Nephron-sparing surgery has been accepted as the ideal treatment of localised RCC, given the similar oncological outcomes to that of radical surgery [13–17]. However, imaging studies may sometimes question the feasibility of NSS mostly because of either the size or the localization of the tumor. Moreover, without a structured and reproducible system for describing the relevant renal mass anatomy, treatment decisions will vary, depending on urologist’s training, biases, comfort levels, and individual experience.
Recently, three different scoring systems were developed to serve as a common vocabulary when discussing anatomic geometry and complexity of renal masses. Preoperative aspects and dimensions used for anatomic (P.A.D.U.A.) classification and R.E.N.A.L. (radius, exophytic/endophytic, nearness, anterior/posterior, location) scoring systems involve similar components and methodology, enabling a comprehensive description of the tumor size, polarity, anterior/posterior location and proximity to the collecting system [7, 8]. The centrality index (C-index), however, is a completely different system that involves a relatively complex mathematical concept and characterizes tumor centrality based on the ratio of the distance between the tumor and kidney center and tumor radius . Currently, there is no clear consensus favoring the utility of any score or index over the others.
Nephrometric scoring systems have also been associated with certain measures of operative complexity such as ischemia time and complication rates. Waldert et al. reported a significant association between high P.A.D.U.A. scores (≥ 10) and increased WIT (22 vs. 34 minutes) . Similarly, Simmons et al. stated that tumors with a C-index of ≤ 1 had a 2.3-fold risk of prolonged WIT (≥ 35 minutes) compared with tumors with a C-index ≥ 1 . Each unit increase in R.E.N.A.L. score was associated with a 35% increased risk of postoperative urine leakage in a study conducted by Bruner et al. . Another study highlighted the significant increase in complication rates with increasing P.A.D.U.A. score among patients undergoing partial nephrectomy. Tumors with scores of 8 to 9 had a hazard ratio of 14.5 for postoperative complications compared with tumors with scores of 6 to 7. Tumors scoring ≥ 10 had a hazard ratio of 30.6 .
Apart from the documented power in predicting perioperative outcomes, these scoring systems also have the potential to influence surgical preferences, which covered the scope of this article. According to our results, mean R.E.N.A.L. and P.A.D.U.A. scores of the tumors managed by robotic assistance were significantly lower than those treated by open surgery. Although being lower in tumors managed by open approach, C-index values did not exhibit a statistically significant difference between open and robotic groups (Table 1). These differences may be due to selection biases, with more peripheral, exophytic, and smaller tumors being scheduled for robot-assisted surgery. However, these are the initial robot-assisted nephron sparing surgeries performed in our clinic by a surgeon who is experienced in open surgery.
Tumor size, estimated blood loss amounts and transfusion rates, which would be considered among the factors indirectly reflecting how “tough” the tumor was, did not differ significantly between open and robotic groups. Robot-assisted nephron-sparing surgeries lasted significantly longer than their open counterparts. However, this difference may not be regarded as a sign of tumoral complexity since recorded time in robotic surgeries included the “docking” maneuvers.
In our practice, the decision to occlude renal pedicle is given during the operation. Therefore, it might be regarded as an “in-vivo” validation of the preoperative morphometric information. Moreover, interrupting renal perfusion temporarily constitutes a major clinical concern since longer warm ischemia time has been associated with acute renal failure, decreased glomerular filtration rate and de-novo chronic kidney disease . In order to clarify this issue we investigated the differences between surgical choice subgroups in terms of morphometric scores. We found a statistically significant difference between robotic off-clamp cases and open clamped cases with regard to each morphometric score (R.E.N.A.L., P.A.D.U.A. and C-index, Figure 1). However, only R.E.N.A.L. score demonstrated significant differences across most of the studied surgical alternative subgroups (robotic off-clamp vs. open off-clamp and robotic clamped vs. open clamped) (Figure 1).
Open conversion during minimally invasive surgery might be regarded as another way of expressing the challenging nature of the tumor being handled. In our series, those tumors, for which open conversion was inevitable, had significantly higher mean R.E.N.A.L. and P.A.D.U.A. scores than those who were successfully treated in a minimally invasive fashion.
On univariate analyses, P.A.D.U.A. and R.E.N.A.L. scores demonstrated sufficient predictive power in determining the route by which tumor was enucleoresected. High R.E.N.A.L. score (cut-off value: 6.5) and high P.A.D.U.A. score (cut-off value: 7.5) predicted the likelihood of an open NSS, with adequate statistical power. This finding was complemented by the significant difference between the number of patients in low, moderate R.E.N.A.L. and low P.A.D.U.A. categories with regard to the route that has been preferred for NSS (Table 2). In a similar study including more than 400 patients undergoing NSS, Canter et al. reported that patients undergoing open surgery had a significantly higher mean R.E.N.A.L. score, than those managed with minimally-invasive NSS (8.19 vs. 6.62, p < 0.0001) . Recently, Rosevear et al. confirmed the predictive power of R.E.N.A.L score in terms of operative approach selection, with significantly more patients with low and high scores undergoing partial and radical nephrectomy, respectively in their cohort of 249 patients .
However, our findings do not mean that we preferentially use the R.E.N.A.L. or the P.A.D.U.A. system for preoperative morphometric assessment. Based on the available literature, there is no clear advantage of one scoring system over the others regarding surgical preferences and perioperative outcome. As the number of patients in each sub-category increase, the C-index differences that were stated as insignificant might gain statistical significance. Therefore, it is hard to draw strict conclusions about the superiority of R.E.N.A.L. or P.A.D.U.A. scoring systems based on our results.
Our study is unique in that it focuses on the utility of morphometric scoring systems in tailoring the surgical approach rather than perioperative outcomes or oncologic results. Moreover, only nephron-sparing surgeries performed by a single surgeon through the open or robot-assisted laparoscopic route were taken into consideration. At last but not the least, we tested the discriminative power of all three scoring systems that are currently being utilized across the globe. However, this data reflects the experience of a single surgeon, who is proficient in open NSS and currently in the initial phase of the learning curve for robotic NSS, which may limit the reproducibility of our findings. Our statistical findings are also handicapped by the retrospective study design and small sample size. Another criticism is the lack of strictly defined indications about the details of the surgical strategy. Route of access and the decision to clamp the renal hilum during enucleoresection depended on a variety of factors such as patient profile (general health status, comorbidities), tumor characteristics (size, location, complexity) and surgeon preference. In the early days of the robotic era in our hospital, smaller (cT1a), cortical and exophytic tumors that were located anteriorly, below or above the hilar plane were scheduled for robot-assisted NSS while more complicated tumors (hilar, cT1b-T2, mostly endophytic) were managed by open NSS in order to ensure a reasonable warm ischemia time. However, as our experience grew, we started doing robot-assisted NSS for more challenging tumors. This selection bias, which can be understandable within the context of the learning curve, should be considered while interpreting our results. Future prospective studies enrolling higher number of patients will more precisely highlight the importance of adapting preoperative morphometric evaluation into routine clinical practice.