Page 58 - An Evidence Review of Active Surveillance in Men With Localized Prostate Cancer

prostate cancer patients with multiple primary cancers were consistently less likely to die of
prostate cancer compared to patients with no multiple primary cancers. No studies provided
information on comorbidities other than multiple primary cancers.

Race/ethnicity. Fifteen studies (12 SEER or component registries, 1 SEER-Medicare, 1
LAC/USC and 1 using data from the NCHS), covering 1969 to 2007 and reported information on
trends in prostate cancer mortality rates (10 studies) or prostate cancer-specific survival (7
studies; 2 studies reported both types of information) stratified by patient race/ethnicity. 16-
19,24,26,27,34,40,44,51,60,76,84,86 Eleven of the studies reported exclusively on non-Hispanic whites or
blacks whereas four reported on other racial/ethnic groups as well.
Overall, studies demonstrated an increase in the mortality rate from the 1980s to the early-
1990s, followed by a decrease from the mid-1990s to 2007 for all racial/ethnic groups. Notably,
the mortality rates among blacks were consistently higher compared to that of non-Hispanic
whites in all studies and across time periods.
Regarding prostate cancer specific-survival, all studies demonstrated improvements in
survival over time, for all racial/ethnic groups during their respective time periods. Six of the
seven studies reporting relevant information demonstrated that blacks were at higher risk for
prostate cancer death compared to non-Hispanic whites, although the difference between the two
groups appeared to decrease over time. One study found no significant difference in the
probability of prostate cancer death (versus nonprostate cancer death) between non-Hispanic
whites and blacks for the years 1988-95, after adjusting for multiple potential confounders
24
(including tumor grade and stage at diagnosis).
Tumor Characteristics
Stage. Six studies (all SEER based) reported on temporal trends in prostate cancer specific
mortality among patients diagnosed with prostate cancer stratified by tumor stage at
diagnosis. 18,24,40,78,81,86 Studies covered 1969 to 2001 and demonstrated that over time the
proportion of deaths due to prostate cancer among patients diagnosed with the disease has
decreased, particularly for patients with early-stage (localized or regional) disease at diagnosis.
One study, using the SEER database, demonstrated that the risk of death due to prostate cancer
based on tumor stage was persistent from 1988 to 1995, with risk of death more than five times
higher for patients diagnosed with distant disease and more than two times higher among those
24
diagnosed with regional disease, as compared to patients with localized disease.
We did not identify any studies reporting trends in population mortality rates stratified by
tumor stage.

Tumor grade. Three studies (all SEER based) including prostate cancers diagnosed between
1973 to 1997, reported information on trends in prostate cancer survival stratified by tumor grade
at diagnosis. 24,40,86 One of the studies demonstrated that the probability of dying from prostate
cancer among patients diagnosed with prostate cancer decreased during the study period (deaths
occurring during 1988-95). Although the decrease was observed for all cancer grades, it was
40
more pronounced among patients with well and moderately differentiated tumors. Another
study, again using data from the SEER database, compared the probability of death by prostate
cancer among patients diagnosed with the disease, stratified by tumor grade. The study
demonstrated that, compared to patients with well differentiated tumors, patients with
moderately differentiated cancers and poorly differentiated disease had a higher probability of

23