Changes in prostate gene expression in men
undergoing an intensive nutrition and
Dean Ornish*†‡, Mark Jesus M. Magbanua§, Gerdi Weidner*, Vivian Weinberg¶, Colleen Kemp*, Christopher Green§,
Michael D. Mattie§, Ruth Marlin*, Jeff Simko?, Katsuto Shinohara§, Christopher M. Haqq§and Peter R. Carroll§
§Department of Urology, The Helen Diller Family Comprehensive Cancer Center, and?Department of Pathology, University of California, 2340 Sutter Street,
San Francisco, CA 94115; *Preventive Medicine Research Institute, 900 Bridgeway, Sausalito, CA 94965;†Department of Medicine, School of Medicine,
University of California, 505 Parnassus Avenue, San Francisco, CA 94143; and¶Biostatistics Core, The Helen Diller Family Comprehensive Cancer Center,
University of California, 513 Parnassus Avenue, Box 0127, San Francisco, CA 94143
Communicated by J. Craig Venter, The J. Craig Venter Institute, Rockville, MD, April 2, 2008 (received for review February 13, 2008)
Epidemiological and prospective studies indicate that comprehensive
lifestyle changes may modify the progression of prostate cancer.
However, the molecular mechanisms by which improvements in diet
and lifestyle might affect the prostate microenvironment are poorly
understood. We conducted a pilot study to examine changes in
prostate cancer who declined immediate surgery, hormonal therapy,
or radiation and participated in an intensive nutrition and lifestyle
intervention while undergoing careful surveillance for tumor pro-
gression. Consistent with previous studies, significant improvements
in weight, abdominal obesity, blood pressure, and lipid profile were
Gene expression profiles were obtained from 30 participants, pairing
RNA samples from control prostate needle biopsy taken before
intervention to RNA from the same patient’s 3-month postinterven-
tion biopsy. Quantitative real-time PCR was used to validate array
observations for selected transcripts. Two-class paired analysis of
global gene expression using significance analysis of microarrays
detected 48 up-regulated and 453 down-regulated transcripts after
the intervention. Pathway analysis identified significant modulation
of biological processes that have critical roles in tumorigenesis,
including protein metabolism and modification, intracellular protein
traffic, and protein phosphorylation (all P < 0.05). Intensive nutrition
and lifestyle changes may modulate gene expression in the prostate.
Understanding the prostate molecular response to comprehensive
tion and treatment. Larger clinical trials are warranted to confirm the
results of this pilot study.
exercise ? lifestyle changes ? prostate cancer ? SHOC2 ? stress management
much lower in parts of the world where people eat a predominantly
low-fat, plant-based diet. We (4, 5) and others (6) have shown
previously that diet and lifestyle interventions in men with early-
stage prostate cancer decrease prostate-specific antigen (PSA) and
decrease the rate of PSA increase. These studies provided some
evidence that comprehensive lifestyle changes may have therapeu-
tic potential in early prostate cancers. However, although these
interventions are associated with decreased circulating insulin-like
growth factor 1 (IGF1) (7), and although serum from men after
intervention has reduced the ability to stimulate prostate cell-line
growth in vitro (4), the actual molecular effects of these interven-
tions in prostate tissue have not been previously examined.
Many men with indolent prostate cancers detected by PSA
screening will not exhibit disease progression during their lifetime;
their treatment and associated side effects are unnecessary (8). We
report here the results of the Gene Expression Modulation by
Intervention with Nutrition and Lifestyle (GEMINAL) study, a
prospective single-arm pilot clinical intervention study in men with
pidemiological evidence (1, 2) and migrant studies (3) indicate
that the incidence of clinically significant prostate cancer is
indolent low-risk prostate cancers, defined by strict clinical and
pathologic criteria designed to minimize the risk for metastatic
disease as a result of study participation (9). The 30 men who
enrolled did not undergo surgery or radiation therapy to treat their
low-risk tumors; rather, they underwent comprehensive lifestyle
changes (low-fat, whole-foods, plant-based nutrition; stress man-
agement techniques; moderate exercise; and participation in a
psychosocial group support). Participants donated serial prostate
needle biopsies at baseline and after 3 months of the lifestyle
intervention, from which nanogram quantities of mRNA were
array platforms were not sensitive to nanogram RNA quantities.
Therefore, a reproducible linear RNA amplification and printed
cDNA array platform was used, as in our previous studies of
melanoma (10), where subsequent studies have confirmed the
validity of the gene expression findings (11, 12). Furthermore,
confirmation of the study results, comparing in pairwise fashion
each man’s postintervention to his own preintervention sample.
This article examines the relationship of comprehensive diet and
lifestyle changes to gene expression in the prostate.
In the GEMINAL study, 273 men were screened, 96 declined to
participate, 146 did not meet inclusion criteria, and 31 were
enrolled. The participants’ demographics included a mean age of
62.3 years (range 49–80) and a mean PSA level of 4.8 ng/ml (range
0.5–21.4) on the day of the initial biopsy. As expected from the trial
eligibility criteria, all patients had a Gleason score of 6. Eighty-four
percent of the men identified their ethnicity as Caucasian, 9%
Hispanic, 3% Asian, and 4% African-American. Two-thirds of the
men were married, and 72% were currently employed.
a history of benign prostatic hyperplasia (BPH)] at the time of
screening. One outlier patient with a history of BPH and a prostate
Author contributions: D.O. and M.J.M.M. contributed equally to this work; D.O., G.W.,
R.M., C.M.H., and P.R.C. designed research; D.O., M.J.M.M., C.K., C.G., M.D.M., R.M., J.S.,
K.S., C.M.H., and P.R.C. performed research; M.J.M.M., V.W., C.G., and C.M.H. analyzed
data; and D.O., M.J.M.M., G.W., V.W., C.K., C.M.H., and P.R.C. wrote the paper.
in the spirit of full disclosure, D.O. writes general-interest books on preventive medicine,
receives lecture honoraria, and consults with food companies to make more healthful
Freely available online through the PNAS open access option.
Data deposition: The data reported in this paper have been deposited in the Gene
‡To whom correspondence should be addressed. E-mail: firstname.lastname@example.org.
This article contains supporting information online at www.pnas.org/cgi/content/full/
© 2008 by The National Academy of Sciences of the USA
June 17, 2008 ?
vol. 105 ?
no. 24 ?
PSAs of 11.6–12.1 ng/ml, meeting eligibility criteria, but had a PSA
of 21.4 ng/ml on the day of initial biopsy. After the 3-month
intervention, this patient’s PSA decreased to 13.9 ng/ml, and his
prostate volume decreased to 75 cc. All other participants had
screening and initial biopsy-day PSA values of ?10 ng/ml.
of all concomitant medications remained stable through the
3-month assessment, with the exception of one participant whose
dosage of a statin drug was reduced. One of 31 study participants
was referred for surgery based on the result of his 3-month
follow-up biopsy, which showed an increased tumor Gleason score
of 3 ? 4 compared with his baseline Gleason of 3 ? 3.
As seen in Table 1, baseline body mass index (BMI), systolic
blood pressure, and LDL cholesterol were all somewhat elevated.
Quality of life scores were near the SF-36 highest-quartile cutoff
score for mental health (?49.9) (24). The psychological distress
reported in this sample was comparable with the level of distress
reported by other men diagnosed with prostate cancer (mean
duration from diagnosis to study entry was 12.3 months) by using
the Impact of Event scale (26).
Patients were able to adhere closely to the lifestyle recommen-
dations. After 3 months, they reported consuming 11.6% (SD ?
3.0) of fat calories per day, exercising ?3.6 h per week (SD ? 1.5),
and practicing stress management 4.5 h per week (SD ? 2.0). A
significant improvement in cardiovascular disease risk factors was
observed, including reductions in BMI, systolic and diastolic blood
pressure, and lipids (Table 1). Waist circumference decreased from
97.2 ? 11.1 cm to 89.5 ? 9.2 cm (P ? 0.001). Triglycerides and
C-reactive protein decreased, although these trends did not reach
(green) in morphologically normal prostate after a comprehensive diet and lifestyle intervention.
Plot of two class pairwise SAM (false discovery rate of ?0.10). Forty-eight transcripts were up-regulated (red) and 453 transcripts were down-regulated
Table 1. Cardiovascular risk factors and psychological functioning at baseline and 3 months
Mean ? SD
Mean ? SD
95% CI of change
Cardiovascular risk factors
Systolic blood pressure, mmHg
Diastolic blood pressure, mmHg
Total cholesterol, mg/dl
C-reactive protein (ln)
Mental component summary
Physical component summary
Impact of Event Scale
26.5 ? 3.6
129.5 ? 15.1
68.2 ? 10.7
121.7 ? 28.2
48.5 ? 12.7
191.7 ? 33.1
2.7 ? 1.0
107.0 ? 72.1
0.12 ? 1.1
23.9 ? 3.0
120.3 ? 12.6
62.8 ? 10.3
87.6 ? 25.1
40.2 ? 12.1
146.5 ? 31.6
2.3 ? 0.8
93.6 ? 46.3
?0.14 ? 1.3
49.7 ? 11.8
54.7 ? 4.9
56.2 ? 5.5
55.4 ? 5.5
6.2 ? 5.6
10.1 ? 6.8
3.7 ? 4.8
7.1 ? 5.8
*Paired-samples t test, two-tailed.
www.pnas.org?cgi?doi?10.1073?pnas.0803080105Ornish et al.
4.8 ? 3.9 to 4.6 ? 3.4 ng/ml; P ? 0.48), although percent free PSA
was improved, from 17.5 ? 7.4 to 18.9 ? 8.3 (P ? 0.05).
Patients reported significant reductions in psychological distress
associated with prostate cancer, as indicated by lower scores on the
intrusive and avoidant thoughts subscales of the Impact of Event
scale. Mental health-related quality of life also improved, with
increases in the Mental Component Summary score of the SF-36,
although physical health-related quality of life was stable.
an experimental biopsy after 3 months of intervention. Of the 31
patients enrolled, 30 were evaluable for gene expression, and one
patient sample was excluded because the biopsy tissue did not contain
man provided his own control tissue, minimized the chance that
contained both pre- and postintervention samples, selected randomly,
Using the significance analysis of microarrays (SAM) algorithm, we
prostate tissue after 3 months of intervention [Fig. 1, Table 2, and
supporting information (SI) Dataset S1 and Dataset S2], com-
pared with the paired normal prostate tissue samples from the
same individual patients at baseline, with the false discovery rate
set at 0.10. The heat map in Fig. 2 demonstrates visually the
substantial modulation of gene expression seen when pre- and
postintervention samples were compared.
We noted that several of the cDNA array clones in the SAM
used the University of California (Santa Cruz, CA) (UCSC)
genome database and software tools (23) to investigate whether
these clones might represent recently recognized genome elements
such as microRNAs, snoRNAs, or scaRNAs to confirm evidence
that clones represent transcribed loci and to exclude clones whose
sequence was composed of repetitive DNA elements. Eighteen
up-regulated transcripts and 388 down-regulated transcripts met
were found. The 10 highest ranked up-regulated and 20 highest
ranked down-regulated transcripts are listed in Table 2.
For the subset of samples with sufficient total RNA remaining
after array analysis, validation QRT-PCR experiments were per-
formed to confirm the accuracy of array gene expression measure-
ments. Results across the 30 study patients are shown in Fig. 3
comparing SHOC2 array and QRT-PCR measurements. Both
methods demonstrated transcript down-regulation after the life-
Table 2. Up-regulated and down-regulated transcripts in paired
pre- and post- diet/lifestyle intervention in normal prostate
ranking Unigene or genome database name
EST chromosome 18
EST chromosome 8
RAN, member RAS oncogene family
Soc-2 suppressor of clear homolog (C. elegans)
Transcribed sequence chromosome 18
Integrin, ? 10
Solute carrier family 35 member D1
Matrix metallopeptidase 9
DENN/MADD domain containing 1B
Ring finger protein 150
Homeodomain interacting protein kinase 1
Nuclear undecaprenyl pyrophosphate synthase 1
Strawberry notch homolog 1 (Drosophila)
Transcribed sequence chromosome 18, 4610527
Glycerol-3-phosphate dehydrogenase 1-like
Transcribed sequence chromosome 1, ZE03F06
ATP citrate lyase
SUB1 homolog (S. cerevisiae)
Kruppel-like factor 6
Cdc2-related kinase, arginine/serine-rich
Fms-related tyrosine kinase 1
Nuclear receptor subfamily 2, group F, member 1
Transcribed sequence BC029658
Zinc finger protein 250
Transcribed sequence chromosome 8
Chromosome 21 open reading frame 131
Chromosome 11 open reading frame 71
Zinc finger protein 160
Transcribed sequence CR627148
Chromosome 6 open reading frame 217
up-regulated transcripts (red in the postintervention samples) and 453 down-
regulated transcripts (green in the postintervention samples) in morpholog-
ically normal prostate after a comprehensive diet and lifestyle intervention.
Ornish et al. PNAS ?
June 17, 2008 ?
vol. 105 ?
no. 24 ?
style intervention, and there was no statistical difference between
the array and QRT-PCR measurements using the Wilcoxon
To gain further insight into the molecular changes occurring in
the prostate tissue, we conducted pathway analysis using the
of biological processes with critical roles in tumorigenesis among
and Dataset S3). Pathways involved in protein metabolism and
modification, intracellular protein traffic, and protein phosphory-
lation were significantly down-regulated (all P ? 0.05) (see Fig. 4
and Dataset S4).
The GEMINAL study was a prospective pilot trial of diet and
lifestyle intervention in men with low-risk prostate cancer. Unlike
patients with other tumor types, who must undergo immediate
resection, radiation therapy, or, rarely, chemotherapy, low-risk
prostate cancers may be observed, providing a unique opportunity
for molecular study when men undergo serial biopsy during active
surveillance. Similar to previous active surveillance clinical studies
(29, 30), our patients reported no adverse events. Referral for
definitive surgery or radiation therapy occurred for only one
of prostate cancer compared with baseline (which was likely due to
sampling variability in only 3 months).
Our observations provide molecular hypotheses that may help
explain some of the effects of comprehensive lifestyle changes. The
imperative to understand these effects has been spurred by the
observation in prospective trials that healthy diet and lifestyle
cancer (31) and breast cancer (32).
We found a set of RAS family oncogenes (RAN, RAB14, and
RAB8A) to be down-regulated. In the prostate, RAN (ras-related
nuclear protein) may function as an androgen receptor coactivator,
and its expression is increased in tumor tissues (33). However,
despite the down-regulation of this AR coactivator, androgen-
regulated PSA in our study was stable comparing baseline with
3-month measurements, both at the RNA level and in serum
measurement of total PSA.
The percentage of free PSA was significantly improved when the
entire group of 30 participants was analyzed. However, the predic-
tive value of the percentage of free PSA has been validated only in
men with a PSA of ?4 ng/ml (34–37). As expected, because of the
small sample size, our results were not statistically significant when
analysis was limited to the 15 of 30 men whose PSA on biopsy day
1 was ?4 ng/ml. Therefore, our percentage of free PSA results
should be considered exploratory.
intervention will be needed to see whether the stabilization in PSA or
and lifestyle intervention trials, it also would be interesting to directly
examine the modulation of the activity of the androgen receptor at its
to examine whether AR occupancy at response elements within the
prostate-specific antigen promoter is modulated by diet and lifestyle
illustrating the down-regulation of these 31 transcripts. Pre- and postcom-
prehensive diet and lifestyle intervention samples are indicated.
Heat map of the gene ontology group ‘‘Intracellular Protein Traffic’’
SHOC2 gene expression comparing microarray (filled bars) to the mean of
three independent replicate QRT-PCR (cross-hatched bars) measurements. No
biopsy RNA was available, precluding measurement.
Waterfall plot of ratio of log (base 2) post- versus preintervention of
Table 3. Overrepresented ontology categories in molecular functions and biological processes (P < 0.05) among genes
down-regulated after a diet/lifestyle intervention
NCBI: Homo sapiens genes,
number of genes
GEMINAL down-regulated genes,
number of genesExpected
Membrane traffic protein
Select regulatory molecule
Protein metabolism and modification
Intracellular protein traffic
www.pnas.org?cgi?doi?10.1073?pnas.0803080105Ornish et al.
synthesis and cell division (38), which suggests that its down-regulation
could have antitumor effects independent of its activity in modulating
gene encodes a protein that is essential in MAPK activation by
been reported previously. SHOC2 is proposed to be an attractive
human malignancies with up-regulated MAPK activity (39).
Several recent studies have examined how nutrition affects gene
expression in the context of obesity and metabolic syndrome (40,
41). These investigations (the FUNGENUT study) profiled gene
expression in subcutaneous adipose tissue and found down-
regulation of IGF pathway genes and genes related to fat metab-
down-regulation of IGF pathway genes [IGF1 receptor (IGF1R),
phosphoinositide-3-kinase, class 2, ?-polypeptide (PIK3C2A), and
forkhead box A2 (FOXA2)] and down-regulation of fat metabolism
genes [acyl-CoA dehydrogenase, long chain (ACADL), and
phytanoylCoA 2-hydroxylase (PHYH)]. In addition, prostate tissue
exhibited down-regulation of carbohydrate metabolism genes
[6-phosphofructo-2-kinase (PFKFB1), glycerol-3-phosphate dehydro-
genase 1-like (GPD1L), and ATP citrate lyase (ACLY)].
Recent studies also indicate that exercise may influence gene
expression (42–44). These studies reported changes in gene expres-
sion in leukocytes after moderate or exhaustive exercise, with
modulation of genes related to oxidative stress and inflammation.
However, GEMINAL results did not include changes related to
oxidative stress or inflammation. Differences in GEMINAL versus
these exercise study observations may relate to lower intensities of
exercise prescribed for our elderly population and tissue-specific
differences in the response to exercise.
Two important challenges and opportunities in conducting re-
search in patients with clinically localized prostate tumors were
evident in our study. First, many men with low-risk tumors are
reluctant to forego definitive treatment, as illustrated by our need
to screen 271 men to enroll 31. Second, our analysis was limited to
normal prostate tissue because tumor tissue was present on the
biopsy specimens of only a minority of patients. Thus, the impli-
cations of this study are not limited to men with prostate cancer.
Because of the microfocal nature of low-risk prostate tumors and
the limitations of ultrasound biopsy guidance, we were unable to
precisely match pre- and postintervention tumor samples for indi-
viduals in our cohort.
It is important to recognize the limitations of this study. Because
a randomized trial in the absence of data first demonstrating that
diet and lifestyle intervention may modulate gene expression.
GEMINAL now provides preliminary evidence that prostate gene
expression may be modulated by diet and lifestyle and provides the
rationale needed to support new randomized controlled trials.
Because only one-third of patient biopsies in our study included
tumor tissue, we were limited to examining the response of the
It will be very important for future work to examine tissue molec-
ular responses to determine whether the normal stroma, tumor
stroma, normal epithelium, tumor epithelium, or a combination of
these tissues respond to diet and lifestyle changes.
Although we observed a decrease in potential oncogenes such as
RAN and SHOC2, suggesting that the effects of the intervention
were beneficial, confirmatory studies are needed. One level of
here are accompanied by parallel changes in protein levels. In
addition, our work predicts that functional studies that lower the
levels of GEMINAL down-regulated genes should reduce tumor
incidence or progression in appropriate animal models.
Future investigations should enroll larger numbers of patients
and should include a control group to rule out the theoretical
possibility (although never documented) that repeat biopsy could
introduce gene expression changes independent of an intervention.
Ongoing, randomized trials of diet and lifestyle interventions in
men with low-risk prostate cancer, including the Molecular Effects
Study (MEAL) studies (45), feature control groups. These studies
are expected to provide important confirmatory data when they
report in future years. With improvements in commercial RNA
amplification and oligonucleotide microarray platforms, future
studies may adopt these standards to facilitate cross-study compar-
isons and to comprehensively examine the human transcriptome.
The intervention used in our study was complex, and future studies
may address which components or combinations of components of
with different genotypes. For example, one P450 cytochrome allele
may metabolize a dietary substrate to a bioactive form, in contrast
and sophistication of clinical trials will become possible.
In conclusion, the GEMINAL study suggests that intensive
prostate. Understanding the mechanisms of how comprehensive
lifestyle changes affect transcriptional regulation may strengthen
efforts to develop effective prevention and treatment strategies for
prostate cancer. Larger randomized controlled clinical trials are
now warranted to confirm and extend the hypotheses generated by
the results of this pilot study and to better understand the relative
contribution of each component of the intervention.
Materials and Methods
Study Subjects and Clinical Assays. Men with low-risk prostate cancer willing to
make comprehensive lifestyle changes gave informed consent under a protocol
prostate cancer for reasons unrelated to this study. Eligibility criteria included:
?6 with no pattern 4 or 5, stage T1 or T2a, ?33% of biopsy cores positive, and
?50% of the length of a tumor core positive. Standard clinical assays were used
for waist circumference, weight, height, blood pressure, serum lipids, C-reactive
protein, and PSA.
Lifestyle Intervention. A 3-month comprehensive lifestyle modification was pre-
scribed (13, 14), comprising a 3-day intensive residential retreat, followed by an
nurse. Lifestyle modifications included a low-fat (10% of calories from fat), whole-
foods, plant-based diet, stress management 60 min per day (gentle yoga-based
stretching, breathing, meditation, imagery, and progressive relaxation), moderate
aerobic exercise (walking 30 min per day for 6 days per week), and a 1-h group
Biopsy Processing. The 18-gauge core needle biopsies were collected, under
ultrasound guidance, from each patient before and after the intervention and
loss of cytokeratin identifies prostate tumor tissue) (15), and 17 sections were
used for RNA preparation on RNEasy columns (Qiagen) guided by the study
pathologist. Only areas of normal prostate peripheral-zone tissue containing
both stroma and epithelial cells were dissected and processed for RNA.
RNA Amplification, Labeling, and Arrays. Briefly,25–100ngoftotalRNA,side-by-
side with an equal quantity of universal human reference RNA (cat. no. 740000;
Stratagene), was linearly amplified through two rounds of modified in vitro tran-
Ornish et al.PNAS ?
June 17, 2008 ?
vol. 105 ?
no. 24 ?
Axon Imager 4000b (Axon Instruments) using GENEPIXPRO3.0. Strategies to mini-
Microarray Data Analysis and Bioinformatics. Gene expression was analyzed
of experiments. The resulting data were then analyzed by using the SAM pack-
age. A two-class paired analysis was used to compare pre- and postintervention
Identification of ontology groups (biological processes/molecular functions)
and pathways were conducted by using PANTHER Version 6.0 (www.pantherd-
Information was chosen as the reference list for comparison using the ?2test for
testing (P ? 0.05 was considered significant).
Clone annotation was updated by using Unigene Build no. 208 of the human
genome accessed through the SOURCE database (22) and UCSC genome data-
a DNA sequence corresponding to repetitive elements or that represented non-
transcribed DNA were removed from the final dataset.
QRT-PCR Analysis. First, 100 ng of total RNA was reverse-transcribed by using the
iScript kit (Bio-Rad). Then, 5 ng of cDNA was subjected to QRT-PCR using Applied
expression value as unitless fold changes in the unknown sample compared with a
were normalized by the relative expression of a housekeeping gene, GUSB, ?-glu-
Quality of Life and Psychological Distress. The SF-36 (QualityMetric), a survey
instrument well validated in the prostate cancer population (24), was adminis-
tered to men at baseline and after the 3-month intervention. Psychological
distress was assessed by using the Impact of Event scale (25), a well validated
array results was analyzed with paired t tests, and the Wilcoxon matched-pairs
ethnicity/race, PSA, serum lipid values, and quality of life.
ACKNOWLEDGMENTS. We thank the men who participated in this study; the
the University of California San Francisco Center for Advanced Technology’s
microarray resource center for helpful advice; Sarah Dumican, our clinical re-
search coordinator, and Eduardo Sosa, Romelyn DeLos Santos, Scot Federman,
Stacey Dunn-Emke, Patty McCormac, Dennis Malone, Christine Chi, Nancy Lau-
and our colleagues at the Preventive Medicine Research Institute and University
of California, San Francisco, for helpful advice and discussion, as well as Nancy
Pelosi and John Murtha. This work was supported, in part, by Department of
Defense/U.S. Army Medical Research Acquisition Activity Grant W81XWH-05-1-
0375; Henry M. Jackson Foundation for the Advancement of Military Medicine
Bahna, Rohde, Talbott, Groppe, Gegax, PepsiCo, California HealthCare, George,
and Alternative Medicine Grant K01AT004199; National Cancer Institute Grant
CA101042–01; the Kerzner Foundation, the Bernard Osher Foundation, and the
Walton Family Foundation.
1. Chan JM, Gann PH, Giovannucci EL (2005) Role of diet in prostate cancer development
and progression. J Clin Oncol 32:8152–8160.
2. Chan JM, et al. (2006) Diet after diagnosis and the risk of prostate cancer progression,
recurrence, and death (United States). Cancer Causes Control 17:199–208.
of the colon, rectum, prostate and breast in Shanghai, China versus the United States.
Int J Epidemiol 20:76–81.
cancer. J Urol 174:1065–1069; discussion 1069–1070.
5. Ornish DM, et al. (2001) Dietary trial in prostate cancer: Early experience and implica-
tions for clinical trial design. Urology 57:200–201.
6. Saxe GA, et al. (2006) Potential attenuation of disease progression in recurrent prostate
cancer with plant-based diet and stress reduction. Integr Cancer Ther 5:206–213.
7. Li Z, et al. (2007) Feasibility of a low-fat/high-fiber diet intervention with soy supple-
from U.S. prostate cancer incidence trends. J Natl Cancer Inst 94:981–990.
prostate cancer with curative intent: Preliminary results. J Urol 167:1231–1234.
10. Haqq C, et al. (2005) The gene expression signatures of melanoma progression. Proc
Natl Acad Sci USA 102:6092–6097.
11. Rangel J, et al. (2008) Osteopontin as a molecular prognostic marker for melanoma.
12. Rangel J, et al. (2006) Prognostic significance of nuclear receptor coactivator-3 over-
expression in primary cutaneous melanoma. J Clin Oncol 24:4565–4569.
J Am Med Assoc 280:2001–2007.
14. Daubenmier JJ, et al. (2006) Lifestyle and health-related quality of life of men with
prostate cancer managed with active surveillance. Urology 67:125–130.
15. Wojno KJ, Epstein JI (1995) The utility of basal cell-specific anti-cytokeratin antibody
(34 beta E12) in the diagnosis of prostate cancer. A review of 228 cases. Am J Surg
16. Baugh LR, Hill AA, Brown EL, Hunter CP (2001) Quantitative analysis of mRNA ampli-
fication by in vitro transcription. Nucleic Acids Res 29:E29.
17. Hughes TR, et al. (2001) Expression profiling using microarrays fabricated by an ink-jet
oligonucleotide synthesizer. Nat Biotechnol 19:342–347.
18. DeRisi J, et al. (1996) Use of a cDNA microarray to analyse gene expression patterns in
human cancer. Nat Genet 14:457–460.
19. Eisen MB, Spellman PT, Brown PO, Botstein D (1998) Cluster analysis and display of
genome-wide expression patterns. Proc Natl Acad Sci USA 95:14863–14868.
20. Wang J, Nygaard V, Smith-Sorensen B, Hovig E, Myklebost OM (2002) Array: Analysing
single, replicated or reversed microarray experiments. Bioinformatics 18:1139–1140.
21. Thomas PD, et al. (2003) PANTHER: A library of protein families and subfamilies
indexed by function. Genome Res 13:2129–2141.
22. Diehn M, et al. (2003) Nucleic Acids Res 31:219–223.
23. Karolchik D, et al. (2008) The UCSC genome browser database: 2008 update. Nucleic
Acids Res 36:D773–D779.
24. Ware J, Dewey J, Kosinski M (2002) How to score version 2 of the SF-36 health survey
(standard & acute forms). Quality Metric.
stress. Psychosom Med 41:209–218.
with early localized prostate cancer. Br J Urol Int 90:56–61.
27. Xie J, et al. (2007) Sno/scaRNAbase: A curated database for small nucleolar RNAs and
cajal body-specific RNAs. Nucleic Acids Res 35:D183–D187.
28. Griffiths-Jones S, Saini HK, van Dongen S, Enright AJ (2008) miRBase: Tools for mi-
croRNA genomics. Nucleic Acids Res 36:D154–D158.
29. Klotz L (2008) Low-risk prostate cancer can and should often be managed with active
surveillance and selective delayed intervention. Nat Clin Pract Urol 5:2–3.
30. Cooperberg MR, Broering JM, Kantoff PW, Carroll PR (2007) Contemporary trends in
low risk prostate cancer: Risk assessment and treatment. J Urol 178:S14–S19.
31. Meyerhardt JA, et al. (2006) Impact of physical activity on cancer recurrence and
32. Chlebowski RT, et al. (2006) Dietary fat reduction and breast cancer outcome: Interim
efficacy results from the Women’s Intervention Nutrition Study. J Natl Cancer Inst
33. Li P, et al. (2002) Heterogeneous expression and functions of androgen receptor
co-factors in primary prostate cancer. Am J Pathol 161:1467–1474.
34. Catalona WJ, et al. (1998) Use of the percentage of free prostate-specific antigen to
enhance differentiation of prostate cancer from benign prostatic disease: A prospec-
tive multicenter clinical trial. J Am Med Assoc 279:1542–1547.
to total-PSA in differentiating benign prostate hypertrophy from prostate cancer.
Prostate Suppl 7:30–34.
36. Egawa S, et al. (1997) The ratio of free to total serum prostate specific antigen and its
use in differential diagnosis of prostate carcinoma in Japan. Cancer 79:90–98.
to total-PSA in differentiating benign prostate hypertrophy from prostate cancer.
38. Sanderson HS, Clarke PR (2006) Cell biology: Ran, mitosis and the cancer connection.
Curr Biol 16:R466–R468.
39. Rodriguez-Viciana P, Oses-Prieto J, Burlingame A, Fried M, McCormick FA (2006)
Phosphatase holoenzyme comprised of Shoc2/Sur8 and the catalytic subunit of PP1
functions as an M-Ras effector to modulate Raf activity. Mol Cell 22:217–230.
40. Dahlman I, et al. (2005) Changes in adipose tissue gene expression with energy-
restricted diets in obese women. Am J Clin Nutr 81:1275–1285.
41. Kallio P, et al. (2007) Dietary carbohydrate modification induces alterations in gene
expression in abdominal subcutaneous adipose tissue in persons with the metabolic
syndrome: The FUNGENUT Study. Am J Clin Nutr 85:1417–1427.
42. Zieker D, et al. (2005) cDNA microarray analysis reveals novel candidate genes expressed
in human peripheral blood following exhaustive exercise. Physiol Genomics 23:287–294.
43. Buttner P, Mosig S, Lechtermann A, Funke H, Mooren FC (2007) Exercise affects the
gene expression profiles of human white blood cells. J Appl Physiol 102:26–36.
44. Connolly PH, et al. (2004) Effects of exercise on gene expression in human peripheral
blood mononuclear cells. J Appl Physiol 97:1461–1469.
patients with favorable-risk prostate cancer. Curr Treat Options Oncol 8:173–196.
46. Kaput J, Perlina A, Hatipoglu B, Bartholomew A, Nikolsky Y (2007) Nutrigenomics:
Concepts and applications to pharmacogenomics and clinical medicine. Pharmacog-
www.pnas.org?cgi?doi?10.1073?pnas.0803080105Ornish et al.