In our study, ERCC1 expression provided both prognostic and predictive information in patients with completely resected bladder cancer. Among patients with transitional cell carcinoma of the bladder treated with cystectomy, high tumoral expression of ERCC1 correlated with longer survival in patients without adjuvant chemotherapy and was associated with shorter survival in those with adjuvant chemotherapy. A statistically significant interaction between ERCC1 expression and adjuvant chemotherapy indicated potential benefits of adjuvant chemotherapy in patients with ERCC1-negative tumors.
To date, the role of adjuvant chemotherapy for bladder cancer has been controversial, with no Level 1 evidence supporting adjuvant chemotherapy. In fact, the available data have not demonstrated a clear benefit of adjuvant chemotherapy. Despite mounting evidence favoring neoadjuvant chemotherapy [1
], physicians are reluctant to adopt its practice as evidenced by only 1.2% of patients with stage III bladder cancer receiving neoadjuvant chemotherapy [10
]. Nonetheless, 10.4% of patients from the same cohort received adjuvant chemotherapy, implying a preference for adjuvant chemotherapy over neoadjuvant chemotherapy despite a paucity of evidence [10
]. Such ubiquitous practice might be attributed to two beliefs: that adjuvant chemotherapy could be given to patients under the most accurate pathologic staging, thereby preventing low-risk patients from unnecessary cytotoxicity; and that up-front surgery increases the chances of curing patients with drug-resistant diseases. In order to strengthen evidence-based practice for adjuvant chemotherapy, further research including rigorous study designs and methodologies are warranted.
A significant strength of this study was the appropriate selection of patients who could potentially benefit from adjuvant chemotherapy based on ERCC1 expression. The target populations in previous adjuvant trials were heterogeneous, ranging from T1-T2 disease to node positive disease, and defined only by pathologic stage. A recent prospective study used p53 expression as a molecular marker for selecting a target population for adjuvant chemotherapy [8
]. Patients whose tumors were p53-positive were randomly assigned to adjuvant chemotherapy or to observation, and those with p53-negative tumors were all assigned to observation. Although this trial failed to demonstrate the prognostic and predictive value of p53 due to a high patient refusal rate or lower than expected event rate, the attempt to use a molecular marker in adjuvant chemotherapy is noteworthy. Further molecularly targeted adjuvant chemotherapy should be investigated.
ERCC1 is a component of the nucleotide excision repair pathway, which is essential for the repair of DNA adducts induced by cisplatin-based therapy. Several studies have shown that high ERCC1 expression is a good prognostic factor in patients without cisplatin-based chemotherapy and also a predictor for poor clinical outcome in patients with cisplatin-based chemotherapy for various tumor types [12
]. ERCC1 was also previously evaluated in metastatic bladder cancer, and high ERCC1 mRNA or protein expression correlated with poor prognosis in patients treated with cisplatin-based chemotherapy [20
]. In muscle-invasive bladder cancer, cisplatin-based chemoradiation therapy showed better efficacy for ERCC1-negative tumors than ERCC1-positive tumors [23
]. Hoffmann et al. demonstrated that high ERCC1 gene expressions were associated with inferior progression-free survival after cisplatin-based adjuvant chemotherapy for locally advanced bladder cancer [24
]. In the present study, the different prognostic values of ERCC1 according to the history of adjuvant gemcitabine plus cisplatin chemotherapy were confirmed in patients with completely resected bladder cancer.
A limitation of our study includes different patient characteristics between two groups with or without adjuvant chemotherapy owing to the retrospective nature of the study. However, such difference and bias which can potentially influence study results were minimized because the value of ERCC1 expression was analyzed independently in each group. In addition, the adjuvant chemotherapy was homogenous with an identical regimen of gemcitabine plus cisplatin.
In summary, we have demonstrated that ERCC1 may potentially be a novel biomarker with clinical predictive and prognostic values in completely resected bladder cancer. Those who have bladder cancer with low ERCC1 expression are more likely to benefit from adjuvant gemcitabine plus cisplatin chemotherapy. Further researches including prospective randomized studies are warranted to confirm our findings.