Abstract
PURPOSE: The accurate selection of patients who are most likely to benefit from neoadjuvant chemotherapy is an important challenge in oncology. Serum AGR has been found to be associated with oncological outcomes in various malignancies. We assessed the association of pre-therapy serum albumin-to-globulin ratio (AGR) with pathologic response and oncological outcomes in patients treated with neoadjuvant platin-based chemotherapy followed by radical nephroureterectomy (RNU) for clinically non-metastatic UTUC. METHODS: We retrospectively included all clinically non-metastatic patients from a multicentric database who had neoadjuvant platin-based chemotherapy and RNU for UTUC. After assessing the pretreatment AGR cut-off value, we found 1.42 to have the maximum Youden index value. The overall population was therefore divided into two AGR groups using this cut-off (low, < 1.42 vs high, ≥ 1.42). A logistic regression was performed to measure the association with pathologic response after NAC. Univariable and multivariable Cox regression analyses tested the association of AGR with OS and RFS. RESULTS: Of 172 patients, 58 (34%) patients had an AGR < 1.42. Median follow-up was 26 (IQR 11-56) months. In logistic regression, low AGR was not associated with pathologic response. On univariable analyses, pre-therapy serum AGR was neither associated with OS HR 1.15 (95% CI 0.77-1.74; p = 0.47) nor RFS HR 1.48 (95% CI 0.98-1.22; p = 0.06). These results remained true regardless of the response to NAC. CONCLUSION: Pre-therapy low serum AGR before NAC followed by RNU for clinically high-risk UTUC was not associated with pathological response or long-term oncological outcomes. Biomarkers that can complement clinical factors in UTUC are needed as clinical staging and risk stratification are still suboptimal leading to both over and under treatment despite the availability of effective therapies.