| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Internal Medicine (J.A.M.J.L.J., H.A.P.P., S.W.J.L.), Urology (M.F.W., K.I., M.J.R., F.H.S.), Clinical Chemistry (B.G.B., R.H.N.v.S.), Erasmus MC, 3015 GD Rotterdam, The Netherlands; and Department of Urology (K.I.), Gunma University Graduate School of Medicine, Showa-machi, Maebashi, Gunma, 371-8511, Japan
Address all correspondence and requests for reprints to: J. A. M. J. L. Janssen, M.D., Ph.D., Department of Internal Medicine, Room D-436, Erasmus MC, Dr Molewaterplein 40, 3015 GD Rotterdam, The Netherlands. E-mail: j.a.m.j.l.janssen{at}erasmusmc.nl.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
To determine whether circulating free IGF-I, total IGF-I, and IGFBP-3 levels can predict future prostate cancer risk, we prospectively studied prostate cancer characteristics in a cohort of men during two rounds (mean interval, 4 yr) of a population-based screening study for prostate cancer. Two hundred one prostate cancer cases were detected at the second-round screening (aged 55–70 yr), and all these subjects were enrolled in the case group for the present study. Prostate cancer had been confirmed by biopsy in all cases. These 201 subjects were matched with the 201 nonprostate cancer cases by age, serum PSA range at the first-round screening (PSA < 2 ng/ml, n = 67; PSA = 2–3 ng/ml, n = 67; and PSA = 3–4 ng/ml, n = 67), and residence area.
At baseline, total IGF-I, free IGF-I, and IGFBP-3 levels and prostate volume of cases with prostate cancer were not different from those of healthy controls. PSA velocity was significantly different between cases and controls (P < 0.001).
Stepwise forward logistic regression analysis showed that only PSA levels at baseline and PSA at round 2 after 4 yr are good predictors of prostate cancer, whereas total IGF-I, free IGF-I, and IGFBP-3 did not predict the development of prostate cancer.
Only one of the 201 subjects with prostate cancer had metastases. Within the subjects with prostate cancer, there were no differences of IGF-I parameters with different tumor node metastasis categories and/or Gleason scores.
Our study suggests that the measurement of serum IGF-I and/or IGFBP-3 concentrations in addition to PSA does not improve the identification of men at high risk to develop early stages of prostate cancer. In addition, our results indicate that the endocrine IGF-I system is not directly involved in the growth of the early stages of prostate cancer.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
IGF-I is produced in large amounts by the liver and secreted into circulation (13). Hepatic production of IGF-I is predominantly dependent on GH secretion. Circulating total IGF-I levels are regulated by nutritional and genetic factors as well (14, 15). IGF-I is involved in the regulation of cell proliferation and differentiation but suppresses programed cell death as well (16). In vitro prostate cancer cells are responsive to the mitogenic effects of IGF-I (17).
Most circulating IGF-I (99%) is bound to IGF binding proteins (IGFBPs). Quantitatively, the most important IGFBP is IGFBP-3 (18). IGFBP-3 regulates IGF-I bioactivity and also independently modulates cell growth and survival. Changes in circulating IGFBP-3 levels have also been linked to prostate cancer. Although circulating IGFBP-3 levels were inversely associated with prostate cancer risk in the study of Chan et al. (1), a recent metaanalysis found a positive association between IGFBP-3 and prostate cancer risk (9).
PSA measurement is an important tool for identifying men at risk for prostate cancer. PSA correlates with the prostate volume in benign prostate hyperplasia and cancer and has been demonstrated to detect prostate cancer at an early stage (19, 20). PSA is an IGFBP-3 protease as well. Cleavage of IGFBP-3 by PSA causes a 10- to 100-fold decrease in the affinity of IGFBP-3 fragments to IGF-I (21). Elevation in serum PSA concentrations is correlated with decreased intact serum IGFBP-3 (22). These data together suggest that IGFBP-3 may also play a role in prostate cancer.
To determine whether circulating free IGF-I, total IGF-I, and IGFBP-3 levels indeed predict future prostate cancer risk, we prospectively studied prostate cancer characteristics in a cohort of men during two rounds of a population-based screening study for prostate cancer with a mean interval of 4 yr. We also studied whether PSA is associated with these parameters of the circulating IGF-I system.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
For the present study, a cohort of 4491 men aged 55–75 yr, all of whom had been randomly assigned to the screening group in the Rotterdam section of ERSPC, were studied (23). None of the participants had a previous diagnosis of prostate cancer. During the study period, screening was discontinued in all participants who reached the age of 75 yr. Men who did not respond to the invitation for screening were excluded from further evaluation in this report. The study was approved by a government commission supervising the compliance with the Dutch law on population screening, and written informed consent was obtained from every participant before randomization.
Screening protocols
First round of screening. The study cohort for the present study was recruited from the 19,970 men who were initially randomized for screening in the Rotterdam section of ERSPC. The first round of screening took place between June 1994 and March 1996. The 4133 men who accepted the invitation to the first screen underwent PSA measurement, digital rectal examination (DRE), and transrectal ultrasound investigation (TRUS). Blood samples were drawn in the nonfasting state. Biopsies were recommended for men whose serum PSA level was 4 ng/ml or greater, or if DRE and/or TRUS result was abnormal. Men who were recommended for biopsy but who either refused a biopsy or could not undergo a biopsy for medical reasons (e.g. because they were receiving anticoagulant therapy) were excluded from further evaluation.
Second round of screening. The second round of screening of the same population took place between June 1998 and December 2000. At this time, 2385 participants who had been screened 4 yr earlier were eligible and took part. By the time the second round of screening began, the screening protocol had been changed (25). In brief, because of low positive predictive value and sensitivity of DRE and TRUS, biopsies were now recommended to all participants with a serum PSA level of 3 ng/ml or higher, regardless of the outcome of DRE or TRUS (24).
To identify individuals with interval prostate cancers, a database from the local Rotterdam Comprehensive Cancer Registry was checked annually. The number of interval cancers (i.e. prostate cancers detected in a screened population outside regular screening) gives an indication of the efficacy of the screening protocol at baseline. In the cohort studied, prostate cancer was diagnosed outside the regular screening in only 18 men during the 4-yr interval and was found coincidentally in most cases (26).
Selection for the present case-control study. Inclusion criteria for the present study were: 1) men aged between 55 and 70 yr at first-round screening; 2) measurement of PSA level at first round; and 3) no prostate cancer at first-round screening.
Two hundred one prostate cancer cases were detected at the second-round screening, and all these subjects were enrolled in the case group for the present study. In all cases, prostate cancer had been confirmed by biopsy. Routine histopathological examination included the determination of stage and Gleason score (27, 28). PSA density (PSAD) and PSA density of the transition zone (PSATZD) were calculated for each subject (29).
Two hundred one participants without prostate cancer were selected from the 2385 men at the second-round screening. These 201 subjects were matched with the 201 prostate cancer cases by age (per 1-yr group), serum PSA range at the first-round screening (PSA < 2 ng/ml, n = 67; PSA = 2–3 ng/ml, n = 67; PSA = 3–4 ng/ml, n = 67), and for social background by place of residence (by postal code).
Blood biochemistry
Blood samples were taken by venipuncture. Serum was separated by centrifugation and stored at –80 C. PSA was measured using the Tandem-E immunoenzymetric assay (Hybritech Inc., San Diego, CA), and blood samples were drawn before the other tests were performed. PSA velocity (PSAV) [i.e. (PSA2 – PSA1/year)] was determined for each subject (30).
Total and free IGF-I and IGFBP-3 were assayed with commercially available immunoradiometric assays (Diagnostic Systems Laboratories, Webster, TX). All assays were carried out in a blinded manner, and quality-control samples were included within assay runs. Interassay coefficients of variation for total IGF-I, free IGF-I, and IGFBP-3 were 8.2, 10.3, and 1.9%, respectively. Intraassay coefficients of variation for total IGF-I, free IGF-I, and IGFBP-3 were 3.4, 10.7, and 3.9%, respectively. From four controls no blood was available for the measurement of parameters of the IGF-I system.
TRUS
TRUS was performed using a Brüel & Kjær model 1846 mainframe (Brüel & Kjær Medical, Naerum, Denmark) and a 7-MHz biplanar endorectal transducer with the participant in the left lateral decubitus position. The prostate volume and transition zone volume were measured using 5-mm step section planimetry.
Statistical analysis
Statistical analysis of the data was performed using SPSS for Windows (version 11.0; SPPS Inc., Chicago, IL). Data are presented as mean ± SD unless otherwise mentioned. P < 0.05 was considered significant. Data were tested for normality of distribution by Kolmogorov-Smirnov test, and, consequently, total IGF-I, free IGF-I, and PSA were log-transformed to obtain normal distribution.
Pearson’s correlation coefficients were calculated to assess associations between variables. Statistical analysis of differences between cases and controls were performed with the two-independent-sample t test if there was a normal distribution, and the Mann-Whitney U test if there was a nonnormal distribution. Independent factors for the prediction of prostate carcinoma were identified by a stepwise forward logistic regression analysis. The following variables were entered into the model log (total IGF-I), log (free IGF-I), IGFBP-3, PSAD, PSATZD, age at baseline, log (PSA visit 1), log (PSA visit 2), and PSAV.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
. PSA values at baseline were not significantly different between cases and controls, whereas there was a significant difference at visit 2. Mean PSA at the first visit was 2.3 ng/ml, and mean PSA at the second visit was 3.7 ng/ml. At baseline, total IGF-I, free IGF-I, and IGFBP-3 levels, prostate volume, PSAD, and PSATZD of cases with prostate cancer were not different from those of healthy controls (Table 2). PSAV was significantly different between cases and controls [0.042 (0.068) ng/ml·yr, mean (SD), vs. 0.009 (0.027) ng/ml·yr; P < 0.001]. Total IGF-I was positively related to free IGF-I (r = 0.47, P < 0.01) and IGFBP-3 (r = 0.51, P < 0.01), and free IGF-I was positively related with IGFBP-3 (r = 0.13, P < 0.01). Total and free IGF-I did not significantly decrease with age (total IGF-I: r = –0.056, P = 0.26; free IGF-I: r = –0.088, P = 0.08), whereas IGFBP-3 did (r = –0.22, P < 0.01). Prostate volume and PSA were not associated with age at baseline (prostate volume: r = 0.060, P = 0.35; PSA: r = 0.039, P = 0.44). Circulating total IGF-I, free IGF-I, and IGFBP-3 levels were not related to prostate volume (total IGF-I: r = –0.035, P = 0.59; free IGF-I: r = 0.05, P = 0.42; IGFBP-3: r = –0.11, P = 0.09). There were also no significant relationships among total IGF-I, free IGF-I, and IGFBP-3 levels and PSA, PSAV, PSAD, PSATZD (data not shown). After stratifying for prostate volume, mean total IGF-I, free IGF-I, and IGFBP-3 levels were not significantly different between prostate cancer cases and controls (Table 3).
|
|
|
There were no differences in IGF-I-related parameters between subjects with prostate cancer with different tumor node metastasis categories (T1–2 vs. T3–4) (Fig. 1). The Gleason scores of the 201 cases with prostate cancer were: Gleason grade 2–6, n = 157 subjects (78.2%) and Gleason grade 7–10, n = 41 subjects (20.5%). For three subjects, the Gleason score was missing (1.5%). There were no differences in IGF-I-related parameters between subjects with prostate cancer with different Gleason grades (Gleason grade 2–6 vs. Gleason grade 7–10) (Fig. 2).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Our study suggests that the measurement of IGF-I and/or IGFBP-3 in addition to PSA does not improve the identification of men at high risk to develop prostate cancer during the subsequent 4 yr. The strength of our study is the population-based design and the number of included prostate cancer cases and controls, all matched at baseline for PSA levels, age, and social background. In addition, cases and controls had comparable prostate volumes at baseline, whereas all subjects with prostate cancer at the first screening round had been excluded from our study. In accordance with our findings also, no differences in circulating total IGF-I and IGFBP-3 levels between prostate cancer patients and healthy controls have been reported in several other studies (5, 6, 8, 11, 33, 34, 35, 36, 37).
However, there are also a number of studies that suggested that increased serum IGF-I concentrations and decreased IGFBP-3 levels were independently associated with prostate cancer risk (1, 2, 3, 4, 7, 8, 9, 38, 39, 40, 41).
What could be the reason for these conflicting results? Most of the protocols of the studies that reported a positive association between IGF-I and/or IGFBP-3 and prostate risk contained several methodological differences compared with our study design. Occurrence of prostate cancer in most previous studies was captured passively (1, 2, 3), whereas in our study we actively screened for prostate cancer. In addition, we included only subjects who after active screening were found to be free of prostate cancer at baseline. In addition, in our study, cases and controls were carefully matched for age, serum PSA range at baseline, and social background, whereas in other studies these variables were not often matched very well. Sample size in several other studies differed from ours as well (1, 2, 4, 9, 38, 40, 41). In many of these studies, prostate volume at baseline was not evaluated (1, 2, 3, 4, 9, 39, 40, 41). Moreover, heparinized whole blood was used to measure IGF-I levels in one study (1). It is known that there may be significant discrepancies in the values of IGF-I and IGFBP-3 measured in serum and plasma when samples collected from the same individuals are compared (42, 43). Furthermore, in the study of Chan et al. (1), baseline PSA levels were already elevated in those individuals who developed prostate cancer during follow-up (see also below). Moreover, the time elapsed between the measurement of IGF-I levels and the actual diagnosis of prostate cancer varied between studies.
Another plausible and probably the best explanation for all these inconsistent results has been given by Chan et al. (34). Although they initially reported a positive association between plasma levels of IGF-I and the risk of prostate cancer within the Physicians’ Health Study, they were not able to confirm these results 2 yr later, when studying a larger sample within the same Physicians’ Health Study (1, 34). However, they still found a strong positive association for plasma IGF-I and an inverse association for plasma IGFBP-3 for extraprostatic and distant metastatic prostate cancers. These associations, however, were absent when tumors were confined to the prostate gland (34). Chan et al. therefore concluded that circulating IGF-I and IGFBP-3 may predict the risk of developing advanced-stage prostate cancer, although the significance for their measurement for screening purposes of localized incident prostate cancer may be limited. Our study confirms this latter conclusion. In our study, based on clinical staging, only 11 of the 201 prostate cancers were not confined within the prostate gland. In addition, serum IGF-I and IGFBP-3 levels did not correlate with the Gleason score in biopsy samples of patients with prostate cancer, in accordance to previous findings (44).
Our observations may have clinical implications because they suggest that the endocrine (i.e. circulating) IGF-I system is not directly involved in the induction of the early growth of prostate cancer cells. However, this still does not exclude involvement of the IGF-I system in prostate cancer, because the previous observations of Chan et al. suggest that as soon there is extracapsular spread of prostate cancer cells, the serum levels of IGF-I and IGFBP-3 may become important, because they may directly influence the growth of prostate cancer cells.
Using stepwise forward logistic regression analysis PSA levels at baseline and at the second round after 4 yr were the only good predictors of prostate cancer in our study. PSA is the most abundant protease produced in the prostate. PSA is also an IGFBP-3 protease. We did not find any correlations among IGF-I, IGFBP-3, and PSA. In two previous reports (22, 45), a negative correlation between PSA and IGFBP-3 has been reported. An explanation for this may be that in our study no patients were included with very high PSA levels associated with advanced metastatic disease. In favor of this latter argument, it has been found by Koistinen et al. (46) that the elevated PSA levels occurring in serum of patients with early prostate cancer do not correlate with the proteolytic cleavage of circulating IGFBP-3. Our findings do not exclude that PSA and other prostatic BP-3 proteases may induce the release of free IGF-I by local proteolysis of IGFBP-3 in the prostate. However, a small study (47) did not find a difference in free IGF-I levels in the prostatic fluid of patients with prostate cancer and normal men.
In conclusion, total IGF-I, free IGF-I, and IGFBP-3 levels are not useful predictors of early stages of prostate cancer in a (population-based) screening setting. Our study confirms the findings of Chan et al., suggesting that the endocrine IGF-I system is not directly involved in the induction of the early growth of prostate cancer cells.
|
Footnotes |
|---|
Received February 9, 2004.
Accepted June 16, 2004.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
1-antichymotrypsin before diagnosis of prostate cancer. Lancet 344:1594–1598[CrossRef][Medline]
This article has been cited by other articles:
 |
|
 |
|
  Z. Wang, R. M. Luque, R. D. Kineman, V. H. Ray, K. T. Christov, D. D. Lantvit, T. Shirai, S. Hedayat, T. G. Unterman, M. C. Bosland, et al. Disruption of Growth Hormone Signaling Retards Prostate Carcinogenesis in the Probasin/TAg Rat Endocrinology, March 1, 2008; 149(3): 1366 - 1376. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. E. Allen, T. J. Key, P. N. Appleby, R. C. Travis, A. W. Roddam, S. Rinaldi, L. Egevad, S. Rohrmann, J. Linseisen, T. Pischon, et al. Serum Insulin-like Growth Factor (IGF)-I and IGF-Binding Protein-3 Concentrations and Prostate Cancer Risk: Results from the European Prospective Investigation into Cancer and Nutrition Cancer Epidemiol. Biomarkers Prev., June 1, 2007; 16(6): 1121 - 1127. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Severi, H. A. Morris, R. J. MacInnis, D. R. English, W. D. Tilley, J. L. Hopper, P. Boyle, and G. G. Giles Circulating insulin-like growth factor-I and binding protein-3 and risk of prostate cancer. Cancer Epidemiol. Biomarkers Prev., June 1, 2006; 15(6): 1137 - 1141. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Wang, G. S. Prins, K. T. Coschigano, J. J. Kopchick, J. E. Green, V. H. Ray, S. Hedayat, K. T. Christov, T. G. Unterman, and S. M. Swanson Disruption of Growth Hormone Signaling Retards Early Stages of Prostate Carcinogenesis in the C3(1)/T Antigen Mouse Endocrinology, December 1, 2005; 146(12): 5188 - 5196. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Meyer, P. Galan, P. Douville, I. Bairati, P. Kegle, S. Bertrais, S. Czernichow, and S. Hercberg A Prospective Study of the Insulin-Like Growth Factor Axis in Relation with Prostate Cancer in the SU.VI.MAX Trial Cancer Epidemiol. Biomarkers Prev., September 1, 2005; 14(9): 2269 - 2272. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. C Larsson, K. Wolk, K. Brismar, and A. Wolk Association of diet with serum insulin-like growth factor I in middle-aged and elderly men Am. J. Clinical Nutrition, May 1, 2005; 81(5): 1163 - 1167. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
