![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Feinberg School of Medicine, Northwestern University, Chicago, IL 60611
2To whom correspondence should be addressed. E-mail: pgann{at}northwestern.edu.
KEY WORDS: • lycopene • tomatoes • biomarkers • prostate
 |
EXPANDED ABSTRACT |
|---|
 TOP EXPANDED ABSTRACT CONCLUSIONSLITERATURE CITED |
|---|
Phase 2 trial in men with high-grade prostatic intraepithelial neoplasia (HGPIN)
Our Phase 2 trial is open to men not receiving hormonal drugs or antioxidant supplements and who have had biopsy-confirmed HGPIN within 2 y of enrollment. Within 1 mo following a baseline 12-core biopsy that is negative for cancer, these men are randomly assigned to placebo or 2 daily tomato oleoresin capsules (LycoMato) standardized to contain 15 mg lycopene/capsule (total daily dose = 30 mg lycopene). Blood and paraffin-embedded tissue samples are stored from baseline. After 6 mo of treatment, subjects are rebiopsied following collection of extraprostatic fluid (EPF) and blood samples. During the study, participants are instructed on how to limit intake of lycopene-containing foods. The primary endpoints for the trial are immunohistochemical and nuclear morphometric changes, as discussed further below. Oxidation markers in peripheral blood and growth factor levels in EPF provide secondary endpoints. The tomato oleoresin form was chosen for the trial because it allows for placebo control yet is a natural product that provides most of the potentially active ingredients of tomatoes in addition to lycopene. The product contains (wt:wt) 6% lycopene, 0.6% phytoene, 0.5% phytofluene, 0.2% ß-carotene, 2% tocopherols, and 0.6% phytosterols. In a recent study with a 30 mg/d lycopene dose for a mean of 24.5 d, the capsules produced a doubling in serum lycopene levels and increases of 
50% in lycopene concentrations in skin and adipose tissue compared to controls (1). Enrolment for this trial is ongoing; the targeted enrollment is 80 men.
Choosing a study design
There are 3 practical designs for tissue-based Phase 2 trials in prostate cancer chemoprevention: 1) preradical prostatectomy (RP), 2) prebiopsy, and 3) biopsy/rebiopsy. The pre-RP design has the advantage of providing abundant tissue for analysis, including fresh tissue for RNA expression analysis. However, its shortcomings include a relatively short exposure period and the difficulty of comparing tissue samples from the diagnostic biopsy and the prostatectomy (in part due to differences in fixation). The prebiopsy design, which would enroll patients before a biopsy, provides a large pool of eligible patients but also has a short exposure period and relatively high endpoint variance. The biopsy/rebiopsy design has greater statistical power (efficiency) due to lower endpoint variance, and it allows for a 3- to 12-mo exposure between biopsies. However, the requirement for 2 biopsies reduces the pool of eligible patients. Decisions about whether to rebiopsy patients with a negative biopsy, and at what interval, have never been very uniform. Recently, however, urologists appear to be moving away from multiple repeat biopsies, perhaps in part due to the widespread use of extended (12-core) biopsies rather than the older sextant procedure and in part because monitoring changes in prostate-specific antigen (PSA) is becoming a more attractive alternative. At least 1 repeat biopsy is indicated for all patients with HGPIN, regardless of PSA change.
HGPIN as an endpoint
In the past few years, pathologists have come to a better consensus regarding criteria for classifying HGPIN on biopsies. Although the propensity to "call" HGPIN still varies considerably, isolated HGPIN (with no cancer) is reported in about 2–12% of all initial biopsies. Because this is not an uncommon event, the question arises: why not use disappearance or reduced volume of HGPIN as an endpoint in Phase 2 trials? Reversal of preneoplasia has been a useful endpoint in studies of cervical, skin, esophageal, and oral cancer chemoprevention studies. But HGPIN is rarely diffuse, and given the absence of imaging techniques for identifying lesions, prostate cancer is the only situation in which blind biopsies are performed. Therefore, due to spatial variation in blind needle samples, HGPIN can "disappear" on repeat biopsy due to sampling error alone. The probability of finding HGPIN a second time on a repeat biopsy has been estimated to be about 33%. The effect of this sampling error on the required study size for a 20% reduction in HGPIN and a given level of statistical power are shown in Table 1. If the agent has no effect and the probability of HGPIN on repeat biopsy is about 30%, as previously estimated, this trial will require >1700 patients. In contrast, the required number of patients for a 20% reduction in expression of a protein marker in normal tissue in a biopsy/rebiopsy study is much smaller, as shown in Table 2. Thus, Phase 2 biopsy/rebiopsy studies of lycopene or tomato products will be far more efficient if they use endpoints such as immunohistochemistry or nuclear morphometry on histologically normal tissue.
|
|
What are the molecular and cytometric characteristics of high-risk "normal" (i.e., histologically benign) tissue? These characteristics, in addition to providing useful clinical predictors for risk in patients with negative biopsies, can provide crucial intermediate endpoint biomarkers for prevention trials. This type of subhistological trait is sometimes referred to as a field effect, or in some literature as "malignancy-associated change." We use 4 types of studies to search for field effects in the prostate: 1) differential expression of a marker by progressive tissue compartments, 2) comparison of "supernormal" tissue (from glands free of cancer) to normal tissue, 3) comparison of normal tissue near versus far from a cancer lesion, and 4) case-control study of marker in previous negative biopsies. There are many emerging options, besides IHC and nuclear morphometry, for characterizing field effects in high-risk normal tissue. mRNA expression, loss of heterozygosity assays, complete genome hybridization (array CGH), fluorescence in situ hybridization/in situ hybridization, and gene-specific methylation assays are now possible using relatively small amounts of formalin-fixed paraffinized tissue. In addition, computer-assisted image analysis is now available to improve the sensitivity, validity, and precision of tissue scoring for some of these methods.
To illustrate, we have conducted such studies for several markers by IHC, including 
-methylacyl-CoA racemase (AMACR) (2), Ki67, minichromsome maintenance protein 2 (Mcm-2), caspase-3, and bcl-2. AMACR is a peroxisomal enzyme that is highly and consistently upregulated in nearly all prostate cancers (3). Expression levels in normal tissue are much lower, but are not always nil. We conducted a case-control study with 23 cases of prostate cancer that had a previous negative biopsy. For each case we selected a control patient, matched on age, who had a negative biopsy at approximately the same time as the case, but was not found to have prostate cancer on at least 3 total biopsies. The earlier negative biopsies were retrieved and stained for AMACR. We found higher AMACR expression in the normal compartment for the cases compared to the controls (P = 0.0006). Normal glands (acini) located near a cancer focus had higher AMACR staining than distant acini (P = 0.0006) and higher than normal glands sampled from prostates without cancer removed due to bladder cancer. These data support the hypothesis that increased AMACR represents a field effect in high-risk normal prostatic tissue.
Ki67 and Mcm-2 are proliferation markers. Mcm-2 is far less well known, but its expression correlates well with Ki67 in normal prostate (r = 0.72), and it is expressed in a far higher percentage of cells, making it easier to work with. We stained prostate tissue for Ki67 in progressive compartments [supernormal, normal low-grade prostatic intraepithelial neoplasia (LGPIN), HGPIN, low-grade cancer, intermediate cancer, and high-grade cancer] and observed a profound shift from basal epithelial cell expression to luminal cell expression occurring between LGPIN and HGPIN. For Mcm-2, the same analysis revealed substantially higher luminal expression in normal versus supernormal and at least an equally profound shift to luminal proliferation at HGPIN. The ratio of luminal:basal cell expression was the strongest discriminator between supernormal and normal tissue, especially for Mcm-2, as shown in Figure 1.
|
We are refining a digital imaging method for obtaining nuclear morphometric scores for normal nuclei based on their size, shape, and chromatin texture. This method digitally extracts nuclei and computes over 40 separate measurements on each nucleus. A continuous, multifeature nuclear grade index can be calculated for each tissue sample. Studies to detect field effects using this assay are ongoing.
Characterizing sources of variation/statistical issues
Prostate biopsy involves spatial variation from core to core within the gland as well as temporal variation when the gland is sampled repeatedly (as in biopsy/rebiopsy studies). It is important to understand the extent of these sources of variation in planning trials: the amount of spatial (core-core) variation affects the number of needle cores that should be assayed to maintain adequate marker variance, and the temporal variation affects the overall number of subjects required for a given statistical power. We use a repeated-measures ANOVA model to estimate variance components and compute the intraclass correlation coefficient—the proportion of total marker variance attributable to between-subject variation. Total variation = Varbetween + Varbetween(visit) + Varwithin, where Varbetween = overall between-patient variation and induces common visit-visit correlation, Varbetween(visit) = core-core variation and induces common core-core correlation, and Varwithin = within-patient variation (residual). This model can be written as follows: Yijk = marker value for ith subject, jth biopsy visit, kth needle core, where Yijk = µ + 
j + 
i(j) + ßk + (ß)jk + 
i(jk); j = jth visit effect, i(j) = random effect for subject i visit j, ßk = kth core effect, (ß)jk = interaction visit x core, and i(jk) = within-subject variation. With this model, the core-core correlation for the marker p27 (cell cycle inhibitor) was 0.71, and the visit-visit correlation was 0.63.
|
CONCLUSIONS |
|---|
|
TOPEXPANDED ABSTRACTCONCLUSIONSLITERATURE CITED |
|---|
|
FOOTNOTES |
|---|
3 Abbreviations used: AMACR, -methylacyl-CoA racemase; array CGH, complete genome hybridization; EPF, extrapostatic fluid; HGPIN, high-grade prostatic intraepithelial neoplasia; IHC, immunohistochemical; LGPIN, low-grade prostatic intraepithelial neoplasia; Mcm-2, minichromsome maintenance protein 2; PSA, prostate-specific antigen; RP, radical prostatectomy.
|
LITERATURE CITED |
|---|
|
TOPEXPANDED ABSTRACTCONCLUSIONSLITERATURE CITED |
|---|
1. Walfisch, Y., Walfisch, S., Agbaria, R., Levy, J. & Sharoni, Y. (2003) Lycopene in serum, skin and adipose tissues after tomato-oleoresin supplementation in patients undergoing haemorrhoidectomy or peri-anal fistulotomy. Br. J. Nutr. 90:759-766.[Medline]
2. Ananthanarayanan, V., Deaton, R. J., Yang, X. J., Pins, M. R. & Gann, P. H. (2005) Alpha-methylacyl-CoA racemase (AMACR) expression in normal prostatic glands and high-grade prostatic intraepithelial neoplasia (HGPIN): Association with diagnosis of prostate cancer. Prostate 63(4):341-346.[Medline]
3. Rubin, M. A., Zerkowski, M. P., Camp, R. L., Kuefer, R., Hofer, M. D., Chinnaiyan, A. M. & Rimm, D. L. (2004) Quantitative determination of expression of the prostate cancer protein alpha-methylacyl-CoA racemase using automated quantitative analysis (AQUA): A novel paradigm for automated and continuous biomarker measurements. Am. J. Pathol. 64:831-840.
This article has been cited by other articles:
|
|
 |
|
  C. D. Davis, C. A. Swanson, R. G. Ziegler, B. Clevidence, J. T. Dwyer, and J. A. Milner Executive Summary Report J. Nutr., August 1, 2005; 135(8): 2014S - 2029S. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
