Aortic valve replacement using an allograft aortic valve has been perf
ormed on 804 patients. From December 1969 to May 1975, 124 patients re
ceived a nonviable allograft valve sterilized by incubation with low-d
ose antibiotics and stored for weeks by refrigeration at 4 degrees C (
series 1). From June 1975 to January 1994, 680 patients received viabl
e allograft valves, now cryopreserved early within 2 hours of collecti
on from transplant recipient donors, 6 hours for multiorgan donor valv
es and 23 hours (mean) for autopsy valves from donor death. The 30-day
mortality was 8.9% +/- 5% (95% confidence limits) for series I and 2.
8% +/- 1% (95% confidence limits) for series II. Actuarial patient sur
vival including hospital mortality at 15 years was 56% +/- 5% for seri
es I and 62% +/- 5% for series II. The probability of a thromboembolic
event was low, freedom at 15 years being 95% +/- 1% for patients rece
iving allografts with or without associated coronary bypass procedures
and 81% +/- 5% for patients having allografts with other associated p
rocedures (eg, mitral valve operations). Actuarial freedom from endoca
rditis was similar for the two series, 91% +/- 3% (series I) and 94% /- 2% (series II) at 15 years. The freedom from valve incompetence, fr
om reoperation for all causes, and from structural deterioration demon
strated clearly the inferiority of the 4 degrees C stored allograft va
lves. For structural deterioration as identified clinically, at reoper
ation and at death, freedom from this event at 15 years was 45% +/- 6%
for series I and 80% +/- 5% for series II (p value for the difference
is 0). The attrition rate appears highest in young patients and in th
ose not receiving a viable cryopreserved valve. An important immunolog
ic response can be unfavorable in some young patients, producing valve
deterioration. But for the majority of patients, the viable cryoprese
rved allograft valve offers low morbidity with a good extended lifesty
le and is superior to the 4 degrees C stored valve.