Objectives. To objectively evaluate the contents of corpus cavernous tissue
in impotent men using an automated novel image analysis system.
Methods. Thirty-three impotent men and 2 normal potent men (controls) under
went corpus cavernous biopsies. The procedures were performed using a Biopt
y gun under local anesthesia. The obtained specimens were stained with Mass
on's trichome technique, and the collagen fiber contents were evaluated by
a computerized morphometric analysis method. In addition, we estimated the
intraobserver and interobserver reliability of this automated image analysi
s system.
Results. No major complication was noted during or after the biopsies. Of t
he 33 impotent patients, 3 were diagnosed as having psychogenic, 11 as havi
ng arteriogenic, 13 as having venogenic, 1 as having neurogenic, and 5 as h
aving idiopathic impotence. The collagen fiber percentages in the 35 men we
re as follows: normal potent: 48.2% +/- 1.4%, psychogenic 55.2% +/- 11.6%,
arteriogenic 73.2% +/- 4,4%, venogenic 66.5% +/- 4.2%, neurogenic 76.9%, an
d idiopathic 77.4% +/- 4.2%. Significant differences were found between the
normal potent and arteriogenic groups (P <0.05) and between the normal pot
ent and idiopathic groups (P <0.05). Patients older than 60 years had a hig
her collagen fiber content (70.4% +/- 3.7%) than those younger than 50 year
s old (58.6% +/- 5.2%). The interobserver and intraobserver variances were
both negligible for this automated image analysis system. This method signi
ficantly reduced the amount of variation introduced by the intra-rater reli
ability of a technician compared with the manual method.
Conclusions. The present automated image analysis system is believed to be
a reliable, accurate quantitative measurement tool for studies of penile ti
ssue. Cavernous biopsy is a rapid, safe, and representative modality to stu
dy penile disease. An increase in cavernous collagen fibers (or corporal fi
brosis) is considered an important factor in impaired erectile function. (C
) 2000, Elsevier Science Inc.