|
The
Very, Very Long and Short of It: A Brief History of Androgen Sensitive
Prostate Cancer Cells (October 2003)
Early cancer is really
never "early" when considered in the context of biologic history. There
considerable consensus that 30 volume doubling times (26 if the tumor is
half stroma) are required for a single cell to achieve a volume of one
cubic centimeter ("On the Growth Rates of Human Malignant Tumors:
Implication for Medical Decision Making, Friberg in J of Surg Oncol,
1997;65:284-297). And once the tumor attains a size that can be
objectively detected on imaging studies further growth follows a smooth
curve of exponential volume increase that translates to linear on semilog
presentation. Although the rate of growth is unique for each heterogeneous
tumor mass, if a doubling time (DT) were figured at 150 days (a
representative possibility for breast cancer) the time from tumor
inception to one cm3 (one billion cells) would be 12 years. Calculated at
that rate, to grow to the first 2 mm mass (10 M cells) it would take 8
years. Most prostate cancers grow more slowly, placing their inception at
10, 20 or 30 years before clinical detection.
The biologic growth
details describing the period at the earliest cancer stages, well before
clinical detection, have been studied by Berges at Johns Hopkins
("Implication of cell kinetic changes during the progression of human
prostatic cancer", Clin Cancer Res 1995, May;1(5):473-80). They made cell
culture measurements of the balance between the daily percentage of cells
proliferating and dying in normal and premalignant prostate cells, and in
malignant cells in the prostate, lymph nodes and bone. The normal cells
showed a balanced rate of proliferation and death at the very low and
balanced rate of <.2%/day. Criticism has been made of such cell culture
studies because of the artificiality of the environment, lacking as it
does, for example, the paracrine influence of nearby stromal cells. But
the observations are none the less generally instructive. In their studies
the glandular cells were replaced every 500 +/- 79 days. The transition to
HGPIN was associated with increased proliferation compared to cell death
leading to a mass doubling time of 154 days. Later in HGPIN the rate of
cell death increased to match the proliferation rate leading to a steady
state with, however, a 6 fold increase in cell turnover (DT 56 days). With
an increase in cell turnover rate there is a greater chance for mutations
to occur. The transition into localized cancer involved no further
increase in proliferation but instead was associated with a decrease in
the cell death rate, leading to a net continuous growth with a
prolongation of the mean DT to > 475 days. PC growth in lymph nodes
and bone showed comparative increased proliferation and further reduction
of cell death yielding a DT of 33 days and 54 days, respectively. No
further increase in proliferation was observed in andgrogen insensitive
cells from lymph node and bone, but, rather, a increase in the death rate
in these androgen independent cells leading to a slowing of growth to DT
126 and 94 days. The remarkably slow proliferation rate of prostate cancer
(i.e <3%/day) may explain the relative insensitivity of PC cells to
chemotherapy and the lengthy gestation time (possibly 30 years) to
pathologically detectable metastatic disease.
Viewed from another
perspective, the growth rate of tumors can be inferred from measurements
of the DT of PSA, and many analyses have been reported based on sequential
observations in patients during "watchful waiting". Each individuals' PSA
relationship to his tumor volume is unique, influenced by prostate size
and the relative proportion of BPH to cancer (BPH secretes PSA at 1/12 the
rate of PC), the tumor histology (PSA secretion decreases with increasing
tumor grade, but growth rate increases), and the amount of the
neuroendocrine component (which secretes chromogranin, but not PSA), and,
of course, other unknown individual biologic variables. In the 12 studies
I have reviewed it was not surprising that the major generalization is the
wide variability in PSA doubling times. The width of this
spectrum makes calculating a "mean PC DT" mathematically interesting, but
biologically rather irrelevant. For example, Choo reported a study from
University of Toronto in which watchful waiting and delayed intervention
was carried out in 134 in men with tumors of T1b-T2b N0 MO, Gleason <7,
PSA < 15. The minimum F/U was 12 months (median 24). The
distribution of PSA DT was: <2 yr 19 patients; 2-5 yr, 46; 5- 10 yr, 25;
10-20 yr, 11; 20-50 yr, 6; >50 yr, 27 ! The median DT, primarily of
arithmetic interest, was 5.1 years and >33% had DT > 10 years. Other
studies have shown a "median" PSA DT of 4 - 5 years.
And then ... with
castration this slowly constructed tumor mass, in one decimating apoptotic
collapse, comes crashing down in less than 21 days (2 or 3 months for
medical castration)! In "Quantitation of Apoptotic Activity Following
Castration in Human Prostatic Tissue in Vivo", The Prostate 54:212, 2003,
Staak presents his evaluation of the apoptotic effect of castration on
human PC cells transplanted under the renal capsules of mice. The baseline
cell death rate was .026%. After castration the apoptosis quickly reached
a maximal rate of 1.54% at 3 days, i.e. 60 X normal, and then gradually
fell back to baseline by 21 days. By this time 87% of the cell mass had
died! But unfortunately, some 13% of the cells remain viable ... the ones
that require our continuing ministrations!
Bottom Line:
The
preclinical history of a prostate cancer is much longer than generally
recognized.
«
Back to Article List |
