SEARCH

SEARCH BY CITATION

Keywords:

  • nephrectomy;
  • renal cancer;
  • radiography

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

Study Type – Prognosis (individual cohort)

Level of Evidence 2b

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

RENAL nephrometry is a quantitative, reproducible scoring system that characterizes RENAL masses and standardizes reporting. Previous work has suggested that the system may be useful in predicting outcomes after partial nephrectomy. This study is the first to correlate RENAL nephrometry score with operative approach or risk of complication in patients undergoing either partial or radical nephrectomy.

OBJECTIVE

  • • 
    To evaluate the utility of the RENAL scoring system in predicting operative approach and risk of complications. The RENAL nephrometry scoring system is designed to allow comparison of renal masses based on the radiological features of (R)adius, (E)xophytic/endophytic, (N)earness to collecting system, (A)nterior/posterior and (L)ocation relative to polar lines.

METHODS

  • • 
    A retrospective review of all patients at a single institution undergoing radical nephrectomy (RN) or partial nephrectomy (PN) for a renal mass between July 2007 and May 2010 was carried out.
  • • 
    Preoperative RENAL score was calculated for each patient. Surgical approach and operative outcomes were then compared with the RENAL score.

RESULTS

  • • 
    In all, 249 patients underwent either RN (158) or PN (91) with average RENAL scores of 8.9 and 6.3, respectively (P < 0.001).
  • • 
    Patients who underwent RN were more likely to have hilar tumours (64% vs 10%, P < 0.001) than patients who underwent PN, but were no more likely to have posteriorly located tumours (50% each).
  • • 
    There were more complications among patients with RN (58%) vs patients with PN (42%, P= 0.02).
  • • 
    RENAL scores were higher in patients with PN who developed complications than in patients with PN who did not develop complications (6.9 vs 6.0, P= 0.02), with no difference noted among patients with RN developing complications (8.9 vs 8.9, P= 0.99).

CONCLUSION

  • • 
    The RENAL system accurately predicted surgeon operative preference and risk of complications for patients undergoing PN.

Abbreviations
BMI

body mass index

ASA

American Society of Anesthesiology

(L)(O)PN

(laparoscopic) (open) partial nephrectomy

(L)(O)RN

(laparoscopic) (open) radical nephrectomy.

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

Due to the increasing use of axial imaging, 66% of renal masses are now detected incidentally [1], resulting in more than 58 240 new cases of renal cell carcinoma diagnosed in 2010 [2]. As the number of incidentally detected tumours has risen, so has the drive for performance of more minimally invasive surgical options. Concomitantly, mounting evidence has shown that nephron-sparing surgery may reduce the morbidity and mortality resulting from renal insufficiency [3]. Because of these factors, surgeons are performing more difficult surgeries on more complex renal lesions [4] with an ever-increasing range of options, including observation, open surgery, robotically assisted surgery, standard laparoscopic surgery, laparoendoscopic single site surgery and percutaneous and laparoscopic ablation. The variety of surgical approaches and types of masses treated makes comparison between studies difficult and often impossible.

In order to help clarify the reporting of renal surgery, Kutikov and Uzzo [5] reviewed the literature for pertinent features that characterize renal tumour anatomical attributes as they relate to resectability and proposed the RENAL nephrometry scoring system, with the goal of producing a ‘structured, reproducible, quantitative scoring system to describe and classify the most surgically relevant anatomical features of solid renal neoplasms’. While the creation of this scoring system for standardizing the reporting of renal surgery outcomes is admirable, it is first important to show that the scoring system has clinical utility in predicting surgeon approach, complications and postoperative outcomes prior to instituting it in prospective studies. To that end, we sought to evaluate the utility of this scoring system in a single institution series of procedures performed for treatment of renal masses.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

After institutional review board approval, a retrospective chart review was conducted identifying all partial and radical nephrectomies performed for a preoperative diagnosis of renal mass at the University of Iowa Hospitals and Clinics from July 2007 to May 2010. Charts were individually reviewed to obtain demographic data including age, gender, presenting symptoms, comorbidities, smoking history, family history of renal cancer and body mass index (BMI). The tumour's diameter, exophytic/endophytic properties, nearness of the deepest portion of the tumour to the collecting system or sinuses, anterior/posterior renal location, location relative to the polar lines and whether the mass touched the main renal artery or vein were determined by preoperative CT or MRI imaging. The RENAL nephrometry score was then calculated using published guidelines [5]. In brief, standardized points (1–3 points per descriptor) are assigned based on tumour size, endophytic/exophytic properties, nearness to collecting system and lesion location relative to polar lines. A further descriptor (a/p) is included to delineate anterior/posterior lesions or lesions that abut the renal hilum. Operative reports were reviewed to determine surgery type (laparoscopic vs open and radical vs partial nephrectomy), side, operating room time, American Society of Anesthesiology (ASA) score, estimated blood loss and intraoperative surgical approach conversion and complications. Complications were further stratified into urological (dialysis, retention, renal failure, recurrence, intraoperative renal injuries, UTIs, urine leak) and non-urological (cardiac, pulmonary, gastrointestinal, thromboembolic, incisional, other), and were classified based on the scoring system by Dindo et al.[6].

Following operative intervention, pathological reports were reviewed to determine tumour histological type, stage, Fuhrman grade, margin status, final dimensions and intrarenal tissue-specific invasion. Length of hospital stay and postoperative dialysis, transfusion and urine leak were determined through thorough chart review. Follow-up medical records were reviewed to determine disease recurrence and additional complications up to the time of most recent surgical follow-up.

All statistical analysis was conducted using SAS Institute's StatView commercially available software, version 5.0.1. Student's t test was used to compare means and a z test was used to compare percentages. Variables were assumed to be normally distributed. P < 0.05 was considered significant.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

A total of 249 patients underwent either partial nephrectomy (PN) or radical nephrectomy (RN) for suspected renal cancer over a 3-year period. When evaluating baseline patient characteristics, there was no statistical difference in operative preference (RN vs PN) based on gender, BMI, family history of renal cancer, ASA score or presence of hypertension, diabetes, coronary artery disease, chronic obstructive pulmonary disease, other cancer history or smoking (Table 1). While not significant, there was a trend toward performing PN in younger patients (P= 0.06). At our institution, we favour attempting PN on all patients for whom surgery is thought technically feasible, as preservation of renal function has been shown to decrease morbidity and mortality [7]. Patients undergoing PN had significantly less intraoperative blood loss and better postoperative renal function as measured by serum creatinine and estimated GFR.

Table 1.  Perioperative characteristics of all patients undergoing surgical intervention divided into PN vs RN
Patients (n)PPN (91)RN (158)
  1. NS, not significant.

Preoperative   
 Mean age (sd), years0.0654 (14)60 (13)
 Male gender, %NS57.160.1
 Mean BMI (range), kg/m2NS32.6 (20–58)31 (14–64)
 Family history, %NS8.810.8
 Creatinine (sd), mg/dLNS1 (0.3)1.1 (0.7)
 MDRD estimated GFR (sd), mL/min/1.73 m2NS76.8 (20.1)71.4 (21.0)
Comorbidities   
 Hypertension (%)NS49.558.9
 Diabetes (%)NS23.121.5
 Coronary artery disease (%)NS1116.5
 Chronic obstructive pulmonary disease (%)NS4.416
 History of other cancer (%)NS17.620.9
 Smoking history (%)NS4455.1
 Mean ASA score (sd)NS2.4 (0.6)2.5 (0.6)
Operative   
 Mean operative duration (sd), minNS241 (77)231 (88)
 Mean estimated blood loss, cc0.02329 (377)579 (1210)
 Conversion (laparoscopy to open), % 7.7NA
 Length of stay (sd), daysNS4 (1.9)4.7 (3.9)
 Creatinine (sd), mg/dL0.041.2 (0.4)1.5 (0.8)
 MDRD estimated GFR (sd), mL/min/1.73 m20.0465.1 (20.6)53.6 (16.0)
Postoperative   
 Follow-up (sd), days0.04182 (226)248 (263)

Using preoperative imaging, RENAL nephrometry scores were determined for all 249 patients (Table 2). For both laparoscopic and open approaches, mean radiographic tumour size (2.9 cm vs 7.3 cm, P= 0.001), mean nephrometry scores (6.3 vs 8.9, P= 0.001) and hilar designation (lesion in contact with a renal vessel, 10% vs 64%, P= 0.001) of resected lesions were significantly lower in the PN group than the RN cohort. In addition to size, scores for the nearness to a sinus/calyx (1 vs 2.8, P= 0.01) and location to polar lines (2 vs 2.8, P= 0.02) were significantly higher for the RN group than the PN cohort.

Table 2.  RENAL nephrometry scores for all patients undergoing surgical intervention divided into PN vs RN
 PNRNP, PN vs RN
  1. NS, not significant.

Patients (n)91158 
Mean radiographic tumour size (cm)2.97.3<0.001
R (size criteria)1.22.20.01
E (exophytic/endophytic)1.31.2NS
N (nearness to sinus/calyx)12.80.01
A (anterior), n (%)45 (49.5)78 (49.4)NS
L (location relative to polar line)22.80.02
Hilar (designation), n (%)9 (9.9)101 (63.9)<0.001
Mean nephrometry score6.38.9<0.001
Total nephrometry score, n (%)   
  420 (22.0)2 (1.3)<0.001
  510 (11.0)3 (1.9)0.01
  623 (25.3)7 (4.4)<0.001
  711 (12.1)6 (3.8)0.03
  816 (17.6)28 (17.7)NS
  98 (8.8)44 (27.8)<0.001
 103 (3.3)66 (41.8)<0.001
 1102 (1.3)NS
 1200NS
 Low (4–6)53 (58.2)12 (7.6)<0.001
 Medium (7–9)35 (38.5)78 (49.4)NS
 High (10–12)3 (3.3)68 (43.0)<0.001

Of the patients undergoing PN, 58.2%, 38.5% and 3.3% had low (4–6), medium (7–9) or high (10–12) nephrometry scores, respectively, compared with 7.6%, 49.4% and 43.0%, respectively, in the group of patients undergoing RN, suggesting that RENAL score predicts approach for patients undergoing PN better than it does for patients undergoing RN. Of the 152 patients who had a laparoscopic procedure, 48 (32%) underwent PN, while 104 (68%) underwent RN. There was a significant difference in mean tumour size (2.3 cm vs 5.6 cm, P < 0.001), R size criteria (1.1 vs 1.9, P < 0.001), nearness to sinus (1.5 vs 2.8, P < 0.001), location relative to polar line (1.9 vs 2.7, P < 0.001), hilar designation (6% vs 53%, P < 0.001) and mean RENAL score (5.6 vs 8.7, P < 0.001) when comparing laparoscopic PN (LPN) vs laparoscopic RN (LRN). Among the group of patients undergoing laparoscopic surgery (either radical or partial), no differences were noted between exophytic/endophytic (1.2 vs 1.2, P= 0.97) or anterior location (29% vs 46%, P= 0.09) compared with patients undergoing either open RN (ORN) or open PN (OPN).

Of the 97 patients who had an open procedure, 43 (44%) underwent PN, while 54 (56%) underwent RN. Patients undergoing OPN had smaller tumours based on overall measurement (3.5 cm vs 11 cm, P < 0.001) and R size criteria (1.3 vs 2.8, P < 0.001) compared with patients undergoing ORN. Additionally, the patients undergoing ORN were more likely to have higher scores for exophytic/endophytic (1 vs 1.3, P= 0.01), nearness to sinus (2.4 vs 2.8, P= 0.01), location relative to polar line (2 vs 3, P < 0.001), hilar designation (14% vs 85%, P < 0.001) and mean RENAL score (7.1 vs 9.5, P= 0.05). As with the laparoscopic analysis, no difference was noted between OPN and ORN based on anterior location (72% vs 56%, P= 0.16).

Overall complications were higher among the patients having RN vs PN (58% vs 42%, P= 0.02), with the majority of complications occurring in the postoperative period (Table 3). Of those patients who had a PN, the average RENAL score was higher for those who developed complications than for those who did not (6.9 vs 6.0, P= 0.02). However, there was no difference in RENAL scores for those patients undergoing RN (8.9 vs 8.9) regardless of complication status. The rate of complications increased with increasing RENAL score for patients undergoing RN and was higher for patients with a RENAL score of 6–8 than for those with a score of 5 or less (Fig. 1). Complication occurrences peaked at RENAL scores 7–9 for PN. Frequencies of complications increased as RENAL score increased for RN.

Table 3.  Complications of all patients undergoing surgical intervention according to operative approach
 LPN, n (%)OPN, n (%)LRN, n (%)ORN, n (%)P, PN vs RN
  • *

    While P for this comparison is significant (P < 0.001), the clinical implication of this is minimal given the bias in doing RN for larger tumours which are anticipated to have worse clinical outcomes.

Patients (n)484310454 
Total complications16 (33.3)37 (86)25 (24)41 (75.9)0.02
Intraoperative5 (10.4)12 (27.9)9 (8.7)9 (16.7)NS
 Visceral injury1 (2.1)5 (11.6)3 (2.9)9 (16.7)NS
 Open conversion2 (4.2)NA6 (5.8)NANS
 LRN conversion2 (4.2)NANANANS
Postoperative7 (14.6)14 (32.6)6 (5.8)11 (20.4)0.01
Haemorrhage requiring transfusion4 (8.3)7 (16.3)3 (2.9)10 (18.5)NS
 Average units of blood (n)3.251.723.3 
Urine leak2 (4.2)6 (14)00Not applicable
Acute renal insufficiency1 (2.1)1 (2.3)3 (2.9)1 (1.9)NS
Dialysis003 (2.9)1 (1.9)NS
Follow-up1 (2.1)2 (4.7)6 (5.8)25 (46.3)0.02
 Surgical site recurrence01 (2.3)02 (3.7)NS
 Metastasis1 (2.1)1 (2.3)6 (5.8)23 (42.6)*
Other urological complications01 (2.3)00NS
Other non-urological complications3 (6.3)10 (23.3)10 (9.6)21 (38.9)NS
Clavien grade     
 I4 (8.3)5 (11.6)4 (3.8)8 (14.8)NS
 II4 (8.3)9 (20.9)3 (2.9)14 (25.9)NS
 III1 (2.1)6 (14)5 (4.8)5 (9.3)NS
 IV01 (2.3)1 (1)0NS
 V0002 (3.7)0.36
image

Figure 1. Percentage of patients with perioperative complications tabulated by RENAL score.

Download figure to PowerPoint

Comparing patients who underwent LRN with OPN, the LRN patients had a higher RENAL score (8.8 vs 7.1, P= 0.01) and had larger masses (5.6 cm vs 3.5 cm, P= 0.01) which were more posterior (71 vs 28, P= 0.03) but there were no differences in nearness to collecting system (2.8 vs 2.8), location relative to polar line (2.7 vs 3), endophytic/exophytic (1.2 vs 1.3) or hilar designation (55% vs 85%).

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

Until recently, no objective system existed to accurately and reproducibly describe renal masses based on preoperative imaging. Kutikov and Uzzo [5] attempted to address this deficiency by developing a system based on a literature review of relevant anatomical features and their effect on outcomes following renal surgery. Their system, RENAL nephrometry, was intended to standardize descriptions of renal masses and allow meaningful comparisons of renal mass cohorts across the literature. In addition to being a valuable research tool, the system has the potential to aid clinical decision-making, specifically surgical approach and risk of complications.

In this retrospective clinical study of 249 patients undergoing renal surgery at a single institution, the RENAL score correlated with renal mass complexity as measured by the surgeon operative preference and complication risk. Patients with lower scores were more likely to undergo PN than RN, consistent with the current surgical paradigm of limiting RN to patients with highly complex renal masses. Using this information, the RENAL nephrometry scoring system could help objectify the indication for PN vs RN and optimize patient care by reducing the number of RNs performed. The RENAL score also correlated with the frequency of overall and postoperative complications for those patients undergoing PN. While the RENAL score predicted the frequency of complications, it did not predict the severity (Clavien grade). Regarding oncological success of the procedure, there was an increased risk of metastasis in patients undergoing RN, indicative of the higher stage of lesions in this segment of patients with large and complex renal tumours. Interestingly, we found no differences between the patients undergoing LPN and LRN with regard to exophytic/endophytic or anterior/posterior location. While this was unexpected given that endophytic posterior lesions appear more technically challenging, we believe this was due to the level of experience of the surgeons operating at our high volume tertiary referral centre.

Comparison of renal surgery cohorts in the literature is difficult because of the variation in reported data. For example, in 1999 Ono and colleagues [8] evaluated 60 patients who underwent LRN and 40 who underwent ORN by the same surgical team. Complication rates were 13.0% and 7.5%, respectively, but the only objective data reported were specimen weight and pathological stage with no comment on proximity to the hilum. Their complication rate was significantly lower than that reported by Dunn and colleagues [9] in 2000 who reviewed 61 LRN and 33 ORN with complication rates of 37% and 55%, respectively, but only reported on tumour size (5.3 and 7.4 cm, respectively), making comparison between these two populations impossible. Similarly, Shuford and colleagues [10] reviewed 41 ORN, 18 hand-assisted LRN and 15 pure LRN. While overall complication rates were similar at 10%, 17% and 12%, respectively, only tumour size was reported so that the indication for each approach was unknown, preventing meaningful translation to a specific patient in terms of operative planning.

Similar to the data from RN series, information on tumour location characteristics is lacking within the PN data. In one of the largest reviews comparing LPN to OPN, Porpiglia and colleagues [11] performed a review of all published literature from 1990 to 2007. Their summary included 1062 LPN and 2756 OPN with 21.4% and 21.3% complications but, again, they only reported average tumour size, 2.7 and 3.2 cm, respectively. Gill and colleagues [12] attempted to add insight by including the description of tumours as ‘central’ or ‘peripheral’ in their report comparing 771 LPN and 1029 OPN. Complications, 18.6% for LPN vs 13.7% for OPN, were higher in the laparoscopic cohort, despite the smaller tumour size (2.6 cm vs 3.3 cm) and lower likelihood of tumours having a ‘central’ location compared with the open cohort (34.4% vs 53.3%), showing that comparison between cohorts regarding outcomes requires a clear understanding of tumour characteristics. Development of standardized preoperative scoring systems, such as the RENAL nephrotomy score, may therefore (i) allow for more direct comparisons between surgical cohorts, (ii) improve planning of operative approach and (iii) help counsel patients on perioperative complications and outcomes.

Two published papers have examined the role of the RENAL score in patients undergoing PN. In the first, Bruner et al. [13] found that RENAL is associated with urine leak after a PN. The second paper by Hayn et al. [14] found that RENAL score was associated with estimated blood loss, warm ischaemia time and length of hospital stay in patients undergoing LPN. Our paper is unique in that it addresses two other areas where RENAL score is useful, namely in determining operative approach and risk of complications, and is also the first to describe the utility of the scoring system for patients who underwent RN.

The present study has limitations that are worth mentioning. Due to the retrospective nature, reporting of complications was non-standardized such that there may be under-reporting of several complications. However, in an attempt to limit this variable, all charts were thoroughly reviewed during the perioperative period to identify specific complications, rather than using billing data which may be more subject to under-reporting. This strict and thorough chart review as well as close adherence to the Clavien system probably explains our increased complication rate compared with similar contemporary reports. Although retrospective series have inherent biases and flaws, the present study allows us insight into surgeon choice of operative approach in a ‘blinded’ fashion. In other words, since the RENAL score had not been instituted in our population of patients during the review period, there is no prospective bias of surgeons choosing a particular approach based on RENAL score. Further, the study found limited utility in comparing patients who underwent ORN vs OPN, possibly because once the decision to proceed with an open approach has been made other factors (i.e. those which are not measured by the RENAL score) become increasingly important. Accordingly, only a retrospective review can show that the RENAL score predicts operative approach. Future prospective studies can use the results from the present paper to develop baseline cut-offs for which patients may be suitable for a particular approach given their preoperative RENAL score.

Due to an increase in the percentage of cases performed using PN over the last several years, average follow-up was significantly shorter in patients undergoing PN. Additionally, the short average follow-up in this contemporary series (6.1 and 8.3 months for PN and RN, respectively) prevents the reporting of intermediate and long-term oncological outcomes. Nevertheless, our goal was not to compare oncological outcomes using the nephrometry score, but rather to look at surgical approach and perioperative outcomes and complications. Future studies with longer follow-up will be required to answer the question of whether the RENAL nephrometry score predicts oncological efficacy of a particular procedure.

The RENAL score was developed as a research tool to make meaningful comparisons between surgical cohorts. Herein, we have also shown that it may have some utility in predicting surgical approach and surgical outcomes. As the RENAL score is only meant to describe complexity of a particular renal mass, it does not take into account other patient variables and comorbidities that may otherwise influence outcomes. Additionally, although overall RENAL score may predict the above variables, modifications of the score, such as variable weighting of certain factors, may improve the specificity for predicting various outcome parameters. Future studies are warranted to evaluate the specific importance of each variable of the renal score and their ability to predict given outcomes. With further refinement future prospective studies will allow for more precise calculation of which patients are more likely to have success with nephron-sparing surgery as well as their risk for operative complications.

The RENAL scoring system has both academic and clinical practical utility. In the academic setting, the reporting of RENAL scores in future publications will allow direct comparison of the renal mass cohort. In the clinical setting, the RENAL scoring system predicts surgeon operative choice as well as risk of complications and can be used as a guide when counselling patients on management of their renal masses. Future refinements in the scoring system will potentially allow for a preoperative ‘risk assessment’ tool to limit perioperative complications.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES