Biological versus mechanical aortic prosthesis? A nineteen-year comparison in a propensity-matched population.

BACKGROUND AND AIM OF THE STUDY

The choice of aortic valve substitutes remains controversial. Malfunction and systemic valve complications affect the results of mechanical and tissue valves. Two devices--the Sorin Monocast (tilting disk) valve and the Hancock Standard valve were compared, the study aim being to determine whether the valve model is a marker or a causal influence of poor outcome after aortic valve replacement (AVR).

METHODS

Between January 1970 and January 1984, patients aged < 70 years and operated on for aortic valve disease were selected. A total of 379 patients received either Sorin (group S) valves (n = 213; median age 51 years) or Hancock Standard (group HcK) valves (n = 192; median age 50 years) (p = NS). Total follow up was 2,471 patient-years (pt-yr) for group S and 2,368 pt-yr for group HcK. Follow up was 98% complete; median duration was 15 pt-yr for group S and 13.2 pt-yr for group HcK. Propensity matching for available patient intrinsic and operative risk factors was ultimately used to investigate whether biological or mechanical valve models impact upon outcome after aortic valve surgery. Patient survival was analyzed according to the 'intention to treat' principle.

RESULTS

The 30-day mortality was 7.5% for group S and 10.9% for group HcK (p = NS). The 19-year Kaplan-Meier freedom from valve-related mortality was 84% (group S) and 82% (group HcK) (p = NS), while overall survival was 42% (group S) and 35% (group HcK) (p = NS). Structural valve deterioration (SVD) was the major cause of reoperation in the HcK group. The 19-year freedom from all valve-related complications was 43% (group S) versus 19% (group HcK) (p = 0.0001). By propensity score, 61% of the valve replacements (247/405) were perfectly matched for available risk factors, with an equal distribution of risk covariates. When SVD and reoperation due to SVD were excluded, survival and freedom from all valve-related complications of the matched patients were identical between the prostheses under comparison.

CONCLUSION

In this relatively young population, the Sorin valve showed a significantly lower valve-related complication rate than the Hancock Standard valve. The latter valve showed a significantly increasing rate of reoperation due to SVD, and thereby a relative inadequacy for use in younger patients. When analyzed according to an 'intention to treat' principle, the 19-year survival and freedom from valve-related complications of patients with the same propensity score for selection of either valve type were similar.