Partial and Radical Nephrectomy Provides Equivalent Oncologic Outcomes in pT3a Renal Cell Carcinoma: A Population-Based Study

Purpose To compare the cause-specific survival (CSS) and overall survival (OS) of patients with localized T3a renal cell carcinoma (RCC) after partial nephrectomy (PN) or radical nephrectomy (RN). Methods We obtained the demographic and clinicopathological data of 7,127 patients with localized T3a RCC and who underwent PN or RN from the Surveillance, Epidemiology, and End Results (SEER) database. These patients were divided into fat invasion cohort and venous invasion cohort for subsequent analysis. Kaplan–Meier analysis (KMA) and univariate and multivariate Cox proportional hazards regression analyses were used to evaluate the effects of PN or RN on OS and CSS. Meanwhile, 65 cases with clinical T1 (cT1) RCC upstaged to pathological T3a (pT3a) who were treated in Tongji Hospital (TJH) from 2011 to 2020 and underwent PN or RN were identified. Results In the study cohort, 2,085 (29.3%) patients died during the 1–172 months’ follow-up, of whom 1,155 (16.2%) died of RCC. In the two cohorts of fat invasion and venous invasion, KMA indicated that the PN group had favorable survival (p < 0.001). However, after propensity score matching (PSM), univariate and multivariate Cox regression analyses showed that the PN and RN groups had comparable CSS in the fat invasion cohort (p = 0.075) and the venous invasion cohort (p = 0.190). During 1–104 months of follow-up, 9 cases in the Tongji cohort had disease recurrence. There was no significant difference in recurrence-free survival between the RN group and the PN group (p = 0.170). Conclusions Our analysis showed that after balancing these factors, patients with localized pT3a RCC receiving PN or RN can achieve comparable oncologic outcomes. PN is safe for selected T3a patients.


INTRODUCTION
Renal cell carcinoma (RCC) accounts for 2%-3% of all adult malignancies (1). In recent years, with the wide applications of imaging examinations, the proportion of early-staged RCC has gradually increased (2), and partial nephrectomy (PN) has therefore played a more important role in the treatment of RCC (3,4).
PN is currently the standard treatment for T1 RCC, which provides similar oncologic control to radical nephrectomy (RN) while reducing the loss of renal function (5,6). The current guidelines for PN recommendations are limited to T1 and selected T2 RCC (combined with solitary kidney or chronic kidney disease if technically feasible) (5,7,8). However, in clinical practice, patients with clinical T1 (cT1) upstaged to pathological T3a (pT3a) RCC underwent PN did not show unfavorable cause-specific survival (CSS) and recurrence-free survival (RFS) than did those who received RN, which led us to think about the safety of PN for selected T3a patients (9,10).
In the past 10 years, the application of PN in T3a RCC patients has been explored, more and more evidences show that PN is safe and feasible for some T3a cases (10)(11)(12)(13), and there are also dissenting voices suggesting that PN is associated with poor oncologic outcome (14). Most of these studies are small-volume and retrospective. To this end, we selected the Surveillance, Epidemiology, and End Results (SEER) database to compare the performance of PN and RN in T3a RCC patients and used propensity score matching (PSM) to control bias; finally, we attached data of cT1 RCC patients upstaged to pT3a from our institution to add new evidence to this controversy.

PATIENTS AND METHODS
SEER is a population-based cancer database found by the National Cancer Institute that collects data on morbidity, treatment, and mortality (15). We obtained data of RCC patients diagnosed from 2004 to 2015 from 18 registries of the SEER database. The process of screening patients was shown below (Figure 1).
The inclusion criteria of the study were as follows: malignant pathological diagnosis; age 15 years and above; histologic type is renal cancer (including clear cell RCC (CCRCC), RCC not otherwise specified (NOS), papillary RCC (PRCC), acquired cystic disease-associated RCC/tubulocystic RCC, chromophobe RCC (CHRCC), clear cell PRCC, collecting duct carcinoma, hereditary leiomyomatosis and RCC-associated RCC, MiT family translocation RCC, mucinous tubular and spindle cell carcinoma, and renal medullary carcinoma); staged T3a; no lymph nodes and distant metastasis; no bilateral or other tumors; the surgical method was PN or RN; follow-up longer than 1 month; and knowable survival status.
The histologic types were divided into CCRCC, PRCC, CHRCC, and Others. According to the tissues of tumor invasion, the study cohort was divided into the fat invasion and venous invasion cohorts. The demographic and clinicopathological data of the RN and PN groups in the two cohorts were analyzed separately. The continuous variables were analyzed by Student's t-test, and the categorical variables were analyzed by Fisher's exact test. Then, a 1:1 PSM was performed between the surgical methods (PN vs. RN); all demographic and clinicopathological features were used as calipers (half of the PN cases in the fat invasion cohort were randomly rejected because of lack of enough RN cases to match).
The Kaplan-Meier analysis (KMA) was used to evaluate the survival difference between the RN and PN groups in the venous invasion cohort and the fat invasion cohort, respectively. Then univariate and multivariate Cox regression analyses were used to analyze the effect of each factor on survival. The variables of multivariate Cox regression were derived from the significant prognostic factors of univariate Cox regression.  Finally, after institutional review board approval, a retrospective review of all patients with upstaged pT3a RCC managed with PN or RN (2010.1.1-2020.12.1) in the Department of Urology of Tongji Hospital of Tongji Medical College, Huazhong University of Science and Technology (HUST) was performed (n = 65). KMA and univariate and multivariate Cox regression analyses were also performed in this cohort. All the above analyses were performed by R version 4.0.3 (Institute for Statistics and Mathematics, Vienna, Austria; www.r-project.org), and p < 0.05 was considered statistically significant.

RESULTS
After the screening procedure, 7,127 T3aN0M0 RCC cases ( .00] mm, p < 0.001). There were also significant differences in histologic types and nuclear grade structure between the two groups (both p < 0.001). The proportion of CCRCC and high-grade RCC in the RN group was higher. The median follow-up time was 56 months (1-172 months); 2,085 cases (29.3%) were confirmed dead at the last follow-up, of whom 1,205 cases (60.4%) died of RCC; the 1-, 3-, and 5-year survival rates were 95.1%, 83.4%, and 73.8%, respectively.
The study cohort was divided into 3,949 cases of fat invasion (including 986 cases of PN and 2,963 cases of RN) and 3,178 cases of venous invasion (including 251 cases of PN and 2,927 cases of RN). The demographics and clinicopathological data of the RN group and the PN group in the two cohorts showed significant differences (Tables 2, 3). KMA showed that the overall survival (OS) and CSS of the PN groups were favorable than those of the RN groups in the two cohorts (both p < 0.001) ( Figure 2). After PSM, there was no significant difference in the characteristics between the PN and RN groups in the two cohorts ( Tables 2, 3). There was no significant difference in OS (p = 0.068) and CSS (p = 0.190) between the PN and RN groups in the venous invasion cohort. The OS in the fat invasion group was favorable in the PN group (p = 0.036), while CSS was comparable in the two groups (p = 0.075) ( Figure 3).
Univariate and multivariate regression analyses showed that the risk factors of OS in the fat invasion cohort were higher age (hazard ratio (HR) 1.05, p < 0.001), greater tumor diameter (HR 1.01, p = 0.007), surgical method was RN (HR 1.35, p = 0.040), and earlier year of diagnosis (HR 1.10, p = 0.032); nuclear grade is also an independent prognostic factor (p = 0.016) ( Table 4). The risk factors for CSS in the fat invasion cohort were higher   and pathological grade was also an independent prognostic factor (p < 0.001) ( Table 5). Higher age was risk factor of both OS (HR 1.05, p < 0.001) and CSS (HR 1.04, p = 0.001) in the venous invasion cohort ( Tables 6, 7). The surgical method had no significant effect on the OS (p = 0.069) and CSS (p = 0.190) of the venous invasion cohort and had no significant effect on the CSS (p = 0.078) of the fat invasion cohort ( Table 8).

DISCUSSION
With the progression of the technology and the in-depth understanding of RCC, the indications for PN are gradually expanding, from initial T1a to T1b and then to the current selected T2 (5,16). It is reasonable to believe that PN will be provided to some appropriate T3a patients in some technologically advanced institutions (12).
Even though urologists do not deliberately provide PN for T3a patients, they will encounter cases where cT1 and cT2 patients undergo PN upstage to pT3a. The postoperative upstaging occurs in approximately 11.35% of cT1 and cT2 patients; studies have shown that RN does not improve the RFS of upstaged patients as compared with PN (9). The study of Russell et al. included 1,955 cases with cT1 RCC who underwent PN, 95 of which upstaged to pT3a, and the PFS and CSS of pT3a patients were significantly worse than those of pT1 patients (both p < 0.01) (17 (14). In our study's SEER cohort and the TJH cohort, the CSS or RFS of pT3a RCC  patients who received PN was not worse than that of patients who received RN. In addition to upstaged pT3a patients, some institutions have also tried to actively perform PN for cT3a patients. The study by Yim et al. included 159 cT3a RCC patients who received robot-assisted PN from multiple centers, of which 64.3% of the cases achieved a trifecta (negative surgical margins, warm ischemia time (WIT) ≤ 25 min, and no perioperative complications), and 37.6% of patients achieved the optimal outcome (trifecta and ≥90% preservation of the eGFR and no stage upgrading of chronic kidney disease). The 5-year RFS, CSS, and OS were 82.1%, 93.3%, and 91.3%, respectively. The downside is that there is no RN patient as a control (12).  Some studies found that patients who received PN had better RFS and CSS than those who received RN and interpreted it as greater renal function preservation, which might be related to better oncologic outcomes (26,27); the same trend was also observed in the TJH cohort in our study. Palacios et al. found that the unfavorable oncologic outcome was more related to the       aggressive characteristics of the tumor itself, rather than the degree of renal function preservation (28). In our study, PN in the pre-PSM cohort was also significantly correlated with better CSS, but it was no longer significant after PSM. Therefore, we should cautiously interpret the trend that CSS and RFS of PN patients are better than those of RN patients in the study, which may be caused by retrospective study design and selection bias.
In view of the fact that some information such as hemoglobin, albumin, BMI, and comorbidities cannot be obtained from the SEER database, there were also some biases in this study, which may affect the accuracy of the conclusion.

CONCLUSIONS
This study proved that PN is safe and feasible in localized T3a RCC patients via a retrospective study with a large sample volume, and the oncologic outcomes of patients who underwent PN were comparable with those of patients who received RN but limited to pT3a patients, and higher-quality research is needed before exploring performing PN for cT3a RCC patients.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the Medical Ethical Committee of Tongji Hospital of Huazhong University of Science and Technology. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

AUTHOR CONTRIBUTIONS
JT: data curation, writing-conceptualization, and original draft. XZ: formal analysis and supervision. JW: data curation and data analysis. JG: data curation and writing-review and editing. CK: software and supervision. WG: data analysis and supervision. ZH: project administration, resources, supervision, and funding acquisition. CY: writing-conceptualization, funding acquisition, and writing-review and editing. All authors contributed to the article and approved the submitted version.