Biologic versus mechanical prosthetic valve replacement in pulmonary valve position, results from a large retrospective study
K Mehdizadeh, E Bitaraf, Y Tolouei Tabar, M Farrashi, M Zarbayani, N Sarisarraf, K Rabeti Moghadam, E Kohansal, Y Mohtasham Kia, S Momeni, Z Khajali, S Saedi, M Shojaeifard, M Gholampour Dehaki, P SadeghipourAbstract
Background
The choice between biologic versus mechanical prosthetic valve for pulmonary position is still controversial.
Purpose
To compare biologic versus mechanical prosthetic valve in patients with Tetralogy of Fallot (TOF)
Methods
We conducted a retrospective cohort study including all consecutive patients with TOF who underwent primary pulmonary valve replacement (PVR) at our tertiary center between May 2002 and October 2017. Main exclusion criteria were patients with history of handmade valved conduits in pulmonary position and those lacking complete documentation. To minimize errors and missing data inherent in long-term retrospective studies, we collected and cross-validated data from multiple sources including hospital admission and discharge notes, imaging, surgical and non-surgical reports, and outpatient clinic notes. We had to process an extensive number of echocardiography reports, which were primarily in free-text format and unsuitable for large-scale analysis. To tackle this, we fine-tuned the OpenAI GPT-4o-mini large language model specifically for structured data extraction from free-text echocardiography reports of congenital heart disorder patients. Main study outcome was any new surgical or percutaneous valve-related reintervention.
Results
In the present study, 932 patients (mean±SD age was 22.5 ± 10 year, and 382 (41%) were female) were included. First PVR was biologic versus mechanical PVR in 311 (33%) and 622 (67%), respectively. Median duration (IQR) of follow up was xx (IQR). Rate of valve-related reintervention at 15 years was 25.7% (95% CI 18.0%–33.4%) in biological PVR and 20.1% (95% CI 13.7%–26.5%) in mechanical PVR. In our Fine–Gray competing risks regression, valve type was not a significant predictor of valve-related reintervention (subdistribution HR 0.78 (95% CI 0.48–1.25). (Figure 1).
Conclusion
In patients with TOF underwent PVR, valve type was not a significant predictor of valve-related reintervention. Rate of valvular dysfunction, all-cause mortality, obstructive prosthetic valve thrombosis, endocarditis and health related quality of life based on EQ-5L questionnaire between two prostheses are pending.Figure 1