Abstract
Background
The pseudocapsule (PC) status of clear cell renal cell carcinoma (ccRCC) is closely associated with tumor aggressiveness and surgical strategy, yet accurate preoperative non-invasive assessment remains challenging. This study aimed to develop and internally validate a contrast-enhanced ultrasound (CEUS)-based prediction model for assessing PC status in patients with ccRCC.
Methods
This retrospective study enrolled 254 patients with pathologically confirmed ccRCC who underwent surgery at Shanghai General Hospital, between August 2021 and December 2024. All patients received preoperative conventional ultrasound and CEUS examinations. Independent risk factors for PC absence or penetration were identified using multivariable binary logistic regression analysis. These factors were then incorporated into a nomogram to construct a predictive model. The model’s discriminatory performance was assessed by the area under the receiver operating characteristic (ROC) curve (AUC). Internal validation was performed using bootstrap resampling to generate calibration curves and evaluate goodness-of-fit via the Hosmer-Lemeshow test. Finally, the clinical utility of the nomogram was determined using decision curve analysis (DCA).
Results
Three CEUS features—incomplete PC sign, peak hypoenhancement, and rapid washout pattern—were independent predictors in the final model. The nomogram showed acceptable discrimination [training AUC: 0.779, 95% confidence interval (CI): 0.692–0.867; validation AUC: 0.749, 95% CI: 0.600–0.898] and calibration (Hosmer-Lemeshow P>0.05). DCA supported its potential clinical utility.
Conclusions
In conclusion, a CEUS-based nomogram was developed and internally validated to preoperatively predict PC status in ccRCC. This practical, non-invasive tool can guide and optimize surgical planning.
Keywords: Contrast enhanced ultrasound (CEUS), clear cell renal cell carcinoma (ccRCC), pseudocapsule (PC)
Highlight box.
Key findings
• We developed and internally validated a contrast enhanced ultrasound (CEUS)-based nomogram for preoperative prediction of pathological pseudocapsule (PC) breaching in localized clear cell renal cell carcinoma (ccRCC).
• Incomplete PC sign, peak hypoenhancement, and rapid washout were identified as independent predictors.
• The nomogram achieved acceptable discrimination (area under the curve 0.779/0.749) and favorable clinical net benefit.
What is known and what is new?
• PC status predicts ccRCC aggressiveness and affects surgical strategy. CEUS enables high-resolution visualization of tumor microperfusion.
• This is the first dedicated nomogram using intuitive CEUS features for noninvasive, individualized prediction of pathological PC integrity.
What is the implication, and what should change now?
• This radiation-free, easy-to-use tool enables preoperative risk stratification to guide surgical planning.
• We recommend incorporating this CEUS nomogram into preoperative evaluation for patients with localized ccRCC.
Introduction
Renal cell carcinoma (RCC) is the 14th most common malignancy worldwide, with more than 435,000 new cases diagnosed annually (1). Clear cell renal cell carcinoma (ccRCC) accounts for approximately 80% of RCCs and represents the most prevalent solid renal malignancy in adults (2,3). A distinctive histopathologic feature of ccRCC is the pseudocapsule (PC), a fibrous structure formed at the tumor-parenchyma interface. Clinically, an intact PC functions as a biological and anatomic barrier that constrains local tumor spread and provides a natural dissection plane for nephron-sparing surgery (4).
From a clinical perspective, an intact PC serves as a natural anatomic barrier that facilitates nephron-sparing surgery by providing a favorable dissection plane, reducing intraoperative blood loss, and lowering the risk of positive surgical margins. It is also associated with confined tumor growth and more favorable oncologic outcomes. In contrast, disruption or absence of the PC reflects loss of barrier function and is closely associated with aggressive tumor behavior, increased risk of local invasion and metastasis, and shorter progression-free survival (5). Therefore, accurate preoperative, noninvasive assessment of PC status is clinically relevant for surgical planning, particularly for selecting candidates for partial nephrectomy or tumor enucleation, and for postoperative risk stratification.
Computed tomography (CT) is important for assessing renal tumor morphology and T staging (6). However, its ability to reliably evaluate PC invasion remains limited. Diagnostic performance is influenced by tumor size, imaging resolution, and reader experience, and subtle PC disruption is often difficult to detect using conventional cross-sectional imaging (7). Contrast-enhanced ultrasound (CEUS) is a radiation-free, real-time imaging technique that visualizes tumor microvascular perfusion using microbubble contrast agents (8,9). Owing to its high temporal resolution and sensitivity to microcirculatory dynamics, CEUS allows detailed visualization of perfusion patterns and tumor-parenchyma interfaces and therefore has theoretical advantages for evaluating PC integrity (10). These advantages have been further consolidated by recent reviews systematically summarizing the clinical applications and standardized protocols of CEUS for renal masses (11,12). Nevertheless, current CEUS research has primarily focused on tumor detection, differentiation, and grading (13-15), whereas systematic investigations specifically addressing its value for preoperative assessment of PC status in ccRCC remain limited.
Nomograms are widely used clinical prediction tools that integrate multiple independent prognostic or predictive variables to generate individualized risk estimates in a graphical and clinically intuitive manner (10). Although several studies have developed nomogram models based on clinical and CT or magnetic resonance imaging-derived radiological features to predict pathological characteristics of RCC, such as nuclear grade, metastasis, or PC invasion (13,14), CEUS-specific microvascular and boundary features have not yet been fully incorporated into such predictive frameworks, despite their potential relevance to PC assessment.
Accordingly, a reliable and clinically applicable nomogram may be constructed using multidimensional features derived from preoperative CEUS to enable noninvasive prediction of PC status in localized ccRCC. The purpose of this study was to retrospectively develop and internally validate a CEUS-based predictive nomogram for preoperative assessment of PC integrity, with the objective of providing surgeons with an objective decision-support tool to optimize individualized surgical planning and risk stratification. We present this article in accordance with the TRIPOD reporting checklist (available at https://tau.amegroups.com/article/view/10.21037/tau-2026-0312/rc).
Methods
Patients
This retrospective study enrolled 254 patients with ccRCC who were admitted to Shanghai General Hospital between August 2021 and December 2024. Conventional ultrasound (US) and CEUS examinations were performed within the month prior to surgery. The inclusion criteria were: (I) pathologically confirmed ccRCC and (II) preoperative clinical T-stage I–II localized RCC. Exclusion criteria included: (I) patients without pathological results; (II) benign lesions, non-ccRCC, renal pelvis lesions, and other rare types of renal tumors; (III) missing or poor-quality CEUS images; (IV) multiple ipsilateral or bilateral ccRCCs; and (V) incomplete clinical data. The study design and patient flow are presented in Figure 1. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Institutional Review Board of Shanghai General Hospital under expedited review (No. 2026-278), and the requirement for written informed consent was waived owing to the retrospective nature of this study.
Figure 1.
Flowchart of patient selection. ccRCCs, clear cell renal cell carcinomas; CEUS, contrast enhanced ultrasound.
Clinical and laboratory data
The clinical data that were extracted from electronic medical records included: (I) demographic information: age, sex, and body mass index (BMI); (II) medical history: clinical symptoms (abdominal pain, flank pain, and gross hematuria) and hypertension (systolic blood pressure ≥140 mmHg and/or diastolic blood pressure ≥90 mmHg); and (III) laboratory indicators: serum creatinine (Scr), albumin levels, C-reactive protein-to-albumin ratio (CAR), neutrophil-to-lymphocyte ratio (NLR), and microscopic hematuria. All data were obtained within the week before surgery.
Conventional US and CEUS examination
The patient was positioned in the supine, lateral, or prone position, as required. The instruments used to acquire the CEUS images in this study included GE LOGIQ E9 (General Electric Co., USA) and Siemens ACUSON Sequoia (Siemens, Mountain View, CA, USA), equipped with a convex transducer probe (GE C1-6 probe, 1–6 MHz; Siemens 5C1 probe, 1–5 MHz. First, patients were positioned laterally for conventional US imaging. The tumor was scanned longitudinally and cross-sectionally using grayscale US, after which color Doppler imaging was applied to evaluate the tumor. Finally, CEUS was performed with patients in the same position, and a section displaying the maximum tumor area was selected for imaging. A low mechanical index (MI <0.1) was used for CEUS examination. A 1.0 mL dose of contrast agent (SonoVue; Bracco, Milan, Italy) was administered via bolus intravenous injection, followed by a 5 mL saline flush. The transducer was manually stabilized, and patients were instructed to take shallow breaths to minimize variation caused by respiration. Continuous dynamic observation of the masses under CEUS was conducted for 2 min, and the images were preserved.
Imaging analysis of US and CEUS
Two ultrasonologists, each with more than five years of renal CEUS experience, retrospectively reviewed the CEUS images in random order. They independently evaluated the CEUS features of ccRCC without access to clinical data or the pathological PC status. If disagreement occurred between the two ultrasonologists, a third ultrasonologist was invited to make the final determination.
For conventional US, the imaging characteristics of each ccRCC case were evaluated: (I) size; (II) boundary (clear/unclear); (III) shape (regular/irregular); (IV) echogenicity (hyper/iso/hypoechoic compared with the cortex); (V) echo halo; (VI) homogeneity (homogeneous/heterogeneous); (VII) growth pattern (exophytic/endophytic); (VIII) calcification; (IX) cystic degeneration; and (X) blood flow signals in color Doppler flow imaging (CDFI) (with or without).
The analyzed CEUS features included: (I) enhancement pattern: based on the arrival time of the contrast agent within the tumor relative to the adjacent renal cortex, wash-in mode (rapid/simultaneous/slow) and wash-out mode (rapid/simultaneous/slow); (II) peak enhancement degree: hyper-enhancement (higher than normal renal parenchyma), iso-enhancement (similar to normal renal parenchyma), or hypo-enhancement (lower than normal renal parenchyma); (III) PC integrity (intact: clear and continuous enhanced rim surrounding the mass during CEUS; incomplete: discontinuous, interrupted, or unclear); and (IV) homogeneity: homogeneous or heterogeneous (a mass completely enhanced without any defects was classified as homogeneous, regardless of variation in enhancement levels, whereas heterogeneous enhancement was defined as a mass with unenhanced areas).
PC status
PC status was assessed and confirmed by a senior genitourinary pathologist. The extent of PC invasion was classified into four grades: grade 1: complete PC, characterized by an intact PC without disconnection or neoplastic infiltration; grade 2: PC involved but not penetrated; grade 3: PC penetration, in which the tumor crosses the PC and invades the peritumoral renal parenchyma; grade 4: PC not visible at any point in all slides. To ensure accurate assessment of the PC, all evaluations were performed with renal parenchyma present on the opposite side. For analysis, patients were stratified into two groups according to PC status: grade 1–2 (defined as the PC-confined group) and grade 3–4 (defined as the PC-breaching group) (16).
Statistical analysis
Statistical analyses were performed using SPSS (version 26.0) and R (version 4.1.0). Continuous variables with normal and non-normal distributions were expressed as mean ± standard deviation and median (interquartile range), respectively, and compared using the independent t-test and Mann-Whitney U test. Categorical variables were presented as numbers (percentages) and compared using the chi-square test.
Variables showing statistical significance in univariate analyses were included in a multivariable binary logistic regression model. A nomogram was constructed based on the final model using the rms package in R.
The model’s discrimination was assessed using the area under the receiver operating characteristic (ROC) curve (AUC) and internally validated using bootstrap resampling (1,000 repetitions). Calibration was evaluated with calibration plots and the Hosmer-Lemeshow test. Clinical utility was quantified using decision curve analysis (DCA). A two-sided P value <0.05 was considered to be statistically significant. Sample size was determined by consecutive enrollment of all eligible patients meeting the inclusion criteria during the study period, rather than by a priori power calculation. Inter-observer agreement between the two primary ultrasonologists for the conventional US and CEUS features was quantified using Cohen’s kappa coefficient (unweighted for dichotomous variables; linear-weighted for ordinal variables), and interpreted using the Landis-Koch criteria (0.41–0.60, moderate; 0.61–0.80, substantial; >0.80, almost perfect).
Results
Baseline characteristics
A total of 254 patients were enrolled, with the following grade distribution: 59 (23.2%) grade 1, 136 (53.5%) grade 2, and 59 (23.2%) grade 3–4. They were randomly assigned to a training cohort (n=177, 70%) and an internal validation cohort (n=77, 30%). The baseline characteristics of both cohorts are presented in Table 1. No significant differences were observed in baseline indicators between the two cohorts (all P>0.05), indicating good comparability for model development and validation.
Table 1. Baseline characteristics of all patients in the training and validation cohorts.
| Variables | Training (n=177) | Validation (n=77) | Z/χ2 | P |
|---|---|---|---|---|
| Maximum diameter, cm | 3.30 (2.20, 5.00) | 3.60 (2.60, 5.00) | −1.182 | 0.24 |
| Age, years | 61.00 (51.00, 69.00) | 62.00 (52.00, 71.00) | −1.368 | 0.17 |
| BMI, kg/m2 | 24.80 (22.49, 26.85) | 24.77 (22.38, 26.81) | −0.243 | 0.81 |
| NLR | 2.15 (1.63, 2.77) | 2.10 (1.73, 2.92) | −0.570 | 0.57 |
| CAR | 0.04 (0.02, 0.09) | 0.04 (0.02, 0.09) | −0.470 | 0.64 |
| Sex | 0.789 | 0.37 | ||
| Female | 46 (25.99) | 16 (20.78) | ||
| Male | 131 (74.01) | 61 (79.22) | ||
| Clinical symptoms | 1.184 | 0.28 | ||
| Without | 146 (82.49) | 59 (76.62) | ||
| With | 31 (17.51) | 18 (23.38) | ||
| Hypertension | 0.091 | 0.76 | ||
| Without | 86 (48.59) | 39 (50.65) | ||
| With | 91 (51.41) | 38 (49.35) | ||
| Scr | 0.269 | 0.60 | ||
| Normal | 158 (89.27) | 67 (87.01) | ||
| Abnormal | 19 (10.73) | 10 (12.99) | ||
| Gross hematuria | 2.216 | 0.14 | ||
| Without | 166 (93.79) | 68 (88.31) | ||
| With | 11 (6.21) | 9 (11.69) | ||
| Microscopic hematuria | 0.495 | 0.48 | ||
| Without | 133 (75.14) | 61 (79.22) | ||
| With | 44 (24.86) | 16 (20.78) | ||
| Growth pattern | 1.009 | 0.32 | ||
| Exophytic | 96 (54.24) | 47 (61.04) | ||
| Endophytic | 81 (45.76) | 30 (38.96) | ||
| Boundary | 0.001 | 0.98 | ||
| Clear | 163 (92.09) | 71 (92.21) | ||
| Unclear | 14 (7.91) | 6 (7.79) | ||
| Shape | <0.001 | 0.99 | ||
| Regular | 147 (83.05) | 64 (83.12) | ||
| Irregular | 30 (16.95) | 13 (16.88) | ||
| Echogenicity | 3.828 | 0.15 | ||
| Hyper | 99 (55.93) | 50 (64.94) | ||
| Iso | 53 (29.94) | 14 (18.18) | ||
| Hypo | 25 (14.12) | 13 (16.88) | ||
| Homogeneity | 0.306 | 0.58 | ||
| Homogeneous | 52 (29.38) | 20 (25.97) | ||
| Heterogeneous | 125 (70.62) | 57 (74.03) | ||
| CDFI | 0.002 | 0.97 | ||
| Without | 57 (32.20) | 25 (32.47) | ||
| With | 120 (67.80) | 52 (67.53) | ||
| Echo halo | 0.007 | 0.93 | ||
| Without | 151 (85.31) | 66 (85.71) | ||
| With | 26 (14.69) | 11 (14.29) | ||
| Cystic degeneration | 0.962 | 0.33 | ||
| Without | 90 (50.85) | 34 (44.16) | ||
| With | 87 (49.15) | 43 (55.84) | ||
| Calcification | 0.007 | 0.93 | ||
| Without | 165 (93.22) | 72 (93.51) | ||
| With | 12 (6.78) | 5 (6.49) | ||
| Wash-in | 0.317 | 0.85 | ||
| Rapid | 39 (22.03) | 19 (24.68) | ||
| Simultaneous | 109 (61.58) | 47 (61.04) | ||
| Slow | 29 (16.38) | 11 (14.29) | ||
| Peak enhancement degree | 0.503 | 0.78 | ||
| Hyper | 125 (70.62) | 55 (71.43) | ||
| Iso | 36 (20.34) | 17 (22.08) | ||
| Hypo | 16 (9.04) | 5 (6.49) | ||
| Pseudocapsule integrity | 0.910 | 0.34 | ||
| Intact | 151 (85.31) | 62 (80.52) | ||
| Incomplete | 26 (14.69) | 15 (19.48) | ||
| Enhancement homogeneity | 0.082 | 0.78 | ||
| Homogeneous | 52 (29.38) | 24 (31.17) | ||
| Heterogeneous | 125 (70.62) | 53 (68.83) | ||
| Wash-out | 4.982 | 0.08 | ||
| Slow | 100 (56.50) | 52 (67.53) | ||
| Simultaneous | 45 (25.42) | 10 (12.99) | ||
| Rapid | 32 (18.08) | 15 (19.48) | ||
| Post-washout intensity | 1.220 | 0.54 | ||
| Hyper | 82 (46.33) | 34 (44.16) | ||
| Iso | 64 (36.16) | 25 (32.47) | ||
| Hypo | 31 (17.51) | 18 (23.38) |
Data are presented as median (interquartile range) or n (%). BMI, body mass index; CDFI, color Doppler flow imaging; CAR, C-reactive protein-to-albumin ratio; NLR, neutrophil-to-lymphocyte ratio; Scr, serum creatinine.
Inter-observer agreement was assessed prior to consensus adjudication; original independent ratings were retained for this analysis. Across the conventional US features evaluated, Cohen’s kappa values ranged from 0.742 (boundary) to 0.968 (calcification). For the CEUS features, kappa values ranged from 0.488 (PC sign) to 0.787 (peak enhancement degree), and the kappa for wash-out pattern was 0.735.
Univariate logistic regression analysis of factors associated with PC status
The comparison of relevant indicators between the grade 1–2 group and the grade 3–4 group is presented in Table 2. The results demonstrated significant differences (P<0.05) between the two groups in terms of boundary (P<0.001), shape (P<0.001), homogeneity (P=0.03), wash-in pattern (P=0.01), peak enhancement degree (P<0.001), PC integrity (P<0.001), enhancement homogeneity (P=0.01), wash-out pattern (P<0.001), and post-washout intensity (P=0.002).
Table 2. Characteristics of PC status.
| Variables | Grade 1–2 (n=134) | Grade 3–4 (n=43) | Z/χ2 | P |
|---|---|---|---|---|
| Maximum diameter, cm | 3.30 (2.20, 5.00) | 3.50 (2.00, 5.00) | −0.349 | 0.73 |
| Age, years | 61.00 (51.00, 69.00) | 61.00 (53.00, 66.00) | −0.029 | 0.98 |
| BMI, kg/m2 | 24.80 (22.58, 26.81) | 24.54 (22.22, 27.04) | −0.166 | 0.87 |
| NLR | 2.15 (1.62, 2.80) | 2.18 (1.65, 2.66) | −0.200 | 0.84 |
| CAR | 0.03 (0.02, 0.07) | 0.05 (0.02, 0.12) | −1.242 | 0.21 |
| Sex | 1.274 | 0.26 | ||
| Female | 32 (23.88) | 14 (32.56) | ||
| Male | 102 (76.12) | 29 (67.44) | ||
| Clinical symptoms | 0.459 | 0.50 | ||
| Without | 112 (83.58) | 34 (79.07) | ||
| With | 22 (16.42) | 9 (20.93) | ||
| Hypertension | 3.208 | 0.07 | ||
| Without | 60 (44.78) | 26 (60.47) | ||
| With | 74 (55.22) | 17 (39.53) | ||
| Scr | 0.837 | 0.36 | ||
| Normal | 118 (88.06) | 40 (93.02) | ||
| Abnormal | 16 (11.94) | 3 (6.98) | ||
| Gross hematuria | 2.856 | 0.09 | ||
| Without | 128 (95.52) | 38 (88.37) | ||
| With | 6 (4.48) | 5 (11.63) | ||
| Microscopic hematuria | 0.016 | 0.90 | ||
| Without | 101 (75.37) | 32 (74.42) | ||
| With | 33 (24.63) | 11 (25.58) | ||
| Growth pattern | 0.667 | 0.41 | ||
| Exophytic | 75 (55.97) | 21 (48.84) | ||
| Endophytic | 59 (44.03) | 22 (51.16) | ||
| Boundary | 13.220 | <0.001 | ||
| Clear | 129 (96.27) | 34 (79.07) | ||
| Unclear | 5 (3.73) | 9 (20.93) | ||
| Shape | 12.979 | <0.001 | ||
| Regular | 119 (88.81) | 28 (65.12) | ||
| Irregular | 15 (11.19) | 15 (34.88) | ||
| Echogenicity | 0.774 | 0.68 | ||
| Hyper | 76 (56.72) | 23 (53.49) | ||
| Iso | 38 (28.36) | 15 (34.88) | ||
| Hypo | 20 (14.93) | 5 (11.63) | ||
| Homogeneity | 4.698 | 0.03 | ||
| Homogeneous | 45 (33.58) | 7 (16.28) | ||
| Heterogeneous | 89 (66.42) | 36 (83.72) | ||
| CDFI | 0.187 | 0.67 | ||
| Without | 42 (31.34) | 15 (34.88) | ||
| With | 92 (68.66) | 28 (65.12) | ||
| Echo halo | 0.695 | 0.41 | ||
| Without | 116 (86.57) | 35 (81.40) | ||
| With | 18 (13.43) | 8 (18.60) | ||
| Cystic degeneration | 0.427 | 0.51 | ||
| Without | 70 (52.24) | 20 (46.51) | ||
| With | 64 (47.76) | 23 (53.49) | ||
| Calcification | 0.572 | 0.45 | ||
| Without | 126 (94.03) | 39 (90.70) | ||
| With | 8 (5.97) | 4 (9.30) | ||
| Wash-in | 8.782 | 0.01 | ||
| Rapid | 29 (21.64) | 10 (23.26) | ||
| Simultaneous | 89 (66.42) | 20 (46.51) | ||
| Slow | 16 (11.94) | 13 (30.23) | ||
| Peak enhancement degree | 32.048 | <0.001 | ||
| Hyper | 104 (77.61) | 21 (48.84) | ||
| Iso | 27 (20.15) | 9 (20.93) | ||
| Hypo | 3 (2.24) | 13 (30.23) | ||
| Pseudocapsule integrity | 22.986 | <0.001 | ||
| Intact | 124 (92.54) | 27 (62.79) | ||
| Incomplete | 10 (7.46) | 16 (37.21) | ||
| Enhancement homogeneity | 6.514 | 0.01 | ||
| Homogeneous | 46 (34.33) | 6 (13.95) | ||
| Heterogeneous | 88 (65.67) | 37 (86.05) | ||
| Wash-out | 21.959 | <0.001 | ||
| Slow | 84 (62.69) | 16 (37.21) | ||
| Simultaneous | 36 (26.87) | 9 (20.93) | ||
| Rapid | 14 (10.45) | 18 (41.86) | ||
| Post-washout intensity | 12.644 | 0.002 | ||
| Hyper | 64 (47.76) | 18 (41.86) | ||
| Iso | 54 (40.30) | 10 (23.26) | ||
| Hypo | 16 (11.94) | 15 (34.88) |
Data are presented as median (interquartile range) or n (%). BMI, body mass index; CDFI, color Doppler flow imaging; CAR, C-reactive protein-to-albumin ratio; NLR, neutrophil-to-lymphocyte ratio; PC, pseudocapsule; Scr, serum creatinine.
Multivariate logistic regression analysis of factors associated with PC-breaching
A multivariate logistic regression analysis was performed with grade 3–4 as the dependent variable and significant univariate factors, including boundary, shape, homogeneity, wash-in pattern, peak enhancement degree, PC integrity, enhancement homogeneity, washout pattern, and post-washout intensity, as independent variables. The results (Table 3) indicated that hypo-enhancement (P=0.007), an incomplete PC (P=0.01), and a rapid washout pattern (P=0.03) were independent risk factors (all P<0.05).
Table 3. Multivariate logistic regression analyses for screening predictors.
| Variables | B | SE | Wald χ2 | OR (95% CI) | P |
|---|---|---|---|---|---|
| Boundary | |||||
| Clear | Reference | ||||
| Unclear | 0.828 | 0.897 | 0.852 | 2.289 (0.395, 13.274) | 0.36 |
| Shape | |||||
| Regular | Reference | ||||
| Irregular | 0.248 | 0.706 | 0.123 | 1.281 (0.321, 5.114) | 0.73 |
| Homogeneity | |||||
| Homogeneous | Reference | ||||
| Heterogeneous | 1.007 | 0.704 | 2.044 | 2.737 (0.688, 10.886) | 0.15 |
| Wash-in | |||||
| Rapid | Reference | ||||
| Simultaneous | −0.295 | 0.549 | 0.290 | 0.744 (0.254, 2.183) | 0.59 |
| Slow | −0.157 | 0.774 | 0.041 | 0.854 (0.187, 3.896) | 0.84 |
| Peak enhancement degree | |||||
| Hyper | Reference | ||||
| Iso | 0.251 | 0.618 | 0.165 | 1.285 (0.383, 4.310) | 0.69 |
| Hypo | 2.459 | 0.903 | 7.406 | 11.690 (1.990, 68.688) | 0.007 |
| Pseudocapsule integrity | |||||
| Intact | Reference | ||||
| Incomplete | 1.610 | 0.626 | 6.615 | 5.005 (1.467, 17.074) | 0.01 |
| Enhancement homogeneity | |||||
| Homogeneous | Reference | ||||
| Heterogeneous | 0.356 | 0.680 | 0.273 | 1.427 (0.376, 5.413) | 0.60 |
| Washout | |||||
| Slow | Reference | ||||
| Simultaneous | 0.663 | 0.676 | 0.963 | 1.941 (0.516, 7.302) | 0.33 |
| Rapid | 2.066 | 0.961 | 4.617 | 7.892 (1.199, 51.947) | 0.03 |
| Post-washout intensity | |||||
| Hyper | Reference | ||||
| Iso | −1.010 | 0.676 | 2.230 | 0.364 (0.097, 1.371) | 0.14 |
| Hypo | −1.271 | 0.975 | 1.701 | 0.281 (0.042, 1.895) | 0.19 |
CI, confidence interval; OR, odds ratio; SE, standard error.
Construction of a nomogram
Based on the independent predictors of PC-breaching (hypo-enhancement, an incomplete PC, and a rapid washout pattern), a nomogram prediction model was developed (Figure 2). In this nomogram, each of the three predictors correspond to a specific point score. By summing the individual points, a total score is obtained, which is then mapped to a predicted probability of capsular absence or penetration. This model enables rapid and individualized estimation of risk probability for each patient.
Figure 2.
Nomogram of the PC-breaching prediction model. PC, pseudocapsule.
Evaluation of the nomogram model
The predictive performance and clinical applicability of the nomogram were comprehensively evaluated using ROC curves, calibration plots, and DCA.
Discriminative power of the nomogram model
The ROC curves for the training and validation cohorts are shown in Figure 3A,3B), respectively. The model demonstrated acceptable discriminatory ability, with an AUC of 0.779 [95% confidence interval (CI): 0.692–0.867] in the training cohort and 0.749 (95% CI: 0.600–0.898) in the validation cohort.
Figure 3.
ROC curves. (A) Training cohort. (B) Validation cohort. AUC, area under the curve; CI, confidence interval; ROC, receiver operating characteristic.
Accuracy of the nomogram model
Calibration plots for both cohorts are presented in Figure 4A,4B. The predicted probabilities showed good agreement with the observed outcomes, as the nomogram calibration curves closely approximated the ideal reference line. This finding was further supported by non-significant Hosmer-Lemeshow test results (training cohort: χ2=2.741, P=0.60; validation cohort: χ2=1.549, P=0.82), indicating good calibration in both datasets.
Figure 4.
Calibration curve for predicting probability of PC status. (A) Training cohort. (B) Validation cohort. PC, pseudocapsule.
Prediction of clinical utility
The DCA curves for the training and validation cohorts are displayed in Figure 5A,5B. In the training cohort, the nomogram provided higher net clinical benefit than the “treat all” or “treat none” strategies across threshold probabilities of approximately 0.13–0.81 and 0.86–0.92. Similarly, in the validation cohort, the nomogram yielded higher net benefit across a wide threshold range of approximately 0.13–0.93. These results suggest that the nomogram may aid clinical decision-making in this setting.
Figure 5.
Decision curve analysis in prediction of PC status. (A) Training cohort. (B) Validation cohort. PC, pseudocapsule.
Discussion
PC is an important pathological feature of ccRCC that represents not only a structural interface between the tumor and adjacent renal parenchyma but also an important indicator of tumor progression and biological aggressiveness (4). Increasing evidence has demonstrated a close association between PC status and surgical margin status, microvascular invasion, local recurrence, and long-term oncologic outcomes (5,17). Therefore, accurate preoperative assessment of PC integrity is of substantial clinical importance for both surgical strategy selection and postoperative risk stratification.
Rationale for PC grouping strategy
In this study, PC absence and PC penetration were defined as PC breaching, based on the concept that the functional integrity of the PC as a local anatomic and biological barrier is more clinically relevant than subtle morphologic differences alone. Tumors with an intact PC or only partial involvement were classified into the PC-confined group, as the barrier function is considered at least partially preserved and the tumor remains relatively confined with a more favorable prognosis. In contrast, tumors with complete PC penetration or complete PC absence were classified into the PC-breaching group, as both conditions represent functional failure or loss of the PC barrier (18). These tumors consistently exhibit higher infiltrative capacity, more aggressive biological behavior, higher microvascular invasion rates, and poorer oncologic outcomes (5,17). This grouping strategy translates morphologic PC heterogeneity into clinically meaningful risk stratification with direct implications for postoperative management and surveillance.
Performance of the CEUS-based predictive model
Based on this PC classification, a CEUS-based nomogram was developed and internally validated for preoperative prediction of PC breaching in ccRCC. The model demonstrated acceptable and stable discriminative ability in both the training and validation cohorts, with AUC values of 0.779 and 0.749, respectively. Calibration curves showed close agreement between predicted and observed probabilities after bootstrap correction, and Hosmer-Lemeshow testing confirmed adequate model fit in both cohorts. DCA further established the model’s clinical utility, showing a consistent net benefit across a wide range of threshold probabilities compared with default strategies. Importantly, all selected CEUS predictors were biologically interpretable and closely associated with established pathological mechanisms of PC disruption, reinforcing the biological plausibility of the model rather than reflecting purely data-driven associations.
Interpretation of CEUS predictors
Multivariable analysis identified three independent risk factors for PC breaching: an incomplete PC sign on CEUS, hypoenhancement at peak time, and a rapid wash-out pattern.
From a structural perspective, PC integrity serves as the most direct imaging surrogate of PC status. A continuous peritumoral hyperenhancement rim on CEUS generally corresponds to an intact PC (Figure 6), whereas a discontinuous or interrupted rim reflects histologic destruction of the PC (10,19,20). In our study, an incomplete PC sign strongly predicted pathological PC penetration or absence. At the tissue level, progressive tumor expansion induces peritumoral ischemia, microvascular collapse, and proteolytic degradation, leading to thinning, fragmentation, and eventual breaching of the PC. These biological changes translate directly into focal or circumferential loss of peritumoral enhancement on CEUS, as disruption of the microvascular rim precedes gross anatomic discontinuity. Unlike conventional imaging, CEUS enables real-time visualization of microvascular perfusion, allowing detection of functional impairment of the PC before overt structural collapse (20,21). Clinically, assessment of PC integrity on CEUS provides incremental value beyond tumor size or morphologic features alone. Identification of an incomplete PC rim may indicate reduced tumor containment and a higher risk of aggressive behavior, with direct implications for surgical planning, particularly in nephron-sparing surgery where enucleation margins depend on PC preservation (22,23).
Figure 6.
Image showing a ccRCC in a 51-year-old man. (A) Conventional US showed a hyperechoic mass on the inferior pole of the right kidney (yellow arrows). (B) An intact PC sign appeared as a rim of perilesional enhancement on CEUS (red arrows). (C) Hyper-enhancement was observed at peak time (25 s) after SonoVue injection (yellow arrows). (D) Histopathology showed an intact PC (red arrows) located between the tumor and adjacent normal renal tissue (H&E staining, ×100). ccRCC, clear cell renal cell carcinoma; CEUS, contrast enhanced ultrasound; NK, normal kidney; PC, pseudocapsule; RCC, renal cell carcinoma; US, ultrasound.
From a microvascular and hemodynamic perspective, highly invasive ccRCC is characterized by structurally abnormal and functionally inefficient neovasculature. Histopathologic studies have demonstrated irregular vessel caliber, endothelial damage, arteriovenous shunting, and loss of normal capillary networks (24,25), with a significantly higher proportion of undifferentiated vessels observed in high-grade compared with low-grade ccRCC (14). These microvascular abnormalities result in heterogeneous and inefficient contrast inflow. On CEUS, this is manifested as reduced peak enhancement (Figure 7), which likely reflects microvascular dysfunction rather than true hypovascularity. Consistent with this interpretation, prior CT-based studies have reported an association between hypoenhancement patterns and incomplete PC status (20). Further, rapid contrast wash-out on CEUS may indicate accelerated outflow caused by arteriovenous shunting, elevated interstitial pressure, and increased vascular permeability, which are characteristic features of a disorganized tumor microcirculation in aggressive ccRCC.
Figure 7.
Image showing a ccRCC in a 59-year-old woman. (A) Conventional US showed a hypoechoic mass on the left kidney (yellow arrows). (B) A discontinuous PC sign was detected at 15 s after SonoVue injection, with short red arrows indicating residual enhanced segments and long red arrows indicating disrupted segments. (C) Hypo-enhancement was detected at 20 s after SonoVue injection (yellow arrows). (D) Histopathological examination revealed PC penetration at the tumor-parenchyma interface (red arrows) (H&E staining, ×100). ccRCC, clear cell renal cell carcinoma; NK, normal kidney; PC, pseudocapsule; RCC, renal cell carcinoma; US, ultrasound.
Thus, these CEUS features reflect a coherent mechanistic cascade from microvascular failure to perfusion impairment, stromal barrier destruction, and PC breaching. This aligns closely with the established pathology of PC invasion, which is associated with higher World Health Organization/International Society of Urologic Pathologists grade, microvascular invasion, and infiltrative growth (26). Therefore, PC breaching should be regarded not merely as a local anatomic phenomenon but also as an imaging biomarker of aggressive tumor biology. The integration of structural and hemodynamic CEUS features into a unified predictive framework provides a biologically grounded rationale for risk stratification, with direct implications for postoperative management and surveillance.
Comparative advantages and clinical integration
The performance of our model is comparable to that of previously reported CT radiomics models but offers several practical advantages. While CT radiomics can achieve high diagnostic accuracy and has been successfully applied to predict renal capsule invasion in RCC and to integrate clinical parameters for preoperative prediction of microvascular invasion or surrounding tissue invasion, these approaches often rely on complex feature extraction and machine-learning pipelines and lack intuitive biological interpretability (27-30). In contrast, the CEUS-based approach in the present study utilizes routinely assessable imaging features that are biologically interpretable. CEUS itself provides a real-time, radiation-free method for assessing microvascular perfusion that is well suited for dynamic evaluation. This combination of diagnostic accuracy, biological interpretability, and clinical practicality facilitates integration into routine preoperative workflows. Beyond the comparison with CT-based radiomics approaches, CEUS also complements conventional cross-sectional imaging. CT is essential for staging but has limited sensitivity to subtle PC disruption, owing to spatial resolution, partial-volume effects, and reader-dependence (7). MRI provides higher soft-tissue contrast but is constrained by longer acquisition times, higher cost, and gadolinium contraindications in advanced renal impairment. In contrast, CEUS provides real-time, high-temporal-resolution imaging of tumor microvascular perfusion and the tumor-parenchyma interface and may detect focal disruption of the peritumoral enhancement rim before gross anatomic changes become apparent on CT or MRI (10,11,21). Because sulfur hexafluoride microbubbles are not renally excreted and involve no ionizing radiation, CEUS is safe and repeatable even in patients with renal insufficiency, making it a useful complement to CT and MRI (31).
Clinical implications for surgical management
CEUS enables real-time visualization of tumor microvascular perfusion and provides dynamic information on enhancement and wash-out behavior that is closely related to tumor pathological characteristics. Accurate preoperative identification of PC breaching is particularly important for nephron-sparing surgery. Tumor enucleation relies on the presence of an intact PC as a natural surgical plane. In PC-confined tumors, enucleation or minimal-margin partial nephrectomy allows maximal preservation of renal parenchyma without increasing the risk of positive surgical margins (32). Conversely, PC breaching is a recognized risk factor for positive surgical margins and local recurrence, and such patients may require wider surgical margins and closer postoperative surveillance (18,33). By providing individualized preoperative estimation of PC breaching risk, the proposed CEUS-based nomogram offers a practical tool for tailoring surgical strategy and informing margin planning in ccRCC.
Limitations and future directions
Several limitations should be acknowledged. First, this was a single-center retrospective study with a relatively limited sample size, particularly the size of the validation cohort. To limit overfitting given the modest sample size, the final model retained only three predictors, yielding around 14 events per variable—above the recommended threshold of 10. Internal validation by bootstrap resampling (1,000 iterations) showed only a small drop in AUC between the training and validation sets (0.779 vs. 0.749), suggesting limited overfitting. Even so, these results should be interpreted with caution, and external validation in larger, prospective, multicenter cohorts is needed to confirm the model's robustness and generalizability. Second, CEUS features were evaluated qualitatively rather than quantitatively, and inter-observer variability may have influenced feature assessment. Inter-observer agreement was substantial-to-almost-perfect for the majority of features, with the PC sign showing only moderate agreement (kappa, 0.488)—reflecting the inherent subjectivity of identifying focal peritumoral rim interruptions and supporting our use of consensus adjudication. Future studies incorporating quantitative CEUS perfusion analysis, radiomics, and artificial intelligence-based automation may further improve diagnostic accuracy and reproducibility. Finally, integration of clinical, laboratory, and molecular biomarkers may further refine individualized PC breaching risk prediction.
Conclusions
In conclusion, a CEUS-based nomogram was developed and internally validated for preoperative prediction of PC breaching in patients with ccRCC. Incorporating three CEUS imaging signatures—PC integrity, peak enhancement degree, and wash-out pattern—the model demonstrated acceptable and stable discrimination and reliable calibration in both the training and validation cohorts, and DCA confirmed clinical net benefit across a wide threshold range. This practical, noninvasive tool may aid preoperative assessment of tumor aggressiveness and support individualized surgical decision-making in ccRCC.
Supplementary
The article’s supplementary files as
Acknowledgments
The authors thank the ultrasonologists and pathologists at Shanghai General Hospital who contributed to image acquisition, image evaluation, and pathological assessment in this study.
Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Institutional Review Board of Shanghai General Hospital under expedited review (No. 2026-278), and the requirement for written informed consent was waived owing to the retrospective nature of this study.
Footnotes
Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tau.amegroups.com/article/view/10.21037/tau-2026-0312/rc
Funding: None.
Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tau.amegroups.com/article/view/10.21037/tau-2026-0312/coif). The authors have no conflicts of interest to declare.
Data Sharing Statement
Available at https://tau.amegroups.com/article/view/10.21037/tau-2026-0312/dss
References
- 1.Siegel RL, Kratzer TB, Giaquinto AN, et al. Cancer statistics, 2025. CA Cancer J Clin 2025;75:10-45. 10.3322/caac.21871 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Young M, Jackson-Spence F, Beltran L, et al. Renal cell carcinoma. Lancet. 2024;404:476-91. 10.1016/S0140-6736(24)00917-6 [DOI] [PubMed] [Google Scholar]
- 3.Rose TL, Kim WY. Renal Cell Carcinoma: A Review. JAMA 2024;332:1001-10. 10.1001/jama.2024.12848 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Feilen T, Rogg M, Andreev G, et al. Proteomic characterization of the pseudocapsule of clear cell renal cell carcinoma in VHL disease reveals a distinct microenvironment at the tumor boundary zone. Neoplasia 2025;68:101214 . 10.1016/j.neo.2025.101214 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Xi W, Hou Y, Hu X, et al. Prognostic significance of pseudocapsule status in patients with metastatic renal cell carcinoma treated with tyrosine kinase inhibitors. Transl Androl Urol 2021;10:4132-41. 10.21037/tau-21-429 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Pinto PVA, Coelho FMA, Schuch A, et al. Pre-operative imaging evaluation of renal cell carcinoma. Rev Assoc Med Bras (1992) 2024;70:e2024S107. [DOI] [PMC free article] [PubMed]
- 7.Toguchi M, Takagi T, Ogawa Y, et al. Detection of a peritumoral pseudocapsule in patients with renal cell carcinoma undergoing robot-assisted partial nephrectomy using enhanced MDCT. Sci Rep 2021;11:2245 . 10.1038/s41598-021-81922-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Roussel E, Campi R, Amparore D, et al. Expanding the Role of Ultrasound for the Characterization of Renal Masses. J Clin Med 2022;11:1112 . 10.3390/jcm11041112 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Atri M, Jang HJ, Kim TK, et al. Contrast-enhanced US of the Liver and Kidney: A Problem-solving Modality. Radiology 2022;303:11-25. 10.1148/radiol.211347 [DOI] [PubMed] [Google Scholar]
- 10.Dai WB, Yu B, Diao XH, et al. Renal Masses: Evaluation with Contrast-Enhanced Ultrasound, with a Special Focus on the Pseudocapsule Sign. Ultrasound Med Biol 2019;45:1924-32. 10.1016/j.ultrasmedbio.2019.03.020 [DOI] [PubMed] [Google Scholar]
- 11.Eusebi L, Masino F, Bertolotto M, et al. Contrast-enhanced ultrasound in the evaluation and management of solid renal lesions based on EFSUMB guidelines. J Med Ultrason (2001) 2025;52:293-303. 10.1007/s10396-025-01546-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Masino F, Eusebi L, Bertolotto M, et al. Contrast-enhanced ultrasound in renal cystic lesions: an update. J Med Ultrason (2001) 2024;51:635-47. 10.1007/s10396-024-01489-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Luo Y, Liu X, Jia Y, et al. Ultrasound contrast-enhanced radiomics model for preoperative prediction of the tumor grade of clear cell renal cell carcinoma: an exploratory study. BMC Med Imaging 2024;24:135 . 10.1186/s12880-024-01317-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Meng X, Yang R, Zhao S, et al. Associations between tumor grade, contrast-enhanced ultrasound features, and microvascular density in patients with clear cell renal cell carcinoma: a retrospective study. Quant Imaging Med Surg 2022;12:1882-92. 10.21037/qims-21-291 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Vas D, Paño B, Soler-Perromat A, et al. Qualitative and Quantitative Analysis of Contrast-Enhanced Ultrasound in the Characterization of Kidney Cancer Subtypes. Diagnostics (Basel) 2025;15:1795 . 10.3390/diagnostics15141795 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Snarskis C, Calaway AC, Wang L, et al. Standardized Reporting of Microscopic Renal Tumor Margins: Introduction of the Renal Tumor Capsule Invasion Scoring System. J Urol 2017;197:23-30. 10.1016/j.juro.2016.07.086 [DOI] [PubMed] [Google Scholar]
- 17.Xi W, Hou Y, Hou J, et al. Pseudocapsule thickness is positively associated with prognosis in patients with clear cell renal cell carcinoma. BMC Cancer 2025;25:279 . 10.1186/s12885-025-13721-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Song T, Yin Y, Liao B, et al. Capsular invasion in renal cell carcinoma: a meta-analysis. Urol Oncol 2013;31:1321-6. 10.1016/j.urolonc.2011.12.019 [DOI] [PubMed] [Google Scholar]
- 19.Wang J, Shi J, Gao L, et al. High-frame-rate contrast-enhanced ultrasound to differentiate between clear cell renal cell carcinoma and angiomyolipoma. BMC Cancer 2024;24:659 . 10.1186/s12885-024-12413-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Wang L, Wu H, Li J, et al. Diagnostic value of pseudocapsule ring hyperenhancement in contrast-enhanced ultrasound in renal focal lesions. Ir J Med Sci 2023;192:1631-6. 10.1007/s11845-023-03320-x [DOI] [PubMed] [Google Scholar]
- 21.Dipinto P, Canale V, Minelli R, et al. Qualitative and quantitative characteristics of CEUS for renal cell carcinoma and angiomyolipoma: a narrative review. J Ultrasound 2024;27:13-20. 10.1007/s40477-023-00852-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Huang X, Yang D, Zhu J, et al. Application value of contrast-enhanced ultrasound in the preoperative evaluation of renal cell carcinoma histological classification and RENAL score. Quant Imaging Med Surg 2024;14:9444-58. 10.21037/qims-24-694 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Liang M, Qiu H, Ou B, et al. Evaluation of contrast-enhanced ultrasound for predicting tumor grade in small (≤4 cm) clear cell renal cell carcinoma: Qualitative and quantitative analysis. Clin Hemorheol Microcirc 2024;88:351-62. 10.3233/CH-231990 [DOI] [PubMed] [Google Scholar]
- 24.Canete-Portillo S, Rodriguez Pena MDC, Wang D, et al. Vascular architectural patterns in clear cell renal cell carcinoma and clear cell papillary renal cell carcinoma. Virchows Arch 2021;479:1187-96. 10.1007/s00428-021-03214-2 [DOI] [PubMed] [Google Scholar]
- 25.Ohe C, Yoshida T, Amin MB, et al. Development and validation of a vascularity-based architectural classification for clear cell renal cell carcinoma: correlation with conventional pathological prognostic factors, gene expression patterns, and clinical outcomes. Mod Pathol 2022;35:816-24. 10.1038/s41379-021-00982-9 [DOI] [PubMed] [Google Scholar]
- 26.Volpe A, Bollito E, Bozzola C, et al. Classification of Histologic Patterns of Pseudocapsular Invasion in Organ-Confined Renal Cell Carcinoma. Clin Genitourin Cancer 2016;14:69-75. 10.1016/j.clgc.2015.07.020 [DOI] [PubMed] [Google Scholar]
- 27.Zhang Y, Zhao J, Li Z, et al. Preoperative prediction of renal fibrous capsule invasion in clear cell renal cell carcinoma using CT-based radiomics model. Br J Radiol 2024;97:1557-67. 10.1093/bjr/tqae122 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Yang L, Gao L, Arefan D, et al. A CT-based radiomics model for predicting renal capsule invasion in renal cell carcinoma. BMC Med Imaging 2022;22:15 . 10.1186/s12880-022-00741-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Xu J, Gao S, Zhu Q, et al. Machine learning-based multiparametric CT radiomics for predicting microvascular invasion before nephrectomy in clear cell renal cell carcinoma. Abdom Radiol (NY) 2025;50:5263-73. 10.1007/s00261-025-04956-2 [DOI] [PubMed] [Google Scholar]
- 30.Wu M, Zhu H, Han Z, et al. Prediction study of surrounding tissue invasion in clear cell renal cell carcinoma based on multi-phase enhanced CT radiomics. Abdom Radiol (NY) 2025;50:2533-48. 10.1007/s00261-024-04712-y [DOI] [PubMed] [Google Scholar]
- 31.Eusebi L, Masino F, Montatore M, et al. Diagnostic evaluation of renal masses in chronic kidney disease: a clinical and radiologic challenge. Abdom Radiol (NY) 2026. [Epub ahead of print]. doi: . 10.1007/s00261-026-05491-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Desai S, Rac G, Patel HD, et al. Imaging Features of Renal Masses to Select Optimal Candidates for Tumor Enucleation Partial Nephrectomy. Curr Urol Rep 2022;23:345-53. 10.1007/s11934-022-01121-w [DOI] [PubMed] [Google Scholar]
- 33.Minervini A, Rosaria Raspollini M, Tuccio A, et al. Pathological characteristics and prognostic effect of peritumoral capsule penetration in renal cell carcinoma after tumor enucleation. Urol Oncol 2014;32:50.e15-22. 10.1016/j.urolonc.2013.07.018 [DOI] [PubMed] [Google Scholar]