Page 60 - An Evidence Review of Active Surveillance in Men With Localized Prostate Cancer

supported a trend toward younger age at diagnosis (the effect was significant in four of the six
studies reporting results of statistical analyses).

Comorbidity. Only two studies (1 CaPSURE, 1 POCS), covering 1997 to 2003, reported trends
in the number of comorbidities present at the time of diagnosis of prostate cancer (Appendix
Table C1.5). 71,77 The CaPSURE database analysis grouped individuals into three groups: those
with no comorbidities, those with one or two comorbidities, and those with three or more
71
comorbidities. The study found no statistically significant difference in the distribution of
patients in these groups, when comparing 1997-99 versus 2000-03. The POCS analysis grouped
individuals into two groups (those with no comorbidity and those with one or more
77
comorbidities) and compared the frequency of each group across two years (1998 versus 2002).
The study concluded that the proportion of patients with no comorbidity has increased over time
(from 78.3 percent to 87.4 percent; P < 0.01).

Race/ethnicity. Eighteen studies (9 SEER, 3 SEER-Medicare, 2 CaPSURE, 2 NCDB, 2 other
databases) covering 1973 to 2003, reported information on trends in the racial/ethnic distribution
of patients with prostate cancer (Appendix Table C1.6). 31,37,46,49,52,53,58,61-63,65,71,74,75,77,80,86,92
Seven of the studies analyzed only whites and blacks; the remaining 11 studies considered
additional racial/ethnic groups. Generally, there was no consistent pattern in the racial or ethnic
distribution of cases over time: some studies indicated that the number of whites increased over
time, others that it remained stable, and others that it decreased. Studies using the same database
often provided discrepant results even for overlapping time periods; thus, no clear conclusion can
be reached. These discrepancies may be due to differences in the selection criteria employed in
each study.

Tumor Characteristics
Stage. Twenty two studies (8 SEER, 2 SEER-Medicare, 6 CaPSURE, 5 NCDB, 1 LAC/USC)
covering 1973 to 2007, reported information on trends in the distribution of prostate cancer stage
at diagnosis (Appendix Table C1.7). 4,16,23,31,37,46,52,56,62,63,65,67,68,71-76,86,90,91 Fifteen of the studies
reported information by grouping cases based on information on tumor size, lymph node status,
and the presence of distant disease (e.g., by grouping patients into localized, regional, and distant
disease stage or by using the American Joint Committee on Cancer staging classification). Six
studies reported information on the distribution of T stage groups only (2 for T1-T4, 1 for T1-T3
and 2 for T1-T2a) and one study reported information only on lymph node status.
Studies reporting on cancer stage consistently demonstrated decreases in the proportion of
patients presenting distant disease and concomitant increases in the proportion of patients with
localized or regional disease, over their respective time periods. All of the studies reporting
information on the distribution of T stage used the CaPSURE data. The two studies reporting on
T1-T4 tumors 4,65 and the two studies reporting on T1-T3 tumor 71,90 s consistently demonstrated
reductions in the proportion of patients presenting with higher T stages (i.e., a shift towards
increasing proportion of patients with T1/2 tumors, even among patients classified as “high
risk”). The two studies reporting on T1/T2 tumors both demonstrated a decrease of T1a/T1b
tumors and T2a tumors and an increase in T1c tumors. 67,68
The study reporting on lymph node status used the SEER database and suggested that the
23
proportion of patients with positive lymph nodes decreased during the study period (1988-96).

25