Bone sialoprotein, bone morphogenetic protein 6 and thymidine phosphorylase expression in localized human prostatic adenocarcinoma as predictors of clinical outcome: A clinicopathological and immunohistochemical study of 43 cases
G. De Pinieux et al., Bone sialoprotein, bone morphogenetic protein 6 and thymidine phosphorylase expression in localized human prostatic adenocarcinoma as predictors of clinical outcome: A clinicopathological and immunohistochemical study of 43 cases, J UROL, 166(5), 2001, pp. 1924-1930
Purpose: Skeletal metastases are the hallmark of advanced prostate cancer a
nd recurrence after local surgery is common. Currently to our knowledge no
biological markers predict the risk of disease progression in individuals w
ith localized prostate cancer. In a search for predictive markers we evalua
ted the expression of bone sialoprotein and bone morphogenetic protein 6, 2
bone related proteins, and the angiogenic factor thymidine phosphorylase.
Materials and Methods: The study population included 43 men who presented w
ith localized prostate cancer treated with radical prostatectomy. Bone sial
oprotein, bone morphogenetic protein 6 and thymidine phosphorylase expressi
on was assessed by immunohistochemical testing. Results were analyzed in re
lation to pathological disease stage, Gleason score and clinical outcome. C
linical followup was 4.3 to 11.4 years after surgery (median 7.9).
Results: Disease did not progress in 17 of the 43 cases, while recurrence a
nd/or metastasis developed in the other 26 at a median of 6.5 and 6.9 years
, respectively. Bone sialoprotein and bone morphogenetic protein 6 expressi
on detected in 28 (65%) and 29 (67%) of the 43 samples, respectively, was s
ignificantly associated (p = 0.0001). Thymidine phosphorylase detected in 2
6 samples (60%) was not related to bone sialoprotein and/or bone morphogene
tic protein 6 positivity. Bone sialoprotein and/or bone morphogenetic prote
in 6 expression correlated with bone metastasis, while thymidine phosphoryl
ase expression was related to local recurrence (p = 0.002 and/or 0.007, and
0.00007, respectively). On multivariate analysis only the correlation of t
hymidine phosphorylase expression with recurrence remained statistically si
gnificant (p = 0.002). Co-expression of the 3 markers was observed in the s
amples of 10 of the 11 patients (90%) with bone metastases and only in 5 of
the 17 (29%) who were disease-free.
Conclusions: This study indicates that the expression of bone sialoprotein,
bone morphogenetic protein 6 and thymidine phosphorylase determined at a c
linically early stage of disease by a simple immunohistochemical technique
would enable subgroups of patients to be identified that are at different r
isks of bone metastasis or recurrence. Detection of such markers would prov
ide additional prognostic information that would be useful for patients wit
h intermediate or low Gleason score or stage disease. These patients would
benefit from a more adapted clinical follow-up.