|
PSA and Prostate Cancer Tumor
Mass:
Physicians and patients watch the PSA with an eagle eye: the basic
assumption being that as the PSA rises and falls there is a corresponding
increase and decrease in the total body tumor burden. Is this correlation
unchanging throughout the disease? What is the biological basis of this
association? Is the association tight enough so that PSA response to
chemotherapy can predict survival?
The clinical aspects of this issue were discussed by Partin in a two part
article in ONCOLOGY, September, 2002: "Prostate-Specific Antigen as a
Marker of Disease Activity in Prostate Cancer." He indicates that early in
the course of the disease there usually is good correspondence between PSA
and tumor mass, but Partin points out that PSA reductions from hormonal
therapy don't always correlate with survival. He refers to the trial of
orchiectomy alone compared to orchiectomy plus flutamide in which the
nadir PSA for the combination was significantly lower (P=<.001) but the
combined treatment showed no survival advantage. David Crawford provided
commentary to this article in which he stated "To date, however, no study
has clearly shown a survival relationship between PSA nadir and survival."
Balk ("Biology of Prostate-Specific Antigen", JCO, January, 2003) did
point up the improved duration of remission predicted by a PSA nadir of
<.4 ng/ml.
A review of the basic biology reveals the complexity of the association of
PSA and prostate cancer tumor mass. My interest in this subject arose from
the article by Denmeade, PROSTATE, Vol.54, 2003, "Dissociation Between
Androgen Responsiveness for Malignant Growth vs. Expression of Prostate
Specific Differentiation Markers PSA, HK2, and PSMA in Human Prostate
Cancer Models." Analysis of this issues leads to a consideration of the
mechanism of activation of the PSA gene. The PSA gene is one of a family
that contain "androgen response elements" (ARE), the very specific points
of attachment of the activated androgen receptor (AAR) to its target site
within the gene. The most forward portion of the gene contains a
"promoter", which in this gene contains two AREs. Four thousand or more
base pairs upstream away from the gene lies the "enhancer" containing
seven AREs. If only the two AREs in the promoter are occupied by activated
AAR, the PSA gene expression is weak. However, if all seven AREs in the
enhancer are also engaged, the strength of PSA secretion is increased 1000
to 3000 times. It remains a puzzle how the distant enhancer structurally
relates to the promoter to cooperate in the initiation of gene
transcription, however, the best explanation is that the DNA, responding
the bonding energies, loops back on itself to allow the enhancer to
contact the transcription machinery assembled on the promoter and augments
the strength of transcription. After transcription is initiated the AAR
quickly detaches and is destroyed and another wave of AARs must find its
way to the gene to continue stimulation. This complex mechanism for gene
expression applies, of course, to all the genes that respond to androgen
stimulation - each with its own set of AREs. This family of diverse genes
codes for a wide variety of transcripts governing secretion of prostate
specific membrane antigen, prostatic acid phosphatase; regulating mitoses
and cell cycle progression; controlling tumor suppressors and DNA repair
and many others. The total effect is further magnified since some genes
that are activated produce transcription factors for yet more genes! The
total number of genes controlled by androgen stimulation is unknown.
However, preliminary data from DNA microarray analysis suggests that
androgen signaling activates 136 genes and silences 215 others! Just this
introductory glimpse into the complexity and amazing specificity of this
entire process invites speculation that the mutational disarray that
results from malignancy could easily uncouple the coordinated expression
of this family of genes. And this is the irregularity reported by Denmeade
who found in his in vitro tests that some prostate cancer cell lines were
responsive to androgen for both growth and marker secretion, some negative
for both, and others mixed in response. Additionally, he found marked
differences in the strength of marker secretion. It would be surprising if
these in vitro irregularities were not duplicated in real life human
prostate cancer.
Fortunately, there seems to be sufficient linking of a decline in PSA to
improved survival exhibited in clinical trials of chemotherapy in hormone
refractory PC so that useful conclusions can be drawn. A consensus
supports the benchmark that a >50% reduction in PSA achieved after 8 weeks
of therapy can serve as an early surrogate for improved survival. It's
interesting that the foundation for this consensus rests primarily on one
very well done study reported by Smith and Pienta, "Change in Serum
Prostatic-Specific Antigen as a Marker of Response to Chemotherapy for
Hormone Refractory Prostate Cancer" (JCO, May 1999). The conclusion of
this study is especially credible because the study took into account many
additional patient factors that potentially could confound the
relationship of PSA and survival. The most important of these were
performance status, hemoglobin level, measurable disease, alkaline
phosphatase, and the relative reduction of PSA at 4 and 8 weeks post
initiation of a uniform therapy, estramustine/VP-16. Those patients who
achieved a >50% reduction of PSA at 8 weeks survived 23 months compared to
9 1/2 months for those who did not (P=.0005). Not surprisingly,
performance status was a strong predictor of survival; hemoglobin <10 g/dL
was of lesser importance. But when PSA decline was adjusted for
performance status and hemoglobin level, PSA retained its independent
strength. A second foundation article by Kelly (JCO, April, 1993) used the
same criterion for evaluating chemotherapy response in 110 patients and
reported a difference in survival of >25 months versus 8.6 months. The
strength of their conclusion was weakened because they couldn't take
account of performance status since the patients were drawn from disparate
studies using a variety of chemotherapy regimens. The consensus for the
>50% PSA reduction benchmark was codified by Bubley (JCO, November 1999)
reporting the agreement of 27 respected researchers. It's interesting that
for these experts the principle value of this criterion was to guide
investigators as to which regimen warranted further study. They cautioned
against its use in assuring patients about the likely outcome of their
treatment.
Bottom Line: Clinicians and patients are fortunate to have a marker as
generally serviceable as PSA to monitor the course of disease. An
understanding of the biology of the relationship of PSA to tumor mass,
however, can add perspective to PSA interpretation.
«
Back to Article List |
