Variation in patterns of practice in diagnosing screen-detected prostate cancer

New variants at 10q26 and 15q21 are associated with aggressive prostate cancer in a genome-wide association study from a prostate biopsy screening cohort

PURPOSE

To identify and examine polymorphisms of genes associated with aggressive and clinical significant forms of prostate cancer among a screening cohort.

EXPERIMENTAL DESIGN

We conducted a genome-wide association study among patients with aggressive forms of prostate cancer and biopsy-proven normal controls ascertained from a prostate cancer screening program. We then examined significant associations of specific polymorphisms among a prostate cancer screened cohort to examine their predictive ability in detecting prostate cancer.

RESULTS

We found significant associations between aggressive prostate cancer and five single nucleotide polymorphisms (SNPs) in the 10q26 (rs10788165, rs10749408, and rs10788165, p value for association 1.3 × 10(-10 ) to 3.2 × 10(-11) ) and 15q21 (rs4775302 and rs1994198, p values for association 3.1 × 10(-8 ) to 8.2 × 10(-9)) regions. Results of a replication study done in 3439 patients undergoing a prostate biopsy, revealed certain combinations of these SNPs to be significantly associated not only with prostate cancer but with aggressive forms of prostate cancer using an established classification criterion for prostate cancer progression (odds ratios for intermediate to high-risk disease 1.8-3.0, p value 0.003-0.001). These SNP combinations were also important clinical predictors for prostate cancer detection based on nomogram analysis that assesses prostate cancer risk.

CONCLUSIONS

Five SNPs were found to be associated with aggressive forms of prostate cancer. We demonstrated potential clinical applications of these associations.

Avoiding pitfalls in applying prediction models, as illustrated by the example of prostate cancer diagnosis

BACKGROUND

The use of different mathematical models to support medical decisions is accompanied by increasing uncertainties when they are applied in practice. Using prostate cancer (PCa) risk models as an example, we recommend requirements for model development and draw attention to possible pitfalls so as to avoid the uncritical use of these models.

CONTENT

We conducted MEDLINE searches for applications of multivariate models supporting the prediction of PCa risk. We critically reviewed the methodological aspects of model development and the biological and analytical variability of the parameters used for model development. In addition, we reviewed the role of prostate biopsy as the gold standard for confirming diagnoses. In addition, we analyzed different methods of model evaluation with respect to their application to different populations. When using models in clinical practice, one must validate the results with a population from the application field. Typical model characteristics (such as discrimination performance and calibration) and methods for assessing the risk of a decision should be used when evaluating a model's output. The choice of a model should be based on these results and on the practicality of its use.

SUMMARY

To avoid possible errors in applying prediction models (the risk of PCa, for example) requires examining the possible pitfalls of the underlying mathematical models in the context of the individual case. The main tools for this purpose are discrimination, calibration, and decision curve analysis.

Assessing individual risk for prostate cancer

PURPOSE

To construct a clinical nomogram instrument to estimate individual risk for having prostate cancer (PC) for patients undergoing prostate specific antigen (PSA) screening, using all risk factors known for PC.

PATIENTS AND METHODS

We conducted a cross-sectional study of 3,108 men who underwent a prostate biopsy, including a subset of 408 volunteers with normal PSA levels. Factors including age, family history of PC (FHPC), ethnicity, urinary symptoms, PSA, free:total PSA ratio, and digital rectal examination (DRE) were incorporated in the model. A nomogram was constructed to assess risk for any and high-grade PC (Gleason score >or= 7).

RESULTS

Of the 3,108 men, 1,304 (42.0%) were found to have PC. Among the 408 men with a normal PSA (< 4.0 ng/mL), 99 (24.3%) had PC. All risk factors were important predictors for PC by multivariate analysis (P, .01 to .0001). The area under the curve (AUC) for the nomogram in predicting cancer, which included age, ethnicity, FHPC, urinary symptoms, free:total PSA ratio, PSA, and DRE, was 0.74 (95% CI, 0.71 to 0.81) and 0.77 (95% CI, 0.74 to 0.81) for high-grade cancer. This was significantly greater than the AUC that considered using the conventional screening method of PSA and DRE only (0.62; 95% CI, 0.58 to 0.66 for any cancer; 0.69; 95% CI, 0.65 to 0.73 for high-grade cancer). From receiver operating characteristic analysis, risk factors including age, ethnicity, FHPC, symptoms, and free:total PSA ratio contributed significantly more predictive information than PSA and DRE.

CONCLUSION

In a PC screening program, it is important to consider age, family history of PC, ethnicity, urinary voiding symptoms, and free:total PSA ratio, in addition to PSA and DRE.

Quantifying the Impact of Prostate Volumes, Number of Biopsy Cores and 5α-Reductase Inhibitor Therapy on the Probability of Prostate Cancer Detection Using Mathematical Modeling

PURPOSE

Previous studies demonstrated a negative correlation between prostate volume and biopsy yield. By decreasing prostate volume 5alpha-reductase inhibitors may enhance cancer detection, which may explain the greater detection of high grade tumors in the finasteride arm of the Prostate Cancer Prevention Trial.

MATERIALS AND METHODS

A mathematical model was constructed to analyze the effects of prostate and tumor volumes, and biopsy core number on cancer detection. The effects of the volume reduction observed with finasteride in the Prostate Cancer Prevention Trial were also modeled, as was the potential reduction in tumor volume needed to explain the observed difference in prostate cancer detection. The model was also applied to the Reduction by Dutasteride of Prostate Cancer Events study.

RESULTS

A higher number of biopsies are required to ensure a detection probability of 0.90 or greater in larger glands or with smaller tumors. In the Prostate Cancer Prevention Trial for a tumor volume of 1 cc a 17% increase in the detection rate in the finasteride arm would be predicted if there was no change in tumor volume, likewise the rate would be 11% to 17% for the dutasteride arm of the Reduction by Dutasteride of Prostate Cancer Events study. The calculated reduction in tumor volume needed to explain the difference in cancer detection between the finasteride and placebo arms of the Prostate Cancer Prevention Trial would be 51% to 66%.

CONCLUSIONS

This model provides guidance on the optimal number of biopsy cores that accord with an earlier model. These findings also suggest that, if there were no reduction in tumor volume, 5alpha-reductase inhibitor therapy could lead to excess cancer detection, including high grade tumors.

Repeat prostate biopsy in the Prostate, Lung, Colorectal and Ovarian cancer screening trial

OBJECTIVE

To determine patterns of repeat prostate biopsy in a cohort of men undergoing prostate cancer screening who have a negative initial biopsy.

SUBJECTS AND METHODS

The Prostate, Colorectal, Lung, and Ovarian (PLCO) cancer screening trial is an ongoing study the prostate component of which consists of six annual screens with measurements of prostate-specific antigen (PSA) level and a digital rectal examination (DRE). The diagnostic follow-up of positive screening results is done by the subject's healthcare provider outside the purview of the PLCO. We analysed the experience of repeat biopsy in men in the PLCO with an initial negative biopsy. Men were divided by indication for initial biopsy into those with suspicious PSA levels and those with suspicious DRE findings.

RESULTS

The probability of having a repeat biopsy within 3 years of initial biopsy was 43% for 1736 men with suspicious PSA levels and 13% for 1025 men with suspicious DRE findings. Rates of third and fourth biopsy after a previous negative biopsy were similar to the initial repeat biopsy rate in PSA-positive men. Most men had a repeat biopsy only after having an additional round of screening. The PSA level and PSA velocity determined after initial biopsy were independent risk factors for a repeat biopsy, both in PSA-positive and DRE-positive men. High-grade prostatic intraepithelial neoplasia was a risk factor for repeat biopsy before any repeat PSA or DRE testing.

CONCLUSION

The experience of this cohort should be generally representative of patterns of care for repeat biopsy in men undergoing periodic screening. These data can provide context to the debate over optimum practices for repeat biopsy.

Optimal Sampling Sites for Repeat Prostate Biopsy: A Recursive Partitioning Analysis of Three-Dimensional 26-Core Systematic Biopsy

OBJECTIVES

To explore an optimal combination of sampling sites to detect prostate cancer in a repeat biopsy setting.

METHODS

A transrectal ultrasound-guided systematic three-dimensional 26-core biopsy (3D26PBx), a combination of transrectal 12 and transperineal 14 core biopsies, was performed in 235 Japanese men with prior negative biopsy. Using recursive partitioning, we evaluated cancer detection of all possible combinations of sampling sites and selected the combination that provides the highest cancer detection rate at a given number of biopsy cores.

RESULTS

Prostate cancer was detected in 87 of the 235 (37%) men. The 3D26PBx improved cancer detection by 89% relative to the conventional transrectal sextant biopsy. Neither Gleason score nor percentage of Gleason 4/5 cancers differed between cancers with and without positive cores within the transrectal sextant-sampling sites. A three-dimensional combination of transrectal and transperineal approaches outperformed either transrectal or transperineal approach alone. Recursive partitioning revealed that a three-dimensional 16-core (transrectal eight cores plus transperineal eight cores) biopsy could detect all the cancers with the minimum number of cores.

CONCLUSIONS

We propose a three-dimensional combination of transrectal eight cores taken from the far lateral peripheral zone and the parasagittal base, and transperineal eight cores taken from the anterior and posterior apex and the transition zone as an optimal set of sampling sites for repeat biopsy.

Dealing with non-cancerous findings on prostate biopsy

Most prostate biopsies do not show malignancy. The proper management of non-cancerous pathologic findings of the prostate is controversial. For this article, we reviewed the current literature for indications for repeat prostate biopsy after initial biopsies demonstrated non-cancerous prostatic tissue or benign prostatic hyperplasia. This review includes discussions of management of asymptomatic prostatitis and how it may affect prostate-specific antigen, and also the management of several potentially premalignant lesions such as atrophy, atypical small acinar proliferation, and high-grade prostatic intraepithelial neoplasia. There is a paucity of randomized trials in this area and, considering the high number of biopsies with non-malignant findings, we conclude that more investigation is warranted in this area.

Making Sense of Prostate Specific Antigen: Improving its Predictive Value in Patients Undergoing Prostate Biopsy

PURPOSE

The clinical usefulness of PSA for prostate cancer screening is unclear, although the test remains in common use. New methods to interpret PSA are needed.

MATERIALS AND METHODS

We examined a cohort of 2,637 men who underwent prostate biopsies for abnormal DRE or PSA between 1999 and 2004. Using risk factors for prostate cancer, including patient age, ethnicity, family history of prostate cancer, previous negative biopsy, voiding symptoms and prostate volume, we developed risk groups for prostate cancer using recursive partitioning modeling independent of PSA or DRE. We then compared prostate cancer probabilities by PSA ranges by risk group.

RESULTS

Of the 2,637 men 1,282 (48.6%) had prostate cancer. Age, ethnicity, family history, previous negative biopsy and prostate volume were predictive for cancer. We constructed 6 risk groups by combining these factors and created tables to assign patients to these groups. Independent of PSA and DRE the probability of cancer ranged from 15% in patients in group 1 to 78% in patients in group 6 (p <0.0001). By adding PSA and DRE to each risk group prostate cancer probabilities were refined from 0% to 100%. Patients in the higher risk groups also had higher grade cancer (p <0.0001).

CONCLUSIONS

We generated 6 risk groups based on simple risk factors for prostate cancer. When used in the right context and patient, PSA is highly accurate for predicting prostate cancer and permitting rational decision making in patients with abnormal PSA.

The Use of Genetic Markers to Determine Risk for Prostate Cancer at Prostate Biopsy

PURPOSE

We examined a panel of 13 polymorphisms in 13 different genes to determine whether specific genotypes can help predict prostate cancer at the time of biopsy among men prescreened with prostate-specific antigen and digital rectal exam.

EXPERIMENTAL DESIGN

We examined 2,088 consecutive men who were referred for prostate biopsy from 1997 to 2003. Thirteen genes were examined, including TNF308, GSTT1, KLK2, endostatin, MCRA, MCRV, tyrosinase, MSR1, CHK2, RNasel, HOGG1-326, HOGG1-11657, and HRAS1. Odds ratio for detection of prostate cancer were adjusted for age, race, prostate-specific antigen, digital rectal exam, family history of prostate cancer, and urinary symptoms.

RESULTS

Of the 2,088 men, 996 (47.7%) had cancer detected. Four genes (TNF308, GSTT1, KLK2, and HOGG1-326) were significantly associated with prostate cancer. The adjusted odds ratios (OR) for prostate cancer for patients with the AA genotype of the TNF308 gene was 1.92 [95% confidence interval (95% CI), 1.0-1.5, P = 0.05], compared with those with the GG genotype, and for patients with the TT genotype of the KLK2 gene, the OR was 1.5 (95% confidence interval, 1.0-2.2, P = 0.04), compared with the CC genotype. The OR for patients with a homozygous deletion of the GSTT1 gene was 0.81 (95% CI, 0.6-1.0, P = 0.06) compared with those with the deletion, and the OR for patients with the GG genotype of the HOGG1-326 gene was 0.68 (95% CI, 0.5-1.0, P = 0.05) compared with the CC genotype. Patients who had all four alleles that were positively associated with prostate cancer had an OR of 9.33 (95% CI, 2.4-35.8, P = 0.0005) for prostate cancer compared with patients with alleles that were negatively associated with prostate cancer.

CONCLUSIONS

Of the 13 polymorphisms, two were found to be positively associated with prostate cancer (TNF308 and KLK2) and two were negatively associated with prostate cancer (GSTT1 and HOGG1-326). Future studies are required to confirm these results.