Predictors of first biopsy cancer detection with suspected local stage prostate cancer
Department of Pathology, University of Mississippi School of Medicine and Section of Urology, Veterans Affairs Medical Center, Jackson, USA. The Journal of Urology
(Impact Factor: 4.47).
04/2000; 163(3):813-8. DOI: 10.1016/S0022-5347(05)67810-X
We determine demographic and tumor related predictors of repeat biopsy cancer detection in men with suspected stage T1c-2 prostate cancer.
The study population included 298 consecutive men with suspected stage T1c-2 prostate cancer who had a benign prostate biopsy at 1 institution between January 1, 1992 and April 1, 1999 and underwent 1 repeat biopsy. Mean age plus or minus standard deviation was 66.8+/-6.7 years for 133 black (55%) and 165 white (45%) patients. Clinical measures included determination of high grade prostatic intraepithelial neoplasia in benign biopsy specimens, Gleason score of malignant biopsy specimens, prostate specific antigen (PSA), PSA density, annualized interbiopsy PSA change, percent free PSA (201 cases) and PSA velocity (171).
Cancer was detected on repeat biopsy in 80 cases (27%). Significant differences between patients with benign and malignant repeat biopsies included age (p = 0.001), PSA density (p = 0.0001), percent free PSA (p = 0.0001) and PSA velocity (p = 0.009). High grade prostatic intraepithelial neoplasia in an initial benign biopsy was not predictive of cancer in repeat biopsy (p = 0.12). Multiple logistic regression analysis of all cases showed that age (p = 0.002) and PSA density (p = 0.0002) were independent predictors of cancer. Subset multiple logistic regression analysis modeled with age, PSA density and percent free PSA demonstrated that age (p = 0.002) and percent free PSA (p = 0.0001) were significant independent predictors of malignancy. Subset multiple logistic regression analysis modeled with age, PSA density, percent free PSA and PSA velocity revealed that age (p = 0.02) and percent free PSA (p = 0.0003) were significant independent predictors of cancer. There were no significant differences between the Gleason scores of cancers detected on repeat biopsy compared to 587 stage T1c-2 cancers detected on initial biopsy during the study period (p = 0.09). PSA, PSA density, percent free PSA and PSA velocity were not significantly different among men without a cancer diagnosis who had high grade neoplasia in 1 or 2 benign biopsies.
Greater than 25% of this population of select patients with suspected stage T1c-2 prostate cancer had malignancy detected on repeat biopsy. Percent free PSA was the most powerful predictor of cancer. High grade prostatic intraepithelial neoplasia was not a predictor of repeat biopsy cancer detection and PSA functions were similar among men without cancer who did and did not have high grade neoplasia in 1 or more benign biopsies. This finding suggests that high grade prostatic intraepithelial neoplasia may not be a reliable indicator of clinically significant existing prostate cancer.
Available from: Nora Pashayan
- "Twenty percent of cancers are detected at the second biopsy session , and a number of parameters have been promoted to reduce the frequency of benign repeat biopsy. Abnormal digital rectal examination (DRE) , PSA , PSA velocity > 0.75 ng/ml/yr , PSA density > 0.15 [5,7] transition zone density > 0.25 [5,8], and free-to-total PSA ratio [5,7], as well as initial histological findings of high grade prostatic intraepithelial neoplasia (HGPIN) [6,8] or suspicious histology  have all been used to identify those men at higher risk of positive repeat biopsy. These individual parameters when taken alone, have poor positive and negative predictive value . "
[Show abstract] [Hide abstract]
ABSTRACT: Little evidence is available to determine which patients should undergo repeat biopsy after initial benign extended core biopsy (ECB). Attempts have been made to reduce the frequency of negative repeat biopsies using PSA kinetics, density, free-to-total ratios and Kattan's nomogram, to identify men more likely to harbour cancer but no single tool accurately predicts biopsy outcome. The objective of this study was to develop a predictive nomogram to identify men more likely to have a cancer diagnosed on repeat prostate biopsy.
Patients with previous benign ECB undergoing repeat biopsy were identified from a database. Association between age, volume, stage, previous histology, PSA kinetics and positive repeat biopsy was analysed. Variables were entered stepwise into logistic regression models. A risk score giving the probability of positive repeat biopsy was estimated. The performance of this score was assessed using receiver characteristic (ROC) analysis.
110 repeat biopsies were performed in this period. Cancer was detected in 31% of repeat biopsies at Hospital (1) and 30% at Hospital (2). The most accurate predictive model combined age, PSA, PSA velocity, free-to-total PSA ratio, prostate volume and digital rectal examination (DRE) findings. The risk model performed well in an independent sample, area under the curve (AUCROC) was 0.818 (95% CI 0.707 to 0.929) for the risk model and 0.696 (95% CI 0.472 to 0.921) for the validation model. It was calculated that using a threshold risk score of > 0.2 to identify high risk individuals would reduce repeat biopsies by 39% while identifying 90% of the men with prostate cancer.
An accurate multi-variable predictive tool to determine the risk of positive repeat prostate biopsy is presented. This can be used by urologists in an outpatient setting to aid decision-making for men with prior benign histology for whom a repeat biopsy is being considered.
BMC Urology 07/2009; 9(1):7. DOI:10.1186/1471-2490-9-7 · 1.41 Impact Factor
Available from: ncbi.nlm.nih.gov
[Show abstract] [Hide abstract]
ABSTRACT: The challenge continues-to find better methods of screening for prostate cancer, of determining who should undergo needle biopsy, and of predicting who will fail initial therapy. Investigators are looking at the value of neural networks and an array of markers to provide improved screening and prognostic information.
Reviews in urology 02/2000; 2 Suppl 4((Suppl 4)):S5-9.
Available from: Steven Bigler
[Show abstract] [Hide abstract]
ABSTRACT: High grade prostatic intraepithelial neoplasia (HGPIN), a premalignant lesion of the prostate gland, is more common in black men than in white men. The influence of HGPIN on the serum prostate specific antigen (PSA) concentration is controversial, and correlations between HGPIN and PSA in black men and white men have not been investigated.
Between January 1992 and December 1998, 411 black men and 639 white men with suspected prostate carcinoma underwent an initial benign prostate biopsy at a single medical center. The presence or absence of HGPIN in the biopsy specimens was determined by one uropathologist.
HGPIN was identified in 8.9% of the specimens. When stratified by PSA concentration (< 4.0 ng/mL, 4.0-9.9 ng/mL, and > or = 10.0 ng/mL), HGPIN was associated with an increased PSA concentration only among men with PSA concentrations < 4.0 ng/mL (P = 0.01). The prevalence of HGPIN in the black and white patients was 13.4% and 5.9%, respectively (P < 0.0001), and was significantly greater in black men than in white men with PSA concentrations < 4.0 ng/mL (P = 0.002). Among the patients with PSA concentrations < 4.0 ng/mL, black race was an independent predictor of an increased PSA concentration when adjusted for patient age, prostate volume, and the presence or absence of HGPIN (P = 0.03).
HGPIN is more common in black men than in white men and may produce an increase in the PSA concentration. However, racial differences in the prevalence of HGPIN may not contribute to racial differences in PSA concentrations among men with no clinical or histologic evidence of carcinoma.
Cancer 04/2001; 91(7):1291-6. DOI:10.1002/1097-0142(20010401)91:73.0.CO;2-J · 4.89 Impact Factor
Data provided are for informational purposes only. Although carefully collected, accuracy cannot be guaranteed. The impact factor represents a rough estimation of the journal's impact factor and does not reflect the actual current impact factor. Publisher conditions are provided by RoMEO. Differing provisions from the publisher's actual policy or licence agreement may be applicable.