Background: Superficial bladder cancer (SBC) is a precursor of muscle‐invasive, potentially life‐threatening bladder cancer. Given the field cancerization effect and the risk of recurrence and progression, SBC appears the most suitable target for bladder cancer systemic chemoprevention. However, new risk biomarkers are demanded to select at‐risk subjects and to conduct efficient clinical chemoprevention trials.

Angiogenesis represents a key step for tumor progression and metastatic spread in solid tumors. Vascular Endothelial Growth Factor (VEGF) is the most known angiogenic factor and is involved in early stages of bladder carcinogenesis. VEGF expression, either as tissue or soluble form, has been consistently reported to be higher in SBC than in invasive bladder cancer.

Purpose: We analyzed the prognostic value of serum VEGF on progression and mortality in patients (pts) with Ta and T1 transitional cell carcinoma of the bladder after a median follow‐up time of 13.7 years.

Methods: We used clinical, DNA flow cytometry and serum biomarker data from a chemoprevention trial of the synthetic retinoid fenretinide in 99 pts with SBC that was initiated at the National Institute for Cancer Research in Genoa in 1993. Serum aliquots at baseline and after 12 months from randomization were available for 62 subjects, 29 allocated in the fenretinide and 33 in the control arm, respectively. Study endpoints were the change of VEGF levels after 1 year of treatment and the prognostic value of baseline VEGF on progression‐free survival (PFS), bladder cancer‐free survival (BCFS), overall survival (OS) and cancer‐free survival (CFS). Results were adjusted for age, sex, smoking habits, treatment, tumor stage, tumor grade, tumor history, DNA index and baseline VEGF levels. A Cox model for OS was used to test the prognostic effect of all factors on study.

Results: No significant difference on the change of VEGF levels were observed between the two treatment arms. Kaplan Meier curves showed that the highest quintile of VEGF (> 350 pg/mL) at baseline defined two groups with different prognosis: at a median follow‐up time of 13.7 years, 8 patients out of 12 (67%) died in the high VEGF group versus 11 of 50 (22%) in the low VEGF group (Log Rank p=0.0008); 6 of 12 (50%) and 7 of 50 (14%) died from any cancer type, respectively (p=0.002); 3 of 12 (25%) and 0 of 50 died from bladder cancer, respectively (p=0.0001).

This effect was observed also in PFS analysis: 3 pts out of 12 (25%) progressed to invasive cancer in the high VEGF group versus 2 of 50 (4%) in the low VEGF group (Log Rank p=0.007); in the Cox model, high VEGF levels (>350pg/mL) and smoking habit (ever vs. never) were the only significant risk factors for OS with hazard ratios of 4.9 (95% CI: 1.7–14.4) and 12.0 (95% CI: 1.5 to 97.4), respectively.

Conclusions: High serum VEGF levels is an independent risk factor for OS, BCS and PFS in our cohort of SBC pts. Given the relevant role of VEGF in early stages of bladder carcinogenesis, it may represent a biomarker of at‐risk subjects and an attractive target to be validated in future intervention trials for chemoprevention in pts with SBC.

Citation Information: Cancer Prev Res 2010;3(1 Suppl):A84.