Nephron Sparing Surgery: A Run after a Chimera?

1Department of Urology, Riga Stradins University Hospital, Latvia 2Department of Urology, Central Hospital of Bolzano, Italy 3Department of Pathology, Central Hospital of Bolzano, Italy 4Department of Hematology and Center of Bone Marrow Transplantation, Central Hospital of Bolzano, Italy 5Department of Internal Medicine V (Haematology and Oncology), Medical University of Innsbruck, Austria 6Chair of Urology, Sigmund Freud University Medical School, Austria


Introduction
Radical nephrectomy (RN) used to be the technique of choice for localized renal cell carcinoma (RCC) [1] with excellent oncological outcomes. The nephron sparing surgery (NSS) was limited to "so called imperative indications" as in case of bilateral RCCs as well as in tumours of a solitary kidney. With abdominal imaging more than 50% of RCCs are nowadays diagnosed incidentally [2,3].
Surgical trends, based on a substantial number of retrospective studies [4][5][6][7][8][9][10] which indicated comparable oncological outcome but less de novo incidence of renal failure for NSS, moved towards renal preservation assuming that, preserving renal parenchyma, the non-cancer related mortality provoked by cardiovascular events could be prevented [11]. This survival benefit could be also the consequence of a selection bias as mentioned by Tomaszewski, et al. [12] because it seems that patients chosen for NSS possess a baseline higher likelihood of long term survival [13]. Currently it is recommended by the EAU guidelines on RCC, that NSS should be performed for T1a and T1b tumours [14]. The indication for NSS, however, was extended also to larger tumours counting on long term benefits with comparable oncological outcome. To date 8 studies in T2 or greater are published and the results are controversal. In stages following the premise to perform NSS whenever technically possible, irrespectively of tumour size.

Material and Methods
This single-institution study ran from April 2000 to May 2016. Demographic and clinical-pathological characteristics of 392 patients, 129 women and 263 men, median age 65 (range 23-88) years, were collected prospectively and introduced in the institutional kidney cancer database. With the approval of institutional review board the data were retrospectively assessed.
Preoperative evaluation of all patients comprised laboratory including glomerular filtration rate (GFR) and cross sectional imaging of the chest and the abdomen to define the clinical tumour stage. A metastatic evaluation was also performed according to the current EAU Guidelines [14]. NSS was offered to all patients irrespectively of the tumour stage whenever it seemed technically feasible otherwise a RN was taken into consideration. All complications within 30 days of surgery were considered perioperative complications, all others as late complications. To classify the complications the Clavien Dindo classification (CDC) was used since 2005 [15].
Chronic kidney disease (CKD) stage was assigned according to the National Kidney Foundation definition [16]. Pathological data were also prospectively collected and included histological subtype, grade diameter and margin status. Tumours were staged according to the WHO classification of the International Society of Urological Pathology (ISUP) [17] and graded according to Fuhrmann until 2015 and according to WHO/ISUP starting in 2016 [17]. Tumour recurrence was defined as a new renal mass in the initial tumor bed in the absence of distant metastases based on CT-imaging.
Follow-up was calculated from the time of surgery until the last follow up (FU) visit or death. Mortality status was obtained through the Provincial Tumour Registry or by death certification of patients outside the province. Postoperative FU of patients included physical examination, laboratory examinations, chest x-ray in two plains and an abdominal sonography every 3 months during the first two years, every 6 months from the second to the fifth year of follow-up and thereafter every year. The first CT control was performed three months after RN/ NSS and used as reference imaging. Thereafter every 6 months a CT scan of the kidneys was performed until the fifth year of follow up, then annually.
If any suspicion of disease progression arose, a complete restaging was performed.

Surgical approach and technique
Performed procedures included laparoscopy and open RN and NSS. Surgical modality and approach were decided by tumour characteristics, anatomical singularities and surgeons' preference. All surgeries were performed from or in the presence of two skilled surgeons (AP, SP).

Endpoints and Statistical Analysis
This observational study was designed to assess cancer control, functional outcome and surgical related complications. The primary outcomes were death or time to progression, with progression defined as local recurrence in either kidney or regional or distant metastases.
The analysis was performed using SPSS ver. 17.0 (SPSS Inc., Chicago, IL). Survival curves were plotted according to the Kaplan-Meier method. Overall survival (OS) was defined as time from diagnosis until death of any cause, cancer specific survival (CSS) as time from diagnosis until cancer related death and disease-free survival time (DSF) from the achievement of a complete remission until first recurrence [18]. Differences of categorical variables on survival were evaluated using the log-rank test. A multivariate analysis by cox regression was used to calculate the prognostic factor of tumour type, grade, stage, lymph node status, vein involvement, metastases, margins, diabetes and hypertension. A p-value of < 0.05 was considered statistically significant. In univariate analysis, whenever possible, Fisher´s Exact test was applied, otherwise the Pearson´s Chi Square test. Median follow-up time for these patients was 48.08 (range 5. 2-194.4) months. Baseline patient´s characteristics showed no significant statistical differences between RN and NSS, although, a greater percentage of the RN group had American Society of Anesthesiologist (ASA) 3 or higher compared to the NSS group. Surgical parameters were similar between the groups even though a higher proportion of NSS was performed as open surgery. Pa-tient´s demographics and surgical data of both groups are given in (Table 1).

Tumour characteristics, progression and survival
Median diameter of the tumours was 3.5 cm (range 0.2-20 cm). Histological subtype and tumour grade were similar between the groups ( Table 1). The NSS group showed a greater proportion of T1 tumours (223 vs. 79) whereas the RN group had a higher proportion of T3/4 tumours (53 vs. 12). In 1.5% the lymphnodes were positive affecting only the RN group. 5.6% of the cases had metastases at the time of surgery, 16 in the RN group and 7 in the NSS group. Oncological outcomes are displayed in (Table 2).

Discussion
To our knowledge, this is the largest case control study to date comparing RN and NSS with the premise to perform NSS whenever technically possible, irrespectively of the tumour size ranging from T1 to T4 stages. It led to NSS being performed for 60.3% of all RCC which is a relatively high percentage compared with the US where still 77.6% of all T1a tumours are treated with RN [17]. Also in Europe, as shown by Aben, et al. [18], 70% of T1b RCC undergo RN. Obviously there are some caveats in considering the EAU and AUA Guidelines.
Our results showed highly significant better CSS and DFS for NSS at five years and at ten years but not for OS after 5 years (96.7% vs. 80%) and at ten years (71.3% vs. 65.3%). The different surgical approaches did not significantly influence OS according to the different disease stages at time of diagnosis. While the survival curves were nearly overlapping for stage T1 and T2, in patients with T3, the OS after NSS was slightly inferior without achieving statistical significance. For T4 tumors the number was too small to draw a reliable conclusion. The statistical tendency, however, is in favour of NSS (Figure 3). The comparison of these results with published data surely indicates some imbalances. First, because our series has In total 88 out of 392 patients (22.4%) died after a median of 90 months  OS in the RN group was 80.0% after 5 years and 65.3% after 10 years, while it was 96.7% and 71.3%, respectively, in the NSS group (p = 0.214) ( Table 2). If OS is stratified according to tumor stages the median OS in T1 was 14.7 years for RN and not reached for NSS, not reached in both for T2, whereas for T3 tumors the median OS in RN was 9.7 years versus 5.6 in NSS. Due to the low number of patients with T4 disease, it was not possible to draw any conclusion (Figure 1). CSS in the RN group was 85.2% after 5 years and 78.4% after 10 years, in the group of patients with NSS it was 98.5% after 5 years and 97.8% after 10 years (p < 0.001) ( Table 2). If CSS is stratified according to tumor stages the median CSS was not reached for T1 and T2 independently of the type of surgery whereas in T3 tumors the median CSS was 9.7 years for RN vs. 5.6 years for NSS. Again in T4 tumors the number was too low to give a reliable evaluation (Figure 2).
In the RN group, 81.8% of the patients were disease free after 5 years and 70.2% after 10 years, in the NSS group, it was 98.0% and 93.5%, respectively (p = 0.011) ( Table 2).
Of the 162 patients who underwent RN, 153 (94.4%) had a negative resection margin, 2 had a positive. 3 out of the 153 patients with negative margin showed recurrence after a median 123 months (range 122-177). 43 (28.6%) of these patients developed metastases after a mean 40 months (1-126). Two patients with a positive resection margin developed metastases within 3 and 4 months.

Functional outcome
In both groups, creatinine levels (1.0 vs. 0.94) and filtration rate (GFR) (62 vs. 67.5) were similar before surgery, while immediately after surgery and at last follow up NSS showed a significantly better functional preservation.
Details are given in (Table 1).  [22]. Both studies have in common that patients with recurrence had significantly larger tumour size, more often T2-T4 tumours and had a higher number of Grade 3-4. Presence of distant metastases was significantly more common after RN than after NSS in all stages. Our findings concerning the progression after RN compared to NSS were probably a result of the selection of patients with smaller and less aggressive tumours for NSS.
Our series failed to evidence an oncological benefit for NSS in higher tumor stages (Figure 3).
We could demonstrate a superior renal function preservation and less new onset of CKDs for NSS in our cohort compared to many other studies [1,3,26]. The percentage decline of GFR from baseline to postoperative follow-up time was significantly smaller in patients treated with NSS (3.7%) compared with RN (12.9%). Both groups experienced a decline in renal function immediately after the operation, then a slight recovery of the GFR with a long lasting stable run, a phenomenon not observed in other studies but one [2]. Although the risk of moderate CKD in the elective setting is higher with RN, the risk of end-stage renal disease (ESRD) is limited after elective RN or NSS. This finding is confirmed by the EORTC 30904 trial [21] where the risk is quantified with 0.8% NSS and 1.2% for RN. The same affirmation cannot be made for patients with preoperative CKD. There the risk of ESRD increases dramatically [12,27].
NSS is considered a riskier operation than RN due to the combination of an ablative surgical phase followed by a reconstructive part. More centrally situated tumours require the reconstruction of the collecting system, the control of potential bleeding sources and the adaption of the parenchyma to prevent bleeding and leakage. All these complications are avoidable with RN.
Several studies showed that when comparing the complication rate between NSS and RN there was a tendency towards a higher complication rate after NSS. [1,2,7,8,21,22,28,29]. In the literature the complication rate for open NSS for T1 RCC ranges between 4.5%-8.7% and increases to 18.8% for tumours > 7 cm, not including the transfusion rate of up to 8.2% [1]. The best evidence was obtained from the prospective study of (EORTC 30904) [19] with an evidence level I-b. There NSS was associated with a higher risk of severe perioperative bleed-no restriction regarding tumour size and comprises all T-stages and second because the evidence in literature is limited to 34 studies fulfilling the PRISMA criteria. All these studies are differing in tumour size and FU available and are contradictory. D'Armiento, et al. [10] found an equal median survival at 98 months FU, [19] reported a statistically not significant survival benefit for RN at 5 years (80% vs. 75%) whereas Lee, et al. [20] demonstrated a tendency in favour of NSS at five years. The only prospective and randomized study from Van Poppel [21] showed inferior 10 years OS for NSS.
Tumour progression was found in 8.9% in patients of the NSS group. Of those, 6 (2.4%) patients developed a recurrence in the previous operated kidney and 16 patients (6.5%) distant metastases after a median duration of 16.5 months. In the literature the local recurrence rate ranges between 4%-6% with few exceptions as in the study from Herr who reports 1.4% local recurrence rate at 10 years [8] or in the Swedish population based study where local recurrences reach 20% [22]. In the Bolzano study the rate of distant metastases was 6.5%, which is in the middle of the in the literature reported range of 5.8% and 8.6% [1] for NSS. Once again, all the studies are biased by different tumour stages, tumour sizes, and different follow-ups.
In the RN group 28.4% of the patients experienced a progression. Local recurrence was seen only in 3 (1.9%) patients. It is similar to the Liebovich study [23] where the rate was 2.2%. The RN group includes in both studies a higher percentage of advanced T-stages and positive lymph nodes in the final pathology, so it could be argued that the local recurrence in NSS could be largely avoided with radical surgery.
Achieving a negative surgical margin (NSM) during NSS is a fundamental principle of oncological surgery and the best prevention of local recurrence, but does a positive surgical margin (PSM) necessarily lead to a recurrence and subsequently to a systemic progression and finally to a cancer related mortality? In the NSS group 10 (4%) patients had a PSM in the final histo pathological report. One (0.4%) of them developed a local recurrence but none of them a distant metastasis which is in accordance with the current literature where the range for PSM is between 4% and 10%. PSM after NSS will not always result in a local recurrence and /or disease progression. It is suggested that a PSM should not be sanctioned immediately by a secondary nephrectomy. A very close follow-up is mandatory [24,25].
The situation is different when it comes to RN. 2 (1.2%) patients had a PSM, none developed local recurrence but both suffered from metastases. It reflects the tumour biology. If RN cannot anymore be performed with NSM it reflects an advanced tumour stage (stage III or IV)   Despite showing several strengths, as e.g. the large study cohort, the long observation period over 15 years, the always identical surgical strategy, the presence of one of two surgeons (AP and SP) in the theatre in all cases, guaranteeing the technical continuity and uniformity of the surgical procedure, the large median tumour size compared to other studies, our study implicates some limitations.
First and most importantly, it is limited by its observational nature and its retrospective design. Therefore, it is suitable to generate a hypothesis but not evidence at high level. Moreover, there may have been other factors impacting survival rates in both groups as selection bias and unobserved but measured confounders. Although the emerging data from the last decades about RN compared with NSS are contradictory, the present study is valuable because it adds the evidence of feasibility of NSS even for large and for all RCC stages, but better CSS and DFS for NSS lead not to longer OS.

Conclusion
This retrospective study, including 392 patients undergoing NSS whenever technically possible, could show that NSS can be safely performed for higher stages of RCC. The surgical complications are higher in the NSS group. Their frequency is acceptable and decreased over time, approaching the complication rate of the RN group over the last years. The risk of local recurrence, even in presence of PSM, is minimal.
Despite a better CSS and DFS in the NSS group, this study could not demonstrate an OS benefit in NSS, especially for higher tumour stages.