Does preoperative percutaneous nephrostomy insertion worsen upper-tract urothelial cancer oncological outcome? A retrospective single center study

Background Physicians doubt percutaneous nephrostomy (PCN) insertion on cancer related hydronephrosis patients causes tumor seeding and worse cancer control. In this article, we attempted to determine if preoperative PCN alters cancer control in upper tract urothelial cancer (UTUC) patients. Methods Retrospective analysis of UTUC patients in a single center from 2005 to 2015. Exclusion criteria included lymph node metastasis, and patients underwent perioperative adjuvant chemotherapy or radiotherapy. There were 664 patients in this analysis, with clinico-pathological data being collected retrospectively for Cox-regression statistical analysis. Outcomes were measured by local recurrence, distant metastasis and cancer-specific death with Kaplan-Meier curves. Results There were respectively 25 and 639 UTUC cancers in the preoperative PCN and non-PCN insertion groups with mean follow-up duration of 37.9 and 48.6 months, respectively. The preoperative PCN group consisted of 17 patients (68%) with tumor located in the ureter, while the PCN-negative group included 236 patients (36%) with tumor located in the ureter being statistically significant. These two groups were comparable in gender, age, follow-up duration, tumor stage, and pathological features of the UTUC. As for the cancer control in the PCN group, 4(16%), 1(4%) and 1(4%) had local recurrence, distant metastasis and cancer-specific death respectively; in the non-PCN group, 101(15.8%), 96(15%) and 72(11.2%) exhibited local recurrence, distant metastasis and cancer-specific death respectively. Statistical analysis showed no difference in oncologic outcomes between these two groups.(p = 0.804, 0.201 and 0.254). Conclusions Preoperative percutaneous nephrostomy on upper-tract urothelial cancer poses little risk on tumor seeding and could be considered as part of treatment strategy if renal function preservation is needed.


Background
Hydronephrosis with compromised renal function is one manifestation of upper tract urothelial cancer [1]. Before definitive upper tract urothelial cancer diagnosis, percutaneous nephrostomy (PCN) may be undertaken for severe infective hydronephrosis or impending renal failure for temporary disease control. Although tumor seeding and invasion through the percutaneous nephrostomy tract has been reported on case reports [2,3], there are no studies comparing locally or systemically driven cancer recurrence rates between preoperative PCN dwelling and non-PCN dwelling patients in preoperative uppertract urothelial cancer patients. This study is the first study to analyze these two groups' of patients in the best of our knowledge. of deteriorating renal function or medical failure infective hydronephrosis with information of possible malignant tumor seeding through the PCN tract to patients. Interventional radiologist performed whole PCN group patients with Fr 10. nephrostomy tube or pigtail. Radical nephroureterectomy procedure cases were enrolled in the study for definite UTUC treatment. Peri-operative positive lymph nodes and distant organ metastasis cases were excluded due to poor prognosis and possible need of further adjuvant treatment; besides, neo-adjuvant or adjuvant candidates were also excluded for comparability of the disease course. Patient basic characteristics and pathological features were collected, such as age, gender, follow up duration, pathological stage, tumor figuration, tumor site, lymphovascular invasion, carcinoma in situ, multifocal, tumor necrosis, smoking and end stage renal disease. The follow-up protocol after nephroureterecotmy was as follows: 1.Computed tomography (CT) abdomen scanning every 3-6 months in the first 2 years and annually thereafter. 2. Cystoscopy every 3 months until 2 years shifted to every 6 months if negative finding. 3. Physical examination over the surgical wound and PCN tract during outpatient clinics visits, and CT scanning if suspicion of disease recurrence. Outcome measurements were classified as bladder recurrence, local recurrence, and distant metastasis. Follow-up cystoscopy showed merely bladder recurrence residual bladder cuffing recurrence was deemed as bladder recurrence. In PCN group, local recurrence was defined as disease recurrence over ipsilateral retroperitoneal space or PCN sites. In PCN negative group, local recurrence was defined as disease over the ipsilateral retroperitoneal space. Distant metastasis was defined as regional/distant lymph nodes metastasis or organ metastasis. We used Coxregression statistical analysis on comparability and significance of tumor recurrence between PCN (+) and PCN (−) groups. Kaplan-Meier survival curve was applied for time to recurrence between these two groups.

Results
In total, 25 patients in the PCN(+) group and 639 patients in the PCN(−) group had mean follow up duration of 37.9 and 48.6 months, respectively. The percentage of ureter tumor in the PCN(+) group(68%) was higher than the PCN(−) group(36%) with significant difference(p = 0.002). The median PCN dwelling time was 18 days [IOR:  Other basic characteristics such as gender, age, tumor stage, tumor grade, lymphovascular invasion, were collected with no significant difference between these two groups ( Table 1.) As for the cancer control in the PCN group, 4(16%), 1(4%) and 1(4%) had local recurrence, distant metastasis and cancer specific death respectively; in the non-PCN group, 101(15.8%), 96(15%) and 72(11.2%) had local recurrence, distant metastasis and cancer specific death respectively. Statistical analysis showed no cancer control difference between these two groups.(p = 0.804, 0.201 and 0.254)( Table 1.) In the PCN group, no patients were found to have skin metastasis or PCN tract tumor invasion and the local recurrence pattern was over the ipsilateral retroperitoneal space, which is similar to PCN negative group. Besides, time to being local-recurrence free and time to being distant metastasis free in Kaplan-Meier survival curve showed no difference between these two groups.(p = 0.804 & 0.201) (Fig 1. & Fig 2.)

Discussion
Tumor seeding after the percutaneous nephrostomy insertion has been reported in case reports but no study has evaluated if PCN procedure alter the risk of tumor seeding or cancer control [2][3][4]. In our study, the demographic and clinic-pathological data were similar between these two groups except for distribution of tumor location (17(68%) and 236(52%) ureter tumors in the PCN and non-PCN groups respectively, p = 0.002). It is reasonable that ureteral tumors are often associated with more event of hydronephrosis and sometimes external drainage was sometimes inevitable. Besides, relieving the post-renal obstruction may be considered earlier than urothelial cancer management when presenting as severe renal function impairment. After standard radical nephroureterectomy, comparable local   Table 2 Summarized the previous case reports of tumor seeding through the PCN tract [3][4][5][6][7][8][9]. Synchronous or metachronous urothelial cancer in upper urinary tract and bladder were the risk factors for tumor seeding [3][4][5][6][7][8][9]. Percutaneous endoscopic resection of the renal pelvis tumor seems another circumstance for tumor seeding [3-5, 8, 9]. However, Goel et al. reported long-term oncological outcomes of upper tract urothelial cancer status post percutaneous endoscopic resection, and concluded risk of tract seeding was low on low grade tumor [10]. Palou et al. also found percutaneous resection of superficial UTUC with good long term cancer outcome [11]. Roupret et al. retrospectively collected 24 patients receiving the percutaneous endoscopic resection of the tumor and reported 5-year disease-specific and tumor-free survival rates were 79.5 and 68%. There was none track seeding in the study as well [12]. Serrano et al. reviewed articles in Medline regarding the cancer disseminating in the PCN tracts and concluded that risk of tumor seeding is very low, and there were still other explanation rather than PCN tract manipulation, like blood stream or lymphatic metastasis [13]. Through the PCN tract, percutaneous resection of upper tract urothelial cancer was an alternative option for preservation renal function with comparable long-term cancer control in low-grade and limited tumor size patients. Meanwhile, risk for the tumor seeding through the PCN tract should be informed but downplayed [10-12, 14, 15]. In addition, Kiss et al. mentioned that internal drainage within urinary tract may increase the risk or urothelial recurrence [16]. According to the reviewed literature, the external drainage carried low risk of recurrence and should be considered for relieving urinary obstruction.
In our study, 25 preoperative PCN patients were initially performed for relieving urinary obstruction and clinical diagnosis instead of percutaneous endoscopic manipulation; hence, the risk of tumor seeding and altered oncological outcomes was minimal after standard nephroureterectomy. In addition, the evidences of neoadjuvant chemotherapy are getting more and more and the oncologic result is inspiring [17,18]; however, impaired renal function is relatively contraindicated for cisplatin-based chemotherapy and such situation is not uncommon in clinical practice especially for UTUC patients [19,20]. The PCN drainage may preserve residual renal function for UTUC patients with hydronephrosis and can be considered as a treatment strategy before neoadjuvant chemotherapy.

Limitation
The article is retrospective with relatively small-sized populations with short PCN dwelling time(Median: 18 days, IQR: 5-26). The prognostic effect of tumor site can not be fully assessed due to such small cohort by

Conclusion
Preoperative percutaneous nephrostomy on upper-tract urothelial cancer poses little risk on tumor seeding and could be considered as part of treatment strategy if renal function preservation is needed.