Detection and Prediction of Bioprosthetic Aortic Valve Degeneration

Background Bioprosthetic aortic valve degeneration is increasingly common, often unheralded, and can have catastrophic consequences. Objectives The authors sought to assess whether 18F-fluoride positron emission tomography (PET)-computed tomography (CT) can detect bioprosthetic aortic valve degeneration and predict valve dysfunction. Methods Explanted degenerate bioprosthetic valves were examined ex vivo. Patients with bioprosthetic aortic valves were recruited into 2 cohorts with and without prosthetic valve dysfunction and underwent in vivo contrast-enhanced CT angiography, 18F-fluoride PET, and serial echocardiography during 2 years of follow-up. Results All ex vivo, degenerate bioprosthetic valves displayed 18F-fluoride PET uptake that colocalized with tissue degeneration on histology. In 71 patients without known bioprosthesis dysfunction, 14 had abnormal leaflet pathology on CT, and 24 demonstrated 18F-fluoride PET uptake (target-to-background ratio 1.55 [interquartile range (IQR): 1.44 to 1.88]). Patients with increased 18F-fluoride uptake exhibited more rapid deterioration in valve function compared with those without (annualized change in peak transvalvular velocity 0.30 [IQR: 0.13 to 0.61] vs. 0.01 [IQR: −0.05 to 0.16] ms−1/year; p < 0.001). Indeed 18F-fluoride uptake correlated with deterioration in all the conventional echocardiographic measures of valve function assessed (e.g., change in peak velocity, r = 0.72; p < 0.001). Each of the 10 patients who developed new overt bioprosthesis dysfunction during follow-up had evidence of 18F-fluoride uptake at baseline (target-to-background ratio 1.89 [IQR: 1.46 to 2.59]). On multivariable analysis, 18F-fluoride uptake was the only independent predictor of future bioprosthetic dysfunction. Conclusions 18F-fluoride PET-CT identifies subclinical bioprosthetic valve degeneration, providing powerful prediction of subsequent valvular dysfunction and highlighting patients at risk of valve failure. This technique holds major promise in the diagnosis of valvular degeneration and the surveillance of patients with bioprosthetic valves. (18F-Fluoride Assessment of Aortic Bioprosthesis Durability and Outcome [18F-FAABULOUS]; NCT02304276)

T he implantation of bioprosthetic heart valves is increasing rapidly due to patient preference, the increasing prevalence of valve disease in an aging population, and the emergence of transcatheter aortic valve implantation (1)(2)(3)(4). In the United States, 90,000 surgical aortic valve replacements are performed annually, with over three-quarters incorporating bioprosthetic valves (3). In addition, >80,000 transcatheter aortic valve implantation procedures have been performed since U.S. Food and Drug Administration approval in 2011 (4). Given the finite lifespan of these valves and the rapid expansion of the population of patients requiring regular surveillance, bioprosthetic valve degeneration will become a major cause of cardiovascular morbidity and health care burden over the coming decades.
The pathophysiology of bioprosthetic valve degeneration is poorly understood. Although calcification appears to contribute to both progressive valve narrowing and leaflet tears (5-7), noninvasive methods for detecting this process have been lacking, and the triggers of valve degeneration and calcification are unknown. Current standard of care relies on serial echocardiography and clinical assessment aimed at detecting the valve dysfunction that occurs only toward the end stages of the degeneration process. Unfortunately, many patients present in extremis with unheralded valve failure due to rapidonset valvular obstruction or regurgitation, with repeat operation a high-risk undertaking. Indeed, emergency repeat aortic valve replacement surgery is associated with a mortality of 22.6% compared with 1.4% for elective repeat surgery (8). Detection of bioprosthetic valve degeneration is therefore highly desirable, allowing at-risk patients to be identified early, offered close tailored monitoring, and optimized timing of repeat elective intervention, thereby avoiding potentially catastrophic valve failure. 18 F-fluoride positron emission tomography (PET) has recently been used to image tissue calcification activity in a range of cardiovascular diseases (9)(10)(11)(12). 18 F-fluoride preferentially binds to areas of developing microcalcification indicative of tissue degeneration (13) that precede the macrocalcification detectable by computed tomography (CT) (14,15).
Given that calcification is one of the key pathological processes underlying bioprosthetic valve degeneration, we hypothesized that increased 18   Cartlidge et al.      Values are n, n (%), mean AE SD, or median (interquartile range). *Incomplete data.
On univariable analysis, the only predictors of deterioration in bioprosthetic valve function (annualized change in bioprosthetic valve peak velocity) were current smoking habit (p ¼ 0.047) and 18 F-fluoride PET uptake, irrespective of whether the latter was considered as a categorical or continuous variable (both p < 0.001) (Online Table 4). On multivariable analysis incorporating age, sex, duration of valve implantation, baseline peak prosthetic valve velocity, and CT findings, 18 Table 6).

DISCUSSION
In this multimodality prospective imaging study, we have identified 18 F-fluoride PET-CT as the first noninvasive technique capable of detecting early  Our study has several major strengths. This is a comprehensive multimodal imaging study using ultrasound, CT, and PET in patients with   abdominal aortic aneurysms, and aortic stenosis (9)(10)(11)(12). We now extend these observations to bioprosthetic valves and confirm that 18