Fertility potential in men with a history of congenital undescended testes: a long-term follow-up study.

Men with a history of congenital undescended testes (UDT) have an increased risk of fertility problems. Despite no definitive proof, current guidelines recommend early surgical intervention because this may have a positive effect on future fertility potential by preventing degenerative changes of the testes in early life. Also surgical intervention facilitates observability of the testes in view of possible malignancy. We evaluated testicular function in adult men with previous UDT treated at different ages before puberty. A long-term follow-up study of men with previous UDT was performed. Andrological evaluation included medical history taking, physical examination, scrotal ultrasound, determination of reproductive hormones, and semen analysis. Findings were compared with those of a control group of men with normal testicular descent. The influence of age at orchiopexy on future fertility parameters was evaluated in a multivariate regression analysis. 62 men were included of whom seven had had bilateral UDT. Twenty-four patients had had their orchiopexy before the age of 24 months of whom eight men had it before 12 months of age. Forty-eight men had had unsuccessful luteinizing-hormone-releasing-hormone (LHRH) nasal spray treatment during childhood, whereas 14 of 24 men operated before 24 months of age had not received LHRH treatment before orchiopexy. Fertility potential in men with a history of UDT is compromised in comparison with controls. We could not detect any influence of age at orchiopexy on fertility parameters. However, the number of patients operated before the age of 12 months is limited. This study does not support the assumption that early orchiopexy results in better fertility potential.

Comparison of 30-day, 90-day and in-hospital postoperative mortality for eight different cancer types.

BACKGROUND: Various definitions are used to calculate postoperative mortality. As variation hampers comparability between reports, a study was performed to evaluate the impact of using different definitions for several types of cancer surgery. METHODS: Population-based data for the period 1997-2008 were retrieved from the Rotterdam Cancer Registry for resectional surgery of oesophageal, gastric, colonic, rectal, breast, lung, renal and bladder cancer. Postoperative deaths were tabulated as 30-day, in-hospital or 90-day mortality. Postdischarge deaths were defined as those occurring after discharge from hospital but within 30 days. RESULTS: This study included 40,474 patients. Thirty-day mortality rates were highest after gastric (8·8 per cent) and colonic (6·0 per cent) surgery, and lowest after breast (0·2 per cent) and renal (2·0 per cent) procedures. For most tumour types, the difference between 30-day and in-hospital rates was less than 1 per cent. For bladder and oesophageal cancer, however, the in-hospital mortality rate was considerably higher at 5·1 per cent (+1·3 per cent) and 7·3 per cent (+2·8 per cent) respectively. For gastric, colonic and lung cancer, 1·0 per cent of patients died after discharge. For gastric, lung and bladder cancer, more than 3 per cent of patients died between discharge and 90 days. CONCLUSION: The 30-day definition is recommended as an international standard because it includes the great majority of surgery-related deaths and is not subject to discharge procedures. The 90-day definition, however, captures mortality from multiple causes; although this may be of less interest to surgeons, the data may be valuable when providing information to patients before surgery.

Trends in incidence, treatment and survival of patients with stage IV colorectal cancer: a population-based series.

AIM: The incidence, patterns of care and survival were determined in patients with stage IV colorectal cancer (CRC) in a population-based series. METHOD: Computer records for patients diagnosed with stage IV CRC diagnosed from 1 January 1995 to 31 December 2007 were retrieved from the Rotterdam Cancer Registry. Surgical resection of the primary tumour, chemotherapy use, hepatic surgery and survival were evaluated according to year of diagnosis, age, gender and primary tumour site. RESULTS: In the southwestern part of the Netherlands, 19 014 new patients with CRC were diagnosed and synchronous metastatic disease was found in 3482 (18%). This proportion increased during the study period, from 16% to 21%. Surgical resection of the primary tumour was performed in approximately 50% of the patients and did not change over time. Postoperative 30-day mortality was 8%. Chemotherapy use increased from 18% in the first period to 56% in the latest period. Liver surgery increased from 4% in the first period to 10% in the latest period. Median survival increased from 7 months to 12 months and 2-year survival increased from 14% to 28%. Two-year survival declined with increasing age and was significantly worse for right-sided tumours (14%). CONCLUSION: Survival of patients with stage IV CRC has improved over time and this is probably a result of the increased use of chemotherapy and the increased numbers of patients who underwent hepatic surgery.

The interobserver variability of digital rectal examination in a large randomized trial for the screening of prostate cancer.

BACKGROUND: To analyze to what extent the percentage of suspicious digital rectal examination (DRE) findings vary between examiners and to what extent the percentage of prostate cancers (PCs) detected in men with these suspicious findings varies between examiners. METHODS: In the first screening round of the European Randomized study of Screening for PC (ERSPC) Rotterdam, 7,280 men underwent a PSA-determination and DRE of whom 2,102 underwent prostate biopsy (biopsy indication PSA > or = 4.0 ng/ml and/or suspicious DRE and/or TRUS). Descriptive statistics of DRE-outcome per PSA-range were used to determine the observer variability of six examiners. Because this analysis did not correct properly for other predictors of a suspicious DRE (PSA-level, biopsy indication, TRUS-outcome, prostate volume and age), a logistic regression analysis controlling for these explanatory variables was performed as well. RESULTS: In 2,102 men biopsied, 443 PCs were detected (PPV = 21%). For all PSA levels the percentage suspicious DRE varied between examiners from 4% to 28% and percentage PC detected in men with a suspicious DRE varied from 18% to 36%. Logistic regression analysis showed that three of six examiners considered DRE significantly more often abnormal than others (ORs 3.48, 2.80, 2.47, P < 0.001). For all examiners the odds to have PC was statistically significantly higher in case of a suspicious DRE (ORs 2.21-5.96, P < 0.05). This increased chance to find PC was not significantly observer-dependent. CONCLUSIONS: Three of six examiners considered DRE significantly more often suspicious than the others. However, under equal circumstances a suspicious DRE executed by each examiner increased the chance of the presence of PC similarly.

Prostate-specific antigen velocity at low prostate-specific antigen levels as screening tool for prostate cancer: results of second screening round of ERSPC (ROTTERDAM).

OBJECTIVES: To study retrospectively whether the prostate-specific antigen (PSA) velocity, that is, the change in PSA level over time, might serve as a screening tool in this PSA range. It is estimated that 40% of detectable prostate cancers are present in men with a PSA level of 4.0 ng/mL or less. Digital rectal examination and/or transrectal ultrasonography have been used as screening tools at these low PSA levels, but this approach is not very efficient. METHODS: The possible predictors (including PSA velocity) for biopsy outcome were studied using univariate and multivariate logistic regression analysis in 774 men who underwent biopsy between November 1997 and January 2002 in the second screening round of the European Randomised Study of Screening for Prostate Cancer (ERSPC). The clinical stage of the tumors was determined, and the Gleason scores of the biopsies were studied. RESULTS: A total of 149 cancers were found (positive predictive value 19.2%). The odds ratio for the PSA velocity determined by univariate logistic regression analysis was 2.2 (95% confidence interval 0.7 to 6.9, P = 0.19) and was 0.73 (95% confidence interval 0.20 to 2.6, P = 0.64) by multivariate analysis. The distribution of the clinical stage of the detected tumors was 64.4% T1c, 32.2% T2, and 3.4% T3. The biopsy Gleason score was 6 in 84.5%, 7 in 14.2%, and 8 in 1.3%. CONCLUSIONS: The number of cancers detected in this study and the distribution of clinical stage and biopsy Gleason score confirmed that a relatively large proportion of potentially curable cancers can be found in the low PSA ranges. The PSA velocity did not appear to be a useful screening tool for the identification of these cancers.

Serendipity in detecting disease in low prostate-specific antigen ranges.

OBJECTIVE: To assess the magnitude of prostate cancer detection by serendipity (the coincidental detection of prostate cancer during the evaluation of an abnormal screening test result) when a digital rectal examination (DRE) and transrectal ultrasonography (TRUS) are used as initial screening tests for prostate cancer in men with low levels of prostate-specific antigen (PSA; 0.0-3.9 ng/mL). PATIENTS AND METHODS: In all, 117 participants of a population-based screening study were diagnosed with prostate cancer after a standard evaluation of an abnormal screening test result; 49 underwent radical prostatectomy. Serendipity was defined as either: (i) the presence of prostate cancer opposite to the side that raised suspicion for cancer on DRE and/or TRUS; (ii) a negative lesion-directed biopsy while cancer was present in one or more of the cores of the sextant biopsy; (iii) a tumour volume of < 0.5 mL on radical prostatectomy. RESULTS: Depending on the definition, 27-63% of prostate cancers detected at low PSA values were detected coincidentally and not as a result of a true-positive test result. The proportion of cancers detected by serendipity was inversely correlated with serum PSA level. CONCLUSION: A relatively high proportion of prostate cancers diagnosed in men with low PSA levels, and in which a biopsy was prompted by a suspicious DRE and/or TRUS, are considered to be detected by chance only. As these cancers are mostly small (< 0.5 mL), with potentially low biological aggressiveness, relying on serendipity seems disadvantageous in prostate-cancer screening. The level of serendipity in prostate cancer detection, the poor performance of the screening test, and high inter-observer variability, casts further doubt on the utility of DRE (and TRUS) as initial screening tests for prostate cancer in population-based screening.

Prostate-specific antigen-based early detection of prostate cancer--validation of screening without rectal examination.

OBJECTIVES: The evaluation of the screening procedures for prostate cancer (PCa) was a part of the protocol of the European Randomized Study of Screening for Prostate Cancer (ERSPC), section Rotterdam, The Netherlands. We sought to establish an improved strategy for the early detection of PCa using a prostate-specific antigen (PSA) cutoff of 3.0 ng/mL or greater as the only indication for prostate biopsy with omission of the digital rectal examination (DRE). METHODS: In June 1996, 8612 men, 55 to 74 years old, were randomized to screening and were screened within the ERSPC Rotterdam by a PSA level of 4.0 ng/mL or greater or positive DRE or transrectal ultrasound findings as the indication for biopsy. Four hundred thirty men had PCa. Those treated by radical prostatectomy provided the tumor characteristics considered essential for a change in the screening strategies. Various options were evaluated and predictions made by logistic regression analyses. The protocol change was implemented in February 1997. Another 7943 men were screened according to the new protocol (PSA 3.0 ng/mL or greater). The resulting data were used to compare the two protocols. RESULTS: The detection rate (proportion of PCa in those screened) turned out to be very similar, with rates of 5.0 and 4.7 at a PSA cutoff of 4.0 ng/mL or greater and 3.0 ng/mL or greater, respectively. This was due to a much larger number of cases of PCa per biopsy in the PSA range of 3 to 3.9 ng/mL than expected. The positive predictive value of the PSA range 3.0 to 3.9 ng/mL in the two protocols was 18.0% and 6.4%, respectively. Tumor characteristics were studied on radical prostatectomy specimens from the original protocol. PCa detected with the new screening regimen had a similar distribution of Gleason scores but a larger proportion of confined disease. Tumor volumes were smaller in patients with PSA levels of less than 2.9 ng/mL; the proportion of "minimal disease" in that group was 50% compared with 28% in the group with a PSA level between 3.0 and 3.9 ng/mL. CONCLUSIONS: Lowering the biopsy indication to a PSA cutoff of 3.0 ng/mL or greater without a DRE improved the positive predictive value from 18.2% to 24.3%. The number of biopsies necessary to detect 1 case of PCa accordingly changed from 5.2 to 3.4. The overall characteristics of the cases detected at that PSA cutoff differed very little from those detected with the regimen based on PSA, DRE, and transrectal ultrasound.

Prostate specific antigen testing and digital rectal examination before and during a randomized trial of screening for prostate cancer: European randomized study of screening for prostate cancer, Rotterdam.

PURPOSE: Worldwide 2 large-scale randomized screening trials for prostate cancer have been initiated. Determining prostate specific antigen (PSA) involves a simple test that may influence the outcome of these trials if frequently done in the control arm or before study enrollment. We quantified PSA and digital rectal examination before and during the screening trial in Rotterdam, The Netherlands and in the general population. MATERIALS AND METHODS: Trial participants were administered study intake questionnaires on tests done before study participation. Data on PSA from the regional general practice laboratory were correlated with participant data. Various sources were used to quantify PSA tests and digital rectal examinations in the general population. RESULTS: Of men 55 to 74 years old 45% underwent digital rectal examination at 1 time and 13% reported that PSA was tested before trial participation. Each rate increased with age. No statistically significant effect of former PSA testing or digital rectal examination on the cancer detection rate was identified. The rate of PSA determination after initial screening and/or randomization in the control arm was 2-fold that in the screening arm (76 versus 33/1,000 person-years). PSA determination initially decreased in the screening arm but increased rapidly after some time. The number of PSA determinations in the general population was estimated to be 45/1,000 person-years at ages 55 to 69 years. CONCLUSIONS: PSA testing was moderate in the control arm but if different men undergo this test each year, the contamination rate may become rather high. In the final analysis of mortality PSA testing should be considered.

Histopathological prostate cancer characteristics at radical prostatectomy after population based screening.

PURPOSE: Although early detection of prostate cancer by prostate specific antigen based screening results in a shift towards more clinically organ confined tumors, changes in prostate cancer characteristics after radical prostatectomy are less clear. MATERIALS AND METHODS: We studied 121 totally embedded radical prostatectomy specimens that were obtained from consecutive participants of the European Randomized Study of Screening for Prostate Cancer who were systematically screened and treated surgically. In each specimen pathological stage, Gleason score and proportion of high grade cancer (Gleason pattern 4 or 5) were determined. Lymph node status at operation, stage and grade were compared to a historical series of 72 surgical procedures performed for clinically localized prostate cancer at our hospital before the introduction of serum prostate specific antigen as a diagnostic tool. RESULTS: Although none of the screen detected cases had positive lymph nodes at surgery, operation was discontinued in 13 (18%) of the 72 historical cases because of positive lymph nodes. Compared with the remaining 59 historical prostatectomy specimens, the screen detected specimens showed a definite increase in the frequency of pathologically organ confined tumors and a relative decrease in Gleason score 8 to 10 tumors. However, 60% of screen detected tumors contained areas with high grade cancer (Gleason pattern 4 or 5) and 50% had a Gleason score of 7. The relative amount of high grade cancer in each tumor was related to volume (Kruskal-Wallis test p <0. 001). CONCLUSIONS: Screening for prostate cancer leads to an increase in surgical treatment for relatively small tumors that have a higher probability of being organ confined. Although the frequency of positive lymph nodes at operation decreases dramatically and the proportion of organ confined tumors after surgery increases, there is a shift from Gleason 8 to 10 tumors towards lower grade tumors at radical prostatectomy. Still, judged by the high frequency of focal dedifferentiation in screen detected tumors, most of them and surgically treated tumors are likely to be clinically important. The relatively large accumulation of these tumors in the Gleason 7 category is a concern because it could lead to a decrease in the clinical usefulness of the Gleason score system.

Identification of genetic markers for prostatic cancer progression.

Despite the high incidence of prostate cancer, only limited data are available on genes or chromosomes specifically involved in its initiation and progression. We have applied comparative genomic hybridization to routinely processed, paraffin-embedded, tissues at different times in prostatic tumor progression to screen the tumor genome for gains and losses. Our panel included specimens derived from 56 different patients: 23 patients with primary, prostate-confined carcinomas; 18 patients with regional lymph node metastases; and 15 patients with distant metastases. Chromosome arms that most frequently showed losses, included 13q (55%), 8p (48%), 6q (43%), 5q (32%), 16q (25%), 18q (20%), 2q (18%), 4q (18%), 10q (18%), and Y (16%). Gains were often seen of chromosome arms 8q (36%), 17q (23%), Xq (23%), 7q (21%), 3q (18%), 9q (18%), 1q (16%), Xp (16%). Furthermore, specific high-level amplifications, eg, of 1q21, 1q25, and Xq12 to q13, were found in metastatic cancers. A significant accumulation of genetic changes in distant metastases was observed, eg, loss of 10q (p = 0.03) and gain of 7q (p = 0.03) sequences. In addition, investigation of a potential biomarker identified in previous studies by our group, ie, extra copies of #7 and/or #8, revealed a high prevalence of 7pq and/or 8q gain in the distant metastases (p = 0.02). Importantly, gains were observed more frequently in tumors derived from progressors after radical prostatectomy, than in nonprogressors (mean time of follow-up, 74 months). Specifically, gain of chromosome 7pq and/or 8q sequences appeared an accurate discriminator between the progressors and nonprogressors. Multivariate analysis showed a significant correlation between progressive disease and the number of chromosomes with gains. This correlation also held true when stage (p = 0.007) or grade (p = 0.002) were taken into account. Likewise, this applied for gain of chromosome 7pq and/or 8q sequences (p = 0.03 and p = 0.005 for stage or grade, respectively). Additionally, an increase in the number of chromosomes with gains per case was related to a decrease in biochemical progression-free survival (Ptrend <0.001). More specifically, the gain of 7pq and/or 8q sequences markedly reduced the biochemical progression-free survival (p < 0.001). In conclusion, this study has, firstly, documented the spectrum of chromosomal alterations in subsequent stages of prostate cancer, a number of which had not been described previously. It allowed us to identify chromosomal regions related to advanced tumor stage, ie, loss of 10q24 and gain of 7q11.2 and/or 7q31 sequences. Secondly, gain of 7pq and/or 8q was identified as a potential genetic discriminator between progressors and nonprogressors after radical surgery.

Quantitative analysis of the decay of immunoreactivity in stored prostate needle biopsy sections.

Application of immunohistochemistry to assess the presence of prognostic tissue markers is used widely. The quantitation of these markers may be hampered by a time-related loss of antigenicity in formalin-fixed paraffin-embedded tissue stored on glass slides. Potential loss of immunohistochemical staining intensity was studied on prostatic needle biopsy sections stored for a maximum of 4 years with antibodies against p27kip1, CD-44s, MIB-1, and androgen receptor (AR). In benign tissue, the positive/total ratio for p27kip1 was determined, while CD-44s staining intensity was assessed semiquantitatively. For MIB-1 and AR, nuclear staining intensity was assessed using computed image analysis. An exponential and significant decay of immunoreactivity was seen for p27kip1, CD-44s, MIB-1, and AR, with half-lives of 587 days, 214 days, and 290 days for p27kip1, MIB-1, and AR, respectively. Immunohistochemical assessment of prognostic tissue markers on stored slides must be considered with care in research and clinical settings.

Prostate cancer detection at low prostate specific antigen.

PURPOSE: At low prostate specific antigen (PSA) the indication for prostate biopsy is usually an abnormal digital rectal examination. We evaluate the diagnostic value of PSA, digital rectal examination, transrectal ultrasonography and tumor characteristics at low PSA (0 to 4.0 ng./ml.). We confirm and add to recent evidence that digital rectal examination has a low predictive value and that many significant cancers at this PSA range may be missed. MATERIALS AND METHODS: From 1994 to 1997 a total of 10,523 participants 54 to 74 years old were randomized to screening in the Rotterdam section of the European Randomized Study of Screening for Prostate Cancer. Of the participants 9,211 (87.5%) had PSA less than 4.0 ng./ml., and underwent digital rectal examination and transrectal ultrasonography. Expected rates of prostate cancer detection were calculated using logistic regression analysis. Radical prostatectomy was performed in about half of the 478 men diagnosed with prostate cancer. Tumors were characterized by pT category, Gleason score and cancer volume in 166 processed radical prostatectomy specimens. In 50 of these cases PSA was 0 to 4.0 ng./ml. RESULTS: The positive predictive value of digital rectal examination and transrectal ultrasonography at PSA 0 to 4.0 ng./ml. was only 9.7%. Positive predictive value strongly depended on PSA. Sensitivity was calculated by using estimates of the prevalence of sextant biopsy detectable prostate cancers. Of 760 detectable cancers 478 (67%) were diagnosed irrespective of PSA in men screened with digital rectal examination, transrectal ultrasonography and PSA. Only 127 of 348 detectable prostate cancers (36.5%) were actually diagnosed in men with PSA 2 to 4 mg./ml. The importance of these missed cancers was evaluated with parameters of tumor aggressiveness within PSA ranges. CONCLUSIONS: Approximately half of the tumors missed with PSA 0 to 4 ng./ml. had aggressive characteristics (Gleason score 7 or greater, Gleason 4-5 components) and were organ confined. These tumors should be diagnosed and treated according to the present understanding of their natural history. More sensitive and selective screening strategies are needed. Presently a wrong "window of opportunity" is used for early detection of prostate cancer.

Prostate-specific antigen: A surrogate endpoint for screening new agents against prostate cancer?

BACKGROUND: An endpoint for clinical trials of prostate cancer which simplifies traditional endpoints (response of measurable lesions, progression rates, and death) is urgently needed. This is especially true for hormone-unresponsive disease, for which many new drugs are presently in a development phase. This paper presents a rationale for the use of prostate-specific antigen (PSA) in clinical trials of progressive prostate cancer under endocrine treatment. METHODS: The study is based on 84 patients who progressed after radical prostatectomy or node dissection, of whom 24 showed increasing PSA levels under subsequent endocrine treatment. An average linear relationship between (log-transformed) PSA and time and a subject-specific deviation from this average relationship were assessed. The predictive value of the subject-specific parameters of the linear fit with respect to time to prostate cancer-specific death was determined. The outcomes of the fitting procedure were used to calculate sample sizes for future studies (duration, 6 months) using PSA increase over time in hormone-unresponsive prostate cancer as a marker for treatment efficacy. RESULTS: The average PSA doubling time in this population was 4 months (corresponding time constant = 0.25). The assessed variance of the time constants equalled 0.04; the overall residual variance equalled 0.265. The subject-specific rate of change of the log-transformed PSA value in hormone-unresponsive prostate cancer was a highly significant predictor of prostate cancer-specific death. This suggests the potential usefulness of PSA as an endpoint in trials of hormone-unresponsive prostate cancer. Depending on conditions chosen (e.g, desired power and changes in log PSA slope), 18-70 participants per arm will be necessary in future phase III studies. A suggestion (algorithm) for the use of PSA in drug development is presented. CONCLUSIONS: Relatively small PSA-based trials in patients with hormone-unresponsive prostate cancer are possible if a similar patient population is utilized. As long as surrogacy is not established, such studies cannot be considered conclusive with respect to effectiveness of treatment, but are likely to be useful as a screening tool for new drugs. Experimental confirmation in human prostate cancer model systems of synergism between PSA decrease and tumor control by a given test treatment is likely to enhance the level of certainty of PSA-based drug evaluation.

The European Randomized Study of Screening for Prostate Cancer (ERSPC): an update. Members of the ERSPC, Section Rotterdam.

Screening for prostate cancer is controversial. While in some parts of the world screening is practised as a healthcare policy, it is strongly rejected in other areas, because solid evidence of effectiveness of screening combined with early treatment with respect to lowering the mortality of prostate cancer has not been shown. It is for this reason that a large European study is installed to establish or rule out the value of screening for this frequent disease. The present paper presents the goal of the study and elaborates on the value of presently available screening tests. Preliminary results with respect to the first round of screening in the Rotterdam area relating to 32,000 randomized men are presented. Evidence of effectiveness of screening through other studies and mechanisms is discussed.

Changing role of 3 screening modalities in the European randomized study of screening for prostate cancer (Rotterdam).

A randomized screening trial was started in Europe to show the effect of early detection and treatment of prostate cancer on mortality (European Study on Screening of Prostate Cancer). In one centre (Rotterdam), the screening protocol initially consisted of 3 screening tests for all men: prostate-specific antigen (PSA), digital rectal examination (DRE) and transrectal ultrasonography (TRUS). A PSA value of >/=4 ng/ml and/or an abnormality on DRE and/or TRUS were taken to indicate that biopsy was required. In this study, we examined the possibilities for a more efficient screening protocol. A logistic-regression model was used to predict the number of cancers for PSA < 4 ng/ml if all men were biopsied (predictive index, PI). Effects of a change in PSA cut-off on the screening results were explored. Weights were applied to procedures and cancers to explore the possibility of expressing differences between protocols in one overall figure. Biopsies in men with PSA < 1 ng/ml and a positive DRE or TRUS were very inefficient. Applying DRE and TRUS only in the PSA ranges 1.5 to 3.9 and 2 to 3.9 ng/ml to determine whether a biopsy was required would result in a decrease of 29 to 36% in biopsies and a decrease of 5 to 8% in cancers. However, the results of DRE and TRUS could not be duplicated entirely. A protocol with only PSA >/= 3 ng/ml as a direct biopsy indicator resulted in a decrease of detected cancers by 7.6% and of biopsies by 12%, also a much simpler screening procedure. Use of the PI would give more efficient protocols, but this should be viewed as a preliminary finding, with the disadvantage of necessitating many additional screening visits. Since the results of DRE and TRUS could not be duplicated, a change in protocol towards PSA >/= 3 ng/ml appears acceptable. If this proves effective, a final judgement about the optimal combination of screening tests may be made. Int. J. Cancer (Pred. Oncol.) 84:437-441, 1999.

Estimation of prostate cancer probability by logistic regression: free and total prostate-specific antigen, digital rectal examination, and heredity are significant variables.

BACKGROUND: Despite low specificity, serum prostate-specific antigen (PSA) is widely used in screening for prostate cancer. Specificity can be improved by measuring free and total PSA and by combining these results with clinical findings. Methods such as neural networks and logistic regression are alternatives to multistep algorithms for clinical use of the combined findings. METHODS: We compared multilayer perceptron (MLP) and logistic regression (LR) analysis for predicting prostate cancer in a screening population of 974 men, ages 55-66 years. The study sample comprised men with PSA values >3 microg/L. Explanatory variables considered were age, free and total PSA and their ratio, digital rectal examination (DRE), transrectal ultrasonography, and a family history of prostate cancer. RESULTS: When at least 90% sensitivity in the training sets was required, the mean sensitivity and specificity obtained were 87% and 41% with LR and 85% and 26% with MLP, respectively. The cancer specificity of an LR model comprising the proportion of free to total PSA, DRE, and heredity as explanatory variables was significantly better than that of total PSA and the proportion of free to total PSA (P <0.01, McNemar test). The proportion of free to total PSA, DRE, and heredity were used to prepare cancer probability curves. CONCLUSION: The probability calculated by logistic regression provides better diagnostic accuracy for prostate cancer than the presently used multistep algorithms for estimation of the need to perform biopsy.

Predictors for biopsy outcome in the European Randomized Study of Screening for Prostate Cancer (Rotterdam region).

BACKGROUND: In the European Randomized Study of Screening for Prostate Cancer (ERSPC, Rotterdam region), men aged 55-74 years are screened for prostate cancer by prostate-specific antigen (PSA) sampling, digital rectal examination (DRE), and transrectal ultrasound investigation (TRUS). All men with a PSA > or =4 ng/ml and/or a suspicious DRE and/or a suspicious TRUS are biopsied. METHODS: Logistic regression analysis was applied to derive a predictive index that equals the chance to find prostate cancer in a biopsy given the outcomes of the screening tests. This model was used to assess the number of cancers that could have been detected if all men had been biopsied (extrapolation). Furthermore, the model was used to study the possibilities for improvement of the current screening protocol. RESULTS: PSA was the dominant predictor for prostate cancer in a biopsy, followed by prostate volume, DRE, and TRUS result. It is assessed that 69% (95% CI, 52-86%) of cancers that could be identified if all men were biopsied are currently detected. Application of the same methods to screening data obtained in Göteborg (the Swedish ERSPC partner) yielded almost identical results. It was found that, in the Rotterdam protocol, a considerable number of men were biopsied according to the screening protocol with an assessed lower chance to have prostate cancer than men who were not biopsied according to the protocol. CONCLUSIONS: The chance to detect prostate cancer in a biopsy can be modeled quite accurately as a function of serum PSA, prostate volume, DRE, and TRUS results. Important improvements in the screening protocol can be achieved by the application of the predictive index.

Evaluation of prostate needle biopsies in a population-based screening study: the impact of borderline lesions.

BACKGROUND: The finding of isolated high grade prostatic intraepithelial neoplasia (PIN) or borderline lesions (lesions suspicious for malignancy) in prostate needle biopsies warrants repeat biopsies. The reported frequency of these lesions in prostate needle biopsies varies considerably. The authors evaluated the frequency and clinical impact of high grade PIN and borderline lesions in sextant prostate needle biopsies obtained from screened participants in the European Randomized study of Screening for Prostate Cancer (ERSPC). METHODS: A total of 8763 participants in the Rotterdam section of the ERSPC ages 55-75 years were screened systematically for prostate carcinoma. Systematic sextant prostate needle biopsies were prompted by an abnormal digital rectal examination and/or abnormal transrectal ultrasonography findings at serum prostate specific antigen (PSA) levels > or = 1.0 ng/mL or a PSA level > or = 4.0 ng/mL. Repeat biopsies were obtained within 6 months after initial biopsy. RESULTS: Of 1824 biopsied men, 384 (21.1%) were found to have prostate carcinoma on initial biopsy. Twelve participants (0.7%) had isolated high grade PIN and 43 (2.4%) had borderline lesions. Repeat biopsies yielded no carcinoma in 7 participants with initial high grade PIN and 15 tumors (38.5%) in 39 participants with borderline lesions. CONCLUSIONS: In prostate needle biopsies obtained from a screened population, indications for repeat biopsy such as high grade PIN and borderline lesions do not represent large diagnostic subsets. Borderline lesions comprise the most important indication for a repeat biopsy. The low frequency of equivocal biopsy diagnoses in the current study supports the clinical applicability of sextant needle biopsies in population-based screening for prostate carcinoma.

Evaluation of the digital rectal examination as a screening test for prostate cancer. Rotterdam section of the European Randomized Study of Screening for Prostate Cancer.

BACKGROUND: The utility of digital rectal examination (DRE) as a screening test for early detection of prostate cancer has not been established. Therefore, we evaluated the usefulness of DRE as a stand-alone screening test and in conjunction with measured serum prostate-specific antigen (PSA) levels of 0-3.9 ng/mL and transrectal ultrasonography (TRUS). METHODS: Our study population consisted of 10,523 men aged 54-76 years who were randomly assigned to the screening arm of the Rotterdam, The Netherlands, section of the European Randomized Study of Screening for Prostate Cancer. The underlying prevalence of detectable prostate cancer was estimated by logistic regression analysis and used for calculating the sensitivity of DRE as a test. Pathologic characteristics of 105 radical prostatectomy specimens were used to determine the aggressiveness of the tumors diagnosed (and missed) by DRE. RESULTS: The overall detection rate for prostate cancer in this population when serum PSA measurement, DRE, and TRUS were used was 4.5%, and the detection rate with DRE alone was 2.5%. The positive predictive value of DRE ranged from 4% to 11% in men with PSA levels of 0-2.9 ng/mL and from 33% to 83% in men with PSA levels of 3.0-9.9 ng/mL or more. Most tumors detected by DRE in men with PSA levels of less than 4.0 ng/mL were small (mean volumes = 0.24-0.83 mL), and most were well differentiated (Gleason scores of 6 or less). Minimal, moderate, and advanced cancers were seen in 42%, 42%, and 16% of men, respectively, with a PSA level of 4.0 ng/mL or less. DRE alone allowed detection of 264 (55.8%) of 473 cancers; 82 (17.3%) of the 473 cancers would have remained undetected by PSA-based screening alone (i.e., no follow-up procedures for PSA values of 0-3.9 ng/mL). CONCLUSIONS: For PSA values of 0-3.9 ng/mL, the positive predictive value and sensitivity of DRE, tumor volume, and tumor grade were strongly dependent on PSA level. DRE has a poor performance in low PSA ranges.

Comparison of prostate-specific antigen corrected for total prostate volume and transition zone volume in a population-based screening study.

OBJECTIVES: To compare the discriminatory potential between prostate cancer and benign conditions of the prostate in a population-based screening study, of serum prostate-specific antigen levels (PSA) and PSA corrected for both the total prostate volume (PSA-D) and the transition zone volume (PSA-T). METHODS: In a randomized population-based screening study (Rotterdam section of the European Randomized Study of Screening for Prostate Cancer), in which 10,865 men have been screened, the biopsy results of 1202 men with PSA levels of 4 ng/mL or more were evaluated. Planimetric and prolate ellipsoid volumes of the total prostate as well as of the transition zone were measured. The measured volumes were compared with the volumes of 57 radical prostatectomy specimens through Spearman's rank correlation coefficient and agreement tests. A receiver operating characteristic (ROC) curve analysis was done of sensitivity and specificity of biopsy indications through PSA and PSA corrected for the volumes measured with transrectal ultrasound. RESULTS: In the 1202 men studied, 361 cases of prostate cancer were diagnosed. Both PSA-D and PSA-T showed a significantly higher area under the ROC curve (0.77 and 0.79, respectively) than PSA alone (area 0.65). There was no significant difference between PSA-D and PSA-T. The use of a PSA-D threshold value of 0. 10 ng/mL/cc would have avoided 28% of biopsies at the cost of 10% of detectable cancers. A PSA-D threshold of 0.15 ng/mL/cc would have avoided 73.8% of biopsies at the cost of not diagnosing 43.8% of detectable cancers. CONCLUSIONS: The planimetrically obtained prostate volume showed a more favorable agreement with the radical prostatectomy volume than the prolate ellipsoid volume. The discriminatory potential of the corrected PSA value is better at predicting the results of needle biopsy of the prostate when compared with PSA alone. The use of the transition zone volume for this correction results in a higher discriminatory potential when compared to the use of the total prostate volume; however, the observed difference was not statistically significant.