Objective: To investigate the ultrastructural and physiologic effects
of exposure of the human corneal endothelium to mitomycin at concentra
tions of 20 mu g/mL and 200 mu g/mL using electron microscopy and in v
itro specular perfusion techniques. Methods: Four pairs of corneas (wi
th one cornea of each pair receiving balanced salt solution [BSS Plus,
Alcon Laboratories, Fort Worth, Tex] and the other receiving BSS Plus
with 20 mu g/mL of mitomycin) suitable for transplantation, except fo
r extremes of age or systemic disease, underwent perfusion with cornea
l thickness measured serially every 15 minutes followed by fixation fo
r electron microscopy. Mean corneal swelling rate was calculated for a
ll four experiments, and the control group that received BSS Plus was
compared with the group that received mitomycin using a paired t test.
Electron micrographs were examined in a masked fashion. Similar studi
es were performed using two pairs of corneas that received 200 mu g/mL
of mitomycin. Results: The mean swelling rate for corneas perfused wi
th 20 mu g/mL of mitomycin (-4.1 mu m/h) was not significantly differe
nt from that seen in tissue perfused with BSS Plus (-4.2 mu m/h). No c
onsistent ultrastructural changes could be attributed to exposure to 2
0 mu g/mL of mitomycin. Perfusions of mitomycin at 200 mu g/ml resulte
d in prompt corneal swelling with marked ultrastructural alterations c
ompared with tisssue perfused with BSS Plus. Conclusion: Human corneal
endothelium may be exposed to undiluted (200 to 500 mu g/mL) mitomyci
n with inadvertent entry into the anterior chamber during dissection o
f the scleral nap bed in trabeculectomy followed by application of mit
omycin. This will result in prompt destruction of the endothelium. Exp
osure to 20 mu g/mL of mitomycin, a level exceeding the concentration
that may be present in the aqueous humor after its proper application,
appears nontoxic in this system.