Centrality angle is a novel nephrometry score to predict tumor complexity and perioperative outcomes for partial nephrectomy

To propose the centrality angle (C-angle) as a novel simple nephrometry score for the evaluation of tumor complexity and prediction of perioperative outcomes in nephron-sparing surgery (NSS) for renal tumors. The analysis was based on 174 patients who underwent robot-assisted partial nephrectomy retrospectively. C-angle was defined as the angle occupied by the tumor from the center of the kidney in the coronal CT images. Other nephrometry scores were calculated and compared with C-angle. Associations between C-angle and perioperative outcomes were examined. Significant differences were found in C-angle between tumors greater and less than 4 cm, exophytic and endophytic tumors, and hilar and non-hilar tumors. C-angle was correlated with other nephrometry scores, including RENAL, PADUA, and C-index. Significant positive correlations with WIT, operation time, and EBL, and significant negative correlations with preserved eGFR. C-angle could predict perioperative complications. Patients with a C-angle > 45° had worse perioperative outcomes, including longer operative time, longer WIT, lower rate of preserved eGFR, and complications. C-angle can be used to evaluate the complexity of renal tumors and predict perioperative outcomes. C-angle can potentially be used for decision-making in the treatment of patients and to guide surgical planning of NSS.


Materials and methods
This was a retrospective observational study conducted in accordance with the ethical standards described in the Declaration of Helsinki.This study was approved by the Research Ethics Committee of Hiroshima University and the requirement for informed consent was waived by the Research Ethics Committee of Hiroshima University (authorization number: E2016-0588).

Patient selection
We reviewed 177 patients who underwent NSS with thin-slice preoperative CT images available at the Hiroshima University Hospital from January 2015 to May 2022.Because vertex of the C-angle and tumor was overlap, three of the 177 patients had not measurable and be excluded.Therefore, 174 patients were included in the study cohort.In all cases, 4 experienced surgeons performed robot-assisted partial nephrectomy (RAPN).Preoperative demographics (age, body mass index, and sex), tumor characteristics (tumor location, tumor size, clinical T stage, exophytic rate, and hilar tumor), nephrometry scores (C-angle, RENAL, PADUA, and C-index), and pathological features were obtained from the medical records and are listed in Table 1.
Postoperative complications were retrospectively collected through chart review by a medical doctor at 90 days after surgery according to the EAU Guidelines Panel recommendations on reporting and grading complications 14 .Postoperative complications were graded according to the Clavien-Dindo system 15 .Renal function was assessed by eGFR using serum creatinine level based on the Modification of Diet in Renal Disease equation.Postoperative renal function was defined as eGFR measured 6 months after RAPN 16 .

Measurement method of C-angle
The C-angle measurement method required only coronal CT imaging.First, a mid-polar reference point (x) corresponding to the center of the kidney was defined using the same procedure as for the C-index 10 .The midpolar reference axis (y) was defined as the line vertical to the coronal CT slice through the mid-polar reference point.C-angle (z) was measured from the angle from the mid-polar reference axis to both outlines of the tumor invading the kidney using the slice with the largest diameter of the tumor (Fig. 1A).In practice, the most ventral and dorsal kidney borders were identified on coronal CT imaging, and the middle slice was determined as the center slice of the most ventral and dorsal slices.In the middle slice, a mid-polar reference point was assigned to the center of an ellipse outlining the kidney.Next, CT imaging scanned the slice with the largest tumor diameter buried in the kidney and measured the C-angle from the mid polar reference axis to both outlines of the tumor (Fig. 1B).
C-angle values were measured and statistically analyzed for inter-observer reliability by 2 experienced urologists.

Imaging protocol
All imaging was performed using a 16-or 64-slice MDCT scanner (LightSpeed or VCT, GE Healthcare).We acquired unenhanced and dynamic CT scans, including the nephrographic and excretory phases of all patients before RAPN.Coronal images were obtained at a slice thickness of 2 mm.A hilar tumor was defined as follows for the present study: (i) located entirely between the polar lines or crossing a polar line, and (ii) a maximum 5-mm distance between the tumor border and the point where the renal artery or vein enters the renal parenchyma 17 .

Statistical analysis
Fisher's exact test was used to compare the distributions of categorical variables.Differences in variables with a continuous distribution across dichotomous categories were tested using the Mann-Whitney U test.Univariate analysis was performed using Spearman correlation coefficients to identify associations between a continuous variable of perioperative outcomes and nephrometry scores, including C-angle, RENAL, PADUA, and C-index.Receiver operating characteristic (ROC) curves and area under the curve (AUC) were used to reveal the predictive ability of complications for nephrometry scores, as described previously 18 .The DeLong test was used to compare ROC curves.Univariate analysis of C-angle dichotomized into the C-angle < 45° group and C-angle ≥ 45° group was compared with clinical data or perioperative variables using Pearson's chi-square and Fisher's exact tests.Statistical significance was set at p < 0.05.All statistical analyses were performed using JMP Pro 14.0.0 (SAS Institute, Cary, NC, USA).
The C-Angle demonstrated no significant differences in gender (p = 0.384) and revealed no correlation with age in the examined population.(r = − 0.095 = 0.211) (Supplemental Fig. 1).

C-angle was significant related to the complexity of the renal masses
The value of the C-angle is larger in the case of large tumors, endophytic tumors, and hilar renal tumors (Fig. 1C-E).When the C-angle was calculated in 174 cases, significant differences were found between tumors of < 4 cm and > 4 cm (p < 0.001), exophytic and endophytic tumors (p < 0.001), and hilar and non-hilar tumors (p < 0.001) (Fig. 1F-H).Together, this suggests that the C-angle is related to the complexity characteristics of the renal masses.To evaluate the clinical utility of C-angle as a nephrometry score for RAPN, each C-angle value was compared to existing nephrometry scores including RENAL, PADUA, and C-index.These scores are related to perioperative parameters, including warm ischemia time (WIT), operative time, EBL, preserved eGFR, and complications.

Comparison of predictive value of perioperative complications between C-angle and other nephrometry scores
To evaluate the capability of C-angle as a predictor of perioperative complications, ROC curves of C-angle for the incidence of perioperative complications were drawn and compared with those of the other nephrometry scores (Fig. 4).A C-angle cutoff value of 67.3° to predict complications was determined by ROC curve analysis, with a sensitivity of 47% and specificity of 84% in our cohort (Youden index = 0.30).The AUC value for C-angle in predicting perioperative complications was 0.688 (95% CI 0.590-0.772).The AUC value showed no significant difference between C-angle and the other nephrometry scores, including RENAL (AUC = 0.676, 95% CI 0.579-0.761,p = 0.757), PADUA (AUC = 0.679, 95% CI 0.580-0.764,p = 0.837), and C-index (AUC = 0.693, 95% CI 0.593-0.777,p = 0.847) (Table 2).

Inter-observer agreement for C-angle
To evaluate the reproducibility of the C-angle, we compared the C-angles of the two observers in a scatterplot.C-angle had a strong correlation between the two observers.(r = 0.791 p < 0.001) (Fig. 5).

C-angle as a representation of renal tumor complexity and association with perioperative outcomes and other nephrometry scores
To evaluate whether C-angle could represent the clinical characteristics of renal tumors, all 174 patients were dichotomized according to a 45° C-angle.Despite the lack of significant differences in patient demographics between the two groups, significant differences were found in tumor size (p < 0.001), clinical T1b (p = 0.004), endophytic status (p = 0.034), and hilar tumor (p = 0.004).Also, relationships between the higher C-angle group and higher RENAL score (p < 0.001), higher PADUA score (p < 0.001), and lower C-index (p < 0.001) were observed.No significant associations were found between C-angle and pathological outcomes.Patients in the greater than 45° C-angle group had worse perioperative outcomes, including longer operative time (p = 0.010), longer WIT (p < 0.001), and other complications (p = 0.028).Regarding preserved renal function, a greater reduction in eGFR was found for C-angle greater than 45° (p < 0.001).In conjunction with these results, C-angle was able to represent the complexity of the renal tumor and was associated with a worse perioperative outcome and lower rate of residual renal function after RAPN (Table 3).

Discussion
In this study, we showed the usefulness of the C-angle, as calculated from CT images, to evaluate the complexity of renal tumors and predict perioperative outcomes.Preoperative evaluation using nephrometry scores, such as RENAL or PADUA, is generally recommended to predict the potential morbidity of NSS 3 .However, these scoring systems have limitations such as insufficient interobserver reproducibility and incomplete quantification of relevant anatomical features.In addition, overlaps exist between some features of nephrometry score parameters that may complicate preoperative evaluation.Consequently, the Simplified PADUA REnal (SPARE) nephrometry system, which excluded polar location and urinary collecting system involvement, was suggested and had similar predictive accuracy to the original PADUA score 19 .Likewise, the existing mathematical-based nephrometry scores might be complicated for clinical use because they require three-dimensional construction, very thin-slice CT imaging, or structural assumptions.Therefore, a simplified mathematical assessment-based nephrometry score might also be useful.C-angle is simply the angle a renal tumor occupies from the center of the kidney and can be measured using only coronal CT imaging without complex calculations.Considering the anatomical features and surgery of the kidney, C-angle is reasonably accessible in terms of the radial anatomical architecture and surgical approach for partial nephrectomy, because the incision is usually performed vertically from the surface along an anatomical architecture 20 .
Because significant associations between C-angle and other existing nephrometry scores were found, C-angle could involve the components of RENAL, PADUA, and C-index.In the present study, C-angle might be useful as a nephrometry score because it correlated with WIT, operative time, EBL, residual renal function, and perioperative complications, as well as with other nephrometry scores.
In our cohort, we observed a high inter-observer reliability in the C-angle.C-angle may serve as a nephrometry score with high reproducibility.
When perioperative outcomes were compared by dichotomizing C-angle into 45° groups, patients with tumors > 45° had worse perioperative outcomes, such as longer operative time, longer WIT, increased complications, and decreased renal function.Therefore, the simple nephrometry score expressed by C-angle was sufficiently predictive of perioperative outcomes and might be an indicator of technical difficulty for RAPN.This study had several limitations.First, this was a retrospective cohort with relatively few cases, and the outcomes should be validated in a larger cohort.In addition, C-angle could not be measured in three cases because it overlapped with the axial polar axis.In such cases, measurement may be possible using axial CT imaging.Furthermore, this research is an initial report and has not been externally verified to ensure the objectivity of C-angle.It will be necessary to increase the reliability of C-angle through external verification.Finally, C-angle represents only some aspects of renal tumor complexity and lacks geographical location information for the kidney.For accurate evaluation of tumor complexity, C-angle should be combined with original CT imaging or other nephrometry scores.

Conclusion
C-angle might be a suitable criterion for evaluating the complexity of renal tumors.This novel nephrometry score is intuitive and can easily be measured using CT imaging alone.C-angle can predict significant perioperative outcomes, including operative time, WIT, and preserved renal function, as effectively as other existing

Figure 1 .
Figure 1.Measurement method of C-angle.(A) Measurement method model of C-angle.x, mid-polar reference point.y, mid-polar reference axis.C-angle was measured from the mid-polar reference axis to both outlines of the tumor invading into the kidney (z).(B) Measurement of C-angle using actual CT imaging.(C) Illustrations of concepts of size (C), endophyticity (D), and hilar tumor (E).Comparison of C-angle according to tumor size (F), endophyticity (G), and hilar tumor (H).*p < 0.001.

Figure 4 .
Figure 4. ROC curves and AUCs of C-angle, RENAL score, PADUA score, and C-index for the occurrence of perioperative complications.

Figure 5 .
Figure 5. Correlation between Observer A and B for C-angle.

Table 1 .
Characteristics of patients, tumors and outcomes.

Table 2 .
Predictive value of perioperative complications between C-angle and other nephrometry scores.

Table 3 .
Association between C-angle and patient demographics, tumor characteristics, and pathological and perioperative outcomes.