The Evaluation of Correlation between the Histologic Composition of BPH and TRUS Findings

Kim, Jung Gyun; Cho, Kyung Shik; Ahn, Hanjong

Abstract

PURPOSE: It has been believed that o-blocker affects the stromal component of BPH, while 5 alpha -reductase inhibitor affects the glandular component. Information on the tissue composition of BPH might be helpful to choose an appropriate medical therapeutic agent. We evaluated whether transrectal ultrasonographic findings could reflect the histologic composition of BPH and the correlation of the composition of BPH and the clinical parameters such as patient's age, peak flow rate, IPSS symptom score, and prostate volume.
MATERIALS AND METHODS
Fifty five patients with BPH treated by transurethral resection were studied. The proportions of stromal and glandular area in the resected prostate were determined by image analyzer. "Stromal hyperplasia"was defined when more than 75% of the resected prostate was composed of stromal tissue and microscopically, abundant stromal tissue with normal or atrophic glands were dominant. "Glandular hyperplasia" was defined when more than 25% of the prostate was composed of glandular tissue and microscopically, hyperplastic glandular cells and dilated lumen with occasional cystic changes were dominant. Ultrasonographically, it is classified as stromal hyperplasia when the echo of central gland is fine and lower echogenic than that of peripheral gland, and classlfied as glandular hyperplasia when the echo of central gland Is coarse and isoechogenic or hyperechogenic with focal echopenic areas. Statistical significance was judged by Student t-test and linear regression analysis. Concordance of ultrasonographic findings and histopathologic findings of BPH was determined by Kappa index.
RESULTS
Sixty nine percent of resected prostate tissue were composed of stromal tissue. The proportion of stromal tissue in stromal hyperplasia and glandular hyperplasia was 85.8+/- 1.6% and 64.8+/- 1.0%, respectively(p<0.001). Ultrasonography accurately reflect histopathologlc type in 50 of 55 BPH. Two of 12 stromal hyperplasia and 3 of 43 glandular hyperplasia were falsely interpreted on ultrasonography. Three of 5 falsely interpreted cases showed marginal stromal tissue composition(60.9%, 72.9%, 73.3%, 76.4%, and 84.2%). Kappa index of ultrasonogrhaphic finding and histopathologic classification was 0.74(fair to good). In the meantime, clinical parameters including patient's age, peak flow rate, IPSS symptom scores did not correlate with tissue composition. Only prostate volume and resected prostate weight showed negative correlation with proportion of stromal tissue(p=0.0953 and p=0.0794, respectively).
CONCLUSIONS
Using our sonographic criteria, transrectal ultrasonography could reflect histologic type of BPH so that choice of medical therapeutic agent may be possible. Larger prostates had less stromal tissue, however, the histologic composition of the prostate was not related to the severity of symptoms or peak flow rate.