Page 63 - An Evidence Review of Active Surveillance in Men With Localized Prostate Cancer

states (and a concomitant decrease in the proportion of patients from Eastern or Southern states).
Because the centers participating in CaPSURE have not remained stable over time, changes in
these distributions may be difficult to interpret.
One study, comparing 1998 to 2002, reported an increase in the number of patients residing
in areas of higher median income. Again, because sampling strategies changed between the
POCS years (and different regions were included), patterns in the distribution of income are
77
difficult to interpret.
Finally, one study, using the NCDB, assessed trends in the distribution of patients by hospital
caseload, over time (1986-87 and 1992). There was little evidence of change over the time period
74
covered.

Trends in Treatment Patterns
Twenty one studies (6 SEER, 3 SEER-Medicare, 6 CaPSURE, 5 NCDB, 1 POCS) provided
information on treatment trends over time (Appendix Table C1.14). 4,27,36,41,47,48,58,66,67,69,71-
th
th
75,77,82,87,88,90,91 Studies were generally large (median sample size = 71,602; 25 -75 percentile
6290-142,340), published between 1994 and 2011, and covered 1973 to 2008. In 12 studies,
patients managed by observational management strategies of no active treatment (AS, WW or
expectant management) were considered in aggregate with patients receiving androgen
deprivation therapy (ADT), or it was unclear whether the data allowed the distinction between
these treatments. Most studies demonstrated decreasing trends in the proportion of patients being
managed with strategies other than surgery or radiotherapy throughout their respective time
periods; studies explicitly reporting on AS/WW-type strategies also indicated decreases in the
proportion of patients receiving such treatments. In all seven studies (6 CaPSURE, and 1 using
POCS data) providing information for years after 2000, the proportion of patients receiving
AS/WW was less than 10 percent; this also held true for subgroups of “low-risk disease”
(typically defined based on T stage, Gleason score and PSA criteria) investigated in two studies
(both using CaPSURE data).

Summary
We reviewed 79 studies based on large epidemiologic databases sourced from the U.S.
population. For all age or race/ethnicity groups investigated, the incidence rate appears to have
peaked in the early 1990s; subsequently, the incidence rate declined between the early 1990s and
1999. Studies consistently demonstrated that early-stage (localized and regional) prostate cancer
cases were responsible for the observed increase in prostate cancer incidence from the mid-1980s
up to the mid-1990s. Studies also demonstrated decreases in the prostate cancer-specific
mortality rate for all age groups between the early-1990s and 1999. Mean age of diagnosis has
decreased over time, both for blacks and whites. Another consistent trend over time has been the
decrease in low- and high-grade (corresponding to Gleason scores 2-4 and ≥7, respectively)
tumors, and a concomitant increase in intermediate grade tumors (corresponding to Gleason
scores 5-6). These trends, and their impact on grade-adjusted mortality statistics, need to be
interpreted cautiously in the light of changes in histopathologic grading practice (causing a
Gleason score shift toward higher values), and the widespread implementation of PSA screening
(which will tend to increase the detection of nonhigh grade tumors). Over time, patients
diagnosed with prostate cancer are less likely to die of the disease (i.e., they are more likely to
die of nonprostate cancer causes); this is particularly true for patients diagnosed at older age.
Most studies demonstrated a decrease over time of the proportion of patients being managed with

28