Beyond Calcium: Imaging Reveals the Hidden Fibrotic Side of Aortic Valve Disease

Calcific aortic valve disease has long been treated, in imaging terms and in drug trials alike, as a story about calcium. A comprehensive new review published in Frontiers in Cardiovascular Medicine argues that this framing has quietly distorted the field. The condition, the authors contend, involves two distinct structural processes—remodeling of the fibrous matrix and deposition of mineral—that contribute unequally from one patient to the next. Two leaflets can restrict blood flow to exactly the same degree while containing completely different proportions of collagen and calcium, and a calcium score alone cannot tell them apart.

The scale of the problem is substantial. Among 5,201 participants aged 65 or older in the Cardiovascular Health Study, aortic sclerosis was present in 26 percent and aortic stenosis in 2 percent. The Tromsø population study found aortic stenosis in 3.9 percent of people aged 70 to 79 and 9.8 percent of those aged 80 to 89. Using Global Burden of Disease 2021 data, one analysis estimated that the number of people living with nonrheumatic calcific aortic valve disease grew from roughly 4.98 million in 1992 to 13.32 million in 2021, an increase of about 168 percent. Notably, the age-standardized prevalence also rose, from 130.82 to 158.35 per 100,000, which means population aging alone does not explain the trend.

Surgical valve replacement and transcatheter aortic valve replacement have transformed the treatment of severe stenosis, yet no pharmacologic therapy has been shown to modify the underlying disease process—a point the 2025 ESC/EACTS guidelines state explicitly. Between the first leaflet abnormalities and functional obstruction, the disease may advance silently for years. The review’s central argument is that this therapeutic impasse may partly reflect a measurement problem: drug trials have almost always tracked calcification or hemodynamics, while the fibrous side of the disease has gone essentially unmeasured over time.

Contrast-enhanced computed tomography angiography, or CTA, has begun to change that. Unlike noncontrast CT, which quantifies accumulated mineral through the Agatston score or calcium volume, CTA can estimate calcified tissue volume, noncalcified tissue volume, and their sum, which the authors call total fibrocalcific volume. The noncalcified compartment includes collagen-rich matrix, other extracellular constituents, and cells—but it is not collagen-specific, a caveat the review stresses repeatedly. Validation against histology has been encouraging at the group level. In a study of 164 patients with aortic stenosis, total fibrocalcific volume correlated with explanted valve weight with a coefficient of 0.72, and it tracked peak velocity more strongly than the calcium score did, particularly in women.

A more granular validation came from a prospective two-center study of 29 participants, analyzing nine histologic sections per valve. The CTA-derived noncalcified fraction correlated closely with the histologic fibrotic fraction, with a correlation of 0.92 and an intraclass correlation of 0.94. But the mean bias was −4.9 percentage points, and the 95 percent limits of agreement spanned −18.5 to 8.7 percentage points—individual-level error wide enough to limit classification of any single valve. The review is candid about this tension: strong group-level agreement coexists with per-patient uncertainty, and the agreement estimates rest on mostly advanced valves, with two-dimensional section sampling and decalcification adding further constraints.

Longitudinal data add a different dimension. In an analysis drawing on the SALTIRE2 trial, total fibrocalcific volume rose by approximately 17 percent over one year, and its change correlated with change in mean transvalvular gradient. Repeated scans showed a mean difference of −1 percent with limits of agreement from −4.5 to 2.8 percent, and baseline total volume predicted subsequent gradient progression, especially in women. The review concludes that total fibrocalcific volume is far better supported as a candidate structural endpoint than as a validated marker of response to antifibrotic therapy—because whether the noncalcified component adds predictive value beyond calcification requires both components modeled jointly, a demonstrated improvement in model performance, and external validation, none of which yet exists.

Sex differences illustrate why composition matters. Among 651 patients evaluated for transcatheter valve replacement, women had lower indexed total fibrocalcific volume than men (230 versus 293 mm3/cm2) and lower indexed calcified volume (85 versus 149 mm3/cm2), but a higher noncalcified-to-calcified volume ratio (1.47 versus 0.83). Crucially, indexed noncalcified tissue volume did not differ significantly between the sexes. The higher ratio in women therefore reflected less calcified tissue, not more fibrous tissue—a distinction the review says is routinely lost when studies report ratios alone. Histologic work points the same direction: at comparable hemodynamic severity, women show lower calcific burden and higher proportions of collagen and dense connective tissue on average, though the distributions overlap substantially and sex cannot predict an individual patient’s tissue phenotype.

The drug-trial record is sobering. SALTIRE randomized 155 patients with calcific stenosis to intensive atorvastatin and found no slowing of calcification or peak-velocity progression. ASTRONOMER randomized 272 patients with mild or moderate stenosis to rosuvastatin with no improvement in hemodynamic progression. SEAS, with 1,873 patients, found that simvastatin plus ezetimibe reduced ischemic events but not valve-related events. Mineral-targeted approaches fared no better: denosumab and alendronic acid in SALTIRE2 produced the expected bone-metabolism responses but no valvular benefit, and in AVADEC, vitamin K2 plus vitamin D left a between-group calcium-score difference of just −17 Agatston units after two years. In DIP-CAVD, evogliptin failed its primary CT endpoint at 96 weeks, though active calcification volume on PET favored treatment in a subgroup—an exploratory signal the review treats cautiously, since the PET and CT analyses covered different time points and participants.

Preclinical work tells a more nuanced story. In mice treated with 2-thioUTP after eight months of disease induction, micro-CT in the same animals showed mineral burden falling from the pretreatment baseline alongside improved valve function—genuine longitudinal regression, though in an animal model. The Reversa mouse model showed that mineralization responses can separate from collagen and valve-opening responses: genetic lipid lowering pushed mineral burden below a contemporaneous baseline while collagen did not fall. Cell studies reveal why responses diverge—matrix stiffness steers valvular interstitial cells toward mineralization or myofibroblast contraction, and the same TGF-β1 stimulus produced a fibrotic response in intact three-dimensional leaflet cultures but calcification in isolated cell cultures. The review’s conclusion is that established mineral lesions and cell states can sometimes be modified, but these responses do not necessarily move together, and human evidence for reversing fibrosis remains limited to cross-sectional associations, such as the link between angiotensin receptor blocker exposure and lower fibrosis scores in excised valves of women.

The practical upshot is a research agenda. The authors call for trials that measure leaflet composition longitudinally in the same patients, report absolute component volumes rather than ratios alone, and handle the statistical trap of patients who progress fastest and undergo valve replacement earliest, thereby disappearing from imaging follow-up. Two ongoing protocols—CHIANTI, testing low-dose colchicine in moderate stenosis with a 24-month calcium-score primary endpoint, and PASSPORT, testing vitamin K1 in mild or moderate disease—already pair CT, PET, and echocardiography, though neither has reported results. Until treatment-induced change in tissue composition is validated against functional outcomes, the review concludes, the fibrocalcific phenotype will remain a hypothesis rather than a clinical tool. For a disease affecting millions and lacking any disease-modifying drug, that measurement gap may be the most consequential finding of all.

Subject of Research: Fibrocalcific tissue phenotypes in calcific aortic valve disease and their assessment by imaging, clinical cohorts, and intervention studies

Article Title: Fibrocalcific phenotypes in calcific aortic valve disease: imaging characterization, treatment timing, and efficacy assessment

Article References: Liu, F., Zeng, Z., Qian, Z., Li, W., Huang, Y., & Liu, Z. (2026). Fibrocalcific phenotypes in calcific aortic valve disease: imaging characterization, treatment timing, and efficacy assessment. Frontiers in Cardiovascular Medicine, 13, 1997440. https://doi.org/10.3389/fcvm.2026.1997440

Image Credits: AI Generated

DOI: 10.3389/fcvm.2026.1997440

Keywords: calcific aortic valve disease, aortic stenosis, valve calcification, fibrosis, CT angiography, calcium score, 18F-NaF PET, statin trials, drug development, imaging biomarkers, valvular interstitial cells, risk stratification