Presentation Authors: Young Hwii Ko, Yeong Uk Kim*, Jae Young Choi, Phil Hyun Song, Daegu, Korea, Republic of, Kangsoo Shim, Andong, Korea, Republic of, Ki Hak Moon, Hee Chang Jung, Daegu, Korea, Republic of
Introduction: Approaches to ipsilateral ureters potentially aggravate tumor spread, as allegedly reported in ureteroscopy (URS) before radical nephroureterectomy (RNU). If URS truly exacerbates intravesical recurrence (IVR), retrograde pyelography (RGP), a less invasive alternative option for URS, might also give rise to IVR. We thus investigated the impact of preoperative RGP on IVR and attempted to reveal the mechanism involved.
Methods: Of 114 patients that underwent RNU for upper urinary tract urothelial carcinoma (UUT-UC) from January 2004 to June 2013, 72 patients without a history of bladder tumor and preoperative URS were selectively enrolled and retrospectively reviewed. Times between RGP and RNU were collected and the length of intra-ureteral tumors which was defined by the maximal dimension on the image were measured. The impacts of clinicopathological factors on initial IVR were analyzed using logistic regression models.
Results: During a mean follow up period of 64.5 months, 32 (44.4%) patients experienced IVR, which was identified at a mean (Â±SD) 22.3Â±18.8 months after RNU. RGP was underwent in 41 (56.1%) patients. Despite a shorter follow up period (31.3Â±23.1 vs. 48.3Â±18.6 months), the tumor recurrence rate was significantly higher in the RGP group (63.4%, 26/41) than in the non-RGP counterparts (19.4%; 6/31, p < .000). The follow variables differed significantly in the IVR and non-IVR groups; age (64.6Â±8.51 yrs vs. 59.6Â±9.65 yrs, p=0.005), lower tumor location (53.1% vs. 20%, p=0.032), invasive tumor (over pT2; 53.1% vs. 17.5%, p=0.011), lower preoperative hemoglobin (12.8Â±1.36 vs. 13.9Â±1.65, p=0.004), higher preoperative creatinine (1.29Â±0.318 vs. 1.11Â±0.221, p=0.009), and preoperative RGP (81.3% vs. 37.5%, p < .000). The time between RGP and RNU and the length of tumor were not associated with IVR (p=0.411 and 0.715). Multivariate analysis showed tumor stage (>pT2; Odds ratio [OR]=4.877, p=0.008) and preoperative RGP (OR=6.715, p=0.001) independently predicted IVR.
Conclusions: RGP before RNU was found to aggravate IVR, regardless of time between RGP and RNU or the size of tumor. These findings caution against the use of RGP as a diagnostic modality for UUT-UC and provide evidence supporting the notion that monoclonal tumor cell spread underlies the pathogenesis of IVR.