Three-Dimensional Reconstruction as a Complementary Tool for Preoperative RENAL Nephrometry Score Assessment in Partial Nephrectomy

1. Introduction

Successful nephron-sparing surgery depends on accurate preoperative characterization of renal tumor anatomy. Partial nephrectomy has become the preferred treatment for localized renal masses whenever technically feasible, offering equivalent oncological outcomes to radical nephrectomy while preserving renal function. Consequently, precise assessment of tumor complexity plays a central role in surgical planning and intraoperative decision-making.

The R.E.N.A.L. nephrometry score is one of the most widely adopted standardized systems for describing renal tumor anatomy and estimating surgical complexity. By incorporating tumor size, depth, relationship to the collecting system, anterior/posterior location, and proximity to the renal hilum, it provides a common framework for communication among surgeons and facilitates preoperative planning.

Despite its widespread use, interpretation of the R.E.N.A.L. score based on conventional computed tomography (CT) remains subject to interobserver variability. The complexity of renal anatomy and the limitations of two-dimensional image interpretation may result in differences in tumor characterization, potentially affecting the consistency of preoperative assessment.

Three-dimensional (3D) reconstruction has emerged as a complementary imaging modality that provides a more intuitive visualization of the spatial relationships between the tumor, renal parenchyma, collecting system, and hilar vessels. Rather than replacing conventional imaging, 3D reconstruction may provide additional anatomical information that contributes to a more consistent interpretation of tumor complexity and facilitates standardized preoperative evaluation.

The primary objective of this study was to evaluate whether the incorporation of 3D reconstruction influences the reproducibility of R.E.N.A.L. nephrometry score assessment compared with conventional CT. As a secondary objective, we compared perioperative outcomes between patients whose surgical planning included 3D reconstruction and those evaluated using CT alone.

2. Materials and Methods

2.1. Study Design and Population

A retrospective observational study was conducted including patients who underwent partial nephrectomy for renal tumors between March 2023 and March 2025. Both laparoscopic and robot-assisted procedures were included. Patients with congenital or acquired renal anatomical abnormalities that could interfere with tumor characterization or nephrometry assessment were excluded.

The study consisted of two complementary analyses. The primary analysis evaluated the reproducibility of preoperative renal tumor characterization using the R.E.N.A.L. nephrometry score on conventional computed tomography (CT) and three-dimensional (3D) reconstruction. The secondary analysis explored whether incorporation of 3D reconstruction into preoperative planning was associated with differences in perioperative outcomes.

For the primary agreement analysis, 55 patients with available preoperative CT and CT-based 3D reconstruction were included. For the secondary clinical analysis, patients whose preoperative planning included 3D reconstruction were compared with a control cohort of patients who underwent partial nephrectomy during the same study period, fulfilled identical inclusion criteria, had comparable CT imaging characteristics, and were planned using conventional CT alone.

2.2. Image Assessment

Three urologists independently evaluated each case using conventional CT images and 3D reconstructions. For each modality, observers assigned a R.E.N.A.L. nephrometry score to each renal lesion, including both the numerical score and the corresponding risk category: low, intermediate, or high.

CT images were assessed using the institutional PACS viewer (ViewPACS), whereas 3D reconstructions were analyzed using Meshmixer software. All observers had at least five years of experience in renal surgery and routinely performed minimally invasive partial nephrectomy. Evaluations were performed independently, and observers were blinded to surgical outcomes and to the scores assigned by the other observers.

2.3. Statistical Analysis

The primary endpoint was agreement in R.E.N.A.L. nephrometry score assessment between conventional CT and 3D reconstruction. Intraobserver agreement between CT and 3D reconstruction was analyzed separately for each observer. Interobserver agreement was assessed for each imaging modality by comparing the three observer pairs.

For the numerical R.E.N.A.L. score, Pearson correlation coefficients were used to evaluate the degree of linear association between measurements. For risk categorization, weighted Cohen’s kappa was used to assess intraobserver agreement, while Fleiss’ kappa was used to evaluate interobserver agreement among the three observers.

2.4. Perioperative Outcomes

As a secondary exploratory analysis, perioperative outcomes were compared between patients whose surgical planning included 3D reconstruction and those evaluated using CT alone. Variables analyzed included operative time, warm ischemia time, estimated blood loss, urinary collecting system entry, blood transfusion, postoperative hematuria, length of hospital stay, catheterization time, readmission, postoperative renal function, surgical margins, and complications according to the Clavien–Dindo classification.

3. Results

The primary analysis demonstrated greater interobserver agreement in preoperative tumor assessment when 3D reconstruction was incorporated into conventional CT evaluation. Agreement for numerical R.E.N.A.L. scores was higher with 3D reconstruction (57%) than with CT alone (43%). Similarly, agreement in risk categorization increased from 62% with CT to 72% with 3D reconstruction (Table 1).

Interobserver agreement for risk categorization, assessed using Fleiss’ kappa, was very good for 3D reconstruction (κ = 0.804) and moderate for conventional CT (κ = 0.562), indicating greater consistency among observers when three-dimensional models were available (Table 3).

Intraobserver agreement between CT and 3D reconstruction was moderate for all observers. Pearson correlation coefficients ranged from 0.430 to 0.560 for numerical score assessment, while weighted Cohen’s kappa values ranged from 0.398 to 0.441 for risk categorization, indicating moderate consistency between imaging modalities (Table 2).

Table 2. Intraobserver interpretation of R.E.N.A.L. score and resection risk.

	Observer	Comparison	Method	Value
Score intraobserver	U1	3D vs TAC	Pearson r	0.43
	U2	3D vs TAC	Pearson r	0.523
	U3	3D vs TAC	Pearson r	0.56
Risk intraobserver	U1	3D vs TAC	Kappa	0.424
	U2	3D vs TAC	Kappa	0.441
	U3	3D vs TAC	Kappa	0.398

Table 2. Intraobserver interpretation of R.E.N.A.L. score and resection risk.

	Observer	Comparison	Method	Value
Score intraobserver	U1	3D vs TAC	Pearson r	0.43
	U2	3D vs TAC	Pearson r	0.523
	U3	3D vs TAC	Pearson r	0.56
Risk intraobserver	U1	3D vs TAC	Kappa	0.424
	U2	3D vs TAC	Kappa	0.441
	U3	3D vs TAC	Kappa	0.398

Table 3. Interobserver correlation analysis of score and risk interpretation.

	Comparison	Modality	Method	Value
Score Interobserver	U1 vs U2 (3D)	3D	Pearson r	0.706
	U1 vs U3 (3D)	3D	Pearson r	0.631
	U2 vs U3 (3D)	3D	Pearson r	0.84
	U1 vs U2 (TAC)	TAC	Pearson r	0.845
	U1 vs U3 (TAC)	TAC	Pearson r	0.878
	U2 vs U3 (TAC)	TAC	Pearson r	0.866
Risk Interobserver	3D	3D	Fleiss Kappa	0.804
	TAC	TAC	Fleiss Kappa	0.562

Table 3. Interobserver correlation analysis of score and risk interpretation.

	Comparison	Modality	Method	Value
Score Interobserver	U1 vs U2 (3D)	3D	Pearson r	0.706
	U1 vs U3 (3D)	3D	Pearson r	0.631
	U2 vs U3 (3D)	3D	Pearson r	0.84
	U1 vs U2 (TAC)	TAC	Pearson r	0.845
	U1 vs U3 (TAC)	TAC	Pearson r	0.878
	U2 vs U3 (TAC)	TAC	Pearson r	0.866
Risk Interobserver	3D	3D	Fleiss Kappa	0.804
	TAC	TAC	Fleiss Kappa	0.562

No statistically significant differences were observed in operative time, warm ischemia time, estimated blood loss, transfusion rate, collecting system entry, postoperative renal function, or complication rates. Hospital stay was slightly longer in the 3D group (2.7 vs 2.2 days; p = 0.04) (Table 4).

Overall, incorporation of 3D reconstruction into preoperative planning was not associated with measurable improvements in perioperative outcomes in this cohort.

4. Discussion

The principal finding of this study is that incorporation of three-dimensional reconstruction into preoperative assessment was associated with greater agreement in R.E.N.A.L. nephrometry score interpretation among observers. Rather than demonstrating superiority over conventional CT, our findings suggest that three-dimensional models may contribute to a more standardized characterization of renal tumor anatomy before partial nephrectomy.

Our findings are consistent with previous reports suggesting that three-dimensional reconstruction enhances anatomical understanding during preoperative planning. Shirk et al. reported improved anatomical perception with virtual three-dimensional models, although without significant differences in perioperative outcomes. Unlike previous studies, which primarily evaluated the impact of three-dimensional visualization on surgical performance, our study specifically focused on the reproducibility of preoperative anatomical characterization using the R.E.N.A.L. nephrometry score.

The greater agreement observed with three-dimensional reconstruction is likely explained by its ability to provide a more intuitive representation of renal anatomy rather than by the imaging modality itself. Surgical planning for partial nephrectomy depends on understanding the spatial relationship between the tumor, the renal sinus, the collecting system, and the segmental renal vasculature [6]. Although contrast-enhanced CT remains the standard imaging modality, interpretation of these relationships requires mental reconstruction of multiple two-dimensional slices, a process that is inherently susceptible to individual interpretation [6]. Three-dimensional models integrate this anatomical information into a single spatial representation, facilitating recognition of tumor depth, endophytic components, hilar proximity, and collecting system involvement [4]. Consequently, they may reduce interpretative variability and promote a more standardized characterization of tumor complexity before surgery [2,4].

Although greater consistency in anatomical assessment did not translate into measurable perioperative benefits in this cohort, reproducible preoperative characterization represents an important prerequisite for surgical planning and communication among surgeons. Anatomy-based nephrometry systems were originally developed to provide standardized and reproducible descriptions of renal tumor anatomy while minimizing interobserver variability. Improving the consistency with which these anatomical features are identified therefore represents a logical extension of their original purpose.

The absence of significant differences in perioperative outcomes should be interpreted within the context of the study design, which was not powered to detect differences in clinical endpoints but rather to evaluate agreement in preoperative assessment. Moreover, the lack of measurable clinical benefit may reflect the high level of surgical expertise at our institution and the careful selection of patients for nephron-sparing surgery, rather than a lack of value of three-dimensional reconstruction itself.

Three-dimensional reconstruction may therefore complement conventional CT by providing additional anatomical information that facilitates a more consistent interpretation of the variables incorporated into nephrometry scoring systems, even when this does not immediately translate into measurable perioperative benefits.

This study has several limitations. Its retrospective design may have introduced selection bias despite the use of comparable inclusion criteria for both cohorts. In addition, the sample size may have been insufficient to detect subtle differences in perioperative outcomes. Finally, long-term functional and oncological outcomes were not evaluated. Nevertheless, the study addresses an underexplored aspect of three-dimensional reconstruction by focusing on reproducibility and standardization of preoperative anatomical assessment rather than on surgical performance alone.

5. Conclusions

Three-dimensional reconstruction may represent a valuable complementary tool for preoperative assessment in partial nephrectomy by improving the consistency of R.E.N.A.L. nephrometry score interpretation. Although this greater agreement was not associated with measurable differences in perioperative outcomes in the present cohort, more reproducible preoperative anatomical characterization may contribute to a more standardized framework for surgical planning and communication among surgeons. Further prospective multicenter studies are warranted to determine whether these improvements translate into clinically relevant benefits.