F. Di Silverio et al., Independent value of tumor size and DNA ploidy for the prediction of disease progression in patients with organ-confined renal cell carcinoma, CANCER, 88(4), 2000, pp. 835-843
BACKGROUND. Greater than 20% of patients with apparently localized renal ce
ll carcinoma (RCC) present with disease progression after surgery. The obje
ctive of the current study was to improve the ability of clinicians to pred
ict prognosis in patients with localized RCC.
METHODS, The authors studied 154 patients with organ-confined RCC classifie
d as pT1 to pT2-pN0-M0 who underwent radical nephrectomy. Follow-up ranged
from 24-128 months (median, 72 months). Several morphologic parameters of t
he tumor were considered. DNA content was analyzed by flow cytometry and tu
mor size was determined from the surgical specimen. A Cox proportional haza
rds regression model was used to identify significant independent prognosti
c factors for disease progression.
RESULTS. At 5 and 10 years of follow-up, disease free survival was found to
be 87% and 86%, respectively. Univariate analysis revealed that DNA conten
t, Furhman grade, and tumor size had a statistically significant predictive
value for disease progression, whereas, with regard to grade, the differen
ce was significant only between patients with Grade 3 tumors and all other
patients with Grade 1-2 tumors (P < 0.0001). Although DNA content was found
to correlate with tumor size (P < 0.0001), multivariate analysis showed th
at these were the only significant independent predictors of disease progre
ssion. The sum of DNA content and tumor size therefore was considered to di
stinguish separate risk groups. For a patient with diploid RCC, the risk of
progression increased from 4% if the tumor size was 3 cm to 43% if the tum
or size was 10 cm. For a patient with nondiploid RCC, this risk was 32% if
the tumor size was 3 cm, increasing to 99% for tumors measuring 10 cm.
CONCLUSIONS, Stratification of organ-confined RCC according to tumor size a
nd DNA content could possibly provide more information that could be useful
in the selection of individuals with significantly different risks of dise
ase progression. (C) 2000 American Cancer Society.