The predictive value of C-reactive protein for prognosis in patients with upper tract urothelial carcinoma treated with radical nephroureterectomy: a multi-institutional study.

Background: Few studies have discussed the prognostic impact of serum C-reactive protein (CRP) level in upper tract urothelial carcinoma (UTUC).

Objective: To investigate whether the perioperative level of CRP provides additional prognostic information following radical nephroureterectomy (RNU).

Design, Setting, And Participants: A total of 564 patients with UTUC from a retrospective multi-institutional cohort were included. The median follow-up was 32 mo.

Intervention: All patients underwent RNU without neoadjuvant chemotherapy, while 106 patients (18.8%) received adjuvant chemotherapy.

Outcome Measurements And Statistical Analysis: Associations between perioperative CRP level and outcome were assessed using multivariate analysis. A serum CRP level >0.50mg/dl was defined as elevated.

Results And Limitations: Preoperative CRP (pre-CRP) level was elevated in 136 patients (24.1%). Multivariate analysis showed that pre-CRP elevation was an independent predictor of subsequent disease recurrence (hazard ratio [HR]: 1.47 for CRP 0.51-2.00; HR: 1.89 for CRP >2.00). Five-year recurrence-free survival rates were 69.2% in patients with pre-CRP levels ≤ 0.50 mg/dl, 54.3% in patients with pre-CRP levels between 0.51 and 2.00 mg/dl, and 35.4% in patients with pre-CRP levels >2.00 mg/dl (p<0.001). Similar results were found in cancer-specific mortality, showing that pre-CRP elevation was an independent predictor of worse outcome (HR: 1.74 for CRP 0.51-2.00; HR: 2.31 for CRP >2.00). In a subgroup analysis of the elevated pre-CRP group, postoperative normalisation of CRP level was an independent predictor of better outcome. This study is limited by its retrospective nature as well as its heterogeneous group of patients and variable follow-up protocols resulting from the multi-institution design.

Conclusions: Serum CRP may become a possible biomarker in UTUC, suggesting that patients with an elevated pre-CRP level could be predicted to have subsequent disease recurrence and cancer-specific mortality, while postoperative normalisation of CRP level was an independent predictor for prognosis.