|
Molecular Diagnostics (A NEW PCa COMMENTARY SERIES): SNPs and Gene
Expression Markers of Clinical Relevance.
(June 2008)
The cellular
heterogeneity within prostate tumors is the Achilles heel of accurate
prediction of future clinical behavior. Currently, for predictive
purposes men are stratified into risk categories, but outcome estimates
within these groups do not provide optimal specificity, based as they
now are on the now standard parameters of PSA level, tumor stage, and
Gleason score, and frequently augmented by data about percent of biopsy
cores positive for cancer, PSA density and PSA velocity. These venerable
predictors have been, and remain, serviceable, but fail to identify
individuals whose cancer will behave distinctly worse or better than
projected for their risk category. The mathematical model of "median"
captures the obvious - in every collective grouping there are those
whose performance lies at the extreme ends of the spectrum.
The future of better
prediction is coming into full view as sophisticated techniques begin to
allow characterization of these outliers based on unique individual
molecular features, gene alterations, and patterns of gene expression.
SNPs ("snips" - single nucleotide polymorphisms) are unique base-pair
exchanges in DNA that don't inhibit translation but do result in a
functionally different protein compared to what is customary for the
gene in question. Gene expression markers include mutations, gene
fusions, or a collective pattern of gene expression that defines unique
biologic consequences.
Although the literature
is burgeoning with reports of SNPs and markers that point to an
increased risk for a diagnosis of prostate cancer, this
series will focus on examples which yield information that would likely
influence the clinical management of an
individual patient.
The series begins with
two candidates for potentially clinically useful markers.
1. TMPRSS2-ERG
- What a mouthful of letters! Jointly they specify an unwelcome gene
fusion, found in the majority of prostate cancers, joining the TransMembrane
PRotease Serine 2 (TMPRSS2) promoter with a
transcription factor coded by the gene, ERG, a member of the
ETS transcription family. This fusion results in a cellular
phenotype of increased aggressiveness, distinct from other prostate
cancers. Research into the clinical consequences of harboring this gene
fusion is intense.
What might be the
potential utility of identifying this gene fusion - in tissue specimens
or urine? Some preliminary data suggests that it offers significant
predictive information. Three examples follow:
Fusion positive HGPIN: TMPRSS2-ERG gene fusion
has been found in a subset of HGPIN, which raises the possibility that
fusion positive HGPIN may be a harbinger of more aggressive disease. Its
presence in a biopsy showing only isolated HGPIN might become an
indication for earlier rebiopsy. (Mosquers JM, Clin Cancer Res,
June 1, 2008)
TMPRSS2-ERG:
Prognostic for Cancer Progression:
Nam RK et al., Cancer Biol Ther. 2007 Jan, reported that among 26
patients operated on for localized, Gleason score 7 prostate cancer,
eleven (42.3%) were gene fusion positive. Nine of 26 experienced
biochemical relapse. However, "Patients with the fusion protein had a
significantly higher rate of recurrence (5-year recurrence rate 79.5%)
compared to patients who lacked the fusion protein (5-year recurrence
rate 37%, p=0.009). ... On multivariate analysis, the presence of gene
fusion was the single most important prognostic factor."
Predictive of
Progression in an Observational Study:
Attard et al. (Oncogene. 2008 Jan) evaluated 445 men in an
observation study analyzed for TMPRSS2-ERG fusion. Men with
"Cancers lacking ERG alterations exhibited favorable cause-specific
survival (90% survival at 8 years)" compared to 25% cause-specific
survival at 8 years. They concluded that the novel category of fusion
alteration they studied "allows stratification of prostate cancer into
distinct survival categories."
2. PCA PROSTATE
CANCER GENE: The PCA3 gene is over expressed 60 - 100 fold by
prostate cancer cells as compared to the PSA gene, and a test is
currently commercially available to evaluate the ratio of urinary PCA3
mRNA to the PSA mRNA from non-cancerous prostate cells following a
prostate massage. (See PCa Commentary under "Diagnostics", May 2007.)
The most recent iteration of the test is more accurate in indicating
early prostate cancer than serum PSA.
The informational
utility of this test has now been reported to allow estimation of both
tumor volume and Gleason score: "PCA3
Molecular Urine Assay Correlates with Prostate Cancer Tumor Volume:
Implication in Selecting Candidates for Active Surveillance", Nakanishi
et al., J Uuol, May 2008. The PCA3 test is scored on a scale of 0
- 100, and expresses the ration of PAC3 mRNA to PSA mRNA. In their study
the median test values for total tumor volumes of <0.5cc was 21.5; for
volumes 0.5cc to <2cc, 41.8; and for volumes >2cc, 48.1.
The test performed best
in designating low volume tumors, and allowed a
significant distinction between of total tumor volumes of 0.5 cc vs. <2
cc tumors (p=0.002); and between the 0.5 cc tumors vs. those > 2 cc
(p=0.001). Although their additional conclusion is preliminary and
requires validation, their data suggests the test results may also
estimate the pathologic tumor grade. For Gleason score 3+3 the median
test value was ~25, versus a median test score of ~55 for tumors with
Gleason scores >7, p=0.005.
The authors concluded:
"The PCA3 score appears to stratify men based on tumor volume and may
have clinical applicability in selecting men who have low volume/low
grade cancer as active surveillance candidates."
An editorial remark by Leonard Marks offered
perspective: "As compelling as the PCA3 story now appears, in the
future, test panels [combining] molecular markers including PCA3 and the
TMPRSS2-ETS gene fusion products will likely prove to yield more
clinical information than single tests."
«
Back to Article List
|
