Take Home Message
The lower complexity scores for kidney tumor surgery with three-dimensional virtual models in comparison to standard two-dimensional scans could boost the use of nephron-sparing techniques and improve the identification of patients at greater risk of postoperative complications.
Keywords: Three-dimensional imaging, Robotic surgery, Renal cell carcinoma, Kidney cancer, Nephron-sparing surgery, Nephrometry scores
Abstract
Background and objective
The aim of our study was to compare assessment of PADUA and RENAL nephrometry scores and risk/complexity categories via two-dimensional (2D) imaging and three-dimensional virtual models (3DVM) in a large multi-institutional cohort of renal masses suitable for robot-assisted partial nephrectomy (RAPN), and evaluate the predictive role of these imaging approaches for postoperative complications.
Methods
Patients were prospectively enrolled from six international high-volume robotic centers, calculating PADUA and RENAL-nephrometry scores and their relative categories with 2D-imaging and 3DVMs. The concordance of nephrometry scores and categories between the two approaches was evaluated using χ2 tests and Cohen’s κ coefficient. Receiver operating characteristic curves were plotted to assess the sensitivity and specificity of the 3DVM and 2D approaches for predicting the occurrence of postoperative complications. Multivariable logistic analyses were conducted to identify predictors of major postoperative complications.
Key findings and limitations
A total of 318 patients were included in the study. There was low concordance for nephrometry scores and categories between the 3DVM and 2D assessment methods, with downgrading of PADUA and RENAL scores on 3DVM assessment in 43% and 49% of cases, and downgrading of the corresponding categories in 25% and 26%, respectively. Moreover, 3DVM assessment showed better accuracy than the 2D approach in predicting overall (p < 0.001) and major (p = 0.001) postoperative complications. In line with these findings, multivariable analyses showed that 3DVM-based nephrometry scores and categories were predictive of major postoperative complications (p < 0.001). Limitations include the risk of interobserver variability in evaluating nephrometry scores and categories, production costs for the 3DVMs, and the experience of the surgeons involved, with potential impacts on diffusion of this technology.
Conclusions and clinical implications
In this multi-institutional study, 3DVMs had superior accuracy to 2D images for evaluating the surgical complexity of renal masses and frequently led to downgrading. This could facilitate an increase in recommendations for kidney-sparing surgery and better identification of cases at risk of postoperative complications.
Patient summary
Our study shows that the use of three-dimensional models gives lower complexity scores for kidney tumors in comparison to standard two-dimensional scans. This can improve surgical planning and may boost the use of kidney-sparing techniques and better identification of cases that are more likely to have postoperative complications.
1. Introduction
Partial nephrectomy (PN) is the gold standard for treatment of T1 renal masses and is preferably performed using minimally invasive techniques [1], [2]. When planning the surgery, nephrometry scores such as PADUA and RENAL can play a role in predicting the surgical complexity of the intervention and postoperative complications [3], [4].
These tools are historically based on contrast-enhanced computed tomography (CT) scans or magnetic resonance imaging (MRI) and allow a description of renal masses in terms of objective characteristics, offering pivotal data for surgical indications, treatment planning, and patient counseling [5].
However, the two-dimensional (2D) nature of this imaging limits a complete understanding of the morphological characteristics of tumors, with potential for inaccurate scoring. Moreover, the mental visualization process needed to understand the anatomy requires experience, which can lead to errors during the PN learning curve. To overcome these issues, automatic rendering software is available for several DICOM viewers, but their quality remains unsatisfactory. To further improve the quality of preoperative imaging, Porpiglia et al [6] created hyperaccurate three-dimensional virtual models (3DVMs) via processing of 2D imaging. These models allow surgeons to more precisely assess the location and characteristics of a tumor, its relationship to vessels and the upper urinary collecting system (UCS), and the perfusion status of the kidney [7], which yields a better understanding of the surgical complexity of PN in individual cases.
Studies assessing the role of this technology in preoperative planning [8] and intraoperative guidance [9] found surgical advantages in terms of minimization of ischemic damage, higher pure enucleation rates, and lower rates of UCS violation [10], leading to lower rates of postoperative complications [11], higher rates of nephron sparing, and better functional outcomes [12]. In particular, a preliminary study on the role of 3DVMs in assessing tumor surgical complexity when planning minimally invasive PN demonstrated their greater precision in comparison to standard 2D imaging [13]. However, despite the growing body of literature on this topic, 3DVMs are not routinely used in clinical practice, primarily because of their production costs and the lack of strong evidence supporting their role [14].
To increase the strength of the current evidence, we conducted a prospective multicenter study to define the role of 3DVMs and nephrometry scores in preoperative planning for robot-assisted PN (RAPN).
2. Patients and methods
2.1. Study design and participants
All patients with a radiological finding of a single organ-confined renal mass diagnosed between June 2019 and September 2022 and suitable for RAPN were included in this multicenter prospective comparative study. Participating centers were selected from those included in the definition and first clinical validation of the European Association of Urology Robotic Urology Section (ERUS) curriculum for RAPN [15].
The study was approved by the ethics committees of the coordinating and participating centers and was registered on ClinicalTrials.gov (NCT05729763).
Each patient enrolled was required to undergo four-phase (unenhanced, corticomedullary, nephrographic, and urographic phases) contrast-enhanced CT within 3 mo before surgery.
Patients were excluded from the study if they had anatomic abnormalities (eg, horseshoe or ectopic kidney), inadequate CT scans (eg, acquisition interval >3 mm), or imaging that was performed more than 3 mo before surgery.
2.2. 3DVM production
Starting from DICOM images, bioengineers from Medics3D Srl (Turin, Italy) produced hyperaccurate 3D models as detailed in previous studies [6], [7], [9].
To guarantee privacy, CT scan images from all the recruiting centers were anonymized and uploaded in DICOM format to a dedicated platform (https://platform.medics3d.com), with scans from each patient labeled with a unique code. After image uploading, the 3DVM was produced and was available for download in .pdf format.
2.3. Assignment of nephrometry scores
CT-based and 3DVM-based PADUA [16] and RENAL [17] nephrometry scores were assigned by two independent expert urologists (>100 renal masses evaluated using nephrometry scores) according to a visual evaluation of the CT scans and 3DVMs. In detail, one urologist from both the promoting and recruiting centers independently scored the same cases using both 2D images and 3DVMs. In all cases, both urologists were blinded to the intraoperative strategy and surgical outcomes. Evaluation of the CT and 3D images was not conducted simultaneously for each case. The first evaluation occurred during uploading of the images to the online platform, while the second took place 72 h later, which is the time required for the bioengineers to produce the 3DVM. In the case of discrepancy between the evaluations, a collegial reassessment was conducted by the two urologists along with a third urologist to reach consensus. The judgments of the evaluators were independent and unaffected by prior assessments.
Finally, 2D- and 3D-based nephrometry risk or complexity categories were assigned on the basis of the PADUA (low risk = 6–7, intermediate risk = 8–9, high risk = ≥10) and RENAL (low complexity = 4–6, moderate complexity = 7–9, high complexity = 10–12) scores (Fig. 1).
Fig. 1.
Examples of tumors for which nephrometry scores were evaluated via three-dimensional virtual models (3DVMs) and two-dimensional (2D) computed tomography (CT). (A) Assessment performed via 3DVM resulted in downgrading of the 2D-based nephrometry scores (both PADUA and RENAL scores and their relative categories). 3DVM was able to show the growth pattern of the tumor more accurately, as well as the relationship to the renal sinus. (B) According to 3DVM assessment, the 2D-based nephrometry scores remained unchanged, indicating a higher rate of concordance between the two imaging tools for high-complexity renal masses. UCS = urinary collecting system.
2.4. 3DVM-assisted surgery
All RAPN procedures were performed robotically by high-volume surgeons (>150 procedures performed) [15] to avoid confounding factors related to differences in surgical expertise. A transperitoneal or retroperitoneal approach was used depending on the tumor location, the patient’s characteristics, and the surgeon’s preference. The vascular pedicle was managed via global clamping, 3D-based selective clamping, or a clampless approach. The tumor was removed according to the principles of enucleation/enucleoresection. Depending on the surgeon’s preference, either one or two running monofilament sutures were used to close the renal defect. Violation of the UCS was managed with sutures. Each specimen was evaluated in each center by a dedicated uropathologist.
2.5. Data collection
For each patient, demographic, perioperative, and pathological data were prospectively collected. Demographic data included age, gender, body mass index (BMI), comorbidities classified according to the age-adjusted Charlson comorbidity index (CCI), and whether or not the patient had a solitary kidney. Preoperative data included clinical tumor size, side, clinical T stage, serum creatinine (sCr), and the estimated glomerular filtration rate (eGFR) as calculated using the Modification of Diet in Renal Disease formula.
2D-based and 3DVM-based PADUA and RENAL scores and risk/complexity categories were compared, with each patient classified as downgraded, unchanged. or upgraded according to the change in score or category after 3DVM assessment.
Intraoperative data included the surgical approach (transperineal vs retroperitoneal), operative time, management of the renal pedicle (global clamping, selective clamping, or clampless), warm ischemia time, resection type (pure enucleation vs enucleoresection), UCS violation, estimated blood loss, intraoperative complications, and conversion to radical nephrectomy.
Postoperative data included length of hospital stay, sCr and eGFR at discharge, and 90-d postoperative complications classified according to the Clavien-Dindo classification. Pathological data comprised tumor stage (8th edition of the TNM classification), tumor grade (World Health Organization/International Society of Urological Pathology 2022 categories), and margin status (R0 or R1).
2.6. Statistical analysis
An appropriate statistical method was chosen for each specific research question in the study. The sample size was calculated using the Casagrande method [18] on the basis of preliminary results from the coordinating center [13] to ensure adequate power for detection of meaningful differences.
To compare categorical variables between groups, frequencies and proportions were used. Student’s t test was used to compare mean values for continuous variables. For comparison of proportions, the Mantel-Haenszel χ2 test was applied to assess differences in categorical data while accounting for potential confounding.
To examine the consistency between the 2D and 3D assessment methods, the weighted Cohen κ coefficient was calculated for the nephrometry scores and categories. This method was selected as it provides a reliable measure of agreement between classifications, which was central to understanding the reliability of the 2D and 3D approaches.
Receiver operating characteristic (ROC) curves were plotted to assess the sensitivity and specificity of the 2D- and 3D-based nephrometry scores and categories in predicting postoperative complications. This analysis addressed the question of whether the 2D or 3D approach is more effective in predicting outcomes. DeLong’s test was applied to determine the statistical significance of differences between the ROC curves.
Univariate and multivariate analyses were conducted to identify preoperative predictors of major postoperative complications. This addressed the research question regarding which variables (BMI, age-adjusted CCI, preoperative eGFR, and each of the 2D- and 3D-based PADUA and RENAL scores and categories) were most strongly associated with the outcomes of interest. The analyses involved eight separate models, with each variable analyzed individually. The threshold for statistical significance was set at p < 0.05. Jamovi v.2.3 was used for statistical analysis.
3. Results
Five high-volume centers in Europe and one center in the USA were included in the study. According to the sample size calculation, which was based on data from the single-center preliminary study [13], 270 patients were needed for 90% power and considering a 10% margin for loss to follow-up loss. Data collection at all six centers was conducted prospectively and a total of 318 patients who met the inclusion criteria were enrolled in the study.
Demographics and preoperative characteristics are reported in Table 1. The median patient age was 65 yr (interquartile range [IQR] 55–72) and 68% of the cohort were male. The mean BMI was 26 kg/m2 (IQR 23–29) and the median age-adjusted CCI was 4 points (IQR 2–5). The cohort included 14 patients (4.4%) with a solitary kidney.
Table 1.
Demographic and preoperative characteristics of the study population (n = 318)
| Parameter | Result |
|---|---|
| Median age, yr (IQR) | 65 (55–72) |
| Male, n (%) | 217 (68) |
| Median body mass index, kg/m2 (IQR) | 26 (23–29) |
| Median age-adjusted CCI, points (IQR) | 4 (2–5) |
| Solitary kidney, n (%) | 14 (4.4) |
| Median clinical tumor size, mm (IQR) | 36 (25–48) |
| Right-sided tumor, n (%) | 153 (48) |
| Clinical T stage, n (%) | |
| cT1a | 190 (60) |
| cT1b | 108 (34) |
| cT2a | 18 (5.4) |
| cT2b | 2 (0.6) |
| Median PADUA score (IQR) | 8 (7–10) |
| Median PADUA risk category (IQR) | 2 (1–3) |
| Median RENAL score (IQR) | 7 (6–9) |
| Median RENAL complexity category (IQR) | 2 (1–2) |
| Median baseline sCr, mg/dl (IQR) | 0.9 (0.7–1.1) |
| Median baseline eGFR, ml/min/m2 (IQR) | 82 (65–97) |
CCI = Charlson comorbidity index; IQR = interquartile range; sCr = serum creatinine; eGFR = estimated glomerular filtration rate.
The median clinical tumor size was 36 mm (IQR 25–48) and 52% of the tumors were in the left kidney. Renal masses were mainly staged as cT1a (190/318, 60%), with a small number of cT2 tumors (20/318, 6.2%) and no cT3 tumors. Regarding the overall surgical complexity of the cohort assessed using the 2D- and 3D-based approaches, the median PADUA score was 8 (IQR 7–10) and the median RENAL score was 7 (IQR 6–9) respectively. At baseline, median sCr was 0.9 mg/dl (IQR 0.7–1.1) and the median eGFR was 82 ml/min/m2 (IQR 65–97).
Table 2 shows the distribution of PADUA and RENAL scores and categories according to the 2D and 3DVM assessments. The distributions of the PADUA scores and risk categories significantly differed between the 2D and 3DVM assessments (both p < 0.001), as did the distributions for the RENAL scores and complexity categories (both p < 0.001). Values for the weighted Cohen κ coefficient were in line with the p values for each variable considered, demonstrating moderate rates of 3DVM-2D concordance at 0.51 for PADUA scores, 0.67 for PADUA risk categories, 0.40 for RENAL scores, and 0.58 for RENAL complexity categories.
Table 2.
Distribution of PADUA and RENAL scores and categories according to 2DCT and 3DVM assessments
| Parameter | Patients, n (%) |
p value | Cohen’s κ | |
|---|---|---|---|---|
| 2DCT | 3DVM | |||
| PADUA score | 0.003 | 0.51 | ||
| 6 | 33 (10) | 64 (20) | ||
| 7 | 72 (23) | 81 (25) | ||
| 8 | 69 (22) | 46 (14) | ||
| 9 | 51 (16) | 64 (20) | ||
| 10 | 42 (13) | 23 (7.2) | ||
| 11 | 20 (6.2) | 19 (5.9) | ||
| 12 | 14 (1.2) | 7 (2.2) | ||
| 13 | 15 (4.7) | 11 (3.4) | ||
| 14 | 2 (0.6) | 3 (0.9) | ||
| PADUA risk category | <0.001 | 0.67 | ||
| Low risk (6–7) | 108 (34) | 147 (46) | ||
| Intermediate risk (8–9) | 119 (37) | 107 (34) | ||
| High risk (≥10) | 91 (29) | 64 (20) | ||
| RENAL score | 0.004 | 0.40 | ||
| 4 | 7 (2) | 16 (5) | ||
| 5 | 37 (12) | 52 (16) | ||
| 6 | 72 (23) | 79 (25) | ||
| 7 | 70 (22) | 50 (16) | ||
| 8 | 36 (11) | 40 (13) | ||
| 9 | 36 (11) | 47 (15) | ||
| 10 | 32 (10) | 24 (7.5) | ||
| 11 | 28 (8.8) | 10 (3.1) | ||
| RENAL complexity category | 0.001 | 0.58 | ||
| Low complexity (4–6) | 116 (36) | 147 (46) | ||
| Intermediate complexity (7–9) | 142 (45) | 137 (43) | ||
| High complexity (≥10) | 60 (19) | 34 (11) | ||
2DCT = two-dimensional computed tomography; 3DVM = three-dimensional virtual model.
Results for changes in the nephrometry scores and categories on 3DVM evaluation are shown in Fig. 2. Downgrading of the nephrometry score at 3DVM assessment occurred in 43% of cases for PADUA scores and 49% of cases for RENAL scores. Similarly, downgrading occurred for 25% of the PADUA and 26% of the RENAL risk/complexity categories. By contrast, a small percentage of patients had upgrading of PADUA and RENAL scores (5.0% for both) or their respective categories (3.4% for both) at 3DVM assessment.
Fig. 2.
Changes in PADUA and RENAL scores and categories from two-dimensional computed tomography (CT) assessment to three-dimensional (3D) assessment using virtual models.
Table 3 reports perioperative and pathological outcomes. Ten surgeons were involved in the study. Transperitoneal RAPN was performed in 93% of the cases, with a median operative time of 114 min (IQR 89–145). Global clamping was performed in 175/318 patients (55%) with a median warm ischemia time of 17 min (IQR 9.0–28), while selective clamping was used in 92/318 cases (29%). It is noteworthy that a clampless approach was adopted in 51/318 patients (16%).
Table 3.
Perioperative and pathological characteristics in the overall population (n = 318)
| Parameter | Result |
|---|---|
| Surgical approach, n (%) | |
| Transperitoneal | 294 (93) |
| Retroperitoneal | 24 (7.5) |
| Median operative time, min (IQR) | 114 (89–145) |
| Management of the renal pedicle, n (%) | |
| Main artery clamping | 175 (55) |
| Selective clamping | 92 (29) |
| Clampless | 51 (16) |
| Median warm ischemia time, min (IQR) | 17 (9.0–28) |
| Extirpative strategy, n (%) | |
| Pure enucleation | 157 (49) |
| Enucleoresection | 161 (51) |
| Opening of the UCS, n (%) | |
| UCS violation | 73 (23) |
| UCS integrity respected | 245 (77) |
| Median estimated blood loss, ml (IQR) | 195 (150–350) |
| Conversion to radical nephrectomy, n (%) | 6 (1.8) |
| Median hospital stay, d (IQR) | 4 (3–6) |
| Median serum creatinine at discharge, mg/dl (IQR) | 1.1 (0.7–1.3) |
| Median eGFR at discharge, ml/min/m2 (IQR) | 75 (58–93) |
| 90-d postoperative complications, n (%) | 63 (20) |
| 90-d major complications (Clavien grade ≥III), n (%) | 12 (3.7) |
| 90-d postoperative complications by grade, n (%) | |
| Clavien grade I | 24 (7.6) |
| Clavien grade II | 27 (8.6) |
| Clavien grade III | 11 (3.5) |
| Clavien grade IV | 1 (0.3) |
| Malignant histology, n (%) | 244 (77) |
| Pathological T stage, n (%) | |
| pT1a | 148 (61) |
| pT1b | 75 (31) |
| pT2a | 14 (5.8) |
| pT2b | 3 (1.3) |
| pT3a | 4 (1.6) |
| ISUP grade group, n (%) | |
| 1 | 73 (30) |
| 2 | 85 (35) |
| 3 | 68 (28) |
| 4 | 7 (2.5) |
| Not applicable | 11 (4.6) |
| Positive surgical margins, n (%) | 10 (4.1) |
IQR = interquartile range; UCS = urinary collecting system; eGFR = estimated glomerular filtration rate; ISUP = International Society of Urological Pathology.
Pure enucleation and enucleoresection rates were similar (49% vs 51%). UCS violation occurred in 23% of the cases. The mean estimated blood loss was 204 ml (standard deviation 207), and 6/318 cases (1.8%) required conversion to radical nephrectomy.
Baseline and postoperative sCr (median 0.9 vs 1.1 mg/dl; p = 0.6) and eGFR (median 82 vs 75 ml/min/m2; p = 0.3) did not differ significantly.
The median length of stay was 4 d (IQR 3–6) and 90-d postoperative complications occurred in 63/318 patients (20%), with only 12 (3.7%) major complications. One patient died of myocardial infarction during the hospital stay (Supplementary Table 2).
Pathology showed malignancy in 76% of the cases, mainly pT1a (61%), with upstaging to pT3a in only four cases (1.6%). The overall positive surgical margin rate for the cohort was 4.1%.
Fig. 3 shows the performance of the 2D-based and 3DVM-based nephrometry scores and corresponding risk/complexity categories in predicting the risk of overall and major postoperative complications. For prediction of overall complications, area under the ROC curve (AUC) values were significantly higher for the 3D-based nephrometry scores versus the 2D-based scores (PADUA: 3D 0.70 vs 2D 0.61, p < 0.001; RENAL: 3D 0.71 vs 2D 0.59, p < 0.001; Fig. 3A). Similar AUC differences were observed for the nephrometry categories (PADUA risk: 3D 0.71 vs 2D 0.60, p < 0.001; RENAL complexity: 3D 0.70 vs 2D 0.56, p < 0.001).
Fig. 3.
Receiver operating characteristic curves showing the performance of each imaging tool for calculating nephrometry scores and the corresponding risk/complexity categories for prediction of the risk of (A) overall and (B) major postoperative complications. 3D = three-dimensional virtual model; CT = computed tomography.
In line with previous findings, prediction of major complications was better with 3D-based versus 2D-based nephrometry scores according to AUC values (PADUA: 3D 0.76 vs 2D 0.66, p < 0.001; RENAL: 3D 0.80 vs 2D 0.58, p = 0.001) and categories (PADUA risk: 3D 0.76 vs 2D 0.67, p = 0.001; RENAL complexity: 3D 0.79 vs 2D 0.53, p = 0.001), as shown in Fig. 3B.
Finally, multivariable analyses (Supplementary Table 1) to assess potential predictors of major postoperative complications revealed that 3D-based PADUA scores (odds ratio [OR] 1.40, 95% confidence interval [CI] 1.08–1.82; p = 0.011), PADUA categories (OR 3.83, 95% CI 1.58–9.26; p = 0.003), RENAL scores (OR 1.80, 95% CI 1.25–2.60; p = 0.002), and RENAL categories (OR 4.36, 95% CI 1.78–10.69; p = 0.001) had significant predictive value.
4. Discussion
Nephrometry scores are commonly used before minimally invasive PN [19], [20], [21], although their prognostic value is still a matter of debate [4] and they have historically been calculated using 2D imaging only [22]. Porpiglia et al [13] attempted to expand the application by evaluating 101 organ-confined renal masses using 3DVMs instead of 2D imaging and demonstrated an overall reduction in surgical complexity and more accurate prediction of postoperative complications. Although preliminary, the study highlighted the potential role of 3DVMs when assessing renal masses.
Our prospective multicenter comparative study strengthens this evidence for the RAPN setting. The distributions of nephrometry scores and risk/complexity categories significantly differed between the 2D and 3DVM approaches (p < 0.001), which was confirmed by the moderate concordance indicated by values for Cohen’s κ coefficient (range 0.40–0.67).
In addition, 3DVM assessment led to a reduction in anatomic complexity according to nephrometry scores (in 43% of cases for PADUA and 49% for RENAL) and risk/complexity categories (in 25% of cases for PADUA risk and 26% for RENAL complexity), in line with previous results.
Rocco et al [23] used a similar technology for preoperative planning of ten RAPN procedures. The authors observed 3DVM downgrading of tumor complexity according to both PADUA and RENAL scores; the lowest agreement between the 2D-based and 3D-based scores was for variables related to protrusion of the mass (ie, exophytic ratio) and its relationship to the hilum. Campos et al [24] compared eight CT/MRI scans and corresponding 3DVMs, with five radiologists and five urologists evaluating each case, for a total of 160 assessments. After 3DVM evaluation, the 2D-based RENAL score differed for 65/80 evaluations (81%), with a lower score found in 44 cases (67%). Bianchi et al [25] calculated different nephrometry scores from 58 3DVMs and observed significant differences in distribution in comparison to the corresponding CT-based scores. The same group developed a novel complexity assessment tool based on morphological and volumetric parameters from 3DVMs for predicting surgical outcomes for 69 renal masses [26]. Yoshitomi et al [27] compared 2D- and 3D-based RENAL scores and found high heterogeneity between the results, and a better understanding of the tumor-UCS interface using the 3D approach.
In terms of clinical implications, our ROC analyses demonstrated better performance of the 3DVM-based nephrometry scores and risk/complexity categories in predicting overall and major complications (Fig. 3) in comparison to the corresponding 2D-based parameters. These findings are supported by separate multivariable models (Supplementary Table 1) that demonstrated significant associations of both 3DVM-based and 2D-based nephrometry scores and risk/complexity categories with major postoperative complications when analyzed individually. These models assessed associations, and not confirmations, and did not compare variables within the same model.
Our results are supported by previous studies [13], [26], [27], [28], [29]. Bianchi et al [25] found that 3D PADUA, RENAL, and ABC scores were more reliable than the corresponding 2D scores in predicting the occurrence of overall complications in PN cases. Mashin et al [28] investigated the role of 3D-based nephrometry scores in predicting intraoperative and postoperative complications in an RAPN cohort. The authors evaluated 264 cases using C-index, PADUA, RENAL, and a newly developed 3D-based nephrometry score that was better than all the others in predicting the likelihood of complications because of the greater accuracy of 3DVMs in representing the anatomy. Li et al [29] evaluated anatomic tumor parameters using 3D models for 232 patients undergoing RAPN. The authors developed the ADDD score (diameter, depth, distance), which was a predictor of postoperative complications (hazard ratio 1.501, 95% CI 1.178–1.914).
Our trial confirms these previous results and, as the largest study and the only one conducted in a multicenter setting, increases the strength of the evidence. Our findings demonstrate that, regardless of the nephrometry score used, each imaging technique has a role in predicting PN outcomes, and 3DVMs are more accurate in predicting postoperative complications than 2D CT imaging and increase surgical accuracy by offering precise anatomic details. In addition, our study underlines once again the need to use the right imaging technology during preoperative planning given the potential impact on both the surgical indication and technique [30].
The downgrading of nephrometry category on 3DVM evaluation for at least 25% of our cohort suggests that a certain number of cases are surgically easier than they appear at CT imaging and can be treated safely with a conservative approach. Conversely, patients for whom 3DVM assessment confirms surgical complexity are more likely to be at risk of postoperative complications and should be more carefully evaluated regarding the surgical indication at the time of planning and patient counseling. With the proper imaging tool used for preoperative evaluation, a non-negligible proportion of renal units may be saved among patients for whom a nephron-sparing instead of a radical approach is indicated [31].
Our study has several limitations. There was a risk of interobserver variability as multiple urologists were involved in the evaluation process. Patients were not included consecutively because not all had imaging characteristics suitable for adequate reconstruction for the study aim. Possible delays or errors in uploading CT scan files to the online platform by researchers also contributed to the nonconsecutive inclusion of cases. In addition, 3DVM production is expensive and highly dependent on the availability of biomedical engineers, which may slow diffusion of this technology beyond the experimental setting. However, studies are currently under way to automate the 3DVM production process using artificial intelligence strategies, with the aim of reducing production times and costs in the near future. If the production techniques are optimized, it should be possible to implement 3DVMs in daily clinical practice outside experimental settings and in nontertiary centers.
Despite its limitations, our prospective multicenter comparative study provides evidence supporting the importance of imaging assessment tools in preoperative planning for RAPN. Specifically, 3DVMs were more accurate than 2D images in assessing the surgical complexity of renal masses, and often revealed lower complexity and predicted a lower risk of postoperative complications in comparison to 2D images. This better accuracy of 3DVMs has the potential to increase the number of cases for which a nephron-sparing approach is indicated in comparison to traditional 2D imaging assessment.
5. Conclusions
Our prospective multicenter study demonstrated differences in nephrometry scores when calculated using 2D CT scans versus 3DVMs, with a reduction in score in most cases on 3DVM assessment. This evidence suggests that the perceived surgical complexity of renal masses is lower when 3DVMs are used. In addition, 3DVMs were more accurate in predicting cases with a higher risk of postoperative complications in comparison to 2D scans. Thus, 3DVMs could potentially support greater adoption of nephron-sparing approaches among candidates with a suitable safety profile.
Author contributions: Daniele Amparore had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Amparore, Porpiglia.
Acquisition of data: Verri, Checcucci, Piana, Basile, Larcher, Gallioli, Territo, Gaya, Piazza, Puliatti, Grosso, Mari, Campi, Zuluaga, Burak.
Analysis and interpretation of data: Amparore, Piramide, Verri.
Drafting of the manuscript: Amparore, Piramide, Verri.
Critical revision of the manuscript for important intellectual content: Ketan, Serni, Capitanio, Montorsi, Mottrie, Fiori, Minervini, Wiklund, Breda, Porpiglia.
Statistical analysis: Amparore, Piramide.
Obtaining funding: None.
Administrative, technical, or material support: None.
Supervision: Ketan, Serni, Capitanio, Montorsi, Mottrie, Fiori, Minervini, Wiklund, Breda, Porpiglia.
Other: None.
Financial disclosures: Daniele Amparore certifies that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: None.
Funding/Support and role of the sponsor: None.
Associate Editor: M. Carmen Mir
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.euros.2025.02.001.
Appendix A. Supplementary material
The following are the Supplementary data to this article:
References
- 1.Ljungberg B., Albiges L., Abu-Ghanem Y., et al. European Association of Urology guidelines on renal cell carcinoma: the 2022 update. Eur Urol. 2022;82:399–410. doi: 10.1016/j.eururo.2022.03.006. [DOI] [PubMed] [Google Scholar]
- 2.Campi R., Grosso A.A., Lane B.R., et al. Impact of trifecta definition on rates and predictors of “successful” robotic partial nephrectomy for localized renal masses: results from the Surface-Intermediate-Base Margin Score International Consortium. Minerva Urol Nephrol. 2022;74:186–193. doi: 10.23736/S2724-6051.21.04601-2. [DOI] [PubMed] [Google Scholar]
- 3.Klatte T., Ficarra V., Gratzke C., et al. A literature review of renal surgical anatomy and surgical strategies for partial nephrectomy. Eur Urol. 2015;68:980–992. doi: 10.1016/j.eururo.2015.04.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Veccia A., Antonelli A., Uzzo R.G., et al. Predictive value of nephrometry scores in nephron-sparing surgery: a systematic review and meta-analysis. Eur Urol Focus. 2020;6:490–504. doi: 10.1016/j.euf.2019.11.004. [DOI] [PubMed] [Google Scholar]
- 5.Porpiglia F., Amparore D., Checcucci E., et al. Current use of three-dimensional model technology in urology: a road map for personalised surgical planning. Eur Urol Focus. 2018;4:652–656. doi: 10.1016/j.euf.2018.09.012. [DOI] [PubMed] [Google Scholar]
- 6.Porpiglia F., Bertolo R., Checcucci E., et al. Development and validation of 3D printed virtual models for robot-assisted radical prostatectomy and partial nephrectomy: urologists’ and patients’ perception. World J Urol. 2018;36:201–207. doi: 10.1007/s00345-017-2126-1. [DOI] [PubMed] [Google Scholar]
- 7.Amparore D., Piramide F., Checcucci E., et al. Three-dimensional virtual models of the kidney with colored perfusion regions: a new algorithm-based tool for optimizing the clamping strategy during robot-assisted partial nephrectomy. Eur Urol. 2023;84:418–425. doi: 10.1016/j.eururo.2023.04.005. [DOI] [PubMed] [Google Scholar]
- 8.Amparore D., Pecoraro A., Checcucci E., et al. 3D imaging technologies in minimally invasive kidney and prostate cancer surgery: which is the urologists’ perception? Minerva Urol Nephrol. 2022;74:178–185. doi: 10.23736/S2724-6051.21.04131-X. [DOI] [PubMed] [Google Scholar]
- 9.Porpiglia F., Checcucci E., Amparore D., et al. Three-dimensional augmented reality robot-assisted partial nephrectomy in case of complex tumours (PADUA ≥10): a new intraoperative tool overcoming the ultrasound guidance. Eur Urol. 2020;78:229–238. doi: 10.1016/j.eururo.2019.11.024. [DOI] [PubMed] [Google Scholar]
- 10.Piramide F., Kowalewski K.F., Cacciamani G., et al. Three-dimensional model-assisted minimally invasive partial nephrectomy: a systematic review with meta-analysis of comparative studies. Eur Urol Oncol. 2022;5:640–650. doi: 10.1016/j.euo.2022.09.003. [DOI] [PubMed] [Google Scholar]
- 11.Pecoraro A., Amparore D., Checcucci E., et al. Three-dimensional virtual models assistance predicts higher rates of “successful” minimally invasive partial nephrectomy: an institutional analysis across the available trifecta definitions. World J Urol. 2023;41:1093–1100. doi: 10.1007/s00345-023-04310-5. [DOI] [PubMed] [Google Scholar]
- 12.Amparore D., Pecoraro A., Checcucci E., et al. Three-dimensional virtual models' assistance during minimally invasive partial nephrectomy minimizes the impairment of kidney function. Eur Urol Oncol. 2022;5:104–108. doi: 10.1016/j.euo.2021.04.001. [DOI] [PubMed] [Google Scholar]
- 13.Porpiglia F., Amparore D., Checcucci E., et al. Three-dimensional virtual imaging of renal tumours: a new tool to improve the accuracy of nephrometry scores. BJU Int. 2019;124:945–954. doi: 10.1111/bju.14894. [DOI] [PubMed] [Google Scholar]
- 14.Bertolo R., Kuusk T., Mir M.C. Three-dimensional models-assisted minimally-invasive partial nephrectomy: looking forward to more evidence. Minerva Urol Nephrol. 2023;75:665–666. doi: 10.23736/S2724-6051.23.05446-0. [DOI] [PubMed] [Google Scholar]
- 15.Larcher A., De Naeyer G., Turri F., et al. The ERUS curriculum for robot-assisted partial nephrectomy: structure definition and pilot clinical validation. Eur Urol. 2019;75:1023–1031. doi: 10.1016/j.eururo.2019.02.031. [DOI] [PubMed] [Google Scholar]
- 16.Ficarra V., Novara G., Secco S., et al. Preoperative aspects and dimensions used for an anatomical (PADUA) classification of renal tumours in patients who are candidates for nephron-sparing surgery. Eur Urol. 2009;56:786–793. doi: 10.1016/j.eururo.2009.07.040. [DOI] [PubMed] [Google Scholar]
- 17.Kutikov A., Uzzo R.G. The R.E.N.A.L. nephrometry score: a comprehensive standardized system for quantitating renal tumor size, location and depth. J Urol. 2009;182:844–853. doi: 10.1016/j.juro.2009.05.035. [DOI] [PubMed] [Google Scholar]
- 18.Casagrande J.T., Pike M.C., Smith P.G. An improved approximate formula for calculating sample sizes for comparing two binomial distributions. Biometrics. 1978;34:483–486. [PubMed] [Google Scholar]
- 19.Mottrie A., Gandaglia G. Do we need a novel nephrometry scoring system in partial nephrectomy? Eur Urol. 2016;69:80–81. doi: 10.1016/j.eururo.2015.08.058. [DOI] [PubMed] [Google Scholar]
- 20.Alvim R.G., Audenet F., Vertosick E.A., Sjoberg D.D., Touijer K.A. Performance prediction for surgical outcomes in partial nephrectomy using nephrometry scores: a comparison of Arterial Based Complexity (ABC), RENAL, and PADUA systems. Eur Urol Oncol. 2018;1:428–434. doi: 10.1016/j.euo.2018.05.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Daza J., Okhawere K.E., Ige O., et al. The role of RENAL score in predicting complications after robotic partial nephrectomy. Minerva Urol Nephrol. 2022;74:57–62. doi: 10.23736/S2724-6051.20.03608-5. [DOI] [PubMed] [Google Scholar]
- 22.Wadle J., Hetjens S., Winter J., et al. Nephrometry scores: the effect of imaging on routine read-out and prediction of outcome of nephron-sparing surgery. Anticancer Res. 2018;38:3037–3041. doi: 10.21873/anticanres.12559. [DOI] [PubMed] [Google Scholar]
- 23.Rocco B., Sighinolfi M.C., Menezes A.D., et al. Three-dimensional virtual reconstruction with DocDo, a novel interactive tool to score renal mass complexity. BJU Int. 2020;125:761–762. doi: 10.1111/bju.15049. [DOI] [PubMed] [Google Scholar]
- 24.Campos T.J.F.L., de V Filho F.E., Rocha M.F.H. Assessment of the complexity of renal tumors by nephrometry (R.E.N.A.L. score) with CT and MRI images versus 3D reconstruction model images. Int Braz J Urol. 2021;47:896–901. doi: 10.1590/S1677-5538.IBJU.2020.0930. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Bianchi L., Schiavina R., Bortolani B., et al. Interpreting nephrometry scores with three-dimensional virtual modelling for better planning of robotic partial nephrectomy and predicting complications. Urol Oncol. 2021;39:836.e1–836.e9. doi: 10.1016/j.urolonc.2021.07.024. [DOI] [PubMed] [Google Scholar]
- 26.Bianchi L., Schiavina R., Bortolani B., et al. Novel volumetric and morphological parameters derived from three-dimensional virtual modeling to improve comprehension of tumor’s anatomy in patients with renal cancer. Eur Urol Focus. 2022;8:1300–1308. doi: 10.1016/j.euf.2021.08.002. [DOI] [PubMed] [Google Scholar]
- 27.Yoshitomi K.K., Komai Y., Yamamoto T., et al. Improving accuracy, reliability, and efficiency of the RENAL nephrometry score with 3D reconstructed virtual imaging. Urology. 2022;164:286–292. doi: 10.1016/j.urology.2022.01.024. [DOI] [PubMed] [Google Scholar]
- 28.Mashin G.A., Kozlov V.V., Chinenov D.V., et al. Nephrometric score based on 3D modeling (3D nephrometry score) for the probability prediction of intra- and postoperative complications for kidney surgery. Urologia. 2022;89:179–184. doi: 10.1177/03915603211036427. [DOI] [PubMed] [Google Scholar]
- 29.Li X., Wu D., Zhang X., et al. A three-dimensional renal tumor anatomy and intrarenal relationship nephrometry (ADDD) for robot-assisted partial nephrectomy: 3D-CT based nephrometry for RAPN. World J Urol. 2023;41:1847–1853. doi: 10.1007/s00345-023-04448-2. [DOI] [PubMed] [Google Scholar]
- 30.Checcucci E., Piana A., Volpi G., et al. Visual extended reality tools in image-guided surgery in urology: a systematic review. Eur J Nucl Med Mol Imaging. 2024;51:3109–3134. doi: 10.1007/s00259-024-06699-6. [DOI] [PubMed] [Google Scholar]
- 31.Bertolo R., Autorino R., Fiori C., et al. Expanding the indications of robotic partial nephrectomy for highly complex renal tumors: urologists’ perception of the impact of hyperaccuracy three-dimensional reconstruction. J Laparoendosc Adv Surg Tech A. 2019;29:233–239. doi: 10.1089/lap.2018.0486. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.