Predicting Adherent Perinephric Fat Using Preoperative Clinical and Radiological Factors in Patients Undergoing Partial Nephrectomy

EUF-820; No. of Pages 7 E U RO P E A N U R O L O GY F O C U S X X X ( 2 019 ) X X X– X X X

available at www.sciencedirect.com journal homepage: www.europeanurology.com/eufocus

Kidney Cancer

Predicting Adherent Perinephric Fat Using Preoperative Clinical and Radiological Factors in Patients Undergoing Partial Nephrectomy Leonardo D. Borregales a, Mehrad Adibi a, Arun Z. Thomas a, Rodolfo B. Reis a, Lisly J. Chery a, Catherine E. Devine b, Xuemei Wang c, Aaron M. Potretzke d, Theodora Potretzke e, Robert S. Figenshau d, Tyler M. Bauman d, Yara I. Aboshady a, Edwin Jason Abel f, Surena F. Matin a, Jose A. Karam a, Christopher G. Wood a,* a

Department of Urology, The University of Texas MD Anderson Cancer Center, Houston, TX, USA;

b

Department of Radiology, The University of Texas MD

Anderson Cancer Center, Houston, TX, USA; c Department of Biostatistics, The University of Texas MD Anderson Cancer Center, Houston, TX, USA; d Division of Urology, Washington University School of Medicine, St. Louis, MO, USA; e Mallinckrodt Institute of Radiology, Washington University School of Medicine, St. Louis, MO, USA; f Department of Urology, University of Wisconsin School of Medicine and Public Health, Madison, WI, USA

Article info

Abstract

Article history: Accepted October 17, 2019

Background: The decision to perform a partial nephrectomy (PN) relies largely upon the complexity of the renal mass and its surrounding anatomy. The presence of adherent perinephric fat (APF) can increase surgical complexity and extend operative times. The accurate prediction of APF may improve surgical planning and aid in decision making for the surgical approach. Objective: We sought to develop and externally validate a score that predicts APF based on preoperative clinical and radiological prognostic factors. Design, setting, and participants: We retrospectively analyzed 495 consecutive patients who underwent open or minimally invasive PN. APF was defined as the presence of “dense,” “adherent,” or “sticky” perinephric fat at the time of dissection by the surgeon, and this did not require subcapsular dissection. Additionally, we analyzed an independent cohort of 285 patients for external validation. Outcome measurements and statistical analysis: A score model was developed using multivariate logistic regression analysis. Calibration of the fitted model was assessed graphically with a plot of the predicted versus the actual probability of APF, and discrimination was assessed by calculating the area under the receiver operating characteristic curve. Results and limitations: Of the 495 patients, 95 (19%) had APF. Patients with APF had longer operative (p = 0.02) and arterial clamp (p = 0.01) times than non-APF patients. On multivariate analyses, diabetes mellitus (p = 0.009), posterior perinephric fat thickness (p < 0.001), and perinephric stranding (p < 0.001) were predictors of encountering APF in PN. A risk score ranging from 0 to 4 was developed based on these three variables to predict APF. The scoring system demonstrated good discrimination of 0.82 and 0.84 for the development and external validation cohorts, respectively. Conclusions: The APF score can accurately predict the presence of APF in patients with a small renal mass who are planning to undergo PN. This score could aid in pre- and intraoperative planning and impact the surgical approach. Patient summary: The presence of “sticky” fat surrounding the kidney in patients undergoing partial nephrectomy has previously been linked to longer operative times, intraoperative complications, and surgical conversion. In our study, we found that this feature is more often presented in patients with diabetes mellitus, and thicker and more inflammatory fat on renal imaging. Based on these findings, we developed a risk score that can accurately predict this feature before surgery, in order to improve surgical planning and better counsel the patients. © 2019 European Association of Urology. Published by Elsevier B.V. All rights reserved.

Associate Editor: Malte Rieken Keywords: Perinephric adhesive fat Sticky fat Partial nephrectomy Score

* Corresponding author. Department of Urology, The University of Texas MD Anderson Cancer Center, 1515 Holcombe—Unit 1373, Houston, TX 77030, USA. Tel.: +1 713 563 7463; Fax: +1 713 792 3474. E-mail address: [email protected] (C.G. Wood).

https://doi.org/10.1016/j.euf.2019.10.007 2405-4569/© 2019 European Association of Urology. Published by Elsevier B.V. All rights reserved.

Please cite this article in press as: Borregales LD, et al. Predicting Adherent Perinephric Fat Using Preoperative Clinical and Radiological Factors in Patients Undergoing Partial Nephrectomy. Eur Urol Focus (2019), https://doi.org/10.1016/j.euf.2019.10.007

EUF-820; No. of Pages 7 2

E U RO P E A N U RO L O GY F O C U S X X X ( 2 019 ) X X X– X X X

1.

Open, laparoscopic, and robotic PN procedures were carried out

Introduction

similar to previously published methods [10–12]. If APF was encountered

The decision to perform a partial nephrectomy (PN) relies largely upon the complexity of the tumor mass and its surrounding anatomy [1]. Well-established tumor morphometric scores such as the RENAL nephrometry score, PADUA score, and Centrality Index have been shown to objectively evaluate the complexity of a renal mass, as well as to predict surgical difficulties and complications [2]. As such, these scores represent a first step in the development of predicting algorithms aiming to benefit treatment decision-making process when opting between an open and a minimally invasive procedure, or between conservative and radical management [3,4]. Patient-specific anatomy can separately affect surgical dissection and tumor exposure in PN [5]. An often encountered intraoperative challenge among urologists performing PN is the adherent perinephric fat (APF) underlying Gerota’s fascia. A growing body of evidence suggests that the so-called “toxic fat” interferes with the isolation of a renal mass during a PN, and may increase the risk of tumor violation as well as operative times and rates of conversion [6–8]. Thus, the accurate prediction of APF may improve preoperative risk assessment and patient counseling. While the underlying pathogenesis of APF is unknown, studies suggest that inflammation and cardiovascular risk factors may account for APF [7]. Findings in small series have shown that imaging-based factors related to inflammation such as perinephric fat stranding, thickness, and density correlate with the presence of APF in PN [6]. A recent study developed a scoring system for APF based solely on radiological features; however, the small cohort of patients examined, single-surgeon experience, constraints in the definition used for APF, and questions regarding the role of clinical predictors in the prediction of APF lead to inherent limitations in the widespread use of this scoring system [9]. To overcome these limitations, we sought to develop and externally validate a score that predicts APF based on preoperative clinical and radiological prognostic factors, using a more commonly accepted definition for APF.

during exposure of the renal tumor, meticulous dissection was carried out to the extent possible and intraoperative ultrasound was routinely used to precisely identify the renal mass. No subcapsular dissection was performed when APF was encountered. Each surgeon routinely noted APF in the operative report when encountered. APF was defined as the presence of “dense,” “adherent,” or “sticky” perinephric fat noted by the surgeon at the time of dissection.

2.2.

Perinephric and tumor characteristics

All patients underwent preoperative cross-sectional abdominal imaging. The computed tomography or T1-weighted magnetic resonance imaging scans closest to the date of surgery were reviewed for tumor characteristics, RENAL nephrometry score, and perinephric fat thickness, and the degree of perinephric fat stranding was determined by an experienced radiologist (C.D.) who was blinded to the APF status. Clinical TNM staging was based on 2010 American Joint Committee on Cancer (AJCC) classification [13]. Posterior perinephric fat thickness was defined as the length of fat tissue in centimeters from the posterior renal capsule to the posterior abdominal wall musculature at the level of the renal hilum, as described previously by Eisner and colleagues [14]. Perinephric stranding was defined as a perinephric linear area of soft tissue attenuation and determined in three grades according to its severity, as published by Davidiuk et al [9] and Aizenstein et al [15] (Supplementary Fig. 1). Pathological tumor characteristics were reviewed by a genitourinary pathologist, histology was annotated according to the 2004 World Health Organization criteria [16], and tumor grade was determined using the Fuhrman grading system.

2.3.

Statistical analysis

The primary objective of the study was to develop and externally validate a model to predict APF among patients undergoing PN, based on preoperative characteristics. Univariate and multivariable logistic regression models were fit to assess the association between clinically relevant patient characteristics and the risk of APF. Initially, variables that were statistically significant (p < 0.05) in the univariate analyses were included in a multivariable model. We then performed stepwise model selection until all variables included in the multivariable model were significant at the 0.05 level. Calibration of the final multivariable model was assessed graphically with plot of the predicted probability of APF versus the actual probability of APF. In order to adjust for the bias associated with evaluating the performance of a model on the same group of patients used to build the model, the assessment of calibration was repeated for 2000 bootstrapped samples. Model discrimination was

Patients and methods

2.

assessed by calculating the area under the receiver operating characteristics curve (AUC). We then developed a scoring system based on the

2.1.

Data collection and demographic characteristics

fitted multivariable model to predict the risk of encountering APF during PN, and the risk score points were created based on the magnitude of

With Institutional Review Board approval, we retrospectively queried the

each covariate’s odds ratio (OR). Finally, we applied the risk score model

MDAnderson Cancer Center PN database for patients who underwent open

to an independent dataset to externally validate its predictive accuracy

or minimally invasive PN from 2009 until 2014. A cohort of 500 consecutive

for APF within each risk score category and overall. All analyses were

patients operated by three experienced urologists (J.A.K., S.F.M., and C.G.

performed using SAS (version 9) and S-plus (version 8.2).

W.) was identified. Five patients of this cohort were excluded due to ipsilateral previous PN or tumor ablation. A total of 495 patients were included in the final analysis. An external validation cohort of 285 patients was used with the collaboration of the Division of Urology at Washington University School of Medicine (St. Louis, MO, USA).

3.

Results

3.1.

Patient characteristics

Relevant clinical information thought to be associated with APF, such as age, gender, body mass index (BMI), hypertension, diabetes mellitus, smoking, alcohol, and estimated glomerular filtration rate were collected for each patient.

Of the 495 patients included in the development cohort, 95 (19.2%) had APF reported during PN. The median age was 60 yr (interquartile range [IQR]: 51.2–68); most of the

Please cite this article in press as: Borregales LD, et al. Predicting Adherent Perinephric Fat Using Preoperative Clinical and Radiological Factors in Patients Undergoing Partial Nephrectomy. Eur Urol Focus (2019), https://doi.org/10.1016/j.euf.2019.10.007

EUF-820; No. of Pages 7 3

E U R O P E A N U R O L O GY F O C U S X X X ( 2 019 ) X X X– X X X

patients were male (57%), Caucasian (71%), and with a median BMI of 29.9 kg/m2 (IQR: 26.4–34.0). Table 1 summarizes the characteristics of this cohort. The majority of the cases were performed open (73%), while 136 (27%) cases were performed robotically. Of those performed robotically, only one was converted to an open case due to significant bleeding. The median tumor size was 2.9 cm (IQR: 2–4 cm); a total of 308 (62.2%) tumors were clear cell renal cell carcinoma (RCC), 91 (18.4%) were papillary RCC, and the remaining 96 (19.4%) had benign histology. Patients with APF were noted to have longer operative times than non-APF patients (median: 130 vs 111 min; p = 0.02).

In addition, renal artery clamp time was significantly longer in patients who presented with APF (median: 15 vs 13 min; p = 0.01). Of note, nephrometry score was controlled for when evaluating these variables. 3.2.

Clinical predictors of APF, development of scoring system,

and validation

On univariable analysis, age, BMI, hypertension, diabetes, posterior perinephric fat thickness, and perinephric fat stranding were associated with an increased likelihood of the presence of APF (Table 2). Based on these factors,

Table 1 – Patient demographics and disease characteristics in the development cohort.

Patients Age (yr), n (%) <50 50–70 >70 Gender, n (%) Male Female Ethnicity, n (%) Caucasian African American Latin American Other BMI (kg/m2), median (IQR) Clinical T stage, n (%) T1a T1b T2a T2b T3a Unknown/NA Smoking history, n (%) No Yes Unknown Hypertension, n (%) No Yes Diabetes mellitus, n (%) No Yes Preoperative creatinine (mg/dl), median (IQR) Preoperative eGFR (ml/min/m2), median (IQR) Hemoglobin (g/dl) median (IQR) Type of imaging CT MRI Tumor size (cm), median (IQR) RENAL nephrometry score, n (%) 4–6 7–9 10–12 Unknown Posterior perinephric fat thickness (cm), median (IQR) Posterior perinephric fat thickness (cm), n (%) 0–1.9 2 Stranding, n (%) 0 = No stranding 1 = Thin rime-like mild stranding 2 = Diffuse thick band stranding

Development cohort

Validation cohort

495

285

102 (21) 305 (61) 88 (18)

58 (20) 193 (68) 34 (12)

282 (57) 213 (43)

165 (58) 120 (42)

351 (70.9) 54 (10.9) 73 (14.7) 17 (3.4) 29.9 (26.4–34)

234 (82.1) 43 (15.1) 6 (2.1) 2 (0.7) 30 (26–35)

353 (71.3) 104 (21) 7 (1.4) 4 (0.8) 27 (5.5) 0 (0)

221 (77.5) 55 (19.3) 7 (2.5) 0 (0.0) 0 (0.0) 2 (0.7)

301 (60.8) 194 (39.2) 0 (0)

115 (40.4) 114 (40.0) 56 (19.6)

204 (41.2) 291 (58.8)

96 (33.7) 189 (66.3)

404 (81.6) 91 (18.4) 0.89 (0.75–1.04) 86.68 (72.7–99.2) 13.8 (12.7–14.8)

227 (79.6) 58 (20.4) 0.90 (0.74–1.09) 81.75 (65.84–96.11) 13.9 (12.8–15.0)

427 (86.3) 68 (13.7) 3 (2.2–4.1)

202 (70.9) 83 (29.1) 2.8 (1.8–3.5)

157 (31.7) 278 (56.2) 60 (12.1) 0 (0) 1.77 (1–2.7)

76 (26.7) 154 (54.0) 40 (14.0) 15 (5.3) 1.9 (1.0–2.7)

283 (57.2) 212 (42.8)

168 (58.9) 117 (41.1)

227 (45.9) 172 (34.7) 96 (19.4)

94 (33) 145 (50.9) 46 (16.1)

BMI = body mass index; CT = computed tomography; eGFR = estimated glomerular filtration rate; IQR = interquartile range; MRI = magnetic resonance imaging; NA = not available.

Please cite this article in press as: Borregales LD, et al. Predicting Adherent Perinephric Fat Using Preoperative Clinical and Radiological Factors in Patients Undergoing Partial Nephrectomy. Eur Urol Focus (2019), https://doi.org/10.1016/j.euf.2019.10.007

EUF-820; No. of Pages 7 4

E U RO P E A N U RO L O GY F O C U S X X X ( 2 019 ) X X X– X X X

Table 2 – Univariable logistic regression analyses of clinical features for the prediction of adherent perinephric fat during partial nephrectomy in 495 patients.

Table 3 – Multivariable adherent perinephric fat risk score model based on the development cohort (N = 495). Variable

Variables Age (yr) <50 50–70 >70 Race Caucasian African American Latin American Other Gender BMI ECOG (1 vs 0) Smoking Alcohol Hypertension Diabetes Nephrolithiasis Tumor size on imaging (cm) Renal sinus fat involvement Histology Clear cell Papillary Chromophobe Other Grade 1–2 3 4 Posterior perinephric fat (cm) 1.9 2 Stranding on imaging Type 0 Type 1 Type 2 Renal score 4–6 7–9 10–12

Odds ratio (95% CI)

OR (95% CI)

p value

Risk score

0.02

1

0.03 <0.001

0 1 2

0.002

0 1

p value

Reference 4.08 (1.71–9.75) 7.08 (2.76–18.15)

0.002 <0.001

Reference 0.43 (0.18–1.06) 0.56 (0.27–1.13) 0.22 (0.03–1.67) 5.73(3.14–10.43) 1.04 (1.01–1.07) 1.26 (0.80–1.96) 1.36 (0.87–2.14) 1.46 (0.92–2.29) 2.12 (1.30–3.48) 2.74 (1.65–4.56) 1.48 (0.78–2.78) 1.09 (0.96–1.25) 1.17 (0.42–3.26)

0.067 0.106 0.143 <0.001 0.034 0.317 0.178 0.105 0.003 <0.001 0.225 0.172 0.751

Reference 0.67 (0.35–1.29) 0.21 (0.03–1.62) 1.44 (0.81–2.59)

0.236 0.136 0.214

Reference 0.95 (0.56–1.59) 1.78 (0.44–7.10)

0.834 0.420

Reference 8.39 (4.8–14.6)

<0.001

Reference 5.19 (2.55–10.55) 19.63 (9.50–40.58)

<0.001 <0.001

Reference 0.92 (0.56–1.50) 0.78 (0.35–1.71)

0.739 0.536

BMI = body mass index; CI = confidence interval; ECOG = Eastern Cooperative Oncology Group. Bold variables were found to be statistically significant on univariate analysis.

multivariable analysis revealed that diabetes, posterior perinephric fat thickness, and perinephric fat stranding were independent predictors of APF in patients undergoing PN (Table 3). The calibration plot of the fitted model is shown in Figure 1, where the x axis is the model-predicted probability of APF and the y axis is the actual probability of APF. The model also demonstrated good discrimination with an AUC of 0.82 (95% confidence interval [CI] 0.78– 0.87). We then developed a scoring system that would be clinically simple to use, yet predictive of the likelihood of encountering APF during PN by assigning points to each of the three previously mentioned factors according to the value of the OR. This resulted in a score ranging from 0 to 4 with corresponding probabilities for the presence of APF (Table 4). Lastly, we sought to externally validate the risk score model using an independent cohort of 285 patients. The observed proportion of APF and the model-predicted

Diabetes 2.01 (1.10–3.70) Stranding Reference None Type 1 2.34 (1.05–5.17) 7.25 (3.06–17.19) Type 2 Posterior perinephric fat thickness (cm) Reference 0–2 2.93 (2.23–5.73) 2

CI = confidence interval; HR = hazard ratio; OR = odds ratio. In the multivariable model, for gender, HR was 2.04 with p = 0.056.

probability of APF for this validation cohort are shown in Table 4. The observed and predicted probabilities of APF corresponding to each score category are also shown graphically in Figure 2 for the development and validation cohorts, respectively. Using the validation cohort, the risk score model for predicting APF also demonstrated good discrimination with an estimated AUC of 0.84 (95% CI 0.78–0.89). Moreover, predicted probabilities of APF based on a 0.5 threshold yielded specificity of 92.3 and 92.2 in the development and validation cohorts, respectively. 4.

Discussion

The presence of APF during PN can influence the technical difficulty of the procedure, often leading to tedious dissection of the renal capsule from the adherent adipose tissue, while taking care not to violate the capsule or tumor [5]. Recent studies have shown that the presence of APF not only leads to higher estimated blood loss and longer operative times, but may also result in an increased risk of perioperative morbidity [7,9,17]. Although this last notion is controversial, there is little doubt that the APF results in a technically challenging case, particularly with minimally invasive PN [1]. Therefore, we sought to identify predictors associated with the presence of APF intraoperatively and develop a scoring system to better aid in preoperative planning. In a large cohort of patients with clinical small renal masses that underwent PN, we found that diabetes mellitus, and posterior perinephric fat thickness and stranding were independent prognosticators of APF. Based on these three predictive factors, we developed a scoring system that can easily be used preoperatively to classify patients according to the likelihood of the presence of APF. APF has been subject to study in recent years as investigators recognized this factor as a surrogate for surgical complexity [6,8,18]. Initially, Bylund and colleagues [6] revealed that patients in whom APF was encountered tended to have longer operative times. Similar findings were confirmed in a study by Khene and colleagues [7], and additionally, APF was associated with an increased risk of bleeding and transfusion, and higher rates of conversion to radical nephrectomy or open surgery. Although we did not find a significant difference in intra- or postoperative complications among patients with APF, our study has revealed

Please cite this article in press as: Borregales LD, et al. Predicting Adherent Perinephric Fat Using Preoperative Clinical and Radiological Factors in Patients Undergoing Partial Nephrectomy. Eur Urol Focus (2019), https://doi.org/10.1016/j.euf.2019.10.007

EUF-820; No. of Pages 7 5

E U R O P E A N U R O L O GY F O C U S X X X ( 2 019 ) X X X– X X X

Fig. 1 – Calibration plot of actual versus predicted probability of adherent perinephric fat in the development cohort. APF = adherent perinephric fat.

Table 4 – Observed proportions of patients with adherent perinephric fat (APF) and the predicted probabilities of APF (95% confidence intervals) based on the risk score model: development and validation cohorts. Risk score

Development cohort

0 1 2 3 4

Validation cohort

Observed proportion of patients with APF

Predicted probability of APF (95% CI)

Observed proportion of patients with APF

Predicted probability of APF (95% CI)

4/166 9/112 29/116 39/81 14/20

2.4% (0.9–6.2%) 8.0% (4.2–14.7%) 25.0% (18.0–33.7%) 48.1% (37.5–58.9%) 70.0% (47.3–85.9%)

0/77 2/69 17/82 17/49 5/8

0.0% (0.0–1.0%) 2.9% (0.7–10.9%) 20.7% (13.3–30.8%) 34.7% (22.8–48.9%) 62.5% (28.5–87.5%)

CI = confidence interval.

Percentage of paents with APF

87.5 70 70.

62.5 48

52.5

34.7

35.

25

17.5

8 2

2.9

0.

0. 0

Development cohort Validaon cohort

20.7

1

2

3

4

Risk score

Fig. 2 – Percentage of patients with adherent perinephric fat according to the risk score model: development and validation cohorts.

Please cite this article in press as: Borregales LD, et al. Predicting Adherent Perinephric Fat Using Preoperative Clinical and Radiological Factors in Patients Undergoing Partial Nephrectomy. Eur Urol Focus (2019), https://doi.org/10.1016/j.euf.2019.10.007

EUF-820; No. of Pages 7 6

E U RO P E A N U RO L O GY F O C U S X X X ( 2 019 ) X X X– X X X

an additionally increased risk of prolonged ischemia time in these patients. One hypothesis that these findings generated was that APF may make selective arterial clamping difficult, and therefore leading to prolonged ischemia time due to incomplete clamping. Although our findings showed evidence of a statistically significant 2-min difference for ischemia time in patients with APF, these findings should be considered of value in the context of other contributing factors associated with complexity. Given that APF can be a cumbersome intraoperative encounter, there have been multiple studies aiming to predict its presence preoperatively. The first study examining these factors found an association of age, gender, tumor size, perinephric stranding, and thickness of perinephric fat with APF in a cohort of 29 patients who underwent PN. However, the small sample size and patient selection limited the utility of this study [6]. In another series of 41 consecutive patients, perinephric fat density had significant capability to predict the difficulty of perinephric fat dissection due to APF during PN with an AUC of 0.87 (p < 0.001) [17]. In the largest series to date, Khene and colleagues [7] retrospectively evaluated 202 patients, of whom 80 (39.6%) presented with APF. Male gender, BMI, and hypertension were associated with the presence of APF in a multivariate analysis. Finally, in a recent study by Davidiuk and colleagues [9], the authors developed a score that predicts the likelihood of APF in a cohort of 100 patients subjected to robotic PN. Their scoring system based on perinephric fat thickness and perinephric stranding had an AUC of 81%. Notably, our study confirms these findings in a large cohort of patients who underwent PN. Furthermore, we report the largest series examining predictors of APF, and highlight the association of diabetes and radiological characteristics with APF, in addition to providing concomitant external validation. The hypothesis of a chronic inflammatory “environment” associated with APF has been highlighted based on the increased presence of this phenomenon in patients with cardiovascular risk factors [7]. The role of adipose tissue in the generation of chronic systemic inflammation has been emphasized in obesity. Particularly in metabolic syndrome, hypertrophied adipocytes activate a cascade of chemokines that lead to the infiltration of macrophages to the visceral fat. In addition, such chemokines increase the plasminogen activator inhibitor type-1, a fibrinolytic factor associated with the development of fibrosis and adhesion of visceral fat [18]. It is thought that these components of inflammation lead to a fibrous adhesion in perinephric fat [18]. Our study reinforces the association of a chronic inflammatory process with the development of APF. Furthermore, our observations may set the groundwork to investigate predictive inflammatory markers for the presence of APF. The development of a score to predict the presence of this APF would benefit surgical planning. The anticipation of this feature would allow surgeons to counsel patients on surgical difficulties and possibility of conversion to a non– minimally invasive approach, or allow them to consider ablative techniques; anticipation of a hostile surgical field

would allow surgeons to opt for a more conservative approach for patients with many options for management. While we do not suggest that our findings are going to drive the decision to a specific surgical approach generally, a clearer understanding of the predictors of APF appears of value during preoperative planning. Our study differs from previous studies in that we have used a different definition for APF based on surgeon’s intraoperative assessment. A range of definitions have been described in the literature, such as perinephric fat requiring subcapsular dissection for exposure of renal parenchyma or perinephric fat adhering to the renal parenchyma that results in bleeding and decapsulation [7,9]. Although this definition is subjective, we believe that APF may not necessarily require additional maneuvers, but rather a more meticulous dissection and use of intraoperative ultrasound to adequately prepare the tumor for resection. In addition, we believe that most experienced renal surgeons would agree on the presence of APF when encountered, as it is rarely a subtle finding. The limitations of this study include its retrospective design. In addition, our definition of APF can be debated given its subjective categorization. One of the major strengths of this study is that we have studied predictors of APF in the largest cohort of patients to date and have externally validated our findings, reinforcing the reproducibility, simplicity, and generalizability of our model. Despite its retrospective nature, the current study clarifies the predictors of APF and extends them to a large cohort likely to be representative of patients seen in clinical practice.

5.

Conclusions

Among 495 patients who underwent PN, we found that presence of diabetes, posterior perinephric fat thickness, and perinephric fat stranding are predictors of APF. This model was externally validated in a large cohort of patients. Based on these clinical and radiological factors, we developed a scoring system that can accurately predict the presence of APF. This model could aid in preoperative planning and patient counseling.

Author contributions: Christopher G. Wood 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: Borregales, Thomas, Adibi, Chery, Reis, Karam, Wood. Acquisition of data: Borregales, Thomas, Adibi, Chery, Devine, A.M. Potretzke, T. Potretzke, Bauman, Aboshady. Analysis and interpretation of data: Borregales, Thomas, Adibi, Chery. Drafting of the manuscript: Borregales, Thomas, Adibi, Chery, Karam, Wood. Critical revision of the manuscript for important intellectual content: Borregales, Adibi, Thomas, Reis, Chery, Devine, Wang, A.M. Potretzke, T. Potretzke, Figenshau, Bauman, Aboshady, Abel, Matin, Karam, Wood. Statistical analysis: Wang. Obtaining funding: None. Administrative, technical, or material support: None. Supervision: Matin, Karam, Wood.

Please cite this article in press as: Borregales LD, et al. Predicting Adherent Perinephric Fat Using Preoperative Clinical and Radiological Factors in Patients Undergoing Partial Nephrectomy. Eur Urol Focus (2019), https://doi.org/10.1016/j.euf.2019.10.007

EUF-820; No. of Pages 7 7

E U R O P E A N U R O L O GY F O C U S X X X ( 2 019 ) X X X– X X X

Other: None.

[7] Khene ZE, Peyronnet B, Mathieu R, Fardoun T, Verhoest G, Bensalah K. Analysis of the impact of adherent perirenal fat on peri-operative outcomes

Financial disclosures: Christopher G. Wood certifies that all conflicts of interest, including specific financial interests and relationships and

of

robotic

partial

nephrectomy.

World

J

Urol

2015;33:1801–6. [8] Davidiuk AJ, Parker AS, Thomas CS, Heckman MG, Custer K, Thiel

affiliations relevant to the subject matter or materials discussed in the

DD. Prospective evaluation of the association of adherent perineph-

manuscript (eg, employment/affiliation, grants or funding, consultan-

ric fat with perioperative outcomes of robotic-assisted partial

cies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: None.

nephrectomy. Urology 2015;85:836–42. [9] Davidiuk AJ, Parker AS, Thomas CS, et al. Mayo adhesive probability score: an accurate image-based scoring system to predict adherent perinephric fat in partial nephrectomy. Eur Urol 2014;66:1165–71.

Funding/Support and role of the sponsor: None.

[10] Cozar JM, Tallada M. Open partial nephrectomy in renal cancer: a feasible gold standard technique in all hospitals. Adv Urol 2008;2008:916463. [11] Gill IS, Desai MM, Kaouk JH, et al. Laparoscopic partial nephrectomy

Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10. 1016/j.euf.2019.10.007.

for renal tumor: duplicating open surgical techniques. J Urol 2002;167(2 Pt 1):469–76, discussion 475–6. [12] Kaouk JH, Khalifeh A, Hillyer S, Haber GP, Stein RJ, Autorino R. Robot-assisted laparoscopic partial nephrectomy: step-by-step contemporary technique and surgical outcomes at a single highvolume institution. Eur Urol 2012;62:553–61. [13] Edge S, Compton C. The American Joint Committee on Cancer: the 7th edition of the AJCC cancer staging manual and the future of

References

TNM. Ann Surg Oncol 2010;17:1471–4. [14] Eisner BH, Zargooshi J, Berger AD, et al. Gender differences in

[1] Hou W, Yan W, Ji Z. Anatomic features involved in technical complexity of partial nephrectomy. Urology 2015;85:1–7.

subcutaneous and perirenal fat distribution. Surg Radiol Anat 2010;32:879–82.

[2] Kutikov A, Uzzo RG. The R.E.N.A.L. nephrometry score: a compre-

[15] Aizenstein RI, Owens C, Sabnis S, Wilbur AC, Hibbeln JF, O’Neil HK.

hensive standardized system for quantitating renal tumor size,

The perinephric space and renal fascia: review of normal anatomy,

location and depth. J Urol 2009;182:844–53. [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;68:980–92. [4] Canter D, Kutikov A, Manley B, et al. Utility of the R.E.N.A.L. nephrometry scoring system in objectifying treatment decisionmaking of the enhancing renal mass. Urology 2011;78:1089–94. [5] Gorin MA, Mullins JK, Pierorazio PM, Jayram G, Allaf ME. Increased intra-abdominal fat predicts perioperative complications following

pathology, and pathways of disease spread. Crit Rev Diagn Imaging 1997;38:325–67. [16] Eble JN, Sauter G, Epstein J, Sesterhenn I. Pathology and genetics of tumours of the urinary system and male genital organs. Lyon, France: IARC Press; 2004. [17] Zheng Y, Espiritu P, Hakky T, Jutras K, Spiess PE. Predicting ease of perinephric fat dissection at time of open partial nephrectomy using preoperative fat density characteristics. BJU Int 2014;114: 872–80.

minimally invasive partial nephrectomy. Urology 2013;81:1225–31.

[18] Hagiwara M, Miyajima A, Hasegawa M, et al. Visceral obesity is a

[6] Bylund JR, Qiong H, Crispen PL, Venkatesh R, Strup SE. Association of

strong predictor of perioperative outcome in patients undergoing

clinical and radiographic features with perinephric “sticky” fat. J

laparoscopic radical nephrectomy. BJU Int 2012;110(11 Pt C):

Endourol 2013;27:370–3.

E980–4.

Please cite this article in press as: Borregales LD, et al. Predicting Adherent Perinephric Fat Using Preoperative Clinical and Radiological Factors in Patients Undergoing Partial Nephrectomy. Eur Urol Focus (2019), https://doi.org/10.1016/j.euf.2019.10.007