Rheumatic Heart Disease: Inflammation, Immune Pathways, Valve Remodeling, and the Promise of SGLT2 Inhibitors

MicroRNAs (miRNAs) are endogenous, noncoding, single-stranded RNAs ranging from 19 to 25 nucleotides in length. They regulate gene expression and play a crucial role in a wide range of physiological and pathological processes. 135 In recent years, substantial evidence has demonstrated the role of miRNAs as diagnostic and prognostic biomarkers in cardiovascular diseases, including acute coronary syndrome, coronary artery disease, and heart failure. 136 , 137 Their application to RHD, however, is comparatively nascent, driven by emerging evidence linking miRNA dysregulation to the autoimmune and inflammatory cascades that underpin valvular injury. 138 The dysregulated miRNAs identified in RHD converge on 3 core pathological axes: innate immune activation, cytokine-driven inflammation and fibrosis, and disease progression. Two miRNAs, hsa-miR-205-3p and hsa-miR-3909, have been specifically implicated in RHD progression through their target genes IL-1β and IL-1 receptor 1, suggesting a pathway-level amplification of IL-1 signaling in the diseased valve. Complementing this, downregulation of miR-101 in cardiac tissue promotes RHD via the upregulation of Toll-like receptor 2 (TLR2), 139 reinforcing the view that impaired miRNA-mediated suppression of innate immune receptors is a proximal driver of valvular injury. A distinct and well-characterized pathway involves the miR-155-5p/Signal Transducer and Activator of Transcription 3 (STAT3) and Sphingosine-1 Phosphate Receptor 1 (S1PR1) signaling network. In a rat model of RHD, miR-155-5p was markedly upregulated in valvular tissue and serum-derived exosomes, accompanied by reduced S1PR1 expression, activation of the SOCS1/STAT3 signaling pathway, and increased IL-6 and IL-17 levels in both valves and serum, consistent with a heightened inflammatory state. Critically, pretreatment with a recombinant adeno-associated virus (AAV-miR155-inhibitor) to inhibit valvular miR-155-5p expression restored S1PR1 levels and suppressed SOCS1/STAT3 pathway activation, leading to reduced valvular inflammation and fibrosis, along with decreased IL-6 and IL-17 levels in both valves and serum. 140 At the level of valvular remodeling, miR-134-5p has emerged as a key regulator of RHD progression. Yang et al 141 reported that upregulation of miR-134-5p correlates with disease severity and adverse prognosis, while experimental inhibition of miR-134-5p attenuated inflammation and oxidative stress in valvular interstitial cells via targeting lysine acetyltransferase 7 (KAT7). These findings highlight the potential of miR-134-5p as both a diagnostic and prognostic biomarker in RHD. In addition, Li et al 142 highlighted the differential expression of miRNAs in RHD, identifying miR-1183 and miR-1299 as playing distinct roles in RHD pathogenesis, particularly in cases accompanied by secondary pulmonary arterial hypertension. This suggests their potential as biological markers for disease progression. 142 This axis is further elaborated by evidence that LINC00707, a long noncoding RNA, can suppress myocardial fibrosis and immune dysregulation in an RHD rat model by modulating the miR-145-5p/S1PR1 axis, 143 establishing S1PR1 as a convergence point for both coding and noncoding RNA regulation in RHD and a plausible therapeutic target. Taken together, the miRNA landscape in RHD reflects a structured set of pathway dysregulations rather than isolated molecular aberrations. Their functional roles spanning innate immune priming, cytokine amplification, and valvular fibrosis, position miRNAs as mechanistically grounded candidates for diagnostic biomarker panels, prognostic stratification tools, and targeted therapeutic agents. Table 3 summarizes the molecular biomarkers implicated in the initiation and progression of ARF and RHD. Table 3 Molecular Biomarkers Implicated in the Initiation and Progression of ARF and RHD S/N Type of Marker Molecular Biomarkers Biological Function Role in RHD Initiation/Progression 1 Endothelial surface proteins ICAM-1 and VCAM-1 Endothelial activation, leukocyte adhesion, inflammatory cell recruitment Upregulated in rheumatic valves and experimental settings, supporting endothelial activation and inflammatory cell trafficking; may be relevant to early valvular injury rather than established clinical diagnosis H 79 A 144 , 145 2 Immune-regulatory and inflammatory cytokines TNF-α and IL-6 Cytokine-driven inflammation, endothelial activation, immune-cell recruitment Elevated inflammatory mediators implicated in ARF and chronic RHD; may reflect inflammatory burden and disease activity, although they are not disease-specific H 52 , 133 E 146 IFN-γ Autoimmunity and chronic inflammation Supports Th1-predominant immune activation in ARF/RHD and ongoing valvular inflammation H 53 TGF-β1 mVIC activation and differentiation Strongly linked to valvular fibrosis and remodeling in chronic RHD; one of the more plausible fibrosis-associated biomarkers H 147 IL-1β Autoimmune response against GAS, and inflammation Implicated in ARF/RHD inflammatory signaling; biomarker utility remains exploratory H 148 IL-17 Chronic inflammation and valve fibrosis Elevated in chronic RHD and may contribute to progressive valvular remodeling and fibrosis H 59 A 58 IL-10 Chronic autoimmune response by mediating CD8 T cells expression May participate in chronic immune dysregulation in RHD, but current evidence is more mechanistic than biomarker-focused E 61 IL-18 Valvular inflammation in ARF Proposed component of macrophage/NLRP3-related inflammation in ARF/RHD; currently better regarded as a mechanistic mediator than a validated biomarker E 64 3 Inflammatory protein hsCRP/CRP Systemic inflammation Elevated in ARF and chronic rheumatic valve disease; may reflect inflammatory activity but lacks disease specificity H 149 4 Oxidative stress markers and index AOPP Reactive oxygen species (ROS) and related to chronic inflammation Increased levels have been reported in chronic rheumatic valve disease and may complement inflammatory markers H 149 MDA, SOD, AOPP, and glutathione Redox imbalance and oxidative stress burden Associated with severity of chronic RHVD; may reflect disease activity rather than disease-specific mechanisms H 150 OSI Composite oxidative stress burden Reported in rheumatic and degenerative valve disease, but specificity for RHD initiation/progression is limited H 81 5 Fibrosis/ECM turnover markers MMPs (particularly MMP-1), TIMP-1, MMP-1/TIMP-1 ratio Matrix degradation and remodeling Dysregulated ECM turnover is linked to progression from inflammation to fibrosis; the MMP-1/TIMP-1 ratio is a plausible marker of active remodeling H 93 , 97 , 102 , 104 PICP Matrix degradation and remodeling Candidate circulating marker of active fibrotic remodeling in rheumatic mitral disease H 97 , 102 sST2 Fibrosis-related inflammatory stress signaling Candidate marker of progressive fibrosis and maladaptive remodeling; promising but not yet established in RHD H 103 6 Valvular calcified process Fetuin-A Glycoprotein that inhibits ectopic calcification An independent predictor of calcium concentration in the mitral valve tissue of RHD patients H 94 7 Proangiogenic/lymphangiogenic markers VEGF Inflammatory and neoangiogenesis proteins Upregulated in both the onset and progression of RHD H 29 LYVE-1, podoplanin Lymphatic remodeling and inflammatory resolution Tissue markers supporting pathological neolymphangiogenesis in progressive RHD; currently more useful as mechanistic/tissue markers than circulating biomarkers H 115 8 Valvular inflammation and EndMT-related markers p-NF-κB, activin A, p-Smad2, p-Smad3, and EndMT-related factors Inflammatory and EndMT factors (fibrosis) Increased in rheumatic valve tissue and support a mechanistic link between inflammation and fibrosis; best regarded as tissue/pathway markers H 80 9 Extracellular matrix proteins Tenascin-C Pleiotropic protein Increased levels may support ARF/RHD diagnosis and remodeling activity, although broader validation remains limited H 151 , 152 Lumican, vitronectin, collagen VI, and Vimentin ECM structural proteins Dysregulated protein expression supports ECM remodeling in RHD, although direct biomarker utility remains exploratory H 12 10 Small noncoding RNAs — Innate immune activation: the IL-1 and TLR2 axes hsa-miR-205-3p, hsa-miR-3909 Regulate gene expression Dysregulation may amplify IL-1 signaling in rheumatic valve disease; proposed as pathway-linked biomarkers of inflammatory progression H 138 miR-101 Regulate gene expression (TLR2) Downregulation may contribute to RHD progression through upregulation of TLR2; supports impaired repression of innate immune signaling H 139 Small noncoding RNAs — Cytokine-mediated inflammation and the STAT3/S1PR1 axis miR-155-5p Regulation of SOCS1/STAT3-S1PR1 signaling and IL-6/IL-17 axis Upregulated in experimental RHD; inhibition attenuates valvular inflammation and fibrosis, supporting a mechanistic role rather than current clinical biomarker use H 140 miR-145-5p/S1PR1 axis Regulate gene expression May participate in myocardial/valvular fibrosis and immune dysregulation; currently more relevant as a mechanistic and therapeutic axis than as an established biomarker A 143 Small noncoding RNAs — Disease progression, fibrosis, and valvular remodeling miR-1183, miR-1299 Regulate gene expression Differential expression in RHD supports potential diagnostic/progression biomarker relevance, particularly in advanced disease phenotypes H 142 miR-134-5p Regulates inflammation and oxidative stress via targeting lysine acetyl transferase 7 (KAT7) Upregulated in RHD and associated with disease severity and adverse prognosis; currently one of the more promising progression-associated miRNA candidates H 141 11 Genetic susceptibility markers HLA-DR7/DR53 Immune susceptibility Associated with increased risk of ARF/RHD; best considered a susceptibility marker rather than a progression biomarker H 153 IGHV4-61 gene Adaptive immune susceptibility Associated with increased risk of RHD in GWAS; informative for susceptibility, not disease monitoring H 154 Evidence coding: H = human RHD tissue and/or clinical data; A = animal or experimental model data; E = extrapolation from non-rheumatic diseases or general pathway literature. AOPP = advanced oxidation protein products; CRP = C-reactive protein; EndMT = endothelial-to-mesenchymal transition; GWAS = genome-wide association study; hsCRP = high-sensitivity C-reactive protein; ICAM = intercellular adhesion molecule; IFN = interferon; IL = interleukin; LYVE-1 = lymphatic vessel endothelial hyaluronan receptor-1; MDA = malondialdehyde; miR = microRNA; mVIC = mitral valve interstitial cell; OSI = oxidative stress index; PICP = carboxy-terminal propeptide of type I procollagen; RHVD = rheumatic heart valve disease; RHD = rheumatic heart disease; ROS = reactive oxygen species; S1PR1 = sphingosine-1-phosphate receptor 1; Smad = Suppressor of Mothers against Decapentaplegic; SOD = superoxide dismutase; sST2 = soluble suppression of tumorigenicity 2; STAT3 = signal transducer and activator of transcription 3; TIMP = tissue inhibitor of metalloproteinase; TLR2 = Toll-like receptor 2; TNF = tumor necrosis factor; VCAM = vascular cell adhesion molecule; other abbreviations as in Tables 1 and 2 . Molecular Biomarkers Implicated in the Initiation and Progression of ARF and RHD Evidence coding: H = human RHD tissue and/or clinical data; A = animal or experimental model data; E = extrapolation from non-rheumatic diseases or general pathway literature. AOPP = advanced oxidation protein products; CRP = C-reactive protein; EndMT = endothelial-to-mesenchymal transition; GWAS = genome-wide association study; hsCRP = high-sensitivity C-reactive protein; ICAM = intercellular adhesion molecule; IFN = interferon; IL = interleukin; LYVE-1 = lymphatic vessel endothelial hyaluronan receptor-1; MDA = malondialdehyde; miR = microRNA; mVIC = mitral valve interstitial cell; OSI = oxidative stress index; PICP = carboxy-terminal propeptide of type I procollagen; RHVD = rheumatic heart valve disease; RHD = rheumatic heart disease; ROS = reactive oxygen species; S1PR1 = sphingosine-1-phosphate receptor 1; Smad = Suppressor of Mothers against Decapentaplegic; SOD = superoxide dismutase; sST2 = soluble suppression of tumorigenicity 2; STAT3 = signal transducer and activator of transcription 3; TIMP = tissue inhibitor of metalloproteinase; TLR2 = Toll-like receptor 2; TNF = tumor necrosis factor; VCAM = vascular cell adhesion molecule; other abbreviations as in Tables 1 and 2 . Although ARF affects male and female individuals equally in many populations, RHD disproportionately impacts female individuals with a relative risk 1.6 to 2.0 times higher than in male individuals. This sex disparity is more pronounced in adolescents and adults than in children. 23 , 155 Several factors contribute to this difference. In low-income settings, limited access to primary and secondary prophylaxis among women and girls increases susceptibility. In addition, greater exposure to Streptococcus pyogenes due to childcare responsibilities may elevate risk. 16 Beyond environmental influences, sex-related differences in immune response also play a critical role. Women exhibit greater susceptibility to autoimmune diseases due to higher immunoglobulin production and X-linked immune genes that regulate immune responses. The X chromosome carries more than 1,000 genes, many influencing immune function, whereas the Y chromosome contains only about 100. 156 In addition, sex hormones play a pivotal role in immune modulation, with estrogen receptor alpha (ERα) stimulation influencing lymphocyte activity. 157 , 158 The higher prevalence of RHD in adult women suggests that hormones may significantly contribute to disease progression. Estradiol immunomodulatory effects and overexpression of X-linked immune genes likely predispose female individuals to heightened immune activation. Moreover, sex hormones regulate lymphangiogenesis; estradiol stimulates hyaluronic acid synthesis—a key ECM ligand for LYVE-1, crucial for lymphatic vessel formation and immune-cell trafficking. 159 , 160 Although estrogens generally confer cardiovascular protection, ERα activation can promote lymphatic endothelial migration and sprouting, which, together with LYVE-1 activity, may cause excessive neolymphangiogenesis in rheumatic mitral valves, aggravating inflammation and disease severity. 161 Consistent with the modulatory effects of estrogen, recent findings by Passos et al 162 identified that Prothymosin Alpha (ProTα) plays a crucial role in RHD sex predisposition by enhancing CD8+ T-cell cytotoxicity via ERα activity. ProTα facilitates the recognition of type I collagen mimic epitopes by CD8+ T cells, thereby triggering autoimmune mechanisms. This finding supports the hypothesis that molecular mimicry between human type I collagen and Streptococcus pyogenes collagen-like surface proteins contributes to the autoimmune response targeting heart valves in RHD. By increasing ERα activity, ProTα further amplifies autoimmune processes, strengthening the link between estrogens, immune regulation, and sex-based differences in RHD susceptibility. 162

Until now, no therapy has been proven to reverse or effectively treat established RHD. Secondary antibiotic prophylaxis may slow disease progression, but its long-term impact on clinical outcomes remains uncertain. 131 , 132 , 146 In the REMEDY (Global Rheumatic Heart Disease Registry) study, lack of penicillin prophylaxis was identified as an independent predictor of 1-year mortality, suggesting a beneficial effect of ongoing penicillin therapy. 14 However, no intervention has been shown to modify the risk or severity of RHD following an ARF episode. 21 Clinical trials evaluating corticosteroids or intravenous immunoglobulins for preventing valve lesions in ARF have demonstrated limited evidence of benefit. 163 Emerging pharmacological approaches, including SGLT2i, 164 , 165 , 166 , 167 , 168 angiotensin-converting enzyme inhibitors, 103 , 169 and angiotensin receptor blockers, have attracted attention for their potential to mitigate inflammation-driven valvular damage. These agents, through modulation of endothelial function and pro-fibrotic signaling, might theoretically interrupt the cycle of inflammation, endothelial dysfunction, and maladaptive ECM remodeling that drives RHD progression. To help interpret the available evidence for SGLT2i in this setting, we organized this section according to the nature of the supporting data. We first discuss evidence from experimental models and non-rheumatic valvular diseases, particularly degenerative aortic stenosis (AS), which may support biological plausibility but requires extrapolation to the RHD setting. We then summarize findings directly derived from human RHD tissue or RHD-specific clinical studies. Last, we highlight mechanisms that remain untested in RHD and are therefore speculative. This structure is intended to help readers interpret the evidence with an appropriate level of caution. The biological rationale for SGLT2 inhibition in valvular disease is primarily grounded in experimental and non-rheumatic data. Subclinical inflammation has been identified as a key driver of SGLT2 overexpression through activation of the AT1R/NADPH oxidase pathway, promoting oxidative stress, immune activation, and adverse tissue remodeling, which are key mechanisms established in cardiovascular contexts beyond RHD. 167 , 170 In our experimental studies of degenerative AS, plasma from patients with severe AS exhibits a distinct proinflammatory phenotype characterized by elevated cytokine levels and enhanced factor Xa activity. This plasma markedly increases oxidative stress in valvular endothelial cells via SGLT2 upregulation, inducing a prothrombotic, proinflammatory, and adhesive phenotype; an effect significantly attenuated by empagliflozin. 171 SGLT2 expression is also upregulated in the myocardium of patients with low-flow, low-gradient AS, correlating with oxidative stress, myocardial fibrosis, and inflammation. 172 In calcified human valves, SGLT2 expression colocalizes with inflammatory and oxidative markers, and experimental data indicate that SGLT2i may attenuate extracellular vesicle–induced valvular injury. 164 These findings establish a plausible mechanistic basis for SGLT2-mediated valvular protection, but derive entirely from degenerative valve biology and cannot be directly transposed to RHD without further investigation. Translating these experimental observations to the clinical domain, emerging data from observational cohorts in non-rheumatic valve disease suggest that SGLT2i may decelerate valve disease progression. In a post hoc analysis of 458 patients treated with SGLT2i vs 11,240 untreated controls, SGLT2i use was associated with a significantly reduced risk of progression from nonsevere to severe AS (HR: 0.61; 95% CI: 0.39-0.94; P = 0.03) over a median follow-up of 3.4 years, with an apparent dose-dependent protective effect (HR: 0.54, 0.48, and 0.27 for 3-, 6-, and 12-month treatment durations, respectively). 173 A retrospective analysis of 1,838 patients undergoing aortic valve replacement (transcatheter aortic valve replacement or surgical aortic valve replacement) reported a significantly lower incidence of bioprosthetic valve dysfunction among SGLT2i users (HR: 0.37; 95% CI: 0.18-0.78; P = 0.008) over a median follow-up of 4.85 years. 174 In our recent large real-world transcatheter aortic valve replacement cohort, SGLT2i use was independently associated with reduced all-cause mortality and lower rates of bioprosthetic valve failure after comprehensive adjustment for baseline comorbidities. 175 These clinical observations are encouraging and lend biological and clinical support to further investigation in RHD, despite the limitations inherent to their observational design, the potential for residual confounding, and their origin in a mechanistically distinct disease process. Direct validation in RHD is now warranted. Direct clinical evidence for SGLT2 inhibition in RHD remains sparse. A single small randomized controlled trial demonstrated that adding dapagliflozin to standard therapy in patients with RHD-related mitral stenosis conferred hemodynamic benefits, evidenced by increased net atrioventricular compliance and reductions in both the mean mitral valve pressure gradient and N-terminal pro–B-type natriuretic peptide levels. 166 Although these findings are encouraging, the study was limited by its small sample size, short follow-up, and focus on hemodynamic rather than structural or disease-modifying endpoints. No human RHD tissue data currently demonstrate SGLT2 expression, valvuloprotective signaling, or histological modification attributable to SGLT2 inhibition in the rheumatic valve. This represents a critical evidence gap. Beyond the hemodynamic effects observed in the study by Asrial et al, 166 several additional pleiotropic mechanisms of SGLT2 inhibition are biologically plausible in the RHD context but remain untested. These include enhancement of endothelial integrity, suppression of NLRP3 inflammasome activation and NF-κB signaling, downregulation of VCAM-1 expression, promotion of reparative macrophage polarization, and attenuation of TGF-β– and MMP-9–mediated fibrotic remodeling. 176 , 177 , 178 Insofar as these pathways are central to the immune-inflammatory cascade driving RHD progression, SGLT2 inhibition could theoretically be most impactful during the inflammation-dominant early or latent stages of disease, before irreversible fibrotic and calcific remodeling is established. The stage most likely to benefit would correspond to patients with echocardiographically detected subclinical RHD or borderline disease, in whom structural progression has begun but valve deformity remains modest and potentially reversible. Table 4 outlines the proposed mechanisms by which SGLT2i may confer cardiovalvular protection in the context of RHD. Table 4 Proposed SGLT2 Inhibitor Effects on RHD-Relevant Pathophysiology Proposed Mechanism/Effect Relevance to RHD Endothelial protection with reduction of oxidative stress and inflammation Supports the biological plausibility that SGLT2i may mitigate endothelial dysfunction, ROS excess, and inflammatory valvular injury relevant to RHD, although current support comes mainly from degenerative aortic stenosis and non-rheumatic cardiovascular models 164 , 170 , 171 Antithrombotic, anti-adhesive, and anti-remodeling effects Suggests that SGLT2i may attenuate prothrombotic endothelial activation, leukocyte/platelet adhesion, extracellular vesicle-mediated injury, and downstream tissue remodeling, but direct extrapolation to rheumatic valves remains unproven 164 , 165 , 178 Hemodynamic benefit in RHD-related mitral stenosis A small, randomized trial showed improved net atrioventricular compliance and reduced mean mitral gradient and NT-proBNP, supporting possible clinical benefit, but not yet structural disease modification 166 Anti-inflammatory disease modification SGLT2i may suppress pathways implicated in RHD progression, including NLRP3 inflammasome and NF-κB signaling, reduced VCAM-1 expression, and more reparative macrophage polarization, but these mechanisms remain untested in human RHD valve tissue or RHD-specific models 177 Anti-fibrotic and anti-remodeling effects SGLT2i may attenuate TGF-β– and MMP-mediated fibrotic remodeling, EndMT, and maladaptive ECM turnover, suggesting a potential role before irreversible fibrosis and calcification are established 80 , 102 , 178 Greatest benefit in early or compensated stages of RHD The most plausible therapeutic window may be subclinical, borderline, or mild-to-moderate compensated RHD, when inflammatory and fibrotic processes may still be modifiable 155 , 166 NF-κB = nuclear factor kappa B; NLRP3 = NOD-, LRR-, and pyrin domain-containing protein 3; NT-proBNP = N-terminal pro–B-type natriuretic peptide; SGLT2i = sodium-glucose cotransporter 2 inhibitor; other abbreviations as in Tables 1 , 2 , and 3 . Note: Evidence coding: Human RHD. Evidence extrapolated from experimental or non-rheumatic valvular disease settings. Mechanisms that remain speculative in RHD. Proposed SGLT2 Inhibitor Effects on RHD-Relevant Pathophysiology NF-κB = nuclear factor kappa B; NLRP3 = NOD-, LRR-, and pyrin domain-containing protein 3; NT-proBNP = N-terminal pro–B-type natriuretic peptide; SGLT2i = sodium-glucose cotransporter 2 inhibitor; other abbreviations as in Tables 1 , 2 , and 3 . Note: Evidence coding: Human RHD. Evidence extrapolated from experimental or non-rheumatic valvular disease settings. Mechanisms that remain speculative in RHD. The therapeutic promise of SGLT2i in RHD must be regarded as a compelling but unproven hypothesis. Realizing this potential will require a structured translational and clinical research program. As a first step, human RHD valve tissue studies should assess whether SGLT2 is expressed and upregulated in rheumatic valves, and whether its expression correlates with inflammatory activity, EndMT markers, or fibrosis burden. Experimental models replicating rheumatic valvulitis should then be used to test whether SGLT2 inhibition attenuates immune activation, endothelial dysfunction, and ECM remodeling in an etiologically relevant context. Therapeutic specificity is essential to any credible translational roadmap. The most biologically plausible window for SGLT2i intervention is the transition from compensated to decompensated valvular heart disease ( Figure 4 ), most likely in patients with mild-to-moderate RHD who have not yet developed irreversible structural remodeling or pulmonary hypertension. The rationale is 3-fold. • At early stages, the myocardium retains plasticity and SGLT2i-mediated effects on mitochondrial energetics, inflammation, and fibrosis are mechanistically actionable. • Patients in advanced stages of disease (severe MR/mitral stenosis with atrial fibrillation or pulmonary hypertension) are less likely to derive incremental benefit from pharmacologic disease modification, as treatment effects are harder to isolate, and these patients are more likely to proceed to interventional or surgical treatment. • In endemic RHD populations, a high proportion of patients are diagnosed in adolescence or young adulthood at compensated stages, making this a strategically accessible target window. Figure 4 Optimal Window for SGLT2i Intervention in RHD Mechanistic rationale, endpoints stratified for low- and middle-income countries (LMICs) feasibility, and phase 2a trial schematic designed for endemic settings. Panel 1: The biologically plausible window for sodium-glucose cotransporter 2 inhibitor (SGLT2i) intervention in RHD occurs when myocardial plasticity is preserved and antifibrotic and anti-inflammatory effects remain potentially actionable. In later-stage disease, irreversible remodeling and the need for valve intervention may limit the incremental benefit of pharmacologic disease modification. Panel 2: Candidate endpoints are organized by mechanistic domain and feasibility in LMIC settings, with N-terminal pro–B-type natriuretic peptide (NT-proBNP) and left ventricular remodeling indices proposed as co-primary endpoints. Panel 3: Proposed proof-of-concept trial in endemic settings enrolling adolescents and young adults with compensated, mitral-predominant RHD. Participants are randomized 1:1 to SGLT2i or placebo for 52 weeks, with serial clinical, biomarker, and echocardiographic assessments. Co-primary endpoints are change in NT-proBNP and left ventricular end-diastolic volume (LVEDV). 6MWT = 6-minute walk test; eGFR = estimated glomerular filtration rate; hsCRP = high-sensitivity C-reactive protein; IL = interleukin; LVEF = left ventricular ejection fraction; other abbreviations as in Figures 1 and 2 . At early stages, the myocardium retains plasticity and SGLT2i-mediated effects on mitochondrial energetics, inflammation, and fibrosis are mechanistically actionable. Patients in advanced stages of disease (severe MR/mitral stenosis with atrial fibrillation or pulmonary hypertension) are less likely to derive incremental benefit from pharmacologic disease modification, as treatment effects are harder to isolate, and these patients are more likely to proceed to interventional or surgical treatment. In endemic RHD populations, a high proportion of patients are diagnosed in adolescence or young adulthood at compensated stages, making this a strategically accessible target window. Optimal Window for SGLT2i Intervention in RHD Mechanistic rationale, endpoints stratified for low- and middle-income countries (LMICs) feasibility, and phase 2a trial schematic designed for endemic settings. Panel 1: The biologically plausible window for sodium-glucose cotransporter 2 inhibitor (SGLT2i) intervention in RHD occurs when myocardial plasticity is preserved and antifibrotic and anti-inflammatory effects remain potentially actionable. In later-stage disease, irreversible remodeling and the need for valve intervention may limit the incremental benefit of pharmacologic disease modification. Panel 2: Candidate endpoints are organized by mechanistic domain and feasibility in LMIC settings, with N-terminal pro–B-type natriuretic peptide (NT-proBNP) and left ventricular remodeling indices proposed as co-primary endpoints. Panel 3: Proposed proof-of-concept trial in endemic settings enrolling adolescents and young adults with compensated, mitral-predominant RHD. Participants are randomized 1:1 to SGLT2i or placebo for 52 weeks, with serial clinical, biomarker, and echocardiographic assessments. Co-primary endpoints are change in NT-proBNP and left ventricular end-diastolic volume (LVEDV). 6MWT = 6-minute walk test; eGFR = estimated glomerular filtration rate; hsCRP = high-sensitivity C-reactive protein; IL = interleukin; LVEF = left ventricular ejection fraction; other abbreviations as in Figures 1 and 2 . As for candidate endpoints for early-phase clinical trials, they should be stratified by feasibility in LMIC settings, where laboratory and imaging infrastructure remain variable. Trial design must account for the epidemiologic and health system realities of RHD-endemic regions, including resource constraints, limited echocardiographic follow-up infrastructure, adherence challenges, and drug affordability. Pragmatic, adaptive trial designs embedded within existing RHD registry networks such as those operating in sub-Saharan Africa and South Asia may offer the most efficient path to generating definitive evidence. 2

The long-term progression of RHD is likely driven by the interplay of persistent immune activation, endothelial dysfunction, maladaptive valvular remodeling, and altered valve biomechanics. Rather than reflecting a single linear process, established disease appears to evolve through reciprocal interactions among inflammatory injury, endothelial and interstitial cell responses, extracellular matrix remodeling, and mechanical stress. The following sections examine these interconnected mechanisms, beginning with valvular endothelial dysfunction as a plausible early driver of progressive rheumatic valve injury. Mitral valve endothelial cells (mVECs) form a specialized endothelial monolayer covering both the atrial and ventricular surfaces of the leaflet, where they act as a dynamic barrier that regulates immune-cell trafficking, senses mechanical cues, and preserves local anti-inflammatory and antithrombotic homeostasis. 74 , 75 , 76 , 77 , 78 In the context of RHD, available evidence supports the concept that endothelial activation is an early component of valve injury, although the full downstream sequence of mVEC dysfunction has not been demonstrated uniformly in human disease. In particular, studies in RHD support inflammatory endothelial activation, including increased expression of adhesion molecules and pathways linked to valvular remodeling. For example, streptococcal–vimentin cross-reactive antibodies have been shown to induce a proinflammatory phenotype in cardiac endothelial cells, including nuclear factor (NF)-κB–related activation and upregulation of intercellular adhesion molecule-1 and VCAM-1, providing a plausible mechanism for enhanced leukocyte recruitment within the valve. 79 In parallel, human RHD tissue studies support a role for EndMT in valvular remodeling, with loss of endothelial markers and acquisition of mesenchymal traits contributing to leaflet thickening and fibrosis. 80 We propose that endothelial dysfunction in RHD extends beyond inflammatory activation alone and may involve a broader cascade including cytokine-driven apoptosis, extracellular vesicle release, impaired Akt/endothelial nitric oxide synthase (eNOS) signaling, and a transition toward a prothrombotic endothelial phenotype. Together, these processes provide a biologically coherent framework linking immune injury to oxidative stress, 81 loss of nitric oxide bioavailability, and progressive valvular remodeling. Supporting this concept, Ci et al 82 reported increased circulating endothelial extracellular vesicles in mitral valve disease and showed that these vesicles impair mitral valve endothelial function via inhibition of the Akt/eNOS-HSP90 pathway. Although this mechanism has not yet been specifically delineated in human RHD, it provides a relevant mechanistic parallel for understanding how endothelial injury may be perpetuated in rheumatic valve disease. Reduced NO signaling, oxidative stress, and thrombogenic endothelial surface transformation may therefore represent key downstream mediators within this proposed model of RHD progression. 83 , 84 , 85 , 86 Beyond immune-mediated injury, the hemodynamic consequences of MR, which is the earliest and most common valvular lesion in acute rheumatic fever, likely impose an additional mechanically driven insult on the valvular endothelium. Regurgitant flow across an incompetent mitral valve generates abnormal patterns of oscillatory and retrograde shear stress, particularly on the atrial surface of the leaflet, thereby disrupting the laminar flow conditions under which mVECs normally preserve homeostatic quiescence. 77 , 87 In vascular and valvular endothelial biology, disturbed flow is well recognized as a driver of NF-κB activation, upregulation of adhesion molecules, increased endothelial permeability, and suppression of eNOS-derived nitric oxide signaling. 88 , 89 In the setting of rheumatic MR, these biomechanical forces may further amplify the inflammatory endothelial phenotype initially induced by GAS-driven autoimmunity and may help sustain EndMT and subsequent valvular interstitial cell (VIC) activation, even in the absence of clinically recognized recurrent ARF episodes. This interaction between immune injury and abnormal valve hemodynamics provides a plausible valve intrinsic explanation for progressive remodeling once MR is established. At the same time, it also highlights an important gap in the current evidence, as the full pathway linking disturbed rheumatic flow, mVEC dysfunction, VIC activation, fibrosis, and later calcific remodeling has not yet been fully demonstrated in human RHD. Consistent with this concept, reciprocal signaling between mVECs and mVICs appears to be critical for leaflet homeostasis: healthy endothelium helps maintain VIC quiescence, whereas VICs restrain excessive EndMT. Disruption of this bidirectional relationship may therefore represent a key convergence point through which immune injury and altered hemodynamics promote progressive leaflet thickening and fibrosis. 87 VICs are the major cells in heart valves, responsible for maintaining ECM homeostasis by regulating collagen synthesis and degradation. 75 , 90 Under normal conditions, VICs remain quiescent; but in RHD, chronic inflammation due to recurrent streptococcal infection activates VICs, promoting their differentiation into myofibroblasts via TGF-β1 signaling. 91 , 92 , 93 Prolonged inflammation also induces osteogenic differentiation of VICs, contributing to calcification and stiffening of the valve leaflets. 94 EndMT further exacerbates VIC activation and valvular fibrosis. 95 , 96 These activated VICs overproduce collagen and remodel the extracellular matrix (ECM); concurrently, increased matrix metalloproteinase (MMP)-1 and decreased tissue inhibitor of metalloproteinase (TIMP)-1 disrupt collagen homeostasis—a process closely linked to RHD progression. 93 , 97 A study on human pathological mitral valves showed that the ECM composition is profoundly altered, with a 3-fold increase in collagen, proteoglycans, and elastin compared with normal valves. 98 In addition, the fragmentation and disorganization of elastin leads to a loss of elasticity in the valvular leaflets. 99 Collectively, these mechanisms drive progressive fibrosis, calcification, and leaflet thickening, resulting in the stiff, noncompliant mitral valves characteristic of chronic RHD. Several reports have emphasized the central role of TGF-β and inflammatory cytokines in the pathogenesis of valvular fibrosis in RHD. 100 Once activated, VICs drive progressive fibrosis and calcification, leading to irreversible structural remodeling of the mitral valve. Although VIC activation and ECM remodeling are shared features across multiple valvular disorders, the remodeling phenotype of RHD is distinct from degenerative valve disease. In contrast to the nodular fibrocalcific remodeling of calcific aortic valve disease (CAVD) or the matrix-disorganization pattern of myxomatous mitral valve disease, RHD is characterized by chronic inflammatory infiltrates, diffuse leaflet fibrosis, prominent neoangiogenesis, commissural fusion, and a mixed stenotic-regurgitant phenotype. These comparative features are summarized in Table 1 . Table 1 Comparative Remodeling Features of Rheumatic, Calcific, and Myxomatous Valve Disease Features Rheumatic CAVD MMVP Primary driver Autoimmune; molecular mimicry (GAS M-protein) Atherosclerosis-like; oxidative/mechanical stress ECM dysregulation; MMP overexpression Inflammatory infiltrate CD4 + T cells, macrophages, plasma cells; Aschoff nodules Macrophages, foam cells; lipid deposits Minimal VIC activation mechanism TGF-β1/Smad2-3 via streptococcal-driven inflammation Runx2/Wnt; BMP-2; LDL oxidation TGF-β/MMP axis; mechanical stretch Fibrosis pattern Diffuse, uniform subendothelial; leaflet thickening Nodular, focal; leaflet base and body Spongiosa expansion; collagen disorganization Calcification pattern Commissures and leaflet tips Nodular, leaflet base/body Absent or rare Commissural fusion Pathognomonic Absent Absent Neo-angiogenesis Prominent Moderate Minimal Primary valve + lesion Mitral, mixed stenosis ± regurgitation Aortic stenosis Mitral regurgitation BMP = bone morphogenetic protein; CAVD = calcific aortic valve disease; ECM = extracellular matrix; GAS = Group A Streptococcus ; LDL = low-density lipoprotein; MMP = matrix metalloproteinase; MMVP = myxomatous mitral valve disease; TGF = transforming growth factor; VIC = valve interstitial cell. Comparative Remodeling Features of Rheumatic, Calcific, and Myxomatous Valve Disease BMP = bone morphogenetic protein; CAVD = calcific aortic valve disease; ECM = extracellular matrix; GAS = Group A Streptococcus ; LDL = low-density lipoprotein; MMP = matrix metalloproteinase; MMVP = myxomatous mitral valve disease; TGF = transforming growth factor; VIC = valve interstitial cell. Rheumatic mitral stenosis is characterized by progressive fibrotic thickening of the mitral valve, leading to structural remodeling and functional impairment. Several circulating biomarkers reflecting cardiac fibrosis in RHD have been identified, including TGF-β1, the circulating carboxy-terminal propeptide of type I procollagen (PICP), and the MMP-1/TIMP-1 ratio. 101 , 102 Among them, TGF-β1 appears to play a pivotal role, as it orchestrates collagen synthesis and ECM remodeling. Histological studies have shown increased deposition of collagen types I and III in rheumatic mitral valves compared with non-rheumatic controls. 97 Notably, TGF-β1 overexpression is strongly associated with valvular fibrosis in rheumatic mitral valve disease, as it activates myofibroblasts and enhances collagen production. 102 In vitro, stimulation of VICs with TGF-β1 induces myofibroblastic differentiation, a process that can be inhibited by fibroblast growth factor-2. 92 In line with this idea, Ambari et al 103 suggested that angiotensin-converting enzyme inhibitors may play a potential role in attenuating cardiac fibrosis in RHD via the IL-33/ST2 axis. In addition, inflammatory cells and myofibroblasts secrete MMPs that regulate ECM turnover; elevated plasma MMP-1 levels and related gene polymorphisms have been linked to increased RHD risk. 93 However, studies on the role of MMPs in the chronic phase of RHD remain limited, as most research has focused on human plasma and myocardium tissues during acute episodes of ARF. 104 Calcification is present in some cases of RHD; and the underlying cellular mechanisms remain to be fully elucidated, with most evidence being extrapolated from other diseases such as vascular calcification and CAVD. Similar to fibrosis, MMPs contribute to valvular calcification by degrading elastin, creating a microenvironment that promotes the differentiation of myofibroblastic VICs into osteoblastic-like cells. 105 In addition, studies suggest that mineralization occurs predominantly in inflamed and neoangiogenic areas, where vascular endothelial growth factor (VEGF) plays a crucial role in attracting endothelial cells and stimulating osteoblast differentiation. 106 Another proposed mechanism involves calcification-competent extracellular vesicles derived from smooth muscle cells, VICs, or macrophages, resembling the inflammatory-driven calcification observed in CAVD. More specifically, inflammatory cells release extracellular vesicles that serve as scaffolds for calcium-phosphate crystal nucleation, while dysregulated phosphocalcic metabolism further promotes calcium precipitation on extracellular phosphate deposits. 107 , 108 , 109 Critically, in RHD, the transition from early inflammatory valvular injury to established fibrosis and calcification represents a clinically meaningful stage shift, with potential biomarkers beginning to emerge. Candidate markers of fibrosis include circulating procollagen type I C-terminal propeptide (PICP), the MMP-1/TIMP-1 ratio, transforming growth factor-β1, and soluble ST2, which may capture the onset of irreversible matrix remodeling. 101 , 102 For calcific transition, computed tomography–derived valvular calcium burden offers a quantitative and reproducible imaging marker, potentially strengthened by integration with circulating osteogenic markers such as osteopontin and bone morphogenetic protein-2. Prospective studies linking these markers to echocardiographic progression, clinical outcomes, and histopathology are needed to validate their role as indicators of stage transition and as potential surrogate endpoints in trials of disease-modifying therapies. In parallel with fibrosis and calcification, chronic inflammation stimulates pathological neoangiogenesis and lymphangiogenesis within the valve tissue. Neoangiogenesis plays a crucial role in immune-cell infiltration, chronic inflammation, and tissue remodeling. Unlike normal heart valves, which are largely avascular, RHD-affected valves develop an extensive immature vascular network, characterized by fragile neovessels with incomplete pericyte coverage. 110 These abnormal vessels facilitate the influx of immune cells, perpetuating inflammation, and fibrosis. This pathological angiogenesis is primarily driven by VEGF, MMPs, hypoxia-inducible factor-1α, and proinflammatory cytokines such as TNF-α and IL-6. 29 , 111 , 112 , 113 A key question is whether neoangiogenesis in RHD is merely a marker of chronic inflammatory injury or an active contributor to disease progression. The immature, pericyte-deficient neovessels observed in RHD valves support the latter possibility. Their structural immaturity may promote persistent permeability, enabling immune-cell extravasation and sustaining a feed-forward cycle in which inflammation drives further angiogenic signaling. In this context, VEGF-driven neovascularization may enhance inflammatory cell recruitment, while hypoxia-inducible factor-1α and TNF-α further amplify VEGF expression. 29 , 111 Pathological angiogenesis may therefore be more than an epiphenomenon of valvular injury and instead represent a mechanistically relevant component of RHD progression. A distinct feature of RHD is its significant neolymphangiogenesis, marked by the presence of lymphatic vessel endothelial hyaluronan receptor 1 (LYVE-1), podoplanin, and lymphatic endothelial cells. 114 , 115 Mechanistically, this lymphatic dysfunction may reflect inflammation-driven lymphatic remodeling, in which newly formed lymphatic vessels fail to achieve functional competence, leading to impaired antigen clearance and defective resolution of valvular inflammation. Although lymphatic vessels normally aid in immune regulation and fluid drainage, their function in RHD appears to be impaired due to persistent inflammation. As a result, lymphatic dysfunction contributes to tissue edema, excessive collagen deposition, progressive fibrosis, and worsening valvular stiffness and dysfunction. 116 , 117 Whether restoration of lymphatic function can attenuate valvular remodeling in RHD remains a testable but as yet unexamined hypothesis. Unlike other forms of valvular disease, such as CAVD, 118 , 119 RHD lacks a well-documented association with intraleaflet hemorrhage, but its extensive neovascularization remains a key factor in disease progression. 29 , 120 The ongoing inflammatory response, driven by infiltrating monocytes, macrophages, and T cells, leads to excessive ECM remodeling, further compromising the mechanical stability of the valve. These pathological changes highlight neoangiogenesis as one of the central drivers of RHD progression, not only sustaining chronic inflammation but also contributing to long-term structural deterioration of the affected valves. 29 , 121 Translating these insights into actionable pathways will require validated measures of neoangiogenesis and lymphangiogenesis. Candidate markers include circulating VEGF, angiopoietin-2, together with tissue-based indices of microvessel density, pericyte coverage, and LYVE-1– or podoplanin-positive vessels. Prospective studies are needed to determine whether these measures track disease severity, progression, and outcomes and can serve as stage-transition markers or surrogate endpoints. The mitral valve operates in a mechanically demanding environment in which shear stress, cyclic strain, and transvalvular pressure gradients continuously modulate the behavior of valvular endothelial cells and VICs. Under physiological conditions, laminar antegrade flow across the mitral orifice sustains a cardioprotective endothelial phenotype characterized by high eNOS activity, anti-inflammatory gene expression, and suppression of VIC-activating signals. 122 , 123 , 124 This hemodynamic homeostasis is profoundly disrupted in RHD. In early disease, MR is the hallmark lesion 16 and represents the principal mechanobiological insult driving this disturbance. MR generates abnormal flow disturbances that expose the atrial surface of the mitral endothelium to oscillatory and retrograde shear stress. Oscillatory shear stress is a well-established trigger of endothelial proinflammatory signaling, promoting NF-κB nuclear translocation, upregulating VCAM-1 and intercellular adhesion molecule-1, reducing nitric oxide bioavailability through eNOS uncoupling, and increasing endothelial permeability and leukocyte adhesion. In the rheumatic valve, these flow-driven signals act in direct synergy with the cytokine-rich, autoimmune inflammatory milieu already established by molecular mimicry and persistent GAS-antigen stimulation. The concomitant presence of immunological and mechanobiological endothelial stressors may help explain the greater severity and more rapid progression of MR-associated RHD compared with isolated MR of other etiologies. Although direct mechanobiological evidence in RHD remains limited, studies from mitral and aortic degenerative valve pathologies suggest that integrins, cadherins, caveolae, ion channels, and the endothelial glycocalyx play key roles in force sensing and signal transduction. 125 , 126 , 127 , 128 , 129 , 130 In the setting of MR-associated disturbed flow, dysregulation of these sensors triggers mechano-transduced signals that directly activate VICs, promoting their myofibroblastic differentiation and driving collagen overproduction, ECM stiffening, and progressive leaflet thickening. Critically, the mechanical burden imposed by MR is not static. As leaflet thickening and commissural fusion reduce valve compliance, the resulting increase in regurgitant fraction amplifies hemodynamic stress and further deranges endothelial mechanosensing, establishing a self-reinforcing biomechanical-inflammatory loop. This loop operates independently of recurrent streptococcal infection and provides a mechanistic explanation for the observation that RHD can progress in the absence of new ARF episodes. Targeting this loop whether through load reduction, endothelial protection, or mechanosensing pathway inhibition may therefore represent a viable therapeutic strategy in early, inflammation-dominant RHD. The mechanistic framework presented here is derived largely from evidence in degenerative and calcific valve diseases. However, the biology of the rheumatic valve, shaped by recurrent streptococcal immune injury, postinflammatory scarring, and a predominantly young patient population, differs fundamentally from that of age-related degenerative valve disease. Mechanotransduction pathways, calcification cascades, and fibrotic mediators characterized in CAVD or degenerative mitral valve disease may therefore not translate directly to RHD. This limits the precision with which mechanistic hypotheses can be formulated and underscores the urgent need for RHD-specific molecular and translational studies. Current models explain the initiation of RHD better than its long-term progression. They do not fully explain why valve damage can continue for years after ARF, or why left-sided valves, especially the mitral valve, are affected most often. 21 , 23 , 25 Although recurrent ARF episodes can aggravate valve injury, registry data from Australia show that progression of established RHD is more frequent than ARF recurrence over 10 years. In addition, secondary antibiotic prophylaxis appears to slow disease progression beyond simply preventing recurrent ARF. 131 , 132 A recent study by de Barros Branco et al 133 demonstrated that borderline subclinical RHD is associated with persistently elevated IL-6 levels, detectable from the earliest stages and sustained over at least 2 years of follow-up. Together, these observations suggest that once initiated, RHD may enter a biologically active phase that is capable of sustaining further remodeling. Figure 3 depicts the dynamic crosstalk linking autoimmunity, chronic inflammation, and cellular-structural remodeling in RHD progression. Figure 3 Proposed Pathogenesis of RHD Initiation and Progression A structurally normal mitral valve exposed to recurrent GAS infection may evolve over years into a thickened, fibrotic, and commissurally fused stenotic valve. (A) Morphologic progression: progressive leaflet thickening, loss of pliability, and commissural fusion culminate in mitral stenosis. (B) Disease initiation: molecular mimicry and autoimmune endothelial injury after GAS infection trigger early valvular inflammation and expose subendothelial collagen. (C) Progressive remodeling: persistent inflammation, endothelial activation, disturbed flow, oxidative stress, and endothelial-to-mesenchymal transition (EndMT) promote valvular interstitial cell activation, extracellular matrix remodeling, fibrosis, neoangiogenesis, and, in some cases, later calcification. Together, these interconnected mechanisms may establish a self-perpetuating cycle that drives progression from early inflammatory injury to irreversible valvular deformity. EC = endothelial cell; eNOS = endothelial nitric oxide synthase; mVEC = mitral valve endothelial cell; RHD = rheumatic heart disease; ROS = reactive oxygen species; TGF = transforming growth factor; VIC = valvular interstitial cell; other abbreviations as in Figures 1 and 2 . Proposed Pathogenesis of RHD Initiation and Progression A structurally normal mitral valve exposed to recurrent GAS infection may evolve over years into a thickened, fibrotic, and commissurally fused stenotic valve. (A) Morphologic progression: progressive leaflet thickening, loss of pliability, and commissural fusion culminate in mitral stenosis. (B) Disease initiation: molecular mimicry and autoimmune endothelial injury after GAS infection trigger early valvular inflammation and expose subendothelial collagen. (C) Progressive remodeling: persistent inflammation, endothelial activation, disturbed flow, oxidative stress, and endothelial-to-mesenchymal transition (EndMT) promote valvular interstitial cell activation, extracellular matrix remodeling, fibrosis, neoangiogenesis, and, in some cases, later calcification. Together, these interconnected mechanisms may establish a self-perpetuating cycle that drives progression from early inflammatory injury to irreversible valvular deformity. EC = endothelial cell; eNOS = endothelial nitric oxide synthase; mVEC = mitral valve endothelial cell; RHD = rheumatic heart disease; ROS = reactive oxygen species; TGF = transforming growth factor; VIC = valvular interstitial cell; other abbreviations as in Figures 1 and 2 . In this setting, autoimmune endothelial injury after GAS infection is likely the initiating event, establishing a persistent inflammatory milieu within the valve. From there, disease progression is unlikely to be driven equally by all downstream pathways. Persistent inflammatory activation of the valvular endothelium appears to be one of the central processes. This is likely reinforced by the abnormal mechanical burden associated with MR, the earliest and most common valvular lesion in ARF, and by the inherently high-pressure environment of the left-sided valves. The mitral valve may also be especially vulnerable to progressive fibrotic remodeling and pathological neoangiogenesis. Supporting this view, mitral leaflet thickening present at the time of ARF diagnosis has consistently been associated with later progression to RHD. 134 As discussed in the sections “Mitral valve endothelial cells in RHD” and “Mechanotransduction in RHD,” disturbed flow across the rheumatic mitral valve may further worsen endothelial dysfunction, reduce protective nitric oxide signaling, and promote EndMT even in the absence of clinically recognized recurrent ARF. These endothelial and hemodynamic disturbances are then translated into structural remodeling through activation of valve interstitial cells and disruption of ECM homeostasis. Activated mVICs promote collagen accumulation, leaflet thickening, and progressive stiffening of the valve. At the same time, pathological neoangiogenesis and lymphangiogenesis may facilitate continued immune-cell entry, cytokine signaling, and ineffective resolution of injury. Over time, this combination of persistent inflammation, endothelial dysfunction, altered mechanics, and matrix remodeling drives fibrosis, commissural fusion, and, in some patients, later calcification. In this way, what begins as an immune-mediated injury may evolve into a chronic and self-perpetuating valve disease. Table 2 summarizes the evidence base supporting the major mechanism implicated in RHD initiation and progression. Table 2 Evidence Base Supporting Major Mechanisms Implicated in RHD Initiation and Progression Mechanism Main Evidence Category Evidence Base Molecular mimicry Human RHD tissue data + animal/model data Human studies support cross-reactivity between streptococcal antigens and cardiac proteins, immune-cell infiltration, and valve inflammation. 39 , 46 Animal models further support a causal role of T-cell–mediated valvulitis. 48 , 49 Neo-antigen/collagen-directed autoimmunity Animal/model data with limited human supportive data Elevated anticollagen antibodies have been reported in patients with ARF/RHD. 69 , 71 Mechanistic support largely comes from experimental studies involving streptococcal collagen-binding proteins and immunization models. 47 , 65 mVEC dysfunction Human RHD tissue data + extrapolation from other diseases Human RHD data support endothelial activation, adhesion molecule upregulation, and EndMT. 79 , 80 However, broader downstream mechanisms such as extracellular vesicles-mediated injury, Akt/eNOS-HSP90 impairment, apoptosis, and prothrombotic transformation are inferred mainly from mitral valve disease or other cardiovascular settings. 82 , 89 mVIC activation/ECM remodeling Human RHD tissue data Human rheumatic valve studies support VIC activation, excess collagen deposition, altered proteoglycan/elastin composition, and dysregulated MMP/TIMP balance. 97 , 102 Additional mechanistic links involve TGF-β–driven myofibroblast transition. 75 , 92 Valve fibrosis Human RHD tissue data Histological and biomarker studies in human RHD demonstrate increased collagen I/III deposition, TGF-β1 upregulation, and fibrosis-associated remodeling of rheumatic mitral valves. 97 , 103 Calcification Extrapolation from other diseases, with limited descriptive human RHD data Calcification is observed in some rheumatic valves, but most mechanistic explanations are extrapolated from vascular calcification and calcific aortic valve disease, including osteogenic VIC transition, elastin degradation, inflammatory extracellular vesicles, and osteogenic signaling. 107 , 108 Altered hemodynamics/abnormal mechanical stress Extrapolation from other diseases with indirect clinical support in RHD Clinical observations strongly support early MR as a major lesion in ARF/RHD. 134 However, the mechanistic link between disturbed rheumatic flow, endothelial dysfunction, VIC activation, and progressive fibrosis is largely extrapolated from valvular and vascular mechanobiology literature. 77 , 88 , 89 Chronic inflammation Human RHD tissue data Human studies support persistent immune-cell infiltration, cytokine activation, macrophage involvement, and chronic inflammatory remodeling of rheumatic valves, even beyond the acute ARF phase. 49 , 79 Neoangiogenesis Human RHD tissue data Histopathological studies of rheumatic valves demonstrate abnormal neovessel formation, immature vascular networks, VEGF-associated signaling, and lymphatic remodeling markers such as LYVE-1 and podoplanin. 29 , 115 ARF = acute rheumatic fever; EndMT = endothelial-to-mesenchymal transition; eNOS = endothelial nitric oxide synthase; LYVE-1 = lymphatic vessel endothelial hyaluronan receptor 1; MR = mitral regurgitation; mVEC = mitral valve endothelial cell; mVIC = mitral valve interstitial cell; RHD = rheumatic heart disease; TIMP = tissue inhibitor of metalloproteinase; VEGF = vascular endothelial growth factor; other abbreviations as in Table 1 . Evidence Base Supporting Major Mechanisms Implicated in RHD Initiation and Progression ARF = acute rheumatic fever; EndMT = endothelial-to-mesenchymal transition; eNOS = endothelial nitric oxide synthase; LYVE-1 = lymphatic vessel endothelial hyaluronan receptor 1; MR = mitral regurgitation; mVEC = mitral valve endothelial cell; mVIC = mitral valve interstitial cell; RHD = rheumatic heart disease; TIMP = tissue inhibitor of metalloproteinase; VEGF = vascular endothelial growth factor; other abbreviations as in Table 1 . From a translational perspective, these pathways should not be viewed as equally targetable. Advanced fibrosis, commissural fusion, and calcification are more likely to represent late and relatively fixed consequences of disease that often require invasive intervention. By contrast, earlier phases marked by persistent endothelial activation, oxidative stress, inflammatory signaling, and MR-related mechanobiological stress may offer a more realistic window for pharmacologic intervention. This framework helps identify where disease-modifying strategies are most likely to succeed, particularly before fibrocalcific remodeling becomes established. It also provides a rationale for exploring therapies aimed at endothelial protection, inflammatory modulation, and interruption of maladaptive valve remodeling, as discussed in the section “Exploring the Role of SGLT2 Inhibitors in the Progression of RHD.”

Despite being largely preventable, RHD remains a leading cause of cardiovascular mortality among young individuals in LMICs. The immune-endothelial-mechanical interplay driving rheumatic valve remodeling and the potential therapeutic actions of SGLT2 inhibitors are summarized in the Central Illustration . Disease progression is driven by persistent immune activation and endothelial dysfunction, which together initiate EndMT, activation of VICs, and maladaptive ECM remodeling. These processes, amplified by pathological neoangiogenesis and lymphangiogenesis, sustain chronic inflammation and ultimately culminate in fibrotic and calcific valve deformities. Current therapeutic strategies remain limited to infection prophylaxis and late surgical or transcatheter interventions, leaving the chronic inflammatory and fibrotic drivers of RHD unaddressed. Central Illustration Mechanisms and Therapeutic Targets in Rheumatic Heart Disease Mechanisms and Therapeutic Targets in Rheumatic Heart Disease The mechanistic architecture of RHD outlined in this review converges on a set of druggable nodes that have yet to be systematically exploited. Among these, SGLT2is represent a particularly compelling candidate class, given their demonstrated capacity to attenuate NF-κB–driven inflammation, restore endothelial nitric oxide bioavailability, suppress TGF-β–mediated profibrotic signaling, and reduce oxidative stress. All these mechanisms map directly onto the core pathological pathways identified in RHD. This mechanistic alignment generates several testable hypotheses. Priorities for future study are to determine whether SGLT2 inhibitors suppress EndMT and VIC activation in human valvular cells, reduce inflammatory and fibrotic signaling in preclinical RHD, and favorably modify valvular remodeling when initiated early in ARF-associated carditis. Addressing these questions will require coordinated investment in mechanistic studies, improved ARF-relevant animal models, and pragmatic trial infrastructure in high-burden settings. Targeting intersecting inflammatory, endothelial, and fibrotic pathways may provide a scientifically grounded strategy to slow progression and reduce the global burden of RHD.

RHD is a severe postinfectious complication of ARF, resulting from an autoimmune response to pharyngitis caused by Streptococcus pyogenes , a bacterium belonging to GAS. 16 ARF is characterized by a variable combination of joint pain and swelling, as well as cardiac involvement, primarily presenting as mitral regurgitation (MR) in the early stages and progressing to mitral stenosis in the later stages. Although most clinical symptoms of ARF resolve over time, cardiac valvular damage can persist, leading to chronic complications. 17 It is estimated that 1% to 3% of individuals with untreated GAS infections will develop ARF, and among these, up to 60% may progress to chronic RHD. 18 The initial episode of ARF typically presents with mild symptoms; however, repeated infections with Streptococcus pyogenes amplify the immune response, exacerbating RHD. 16 , 19 , 20 , 21 , 22 , 23 Epidemiological data suggest that, once established, RHD can continue to progress even in the absence of new ARF episodes. Consistently, data from an Australian registry indicated that RHD progression occurred more frequently than ARF recurrence, with a 10-year progression risk of 51.9% compared with an ARF recurrence rate of only 19.8%. 24 This observation may reflect unrecognized recurrent episodes, repeated subclinical streptococcal exposure, persistent antigenic stimulation, trained immunity, or self-sustaining inflammatory and remodeling processes. This may partly explain why RHD often begins in childhood (ages 5-14) but reaches its peak clinical burden in adulthood, typically between 25 and 45 years of age. 23 , 25 Mitral valve involvement is the hallmark of RHD. MR is the most common early valvular lesion and may progress over time to mitral stenosis through persistent valvulitis, leaflet thickening, and commissural fusion. Aortic regurgitation is also common, often in association with mitral disease, whereas tricuspid regurgitation is usually secondary to pulmonary hypertension and advanced left-sided valve disease. Isolated right-sided valve involvement is uncommon. 26 , 27 , 28 The predominance of mitral and aortic valve injury raises the possibility that, beyond systemic immune activation, local endothelial and biomechanical factors may also influence valve-specific susceptibility. In advanced mitral stenosis, commissural fusion, leaflet fibrosis, subvalvular thickening, and chordal shortening are common pathological features. Calcification may be present in some cases, but fibrosis and chronic inflammatory remodeling remain the predominant pathological features of rheumatic valve disease. 29 , 30 , 31 , 32 , 33 ARF results from an abnormal immune response to GAS infection, leading to autoimmune reactions that affect the heart, brain, joints, and skin. The immune response originates in the pharyngeal epithelium, where innate immune cells present bacterial antigens (M, T, R proteins and N-acetylglucosamine [GlcNAc]) to both B and T lymphocytes. 34 This interaction triggers the production of antibodies from B cells, some of which may cross-react with host tissues. 20 , 35 , 36 In genetically susceptible individuals, this immune reaction mistakenly targets host tissues through a process called molecular mimicry, where antibodies and T cells cross-react with host antigens, causing inflammation and tissue damage. Among the bacterial antigens, the M-protein plays a central role. Extensive studies have identified the M-protein as the most virulent and pathogenic antigen associated with RHD. 37 , 38 Its structural resemblance to key cardiac proteins, including myosin, laminin, vimentin, and tropomyosin, provokes a robust autoimmune response, which targets and damages heart tissues. Molecular mimicry is a key factor in the immune-mediated injury seen in RHD. 39 , 40 Another key antigen is GlcNAc, a carbohydrate component on the cell surface of Streptococcus pyogenes . 41 GlcNAc is involved in RHD pathogenesis, as persistently elevated anti-GlcNAc antibody levels have been strongly linked to valvulitis development. 42 Host antigens are expressed in myocardial and valvular tissues, leading to pancarditis in the early, potentially reversible stages, followed by chronic valvular damage in the later stages of the disease. 39 , 43 , 44 , 45 Valvular damage begins when circulating autoantibodies bind to the endothelial surface of the valve, triggering the upregulation of vascular cell adhesion molecule (VCAM)-1. This endothelial activation allows the infiltration of T lymphocytes, primarily CD4+ T cells, into the valvular subendothelium, initiating a T-helper 1–mediated immune response. This response leads to oedema and elongation of the chordae tendineae. 46 As tissue injury progresses, extracellular matrix (ECM) components, including collagen type I, become exposed, which stimulates the production of anticollagen antibodies. 47 Antibodies accumulate within the valve tissue, sustaining a proinflammatory environment and making the heart valves particularly vulnerable to immune-mediated damage. Notably, there is currently no evidence indicating that autoantibodies alone are sufficient to cause valve lesions, suggesting the crucial role of cellular immune responses alongside humoral factors. However, emerging research highlights that T cells play a key role in this process. Studies have demonstrated that transferring T-cell lines from M-protein–vaccinated Lewis rats to naïve animals induces valvulitis in the recipients. This finding emphasizes that T cells, on their own, can drive inflammation and valvular damage without the necessity of cross-reactive antibodies to initiate the response. 48 The mechanisms and components involved in this process are summarized in Figure 1 . Figure 1 Immunological Mechanism of Rheumatic Heart Disease Group A streptococcal (GAS) throat infection primes macrophages and dendritic cells to present streptococcal antigens, activating cross-reactive T and B cells that circulate to the heart. Autoantibodies and Th1/Th17 cells bind VCAM-1–upregulated valvular endothelium, breach the layer, and secrete IFN-γ, TNF-α, IL-2, and IL-17. The resulting EndMT, leukocyte infiltration, and ECM remodeling fuel chronic inflammation, fibrosis, and commissural fusion of the mitral leaflets. BCR = B-cell Receptor; EndMT = endothelial-to-mesenchymal transition; GAS = group A Streptococcus ; IFN-γ = interferon-gamma; IL = interleukin; MHC II = major histocompatibility complex class II; TCR = T-cell receptor; Th1 = T-helper 1 cell; Th17 = T-helper 17 cell; TNF = tumor necrosis factor; VCAM = vascular cell adhesion molecule. Immunological Mechanism of Rheumatic Heart Disease Group A streptococcal (GAS) throat infection primes macrophages and dendritic cells to present streptococcal antigens, activating cross-reactive T and B cells that circulate to the heart. Autoantibodies and Th1/Th17 cells bind VCAM-1–upregulated valvular endothelium, breach the layer, and secrete IFN-γ, TNF-α, IL-2, and IL-17. The resulting EndMT, leukocyte infiltration, and ECM remodeling fuel chronic inflammation, fibrosis, and commissural fusion of the mitral leaflets. BCR = B-cell Receptor; EndMT = endothelial-to-mesenchymal transition; GAS = group A Streptococcus ; IFN-γ = interferon-gamma; IL = interleukin; MHC II = major histocompatibility complex class II; TCR = T-cell receptor; Th1 = T-helper 1 cell; Th17 = T-helper 17 cell; TNF = tumor necrosis factor; VCAM = vascular cell adhesion molecule. In RHD, CD4 + T cells play a central role in driving inflammation and tissue injury 49 , 50 through the release of key cytokines such as tumor necrosis factor (TNF)-α, interferon-γ, interleukin (IL)-1, IL-2, and IL-6. T-helper 1 cells dominate the acute phase, promoting inflammatory damage and B-cell activation, while reduced IL-2 levels impair regulatory T-cell function, exacerbating immune dysregulation. 51 , 52 , 53 , 54 , 55 , 56 , 57 In the chronic phase, T-helper 17 cells and their cytokine IL-17 become prominent, inducing EndMT and macrophage polarization, which contribute to progressive valvular fibrosis. 57 , 58 , 59 CD8 + T cells also increase in later stages, potentially driven by IL-10, which enhances their activation and expansion. 52 , 60 , 61 Meanwhile, proinflammatory M1 macrophages amplify tissue injury via NLRP3 inflammasome activation and the release of IL-1β, IL-18, and TGF-β, promoting fibroblast proliferation and matrix remodeling. 62 , 63 , 64 Beyond the classical molecular mimicry hypothesis, recent studies propose a “neo-antigen” theory linking RHD pathogenesis to collagen-directed autoimmunity. During GAS infection, the bacterial peptide associated with rheumatic fever binds to the C3B region of type IV collagen in the basement membrane, rendering it immunogenic. This triggers the production of anticollagen autoantibodies, which may cross-react with type I collagen in heart valves, leading to endothelial injury, inflammation, and progressive valvular fibrosis. 47 , 65 , 66 Figure 2 summarizes the mechanisms of collagen-directed autoimmunity. Figure 2 Collagen Autoimmunity in Rheumatic Heart Disease The peptide associated with rheumatic fever (PARF) motif on streptococcal M-protein binds the CB3 domain of collagen type IV, creating a neo-antigen that elicits antibodies. These antibodies cross-react with exposed collagen (types IV and I) on valve endothelium, recruit inflammatory cells, and trigger the systemic inflammation that culminates in rheumatic valvular injury. CB3 = collagen-binding region 3 of type IV collagen; ECM = extracellular matrix; PARF = peptide associated with rheumatic fever. Collagen Autoimmunity in Rheumatic Heart Disease The peptide associated with rheumatic fever (PARF) motif on streptococcal M-protein binds the CB3 domain of collagen type IV, creating a neo-antigen that elicits antibodies. These antibodies cross-react with exposed collagen (types IV and I) on valve endothelium, recruit inflammatory cells, and trigger the systemic inflammation that culminates in rheumatic valvular injury. CB3 = collagen-binding region 3 of type IV collagen; ECM = extracellular matrix; PARF = peptide associated with rheumatic fever. These mechanisms resemble those in Goodpasture and Alport syndromes, suggesting a shared subendothelial collagen-targeted immune pathway. 67 , 68 Elevated anticollagen antibodies in patients with ARF as well as in mice immunized with streptococcal protein support this hypothesis. 69 , 70 , 71 Notably, epidemiological and molecular evidence indicates that in some endemic regions, such as Northern Australia, rheumatic fever may involve groups C and G streptococci in addition to GAS. 72 , 73 This evidence further reinforces the concept that the pathogenesis of RHD may not be exclusively attributable to GAS pharyngeal infections, but rather to a broader pathogenic mechanism involving collagen-binding bacterial adhesins shared across multiple streptococcal groups. Taken together, the immune-initiation mechanisms outlined previously likely explain early valve injury in ARF but differ in their ability to account for chronic RHD progression and left-sided valve predominance. With respect to chronic progression in the absence of recurrent ARF, molecular mimicry alone appears insufficient. By contrast, collagen-directed autoimmunity, valve neo-antigen formation, and persistence of valve-resident memory T cells provide a plausible framework for self-sustaining inflammation independent of renewed bacterial exposure. With respect to left-sided valve predominance, these immune mechanisms are systemic and do not by themselves explain anatomical selectivity. This pattern may be influenced by hemodynamic amplification, whereby greater mechanical stress and endothelial disruption in the mitral and aortic valves could facilitate local immune injury. Thus, RHD may be best understood as the result of systemic immune initiation coupled with valve-specific susceptibility.

Dr Morel has received grants in support of investigator and investigator-initiated studies from AstraZeneca, Medtronic, and Boehringer Ingelheim, all outside of the submitted work. Dr Morel has been awarded grants by “Fondation Cœur et Recherche” and “Endofrance,” two reputable charities in France committed to advancing research initiatives in cardiovascular disease in endometriosis. Dr Mai has received a France Excellence Scholarship from the French Embassy. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.