Background. In the case of prostate carcinoma, radiation therapy is a
locally applied treatment modality in a malignancy known for systemic
dissemination. Because significant efforts and resources currently are
being consumed to improve local tumor control, failure patterns and p
otential curative gain deserve appropriate assessment. Methods. From 1
975-1989, 647 patients with clinically localized prostate carcinoma we
re definitively irradiated for biopsy-proven adenocarcinoma of the pro
state. Failure patterns were examined, and survival advantage based on
improvement in either local or distant disease control was calculated
. Distant metastatic rate and cause-specific survival analyses were us
ed as parameters by which to compare the outcome for patients in whom
local tumor control was achieved with patients who experienced local f
ailure, thereby assessing further the importance of the effectiveness
of locally applied therapy. Results. Three hundred ninety-two (61%) pa
tients at the time of this writing were clinically disease free. Sixty
-two (10%) patients failed locally only, 133 (20%) distantly only, and
60 (9%) developed local and distant recurrent disease. Both local and
distant failure rates were higher in patients with more advanced stag
e lesions at presentation, and distant failure rates significantly inc
reased in patients with less differentiated tumors. Pretreatment prost
ate-specific antigen was found to be useful in predicting recurrence p
atterns. Overall, there appeared to be more potential for improvement
in survival secondary to reducing distant metastasis. The distant surv
ival advantage (DSA) of reducing distant metastases, compared with the
local survival advantage (LSA) of improving local control, was 26 ver
sus 14%. Although DSA was greater than LSA within each stage category,
the potential to improve survival was most significant in the Stage C
group, where DSA was 35% and LSA 16%. Although LSA varied little acco
rding to tumor grade, DSA was dependent on tumor grade and varied from
13% for well differentiated lesions to 38% for poorly differentiated
lesions. Distant failure free survival at 10 years was 63% for patient
s with local control and 45% for those with local failure (P = 0.01).
Similarly, 10-year cause-specific survival was 75% in locally controll
ed patients compared with 48% for those With local recurrence (P < 0.0
01). Conclusions. Although better local tumor control should translate
into at least modest survival gain for patients with prostate carcino
ma, additional advantage may be seen with improved systemic therapy or
perhaps earlier diagnosis to reduce further the distant metastasis ra
te.