Dear All,
Somebody brought up a topic a few days ago which may have very
significant implications for a lot of people, so I think it deserves
its own post. I am referring to a msg which referenced an abstract
from Urology Times about Gleason scores trending upward over the
years. The article referred to a study by the NCI in which researchers
took slides of diagnostic prostate tissue from former patients and
examined how they had been classified originally (between 1990-92),
and then gave them to pathologists to grade in 2002-2004 The study
showed that the average gleason score had gone from a 5.9 to a 6.8, a
difference of .85, almost a whole percentage point.
This is what some researchers call "grade migration" or "grade
inflation". In other words, the biological composition of the cancers
that are being examined has not changed, but the ratings have. This
phenomenon is real, and it has implications in a number of areas, most
relating to newly diagnosed patients. It may concern:
* your long-term prognosis, possibly
* the way you and your doctor evaluate treatment decisions
* your decision whether to have treatment at all
* the information from the studies you and your doctor rely on
I got my info from an article in the Journal of the National Cancer
Institute, which originally reported this study. ("Prostate Cancer
and the Will Rogers Phenomenon", Albertsen, et al, Journal of NCI,
Vol. 97, No. 17, 9/7/05.)
http://www.ncbi.nlm.nih.gov/entrez/q...&dopt=abstract
This paper outlines the basic facts. But what I would like to
communicate to you is info from an editorial in that same issue of the
NCI Journal about the subject discussed. (Editorial, "Stage
Migrations and Grade Inflation in Prostate Cancer: Will Rogers Meets
Garrrison Keillor", Thompson, et al, JNCI, 2005...)
http://jnci.oxfordjournals.org/cgi/c...ull/97/17/1236
Here are some conclusions that from the NCI study:
1) A subset of people, mostly those currently diagnosed with G6 or
(to a lesser extent) G7-grade tumors, will have a more optimistic long-
term prognosis when using older nomograms, such as the Partin Tables.
2) Because "grade inflation" has occurred, the clinical outcomes of
treatments reported for Pca patients today may be misleading. For
example: a tumor that 30 years ago was rated a 5 is now most likely
rated a 6 (Grade 5's almost never appear in clinical practice today,
acc. to the authors), *in spite of the tumor being histologically
equal* (fancy term for the tumors being the same), this will distort
the picture, because there will be an equal long-term survival rate
for a G6 and a G5. Therefore, it would appear that outcomes are
improving. However, this is an artificial result.
3) The good news is that survival rates for *all tumor grades* have
increased. I am looking at a chart of these outcomes, and I see a
significant increase in the 8's and a dramatic increase in long-term
survival for the 10's.
Where I am going with all this is that Gleason grade inflation and
it's consequences have led to an "insidious" overdiagnosis and
overtreatment of patients, according to the authors of the JNCI
editorial. I will let the experts do the talking. Here are the
excerpts (my comments in brackets).
[Artificially High Outcomes for Recent Tx]
Why does grade inflation in prostate cancer matter? There are two
reasons. First, although this methodology is utterly invalid,
investigators frequently compare results of treatments between two
series of patients, whether examining improvements over time with a
specific therapeutic modality or making comparisons between
modalities.
**If assignment of Gleason score is changing over time, then, given
that tumor grade has an overwhelming impact on outcome, it is obvious
that any more *recent* treatment will demonstrate improved outcomes
due simply to grade inflation. Thus, if there are any differences in
*years of diagnosis* of patients, such a comparison will be fatally
flawed.**
[Overdetection and Overtreatment]
..**We are more concerned that grade inflation is a component of the
more insidious phenomena of overdetection and overtreatment of
prostate cancer.**
Currently, about 50% of men in the United States have a prostate-
specific antigen (PSA) test annually, and about 75% of men have had a
PSA test. [From NCI report: Since the arrival of the PSA test, "the
reported incidence of low-grade cancers has declined. And even though
most men now present with localized disease, their tumors are rarely
graded < 6 ". So tumor grades are rising even though people are
being diagnosed (much) earlier.]
*Despite a 3%-4% lifetime risk of prostate cancer death, more than 17%
of men in the United States will be diagnosed with prostate cancer
during their lifetime.*
By contrast, the lifetime risk of being diagnosed with prostate cancer
in the 1970s was about 10%.
What has fueled this dramatic increase in diagnosis?
[Are tumors significant?]
Certainly, since 1985, the primary impetus has been PSA testing.
Previously it was axiomatic that, whereas autopsy studies showed high
rates of prostate cancer in men who died of other causes, prostate
biopsy simply did not detect these small tumors.
**With the publication of the results of the PCPT study ("Prevalence
of Pca Among Men with PSA <4", NEJM, 2004, abstract is free), ... most
authorities recognized that clinically insignificant cancers are
indeed found with prostate biopsy **(10).
[90% of Men with Organ-Confined Pca Get Treatment]
** One large cohort study has found that more than 90% of men with
organ-confined prostate cancer currently opt for treatment (11).
[**5 out of 6 men may not need treatment**]
* With growing data that as many as five of every six men diagnosed
with prostate cancer (i.e., a 3% risk of death but a more than 17%
risk of diagnosis) may not need treatment and the evidence that
treatment adversely affects quality of life, why is it that so many
men opt for treatment?*
*One reason may be our risk-averse society. (We put labels on to-go
coffee cups that say "don't spill this on you-this beverage is hot.")
[Watchful Waiting Comparing Historical Gleason Scores is Inaccurate?]
* Another reason, however, may be the application of outcomes of
watchful waiting for prostate cancers of decades ago to a patient's
tumor today with its current Gleason score.
For example, a common reference in counseling patients is a previous
report from Albertsen (12). In that study, 767 men diagnosed with
prostate cancer between 1971 and 1984 were watched without treatment.
The primary determinants of risk of cancer death were age and Gleason
score, with risk of cancer death for a Gleason 2-4 tumor being 4%-7%
at 15 years compared with 42%-70% for a Gleason 7 tumor (12).
[I'm not sure what they mean. Are the old tumor grades still
meaningful? Practically speaking, the 2-5 category does not exist
today. The actual average increase over time for *all* tumor grades
has been .85; they have increased by almost a percentage point.
However, "this does not" apply across the board; mostly it affects the
lower-grade ca's))]
Similar conclusions have been reached in other series (13)
*The application of these historic outcomes to today's patient almost
certainly leads to a greater propensity for active treatment in lieu
of surveillance.*
[Unreliable Info?]
* Albertsen et al. have successfully drawn our attention to the
complexity of interpreting prostate cancer outcome data with their
demonstration that comparisons of outcomes among different groups of
patients receiving different treatments at different institutions and
over different periods will be fraught with very serious errors and
are fundamentally unreliable.
[Scarcity of Information for Patients Dx]
What is the answer to this morass of data that is faced by a quarter
of a million men in the United States annually? The mundane question
of "Doctor, what is the best treatment for my cancer?" (a question
that will not go away in the next 20 years) will be answered only
through clinical trials.
[Improved Outcomes for RP Px]
*One of these trials was recently completed; it demonstrated improved
survival and reduced risk of prostate cancer death with radical
prostatectomy compared with surveillance *(14).
These benefits must be balanced against changes in urinary and sexual
function with treatment
**and the recognition that 10-year data show that the number of
individuals needed to treat to prevent one adverse outcome ranges from
4 to 20.
[If I understand this correctly, it is mindblowing! It means that to
save one man's life (prevent adverse outcome), 4-20 men will be
needlessly treated.]
[Most Studies Focus on *High-Risk* Subjects]
Other trials are ongoing, but many more are needed because most have
focused on high-risk subjects-an important group-but one that
represents the **minority of patients currently diagnosed in the
United States.
*One key focus must be the development of large studies to track men
over time, merging pathologic and clinical data with
[The Future: Biomarkers to Pinpoint Pts Who Will Benefit from
Treatment]
What they really need to focus on is identifying ** biomarkers of
tumor aggressiveness to ultimately predict which tumor requires
treatment and which treatment is optimal for a specific tumor. **
Will Rogers mused about the effects of Okies migrating to California
on the average intelligence in both states. Garrison Keillor reflected
on Lake Wobegon, "where all of the women are strong, the men are good
looking, and the children are above average." Against the backdrop of
the growing challenges of prostate cancer, perhaps our current
assessment might be,
***"That's where we are in the United States today, where all the
biopsies are necessary and all cancers require treatment, as all have
Gleason scores above 5."***
Good Luck to you All.
Leah