No abstract

The simplified Bernoulli relationship appears to be quite accurate for predicting gradients across discrete valvular obstructions. Controversy exists about how accurately it predicts the severity of disease in longer segment obstructions. In this study we constructed a pulsatile model of subvalvular pulmonary stensosis in vitro to study nine custom-made subvalvular tunnels 2, 4, and 7 mm in length with flow cross sections of 0.5 to 1.5 cm2 and with the stenotic segment proximal to a nonstenotic bioprosthetic valve, and a pulsatile model in vitro of a 16 mm long tunnel-like ventricular septal defect (VSD) of varying cross-sectional area (0.20 to 0.64 cm2). We also compared the observations in vitro with those in an open-chest dog preparation with a tunnel-like interventricular communication. In the subpulmonic stenosis model, for each individual tunnel, 10 instantaneous peak gradients between 15 to 105 mm Hg were available. The pressure gradients across the tunnel alone, measured in the subvalvular area, were consistently higher than the measured gradients across the tunnel plus valve, suggesting some relaminarization of flow (i.e., a decrease in velocity) and pressure recovery (i.e., an increase in pressure) distal to the obstruction. Continuous-wave Doppler velocities across the 4 and 7 mm tunnels for the highest gradients were slightly lower than for the 2 mm tunnel at the same gradients, and it was only for the 0.5 cm2 cross section, 4 and 7 mm tunnels that there was a suggestion of minor viscous energy loss. For all the subvalvular tunnels studied, the Bernoulli relationship accurately predicted the results of the pressure drop across the tunnel only, while the gradient across tunnel plus valve was consistently lower. For the VSD tunnel model in vitro, the Doppler-derived gradients were approximately 40% higher than the measured gradients. The findings for the subvalvular and VSD tunnels in vitro and similar findings in the open-chest dogs with VSD suggest that relaminarization of flow and recovery of pressure occurred distal to the tunnel orifice, whereas continous-wave Doppler findings correlate with the highest instantaneous gradients measured in the lowest pressure areas at the vena contracta of the tunnel.

After in vitro testing (confirmed in vivo) of three contemporary valve designs (St. Jude, Björk-Shiley and Carpentier-Edwards) demonstrated that the St. Jude valve possessed the most favorable hydrodynamic performance characteristics, a limited clinical trial was begun in high risk patients who might benefit from a prosthesis with improved hemodynamics. Between March 1978 and March 1984, 419 St. Jude prostheses (157 aortic, 156 mitral and 53 double aortic-mitral) were implanted in 366 patients. Ninety-six percent were in New York Heart Association functional class III or IV preoperatively. Early (30 day) mortality was 10.4% overall, and was lower after aortic (5.7%) or double (7.5%) than after isolated mitral valve replacement (16.0%). Forty-four prosthetic mitral valve recipients with severe ischemic mitral regurgitation experienced a 32% early mortality rate; without this group, mitral valve replacement carried a 10% early mortality rate (p less than 0.01). Multivariate logistic regression analysis confirmed that early death was strongly associated with three preoperative patient characteristics (p less than 0.05): ischemic mitral valve disease, depressed left ventricular function (ejection fraction less than 0.55) and advanced functional class (class IV). Late follow-up (7,055 patient-months, mean 22) was 99.7% complete (1 patient lost). Actuarial survival at 4 years was 80, 80 and 79% after aortic, mitral (nonischemic) and double valve replacement, respectively; in the subset with ischemic mitral regurgitation, actuarial survival was 34% (p less than 0.01). Eighty-six percent of survivors were in functional class I or II.(ABSTRACT TRUNCATED AT 250 WORDS)