRNAs into the broader framework of cardiac inflammation and metabolic remodeling will be essential for fully understanding disease progression and developing novel therapeutic interventions for HF.

The regulation of calcium (Ca²⁺) homeostasis is fundamental to cardiac function, and its dysregulation is a hallmark of HF. Recent research has unveiled a new frontier: the direct role of circRNAs in modulating Ca²⁺ handling, offering a novel perspective on the molecular drivers of HF.

A key mechanism involves circRNAs that are derived from, and subsequently regulate, major Ca²⁺ handling genes. circRYR2^[75], for instance, is a circRNA derived from the ryanodine receptor 2 (RYR2) locus that is downregulated in pathological hypertrophy. Its deficiency perturbs the expression of key proteins such as sarcoplasmic/endoplasmic reticulum Ca²⁺-ATPase 2a (SERCA2a) and phospholamban (PLN), impairing Ca²⁺ reuptake and prolonging Ca²⁺ transients. Conversely, restoring circRYR2 expression improves Ca²⁺ handling and cardiac function, establishing a direct link between a circRNA and myocardial contractility. Similarly, circSlc8a1^[76], derived from the SLC8A1/NCX1 (solute carrier family 8 member A1/sodium-calcium exchanger 1) locus, is crucial for maintaining Ca²⁺ efflux. Its specific silencing leads to severe HF, while its forced expression is protective. Mechanistically, circSlc8a1 deficiency disrupts NCX1-mediated Ca²⁺ efflux, triggering diastolic Ca²⁺ overload and adverse metabolic remodeling, highlighting its essential role in preventing HF progression.

Beyond direct modulation of transport proteins, circRNAs also indirectly influence Ca²⁺ dynamics by targeting upstream signaling enzymes and pathways. For example, some circRNAs can alter cyclic adenosine monophosphate (cAMP) signaling and PKA/PKC (protein kinase A / protein kinase C) kinase activity by sponging miRNAs that target key enzymes such as adenylate cyclase^[34]. Furthermore, in the context of ischemia/reperfusion injury, circRNAs have been shown to modulate neuropeptide signaling^[77], which in turn alters the phosphorylation state of Ca²⁺ handling proteins (e.g., RYR2 and PLN), ultimately impacting cardiomyocyte survival.

In summary, circRNAs form a hierarchical, deeply integrated regulatory network governing myocardial Ca²⁺ homeostasis. By acting on channels, transporters, and signaling enzymes, they fine-tune Ca²⁺ dynamics and electromechanical coupling. This emerging perspective does not merely supplement our understanding of hypertrophy and fibrosis; it establishes the circRNA-Ca²⁺ axis as a new, fundamental pillar of cardiac regulation. This provides a solid theoretical basis for developing novel therapeutic strategies aimed at restoring Ca²⁺ homeostasis by targeting specific circRNA.

As a final summary, this article systematically reviews the roles of circRNAs in myocardial remodeling and HF. The reported key circRNAs, along with their host genes, mechanisms of action, downstream signaling pathways, and associated pathological processes, are comprehensively summarized in Table 4^[78,79]. Collectively, these findings provide a panoramic overview of the multi-level regulatory functions of circRNAs in myocardial hypertrophy, cell death, fibrosis, metabolic remodeling, calcium homeostasis, and myocardial regeneration, highlighting their significance in the pathogenesis and potential therapeutic targeting of cardiovascular disease.

An ideal biomarker should be a measurable molecule that reflects disease outcome, course, or treatment response. In HF, classic markers such as BNP and N-terminal pro–B-type natriuretic peptide (NT-proBNP) are widely used for this purpose. circRNAs, however, present a next-generation opportunity. Their covalently closed structure confers exceptional resistance to exonuclease degradation, resulting in stability in blood and other bodily fluids that far surpasses linear RNAs^[80,81]. This inherent stability, combined with their tissue-specific expression, makes them naturally suited to be robust clinical biomarkers.

Growing evidence now substantiates the clinical value of circRNAs in the diagnosis and risk prediction of HF. Multiple studies report that specific circRNAs are abnormally expressed in the peripheral blood of HF patients^[82]. For instance, circ_0040414 is significantly elevated in patients with chronic heart failure (CHF) and participates in pathological signaling via the ceRNA mechanism^[83]. Among the circRNAs investigated as prognostic biomarkers, MICRA (myocardial infarction-associated circRNA) currently represents the most prominent example approaching clinical applicability. A prospective cohort study of acute myocardial infarction (AMI) patients demonstrated that low MICRA expression was a significant and independent predictor of subsequent left ventricular ejection fraction (LVEF) decline. To date, no other circRNA candidates have achieved comparable clinical validation or been evaluated in large patient cohorts, highlighting MICRA’s unique role in bridging basic circRNA research and patient-centered prognostic assessment.

With the advancement of high-throughput sequencing, the capacity to identify and quantify specific circulating circRNAs continues to improve. This opens the door for their use in non-invasive diagnosis, disease subtyping, risk prediction, and efficacy monitoring, laying the foundation for a precision medicine approach to HF.

The therapeutic potential of circRNAs is rapidly gaining attention. Their high stability and ability to be packaged into extracellular vesicles (EVs) for intercellular communication^[84] provide a feasible pathway for “cell-free” therapeutic strategies^[85]. circRNAs released from cardiac cells can be delivered via EVs to target cells, where they can modulate key pathological pathways, including proliferation, apoptosis, and fibrosis^[86]. This mechanism offers several advantages for therapeutic delivery, including natural nuclease resistance and the intrinsic targeting capabilities of EVs.

Functional validation for this concept already exists. For instance, EV-delivered circRNAs have been shown in preclinical models to inhibit cardiac fibrosis (e.g., circ_0036176)^[36] or to promote cardiomyocyte proliferation and improve cardiac function post-MI (e.g., circWhsc1)^[50]. Beyond EV delivery, synthetic circRNA mimetics or antisense oligonucleotides (ASOs) designed to silence pathogenic circRNAs represent another promising therapeutic avenue, offering the potential to reverse pathological remodeling at the molecular level.

However, several technical limitations currently constrain the clinical translation of EV-based circRNA therapies. EV isolation and purification lack universally standardized protocols, and methodological variability can significantly affect vesicle composition and yield. In addition, in vivo targeting efficiency remains suboptimal, as systemically administered EVs are often rapidly cleared and accumulate off-target in organs such as the liver and spleen. Furthermore, large-scale production under good manufacturing practice (GMP) conditions, together with batch-to-batch consistency and rigorous quality control, presents substantial logistical and regulatory challenges. Addressing these issues will be essential to ensure the reproducibility, safety, and scalability of EV-mediated circRNA delivery.

In summary, circRNAs can regulate cardiovascular pathology at multiple levels. Their stability and druggability, particularly via EV-based strategies, provide actionable pathways from molecular mechanism to clinical intervention. By optimizing delivery and ensuring long-term safety, circRNA-based therapies are poised to become a next-generation tool for treating cardiovascular diseases.

Despite their immense potential, the clinical translation of circRNAs faces several key challenges. First, detection and analysis methods lack full standardization, with differences in sample preparation^[87], algorithms^[88], and quantification strategies hindering cross-study comparisons. Second, the tissue-specific and dynamic expression patterns of circRNAs are not yet fully mapped, making it difficult to define optimal detection and treatment windows^[89].

Furthermore, the majority of current mechanistic studies rely on rodent models, whose physiological differences from humans create uncertainty for direct clinical translation. Future work must therefore extend to large animal models and human tissues. Simultaneously, the safety, specificity, and immunogenicity of any circRNA-based therapeutic, whether delivered via EVs^[90] or as synthetic molecules, require rigorous preclinical evaluation.

Looking ahead, the combination of high-throughput sequencing, single-cell analysis, and spatial transcriptomics will undoubtedly provide a much clearer picture of circRNA biology in disease. This will not only refine their use as biomarkers but will also lay a robust theoretical foundation for precision therapeutic strategies. Ultimately, circRNA-based diagnostics and therapeutics hold the promise of becoming a new, cell-free, and highly targetable modality for the clinical management of HF.