RENAL nephrometry score predicts surgical outcomes of laparoscopic partial nephrectomy

Authors


Hyung L. Kim, Division of Urology, Cedars-Sinai Medical Center, Los Angeles, CA 90048, USA. e-mail: kimhl@cshs.org

Abstract

Study Type – Therapy (case series)

Level of Evidence 4

What’s known on the subject? and What does the study add?

For small renal tumours, partial nephrectomy provides excellent cancer control and preserves renal function. The RENAL Nephrometry Score is useful for quantifying anatomic features relevant to surgical decision-making.

In patients undergoing laparoscopic partial nephrectomy, this study shows a correlation between RENAL Nephrometry Score and estimated blood loss, warm ischemia time, and length of hospital stay, suggesting that the RENAL Nephrometry Score may be useful for predicting the technical challenge posed by a renal tumour.

OBJECTIVE

• To assess the use of the RENAL Nephrometry Score (RNS), which has been proposed as an anatomical classification system for renal masses, aiming to predict surgical outcomes for patients undergoing laparoscopic partial nephrectomy (LPN).

MATERIALS AND METHODS

• In the present study, 159 consecutive patients who underwent LPN were reviewed and RNS was calculated for 141 patients with solitary renal masses who had complete radiographic data.

• Renal tumours were categorized by RNS as low (nephrometry sum 4–6), intermediate (sum 7–9) and high (sum 10–12).

RESULTS

• Of the 141 patients, there were 43 (30%) low, 91 (65%) intermediate and seven (5%) high score lesions. There was no statistically significant difference in the demographics of the three groups.

• There was a significant difference in warm ischaemia time (16 vs 23 vs 31 min; P < 0.001), estimated blood loss (163 vs 312 vs 317 mL; P= 0.034) and length of hospital stay (1.2 vs 1.9 vs 2.3 days; P < 0.001) between the low, intermediate and high score groups, respectively. There was no difference in overall operative time (P= 0.862), transfusion rate (P= 0.665), complication rate (P= 0.419), preoperative creatinine clearance (P= 0.888) or postoperative creatinine clearance (P= 0.473) between the groups.

• Sixty-one lesions (43%) were anterior and 80 (57%) were posterior. No difference was found among any intra-operative, pathological or postoperative outcomes when comparing anterior vs posterior lesions.

CONCLUSIONS

• In patients undergoing LPN, a higher RNS was significantly associated with an increased estimated blood loss, warm ischaemia time and length of hospital stay.

• The RNS may stratify tumours based on the technical difficulty of performing LPN.

Abbreviations
EBL

estimated blood loss

LOS

length of hospital stay

LPN

laparoscopic partial nephrectomy

NSS

nephron-sparing surgery

PADUA

Preoperative Aspects and Dimensions Used for an Anatomical

RNS

RENAL Nephrometry Score

WIT

warm ischaemia time

INTRODUCTION

Nephron-sparing surgery (NSS) is the treatment of choice for localized renal tumours. In the past, radical nephrectomy was considered the standard-of-care [1]. However, patients undergoing radical nephrectomy are at increased risk for developing chronic kidney disease, which has been associated with higher risk of cardiovascular disease and death [2,3]. NSS such as partial nephrectomy produces equivalent cancer control compared to radical nephrectomy and minimizes the risk of renal insufficiency [4,5].

When considering surgical options, the anatomical features of the renal tumour and the technical skill of the surgeon are important factors. The contribution of each factor to treatment choice and outcome are particularly relevant because urologists integrate technically challenging procedures such as laparoscopic and robotic partial nephrectomies into their practice. For example, to fully appreciate the learning curve for a new procedure, the technical challenge of performing the procedure needs to be quantified. Therefore, Kutikov and Uzzo [6] recently proposed the RENAL Nephrometry Score (RNS) for quantifying anatomical features relevant to surgical decision-making.

The RNS may be useful for characterizing tumours in surgical series and as a tool for selecting the optimal surgical treatment. When performed laparoscopically, partial nephrectomy is a more challenging procedure than radical nephrectomy. Given the well documented benefits of NSS, the risks and challenges of performing laparoscopic partial nephrectomy drive the decision-making process. For the present study, it was hypothesized that, if the RNS does reflect the technical demands of performing a laparoscopic partial nephrectomy (LPN), the RNS should correlate with clinically important outcomes associated with the procedure. Therefore, the present study aimed to assess the correlation between RNS and surgical outcome in patients undergoing LPN.

MATERIALS AND METHODS

Between April 2005 and July 2009, 159 patients with renal tumours underwent LPN by a single surgeon (H.L.K.). Data were obtained from a prospectively maintained database. Data collection and review were approved by the Institutional Review Board. All patients underwent preoperative imaging with contrast enhanced CT or MRI. Tumour size and location, depth of invasion into the renal parenchyma, and relationship to the hilar structures were determined from preoperative imaging.

RNS was calculated based on the criteria proposed by Kutikov and Uzzo [6] (Appendix 1). The RNS is based on five criteria: R, radius (maximum tumour diameter); E, exophytic/endophytic properties of the tumour; N, nearness of the tumour to the collecting system; A, anterior (a)/posterior (p) descriptor; L, location relative to the polar line (L). Of these five criteria, four (R, E, N, L) are used to assign points on a scale of 1–3 for each criterion. The anterior/posterior criterion is used to describe the location of the tumour; however, this descriptor is not used to assign a score. In addition to the RNS, each renal tumour was classified as central or peripheral based on the definition proposed by Frank et al. [7]; tumours abutting the collecting system or renal sinus fat on preoperative CT or MRI were defined as central.

On the basis of the RNS, patients were divided into low (nephrometry sum 4–6), intermediate (sum 7–9) and high (sum 10–12) groups [6]. The preoperative characteristics assessed were age, gender, tumour location, laterality and estimated GFR. Intra-operative characteristics assessed were operative time, warm ischaemia time (WIT), estimated blood loss (EBL), complications and hospital length of stay. Postoperative outcomes assessed included surgical margin status, pathological results, GFR, bleeding complications and urine leak. Urological complications were classified as intra-operative (during surgery), immediate (<24 h after surgery), delayed (>24 h until discharge) and late (after discharge until 30 days postoperatively). GFR was estimated using the Cockcroft – Gault model based on actual weight, which has been found to accurately estimate creatinine clearance in patients with a renal tumour [8].

Laparoscopic partial nephrectomy was performed using an i.p. approach in all patients. The surgical technique employed has been described previously [9]. Briefly, hilar control was achieved using a laparoscopic Satinsky. After excision of the tumour using cold scissors, the tumour bed was compressed using interrupted, figure-of-eight sutures passed just beneath the cut surface of the entire tumour bed and secured with LapraTy clips (Ethicon, Somerville, NY, USA). This standardized technique was applied regardless of the size or location of the tumour. Statistical analysis was performed using Fisher’s exact test for categorical variables and one-way ANOVA for continuous variables. Univariate and multivariate linear regression analysis was used to quantify the relationship between the RNS and continuous outcome measures. P < 0.05 was considered statistically significant. Statistical tests were performed using STATA®, Version 11.0 (StataCorp, College Station, TX, USA).

RESULTS

Of the 159 patients who underwent LPN, RNSs were calculated for 141 patients with solitary renal masses and complete radiographic data. The median (range) RNS was 7 (4–10). There were 43 (30%) low, 91 (65%) intermediate and seven (5%) high score lesions. There was no statistically significant difference in the demographics of the three groups (Table 1). Higher RNS groups were associated with significantly larger tumours, greater depth of tumour invasion and more central tumours (P < 0.001 for all). This was expected because these parameters were used to calculate the final RNS. Overall, sixty-one lesions (43%) were anterior and 80 (57%) were posterior. When comparing anterior vs posterior lesions, there was no significant difference for intra-operative, pathological or postoperative outcomes (data not shown).

Table 1.  Patient characteristics
CharacteristicsLow (n= 43)Intermediate (n= 91)High (n= 7)P
  1. RNS, RENAL nephrometry score.

Clinical based on RNS    
 Mean age (years), range)59 (30–77)61 (34–89)56 (45–81)0.470
 Males, n (%)32 (74)58 (64)4 (57)0.408
 Body mass index (kg/m2), mean (range)30 (21–41)31 (19–63)30 (24–43)0.549
Preoperative tumour, n (%)    
 Left kidney23 (53)54 (60)1 (14)0.064
 Anterior19 (44)38 (42)3 (43)0.989
 Anterior – superior6 (32)6 (16)0 
 Anterior – middle10 (53)27 (71)3 (100) 
 Anterior – inferior3 (15)5 (13)0 
 Posterior24 (56)53 (58)4 (57)0.989
 Posterior – superior8 (33)18 (34)0 
 Posterior – middle9 (38)25 (47)4 (100) 
 Posterior – inferior7 (29)10 (19)0 
 CT tumour size (cm), mean (range)2.1 (0.9–3.9)3.24 (1.3–6.2)3.79 (1.5–4.8)<0.001
 CT depth of invasion (cm), mean (range)0.86 (0.2–2)2.08 (0.7–3.9)3.13 (1.5–4)<0.001
Central (%)3 (7)74 (81)7 (100)<0.001

The operative and postoperative characteristics for the RNS groups were compared (Table 2). The mean operative times were similar for all groups. The WIT was significantly longer in higher RNS groups (16 vs 23 vs 31 min; P < 0.001). However, there was no significant difference in the preoperative or postoperative GFR. Higher RNSs were associated with higher EBL (163 vs 312 vs 317; P= 0.034), although there was no significant difference in the transfusion rates of the three groups. Among low, intermediate and high score groups, the mean length of hospital stay (LOS) was 1.2, 1.9, and 2.3 days, respectively (P < 0.014). Among patients with RCC, there was no significant difference in RNS based on pathological stage (P= 0.149), nuclear grade (P= 0.776) or surgical margin status (P= 1.000). Table 3 lists the intra-operative and postoperative complications. The overall complication rates were not significantly different between the groups (P= 0.657). There was no significant difference even when complications were further stratified as intra-operative, immediate postoperative, delayed postoperative and late postoperative.

Table 2.  Surgical and pathological characteristics
Characteristics based on RNSLow (n= 43)Intermediate (n= 91)High (n= 7)P
  1. EBL, estimated blood loss; LOS, length of hospital stay; RNS, RENAL nephrometry score; WIT, warm ischaemia time.

Surgical, mean (range)    
 Surgery time (min)210 (96–1555)197 (100–318)202 (144–275)0.862
 WIT time (min)16 (0–40)23 (0–51)31 (24–50)<0.001
 Preoperative GFR (mL/min/m2)100 (47–182)95 (25–237)96 (35–173)0.704
 Postoperative GFR (mL/min/m2)94 (40–140)84 (18–171)96 (43–142)0.278
 EBL (mL)163 (10–1700)312 (10–2100)317 (100–700)0.0343
 Patients transfused, n (%)1 (2)3 (3)01.000
 LOS (days), mean (range)1.2 (1–5)1.9 (1–11)2.3 (1–4)0.014
Pathological, n (%)    
 Renal cell carcinoma32 (74)71 (78)6 (86)0.885
 Clear cell24 (75)54 (76)5 (83)0.790
 Papillary6 (19) 11 (15)0 
 Chromophobe2 (6)6 (8)1 (17) 
 Benign lesion 11 (26)20 (22)1 (14) 
 Pathological tumour size (cm), mean (range)1.96 (0.6–5)2.96 (1.3–7)3.59 (1.5–5.4)<0.001
T stage, n (%)    
 pT1a31 (97)57 (80)5 (83)0.149
 pT1b1 (3)7 (10)1 (17) 
 pT307 (10)0 
Nuclear grade, n (%)    
 13 (9)10 (15)00.776
 218 (56)36 (53)4 (67) 
 311 (34)18 (26)2 (33) 
 404 (6)0 
Positive surgical margin status02 (3)01.000
Table 3.  Surgical complications
ComplicationsLow (n= 43)Intermediate (n= 91)High (n= 7)P
  1. *Repaired intracorporally. †Haematuria requiring clot evacuation and angiographic embolization.IVC, inferior vena cava; LRN, laparoscopic radical nephrectomy; OPN, open partial nephrectomy. Immediate postoperative = within 24 h of surgery; delayed postoperative = more than 24 h after surgery until discharge; late postoperative = after discharge until 30 days postoperative.

Intra-operative, n (%)    
 IVC laceration*1 (2.3)00 
 Conversion to LRN1 (2.3)1 (1.1)0 
 Conversion to OPN03 (3.3)0 
 Diaphragmatic injury*01 (1.1)0 
 Trochar injury to small bowel01 (1.1)0 
 Total2 (4.6)6 (6.6)01.000
Immediate postoperative, n (%)    
 Atrial fibrillation1 (2.3)00 
 Bleeding requiring blood transfusion01 (1.1)0 
 Total1 (2.3)1 (1.1)00.585
Delayed postoperative, n (%)    
 Pulmonary embolus1 (2.3)00 
 Ileus01 (1.1)0 
 Bleeding requiring blood transfusion01 (1.1)0 
 Haematuria02 (2.2)0 
 Total1 (2.3)4 (4.4)01.000
Late postoperative, n (%)    
 Ventral hernia1 (2.3)1 (1.1)0 
 Pseudoaneurysm1 (2.3)2 (2.2)0 
 Bowel obstruction01 (1.1)0 
 Total2 (4.6)4 (4.4)01.000
Overall complications, n (%)6 (14.0)15 (16.5)00.657

To further explore relationships between RNSs and outcomes, pairwise comparisons and regression analyses were performed. When comparing WIT, there was a significant difference between low and high score tumours (P < 0.001) and between low and intermediate score tumours (P < 0.001); however, the difference between intermediate and high score tumours did not reach statistical significance (P= 0.060). Pairwise comparisons for EBL showed a significant difference between low and intermediate score groups (P= 0.032). Similarly, LOS was significantly difference between low and intermediate score groups (P= 0.018).

Scatterplots of the total RNS vs WIT, LOS or EBL were used to estimate the linear regression lines and associated CIs (Fig. 1). On univariate analysis, there was significant positive correlation for RNS vs warm ischaemia time (P < 0.001) or LOS (P < 0.001). The linear correlation between RNS vs EBL was not statistically significant. To determine whether WIT and LOS could have been predicted with equal accuracy by simply using tumour size and knowing whether the tumour was central, a multivariate analysis was performed controlling for these two variables (Fig. 1). In a multivariate model, the positive correlation between RNS and WIT (P= 0.024) or LOS (P= 0.001) remained statistically significant, whereas the correlation between RNS and tumour size or central/peripheral descriptor was not significant.

Figure 1.

Estimate of linear relationships between renal nephrometry score and select clinical outcomes. Scatterplots were used to estimate the linear regression lines and 95% CIs for renal nephrometry score vs (A) warm ischaemia time, (B) length of hospital stay and (C) estimated blood loss. The results of multivariate analysis controlling for tumour size and central location are provided.

DISCUSSION

Tumour size and location are used to help determine the optimal surgical treatment for renal masses. However, most contemporary series that describe surgical outcomes provide only limited anatomical details. Commonly reported information includes size [10,11], anterior or posterior location, depth of penetration [12], proximity to the renal hilum [13,14] and central vs peripheral location [7,15–17]. Unfortunately, there is no uniform definition for important anatomical features. For example, central tumours have been defined as tumours that are completely surrounded by normal renal parenchyma [16], at least 60% endophytic [18], extending into the renal sinus [19], located less than 5 mm from the pelvicaliceal system [15] or contacting the collecting system or renal sinus [7]. Recently, Simmons et al. [20] devised the centrality index to provide a quantitative description of a central tumour based on two-dimensional cross-sectional CT images. The lack of standard definitions hinders meaningful comparisons between studies.

Previous studies have reported clinical outcomes based on anatomical features of renal masses. Nadu et al. [17] and Frank et al. [7] both compared central vs peripheral tumours treated with laparoscopic partial nephrectomy and reported increased WIT and LOS associated with central tumours; there was no difference in overall complication rates. This is consistent with the findings of the present study. It is possible that tumour size is less important predictor of surgical outcome during open surgery. However, in a multi-institutional series of open NSS, Patard et al. [10] compared renal tumours ≤4 cm and >4 cm and found no difference in WIT, LOS or surgical complications. The differences in reported outcomes may reflect a difference between open and laparoscopic approaches. However, it may also reflect differences in way tumours were stratified and compared. This illustrates the difficulty in reaching meaningful conclusions when comparing surgical series.

Kutikov and Uzzo [6] proposed the RNS as a reproducible system for describing a renal mass. The RNS incorporates information about five features of the tumour that have been reported to affect the complexity of surgery [6]. If the RNS truly reflects relevant tumour anatomy and surgical complexity, then surgical outcomes should correlate with the RNS. To the best of our knowledge, the present study is the first to assess operative outcomes based on RNS; the present study assessed the use of the RNS to predict surgical outcomes in patients undergoing LPN. It was found that patients with higher RNSs had significantly increased WIT, EBL and LOS. Pairwise comparisons identified significant differences between groups with low and intermediate scores. When the RNS was treated as a continuous variable, there were significant positive correlations with WIT and LOS on both univariate and multivariate analyses.

There was no significant difference in complication rates when comparing RNS groups. The small number of high score tumours in our series precluded a comparison of this group with other groups. However, the focus of the present study is the population of patients undergoing laparoscopic partial nephrectomy and, in this group, high score tumours are expected to be rare. Furthermore, a limitation of the RNS is that high score tumours require unusual combinations of characteristics [6]. For example, a RNS of 12 would require a tumour ≥7 cm to be completely endophytic. Therefore, a high RNS is expected to be rare in any RCC series.

We found a statistically significant difference in WIT and EBL among RNS groups. However, the difference may not be clinically significant because there was no difference in preoperative or postoperative GFR among RNS groups. The transfusion rates between RNS groups were also not significantly different. However, even small differences in EBL and WIT may be useful when comparing surgical series that have been matched by RNS. In the research setting, it may be useful to measure the effects of incremental modifications to surgical technique or technology. The difference in LOS among RNS groups was also statistically significant, and this relationship may be useful during preoperative counseling and patient teaching. A common reason for higher LOS includes greater postoperative pain. It is possible that larger incisions required for larger tumours may partly explain this observation.

One of the components of the RNS is not assigned a numerical score but, instead, is assigned an anterior (a) or posterior (h) descriptor. For this reason, an analysis was performed comparing anterior (n= 61) and posterior (n= 80) lesions. No difference was found for any preoperative, pathological or postoperative characteristics. This observation suggests that the RNS appropriately excludes the anterior/posterior designation in determining the final score. Anterior/posterior designation of renal masses, however, remains an important descriptive component of the RNS because it can impact the surgical approach for minimally invasive partial nephrectomy [21,22] and ablative therapies [23].

The Preoperative Aspects and Dimensions Used for an Anatomical (PADUA) classification score was recently proposed as a standardized tool for comparing renal tumours [24]. Similar to the RNS, the PADUA score uses several anatomical parameters: (i) anterior or posterior location; (ii) longitudional location; (iii) rim location; (iv) relationships with sinus; (v) relationships with the collecting system; (vi) percentage of tumour within into the kidney; and (vii) maximal diameter. In a study of 164 consecutive patients undergoing open partial nephrectomy, the PADUA scores independently predicted the risk of complications. Perioperative parameters such as WIT, EBL, LOS and renal function were not assessed [24].

There are important limitations to the present study. We utilized the RNS categories proposed by Kutikov and Uzzo [6], which limited the number of patients in the high RNS. In the present study, the RNS groups were used to summarize and describe our dataset. In the univariate and multivariate regression analyses, the RNS was treated as a continuous variable and thus group assignment was not relevant. The results obtained represent outcomes at a tertiary referral centre and all procedures were performed by a single surgeon who had fellowship training in laparoscopic surgery. Therefore, the results obtained may not be applicable to a general urology practice. The present series comprised men who underwent LPN, and the findings obtained may not be applicable to men who undergo open NSS or radical nephrectomy. Despite these limitations, the RNS may help stratify renal tumours and predict the complexity of performing a LPN. These results suggest that RNS may be useful when considering various surgical options for treating renal tumours.

In conclusion, the RENAL nephrometry score correlated with warm ischaemia time, EBL and LOS. This suggests that RENAL nephrometry score reflects the technical complexity of performing laparoscopic partial nephrectomy.

CONFLICT OF INTEREST

None declared.

Appendix

APPENDIX 1. RENAL NEPHROMETRY SCORE

Criteria1 point2 points3 points
(R)adius (maximum diameter, cm)≤4>4 but <7≥7
(E)xophytic/endophytic properties≥50%<50%Entirely endophytic
(N)earness of the tumour to the collecting system or renal sinus (mm)≥7>4 but <7≤4
(A)nterior/posteriorNo points assigned. Mass given a, p, or x descriptor
(L)ocation relative to the polar linesEntirely above the upper or below the lower polar lineLesion crosses the polar line>50% of mass is across the polar line, or masses crosses the axial renal midline, or mass is entirely between the polar lines

Ancillary