Thrombosis, fibrosis and microcalcification in bioprosthetic aortic valves

Background Structural valve degeneration limits the durability of bioprosthetic aortic valves, exposing patients to repeat interventions and their associated risks. The pathogenesis of structural aortic valve degeneration remains incompletely understood but appears to involve thrombosis, fibrosis and calcification. We aimed to determine the presence of activated platelets, fibroblast activity and calcification in patients with bioprosthetic aortic valves and their relationship with structural valve dysfunction. Methods In a prospective single-centre observational cohort study, patients with prior surgical or transcatheter bioprosthetic aortic valve implantation underwent hybrid positron emission tomography and computed tomography angiography with [18F]GP1 (activated platelet marker), [18F]-sodium fluoride ([18F]NaF; active calcification marker) and [68Ga]-fibroblast activation protein inhibitor-46 ([68Ga]FAPI-46; actived fibroblast marker) radiotracers at three distinct timepoints. Radiotracer uptake was quantified using the maximum target-to-background ratios (TBRmax). Features of valve dysfunction were assessed using computed tomography and transthoracic echocardiography1,2. Stage 1 dysfunction was described as gross morphological abnormalities seen on computed tomography or echocardiography, without haemodynamic changes. Stage 2 was described as moderate obstruction, moderate regurgitation, or a combination of the two on haemodynamic assessment. Stage 3 was described as severe obstruction or regurgitation. Results Thirty-one participants were recruited between October 2019 and March 2021. Twenty-six (84%) had undergone surgical and 5 (16%) transcatheter valve replacement. All patients had valvular [18F]GP1 uptake (median TBRmax 2.76 [interquartile range 2.40 to 3.47]) at median of 45 [13 to 80] months following valve replacement. After a median of 75 [48 to 111] months, 24/31 (77%) patients had [18F]NaF uptake (median TBRmax 1.71 [1.55 to 1.83]), and by a median of 88 [59 to 121] months, all 28 patients who completed follow-up had diffuse [68Ga]FAPI-46 valve uptake (median TBRmax 1.37 [1.25 to 1.43]) (Figure 1). There were no significant correlations between radiotracer activity. At visit 3, 17/28 (61%) patients had evidence of valve dysfunction; 2 patients (7%) with stage 2 dysfunction and 3 (11%) with stage 3 dysfunction. Only [18F]NaF activity demonstrated an association with the stages of valve dysfunction at visit 3 (p = 0.02; Table 1). Conclusions Platelet activation, fibroblast activity and calcification commonly occur in bioprosthetic aortic valves. Valve leaflet fibroblast activation was observed in all patients 88 months following valve implantation. No significant in-vivo temporospatial relationship was seen between these processes on positron emission tomography. Only active calcification appears to associate with future bioprosthetic valve dysfunction.