Prognostic factors in patients with metastatic (stage D2) prostate cancer: experience from the Scandinavian Prostatic Cancer Group Study-2

Jørgensen T, Kanagasingam Y, Kaalhus O, Tveter KJ, Bryne M, Skjørten F, Berner A, Danielsen HE

J Urol. 1997



Nuclear texture reflects the overall structures of the chromatin organization. We recently reported the principles and prognostic importance of image analysis of nuclei from metastatic prostate cancer. Immunohistochemical up regulation of the adhesion molecule sialyl Lewis(x) is also reported to be a prognostic parameter. Presently we analyzed statistically the prognostic impact of these 2 new parameters compared to well-known clinical parameters in metastatic prostate cancer.


Prognostic factors, such as sedimentation rate, alkaline and acid phosphatases, hemoglobin, testosterone, performance status, pain due to metastasis, T category, histological grade and patient age, were included in a multivariate Cox proportional hazards regression analysis based on 262 patients from the Scandinavian Prostatic Cancer Group Study-2. Extent of bone lesions, deoxyribonucleic acid ploidy, texture analysis and sialyl Lewis(x) molecules based on subsets of these 262 patients were also analyzed in the same multivariate model.


This test identified chromatin texture as the most important factor (p < 0.001), followed by reaction of the oligosaccharide sialyl Lewis(x) (p < 0.01). Among the routine clinical and laboratory data, sedimentation rate, alkaline phosphatase and hemoglobin (p < 0.05) showed prognostic importance. Performance status, pain due to metastasis and extent of bone lesions showed prognostic value in the univariate analysis (p < 0.05).


These data indicate that computerized nuclear texture analysis as well as up regulation of sialyl Lewis(x) molecules may be new important prognostic factors in metastatic prostate cancer. Furthermore the prognostic importance of sedimentation rate, alkaline phosphatase and hemoglobin was confirmed.