Page 57 - An Evidence Review of Active Surveillance in Men With Localized Prostate Cancer

Tumor grade. Five studies (all using the SEER database or its component registries) covering
1973 to 2005 stratified prostate cancer incidence by tumor grade (level of differentiation or
Gleason score). 17,21,42,52,56 Studies generally indicated that the increase in prostate cancer
incidence observed from the mid-1980s to early-1990s was mainly due to an increase in the
incidence rate of moderately differentiated tumors (or tumors of Gleason score 5-7). A single
study (of SEER) analyzed data after 2000 and reported a continued increase in incidence rate of
tumors with Gleason score 5-7 from 1988 to 2005 and a concomitant decrease in the incidence
42
rate of tumors with Gleason score 2-4 (P < 0.001 for Gleason group-specific changes).

Trends in Prostate Cancer Mortality and Survival Rates
For the overall U.S. population, the NCI’s Cancer Trends Progress Report (2009/10)
indicates that after increasing from 1975 to 1991, prostate cancer death rates fell from 1994 to
a
2007. The baseline prostate cancer mortality in 1975 was 31 deaths per 100,000 men per year
and has declined to the current level of approximately 24 deaths per 100,000 men per year.
Among the studies we reviewed, 21 provided information on trends in prostate cancer
mortality or changes in survival rates of patients with prostate cancer, stratified by the factors
relevant to Key Question 1 (Appendix Table C1.3). Nine of the studies provided information
stratified by age, 17,18,24,28,30,33,39,40,86 15 by race/ethnicity, 16-19,24,26,27,34,40,44,51,60,76,84,86 one by
24
comorbidity status, six by tumor stage, 18,24,40,78,81,86 and three by tumor grade. 24,40,86 No studies
provided information stratified by any of the other factors relevant to Key Question 1. Of the
studies considered eligible, 19 utilized the SEER database, one the SEER-Medicare database and
th
th
one LAC/USC. Studies were generally large (median sample = 60,494; 25 -75 percentile
40,548-229,556), were published between 1990 and 2011, and covered years 1969 to 2007.
Patient Characteristics
Age. Nine studies (all SEER-based) covering 1969 to 2004, reported information on population
mortality rates (5 studies) or prostate cancer-specific survival among patients diagnosed with
prostate cancer survival (4 studies), stratified by age. 17,18,24,28,30,33,39,40,86 Studies of population
mortality rates demonstrated decreases in the mortality rate for all age groups between the early-
1990s and 1999. One study presented information for years after 2000 and showed that
population mortality continued to decline significantly for all age groups in recent years (up to
2004).
One study of prostate cancer-specific survival, using the SEER database, demonstrated that
over time (for deaths occurring from 1988 to 1995) the proportion of patients diagnosed with
prostate cancer who died of their cancer has decreased (i.e., patients with prostate cancer have
40
increasingly died of other causes) across all age groups considered (> 50 years old). Another
study, also using the SEER database, demonstrated that among patients with prostate cancer,
increasing age is associated with death from nonprostate cancer causes and that this effect held
24
true throughout the study period.

Comorbidity. A single study using the SEER database reported information on temporal trends
in prostate cancer mortality stratified by whether patients had been diagnosed with multiple
24
primary cancers. The study demonstrated that throughout the study period (1988 to 1995)

a Available at http://progressreport.cancer.gov/; last accessed September 30, 2011.

22