Serum prostate-specific antigen, clinical stage, pathologic grade, and the incidence of nodal metastases in prostate cancer.

OBJECTIVES

To evaluate the potential gains in using serum prostate-specific antigen (PSA) levels for predicting the incidence of pelvic nodal metastases in patients with otherwise localized prostate cancer.

METHODS

We reviewed 569 patients undergoing staging lymphadenectomy, and evaluated the correlation between clinical stage, tumor grade, presurgical PSA level, and the incidence of nodal metastatic disease using univariate and multivariate regression.

RESULTS

In univariate analysis stage, grade, and PSA level were highly significant covariates with nodal metastasis. Nodal metastatic rates increased as stage increased: Stage T1 (A2), 6 of 127 (5%); Stage T2 (B), 41 of 243 (17%); Stage T3 (C), 95 of 199 (48%), (p < 0.0001). Likewise metastatic rates increased as grade increased: Gleason grades 2 to 4, 6 of 124 (5%); Gleason grades 5 and 6, 52 of 238 (22%); Gleason grade 7, 41 of 122 (34%); Gleason grades 8 to 10, 43 of 84 (51%) (p < 0.0001). The relationship between PSA level and nodal metastases was not linear and we selected the following groupings that correlated with nodal disease: 4 or less ng/mL, 4 of 104 (4%); more than 4 to 20 or less ng/mL, 73 of 335 (22%); more than 20 to 40 or less ng/mL, 35 of 85 (41%); more than 40 ng/mL, 30 of 45 (67%) (p < 0.0001). Using multivariate logistic regression, stage, grade, and PSA were independently predictive of nodal status.

CONCLUSIONS

The gains in predictive accuracy from PSA beyond that obtained from stage and grade were small and in practice would benefit fewer than 15% of our patients. Staging pelvic lymphadenectomy remains the only satisfactory method for elucidating nodal status in the majority of patients with prostate cancer.