Impact of nephron sparing surgery on survival in upper urinary tract urothelial carcinoma patients stratied by tumor characteristics

Background: We aimed to evaluate the impact of nephron sparing surgery (NSS) on survival in upper urinary tract urothelial carcinoma (UTUC) patients stratied by tumor grade and stage compared to radical nephroureterectomy (RNU) or no surgery. Methods: Overall, 10319 UTUC patients between 2004 and 2015 were extracted from the SEER database. Patients were divided into no surgery group, NSS group and RNU group. Kaplan-Meier plots illustrated survival rates according to surgical method. Multivariable Cox regression analyses assessed the effect of different surgical methods on survival rates. Result: Among the 10319 UTUC patients. Kaplan-Meier plots showed that NSS was associated with a higher overall survival (OS) and cancer-specic survival (CSS) in grade (cid:0) and (cid:0) and AJCC stage (cid:0) patients (all p <0.05). In multivariable Cox regression analyses, NSS was prone to a higher OS or CSS in grade (cid:0) ( p > 0.05). Moreover, NSS predisposed to a higher OS only in AJCC stage (cid:0) ( p <0.05). Conclusion: The benecial effect of NSS was evident in grade (cid:0) or AJCC stage (cid:0) . NSS can be considered in grade (cid:0) or stage (cid:0) , for other grades or stages remains to be seen, both of which will be veried by further prospective research.


Background
Urothelial carcinomas are the fourth most common tumor, most of which are located in the lower urinary tract and upper tract urothelial carcinoma (UTUC) accounts for only 5% of urothelial tumors and 8% of renal tumors [1,2] . Radical nephroureterectomy (RNU) with bladder cuff excision is the gold standard treatment for UTUC. However, solitary kidney after RNU may lead to higher risk of dialysis, cardiovascular morbidity and overall mortality [3,4] .In an effort to decrease the potential risk, nephron sparing surgery deserves attention greatly in elaborately chosen patients with UTUC with the intention of obtaining acceptable oncological results [5][6][7] .
Although the stage of UTUC is usually determined by endoscopic biopsy during ureteroscopy, inadequate tissue sampling and concerns about ureteral perforation make accurate staging challenging [8,9] . Therefore, due to concerns of cancer progression, many clinicians are willing to perform radical eradication therapy even in patients with early-stage UTUC [10,11] . However, some scholars have suggested that NSS may play a curative role especially in low grades and early stages [12,13] . Nevertheless, evidencebased data are scarce to support the oncologic e cacy of NSS for its rarity.
Originally, NSS was exclusively considered for UTUC patients who are intolerance of RNU or those with imperative indications such as anatomically or functionally solitary kidneys. Recently, It is that elective risk-based indications were introduced for patients with bilateral functional kidneys when obtaining effective oncologic control for these emergency cases [14] . As a result, all the contemporary reports comparing NSS versus RNU have included a mix of patients with imperative and elective indications, and the available data are therefore highly heterogeneous [15] , which makes it di cult to draw any reasonable conclusions from the current literature. To clarify this issue, we evaluated and compared the treatment of NSS on overall survival (OS) and cancer-speci c survival (CSS) when compared to RNU or no surgery.  , 51-65, 66-80 and >80), sex (male and female), race (white, black and others), marital status (married, unmarried and unknown), urban-rural residence (metropolitan and nonmetropolitan), summary stage (localized, regional, distant), radiotherapy (yes or no) and chemotherapy (yes or no). Household income was divided into three group: lower group (<4219), middle group (4219-5210) and upper group (>5210). OS and CSS are the primary endpoints of our study.

Statistical analysis
Statistical Package for the Social Sciences software (version 24.0; SPSS Inc, Chicago, IL, USA) was used for all statistical analyses. The chi-square test was used to analyze factors related to the surgical method. Kaplan-Meier curves analyses using a log-rank test was used to analyze the OS and CSS rate.
Multivariate Cox regression was used to analyses factors associated with OS and CSS. P value≤0.05 was considered statistically signi cant.

General characteristics
A total of 10,319 eligible UTUC patients were identi ed in our study cohort between 2004 and 2015. Among them, 963 (9.3%) patients did not received surgery, 1,273 (12.3%) received NSS and 8,083 (78.3%) received RNU. Table 1 demonstrated the demographic and clinicopathological characteristics of patients with UTUC. Chi-square test showed some risks had an impact on therapeutic decision making.
Of the 10,319 patients, the proportion of NSS remained stable from 2004 to 2015, while the proportion of NSS decreased in pelvis patients ( Figure 1).

Survival related variables
Kaplan-Meier curves analysis the in uence of risk factors on OS and CSS of UTUC patients ( Multivariate cox regression analysis revealed some risk factors in UTUC patients (Table 3). After adjusting for risk factors, The results shown that compared with the patients who receive RNU treatment , no surgery patients (HR:3.78, 95% CI: 3.51-4.08, p < 0.001) may be at increased risk. In term of CSS, no surgery patients (HR: 2.46, 95% CI: 2.18-2.77, p < 0.001) associated more commonly with low-risk CSS, while the people performed with NSS (HR: 0.78, 95% CI: 0.67-0.91, p =0.002)may also associated with signi cantly greatest odds of CSS.

Subgroup analysis strati ed by AJCC stage and WHO grade
Based on summary stage and WHO grade, we further discussed the difference between surgical method and prognosis among the subgroups of patients (Table 4). We found that for OS and CSS, surgical method was still associated with them among different stage and various grade subgroups patients. Kaplan-Meier plots showed that NSS was associated with a higher overall survival (OS) and cancerspeci c survival (CSS) in grade and and AJCC stage patients (all p<0.05) ( Figure 2, Figure 3). In multivariable COX regression analyses, NSS patients had a higher OS rate in stage IV group (OS: HR: 0.74, 95% CI 0.60-0.90, p=0.003) and a higher OS or CSS rate in grade III group compared to RNU patients.
However, for other subgroup patients, the surgical method had no obvious effect in terms of OS ,as well as CSS (p > 0.05). When classi ed according to the primary organ, the related results are shown in the supplementary materials (Table S2-S3). Supplemental Figure 1-4 was performed to verify the NSS group was prone to a higher OS or CSS in the former periods.

Discussion
Our study was based on a large number of population database to quantitatively compare the impact of NSS on survival in patients with UTUC compared to RNU or no surgery. UTUC is a rare disease with an annual incidence of one to two cases per 100 000 individuals in Western countries. We explored the effect of surgical methods on OS and CSS in UTUC patients and discovery that surgical method was an independent prognostic factor. Moreover, the mean survival time was higher in NSS group than in RNU patients. Kaplan-Meier plots showed that NSS was associated with a higher OS and CSS in grade and and AJCC stage patients. In multivariable Cox regression analyses, NSS was prone to a higher OS or CSS in grade , and NSS predisposed to a higher OS only in AJCC stage .
In our research, the OS and CSS of NSS and RNU were not signi cant for low-grade or early-stage with the selected UTUC patients. This result is similar to the previous report. Jay Simhan et al [16] reported that patients with similar situation when treated with NSS have analogous cancer-speci c mortality (CSM) rates to those patients managed with RNU. Jeldres et al. [17] thought segmental ureterectomy (SU) may be operated to UTUC patients when it is technically feasible, especially when detected in T3 or even T4 lesions. Moreover , Giovanni Lughezzani [18] report a stage and grade migration at NU or SU towards more aggressive disease among surgically treated patients between 1983 and 2004. But the CSM rates are not higher than former that validates the two surgical ways as effective treatments for UTUC. However , In another research Chen GL [12] reported acceptable oncological outcomes were observed in 23 patients undergoing laser ablation with strict ureteroscopic surveillance. In another side, the use of NSS has become an acceptable alternative in select patients with non-high grade, early-stage UTUC who are at low risk of disease progression [19,20] , However, the risk of cancer progression or recurrence is estimated to be as high as 30% within 5 years [21] . A large part of the difference in these results is due to competing model factors for the selected patient sample as well as aspects of the patient pro le. In addition, radiotherapy is usually administered in symptomatic patients or patients unable to undergo radical surgery [22] , however, Li Xiaoying [23] concluded radiotherapy can reduce local recurrence rate in patients with radical nephroureterectomy. In our research, we also found radiotherapy may bene t in UTUC patients. It may result from the population with surgery and high grade occupy most of the population, and radiotherapy may further improve the survival rate of patients. Meanwhile, there are studies focusing on the role of chemotherapy on survival for UTUC patients, and the therapeutic effects differ for different patients. In Zhai Ting-Shuai's research [24] , The bene cial effect of chemotherapy on survival probability was evident in pT3/pT4 or patients with lymph node metastasis, this may be involved more population with distant metastasis and more people are may have strong will of survival, so they have to face more treatment including receive extra chemotherapy. In other side, we observed that NSS has higher a OS or CSS when compared to RNU methods especially in Grade III and AJCC stage IV. In a multi cental research, Kaag et al. [25] identi ed a reduction in mean estimated GFR by ≈24% after RNU in tumor patients. Thus, patients when operated with RNU have notable decreases in renal function and leading to chronic kidney disease [26] . Importantly, such a decline in renal function after radical surgery in patients with UTUC may affect eligibility in the event of disease progression [27] . We concerned more about NSS surgery because NSS could protect the nephron, and it could better maintain the life of the patient with solitary kidney, thereby prolonging the survival period. Previous studies examining the oncological e cacy in patients managed conservatively were limited by small sample sizes, lack of generalizability, and inclusion of patients with aggressive disease characteristics. Using national registry data, the present multicenter large-sample study might be more convincing.
There are limitations in our research. First, the SEER database is a retrospective dataset. Second, the patient's physical condition was unclear, patients with much complications might pursue more conservative treatment. Third, patients receiving NSS may have a selection bias relative to RNU. In addition, for chemotherapy and radiotherapy, we didn't distinguish whether adjuvant or neoadjuvant therapy, and there is no information on speci c RT technique (dose, fractionation, beam energy) or chemotherapy regimens.

Conclusions
To conclude, we demonstrated that the bene cial effect of NSS was evident in grade or AJCC stage . NSS can be considered in grade or stage , for other grades or stages remains to be seen, both of which will be veri ed by further prospective research. For the institutional cohorts, data were extracted from the Surveillance, Epidemiology, and End Results database. This article does not contain any studies with human participants performed by any of the authors. For this type of study, formal consent is not required.

Consent for publication
Not applicable.

Availability of data and material
The datasets used and/or analyzed during this current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests. Authors' contributions L J contributed to study designation, data analysis, methodology, data validation and original draft writing. TS Z contributed to data collection, manuscript preparation and edition. Z Z contributed to data collection, formal analysis, investigation and visualization. X L contributed to formal analysis, methodology and validation. JX Z contributed to formal analysis and methodology. XC N contributed to formal analysis and methodology. HQ L contributed to data collection. XD Y contributed to conceptualization and supervision. LM F contributed to conceptualization, supervision and editing. L Y contributed to conceptualization, funding acquisition, resources, supervision and writing-review and editing. All authors have read and approved the manuscript.