1 Background

The rate of detection of renal masses has been increased due to improved imaging techniques [1]. This raised the need for treatment modalities other than radical nephrectomy (RN) to preserve renal function as possible including partial nephrectomy (PN) [1,2,3,4,5]. It is indicated in localized renal masses (cT1) [6]. Furthermore, PN can safely be performed in cT2 renal masses with acceptable technical and oncological outcomes and with maximal parenchymal preservation, especially in solitary kidney or bilateral tumors [2, 7, 8]. However, PN is associated with increased risk of overall complications [2, 4, 9].

Many studies were conducted to evaluate the outcome of PN, but it was difficult to compare them as there were no standardized criteria to unify the description of different renal masses. This problem led to the development of nephrometry scoring systems (NS) to unify as possible the description of the renal masses. There are many scoring systems including the PAUDA [10], the centrality index (C-index) method [11] and the R.E.N.A.L. NS [12]. Another problem in the management of renal masses using PN was the bias in decision which was based mainly on surgeon preferences and experience without standard criteria [12]. Consequently, these NS were tried to predict the outcome of PN [13]. This could help in decision making by proper selection of the surgical procedure.

R.E.N.A.L. is one of these nephrometry scoring systems that was thoroughly evaluated. However, some studies failed to find any relation between R.E.N.A.L. with operative parameters or postoperative outcome [13,14,15,16,17,18]. Thus, the need for continuous improvement in these NS is obvious for better prediction of the outcome and proper selection of the surgical technique.

Our aim was to evaluate our designed newly modified nephrometry score (MNS) in prediction of the outcome following PN for renal masses and to compare its predictability versus original R.E.N.A.L.

2 Methods

2.1 Study design and inclusion criteria

This was a prospective study for adult patients (≥ 18 years old) who presented consecutively between September 2014 and December 2016 with cT1-2N0M0 renal masses amenable for PN. Indications for PN were patients with solitary kidney, bilateral renal masses, chronic kidney disease or risk of future renal impairment as in cases of hypertension or diabetes, to preserve renal function. Additionally, patients with small renal masses < 4 cm were included even with normal contralateral kidney. We excluded patients with previous renal or abdominal surgeries. The study was approved by institutional review board, and written informed consent was obtained from all patients.

2.2 Clinical assessment

Routine preoperative investigations were performed including CBC, coagulation profile and serum creatinine. Estimated glomerular filtration rate (eGFR) was calculated using the Modification of Diet in Renal Disease (MDRM) equation [19]. Contrast-enhanced cross-sectional imaging of chest, abdomen and pelvis (CT or MRI) for each patient was assessed by two uro-oncologists familiar with the Renal Pelvic Score (RPS) [20] and the R.E.N.A.L [12]. If the two members disagreed in the scoring, a third opinion was obtained.

2.3 Surgical management and postoperative care

After positioning of the patient in the lateral decubitus position, all patients underwent open PN with a flank incision performed with excision of the 11th rib to gain access to the targeted kidney. Trans-gerotal approach was adopted in all cases, where dissection of the peri-renal fat was done except the area surrounding the mass. Site of the mass was identified intraoperatively by either direct visualization of the mass or ultrasound guidance, especially if the mass was totally endophytic. After intraoperative mass assessment, we decided to proceed with either warm ischemia or no ischemia. The incision line for the mass, outside its capsule and surrounding a rim of normal parenchyma, was outlined then deepened using a combination of blunt and sharp dissection, then the mass was removed with its surrounding fat. A sample from the deepest point of the bed was obtained for histopathological examination. Bleeding vessels and injuries to pelvi-calyceal system were repaired. Reconstruction of the renal bed was done by placement of hemostatic materials in the form of an oxidized cellulose polymer and/or oxidized regenerated cellulose. Parenchymal closure with the covering capsule was done using interrupted absorbable 2/0 sutures with the aid of oxidized cellulose polymer to support the sutures. We didn’t routinely insert a ureteric stent. Urethral catheter was removed on (day 1) while the drain was left for an average of 3 days then removed with discharge of the patient.

Follow-up visits were at 1, 4 and 12 week then at 6 and 12 months postoperatively. In each visit, evaluation was done by serum creatinine, eGFR, urine analysis and US. Additionally, CT or MRI was performed at 6 and 12 months.

2.4 Data collection

Different outcome parameters were reported including warm ischemia time (WIT), operative time, estimated blood loss, blood transfusion, renal vascular injury, ureteric or renal pelvis injury, pelvi-calyceal system (PCS) entry, conversion to RN, urinary leakage, length of hospital stay, postoperative renal function and pathology findings.

2.5 Newly modified nephrometry score (MNS)

We compared different outcome parameters according to R.E.N.A.L. and RPS. We modified R.E.N.A.L. with incorporation of new items in the score including hilar position of the mass as a numerical score not only a descriptive character (scored 0 or 5) and RPS (scored 1 or 3 according to the percentage of renal pelvis area contained inside the volume of the renal parenchyma whether less or more than 50%; respectively). Furthermore, we changed the score points given to some items of the original R.E.N.A.L. according to our point of view regarding their importance in describing the complexity of the mass. We kept the score of the radius “R” and relation to polar lines “L” as 1, 2 or 3, but we changed the score of the nearness to PCS and sinus “N” to 1, 2 or 4 and the score of the endophytic nature of the mass “E” to 1, 2 or 5. Accordingly, the new score is ranging from 5 to 23. It is classified into mild complexity (5–9), moderate complexity (10–14) or high complexity (15–23). A renal mass can be described using the MNS as (R + E + N + apx + L + H + RPS). The newly modified nephrometry score was used to compare the different outcome parameters to evaluate its predictability and whether it was better or not than R.E.N.A.L.

2.6 Statistical analysis

The Statistical Package of Social Science (SPSS) Software program, version 20, was used for statistical analysis. Nominal values were compared using Chi-square (χ2) test or Fisher-exact test as appropriate. For comparing numerical values of the three complexity groups, the one way ANOVA (with Post Hoc analysis) or Kruskal–Wallis test were used. p values < 0.05 were considered statistically significant.

3 Results

3.1 Demographic and perioperative data

Fifty-one consecutive patients with renal masses were included. The baseline characteristics are shown in Table 1. Perioperative data are presented in Table 2. Perioperative complications were detected in 17 (33.3%) patients. Injury to the renal artery was reported in 2 patients. Exploration of the renal artery revealed an intimal crack in the first patient; which was repaired, and a blood clot in the second patient; which was dislodged by Fogarty catheter. Postoperative hematuria that failed to respond to intravenous fluid and anti-bleeding measures were detected in 2 patients on day 5 and day 7, respectively. Diagnostic angiography was done and showed AV fistula formation which was controlled by injection of histoacryl. There were 6 cases of postoperative urine leakage. This was associated with perinephric collection in 2 patients. Four cases required double-J stent insertion, while the other 2 cases were successfully managed conservatively by urethral catheter insertion or extending the duration of the already inserted catheter. No pre- or intraoperative stenting was done except for 1 patient who had ureteric injury during dissection of a lower polar mass. Conversion to RN was mandated in 4 (7.8%) patients due to inadequate residual renal tissue after PN (1), severe injury to PCS after excision of the mass with impossible repair (1), failure to localize the site of a completely endophytic renal mass close to renal hilum even after the use of ultrasound (1) and involvement of the hilar vessels by the renal mass in the 4th patient.

Table 1 Preoperative demographic data of the patients and characteristics of the renal masses
Full size table
Table 2 Perioperative data, complications and follow-up
Full size table

Six patients developed postoperative renal function deterioration. Only one of these six patients had normal preoperative renal functions. The preoperative eGFR was (91 ml/min). The creatinine raised from 1 to 3.5 mg/dl immediately postoperatively. On follow-up after 3 months, eGFR improved to 88 ml/min; which was very close to preoperative reading. No patient developed recurrence after a 1 year follow-up.

3.2 Pathological data

The clear cell type was the most dominant pathological finding (26 cases) (Table 2). The cases of positive surgical margins (5.9%) were followed up every 3 months with cross-sectional imaging.

3.3 Effect of complexity on outcome according to R.E.N.A.L. and MNS

Different perioperative factors were compared according to complexity level in R.E.N.A.L. (Table 3) and MNS (Table 4). There was no statistically significant difference between the complexity groups in the preoperative characteristics in both scores. In R.E.N.A.L., the only detected variable that was affected significantly (p = 0.039) by the complexity level was the need for secondary intervention. The hilar position (H) and RPS showed statistically significant difference only according to MNS as they were not included in the original R.E.N.A.L. In MNS, the complication rate was significantly higher with increase in the complexity level (p = 0.037). Furthermore, the rate of complications was nearly doubled while moving from a low complexity level to a higher level (13.3% vs 30% vs 56.3%) which was completely different from the original R.E.N.A.L. The rate of positive surgical margin showed statistically significant difference according to MNS (p = 0.049) and was found only in the highest complexity level while it showed no significant difference according to the original R.E.N.A.L in which it was found in both moderate and high complexity levels. Although secondary intervention showed statistically significant difference in the original score (p = 0.039), it was higher in the low complexity level (18.2%) than in the high complexity level (5%). The relation of the rate of secondary intervention became better associated with the complexity level in the MNS although the significant difference was abolished (p = 0.757). Although the rate of PCS entry did not show statistically significant difference according to complexity level in both scores, it was more associated with the complexity level in MNS with nearly doubling when comparing the low and high levels (33.3% vs 62.5%, respectively) while in R.E.N.A.L, the rate of PCS entry was higher in the low than the moderate complexity. This more association with the level of complexity, in the absence of significant difference, was also observed in MNS when compared with the original score in many other perioperative outcome variables including operative time, WIT, blood loss, hemoglobin loss, blood transfusion, ureteric injury and conversion to RN.

Table 3 Renal mass complexity according to R.E.N.A.L. NS in relation to outcome parameters
Full size table
Table 4 Renal mass complexity according to newly modified score in relation to outcome parameters
Full size table

3.4 Predicting factors for postoperative complications

This was confirmed by analysis of predicting factors for postoperative complications (Table 5). The detected significant predicting factors were the complexity level according to MNS (p = 0.037), MNS (p = 0.014) and hilar position (p = 0.001).

Table 5 Predicting factors for postoperative complications
Full size table

4 Discussion

R.E.N.A.L. is one of the nephrometry scoring systems that proved its reproducibility and minimal inter-observer variability [21, 22]. It was used to predict the outcome of PN based on the complexity of renal mass in multiple studies; most of them were retrospective. In our prospective study, we aimed at evaluating the ability of R.E.N.A.L., and our MNS in prediction of the risk of perioperative complications, conversion to RN and other outcome variables. We had 17 (33%) patients with perioperative complications. This was in the range of reported complications in other studies (24.24–38.46%) [2, 13, 23,24,25,26]. We had 4 cases (7.8%) of conversion to RN which was similar to that found in different studies (6–16.3%) [17, 27, 28].

Similar to our study, many studies showed no significant difference in the rate of complications [13,14,15,16,17, 27], estimated amount of blood loss [15,16,17,18, 27] transfusion rate [13, 14, 18, 27], ischemia time [13,14,15,16], length of hospital stay [13, 15, 16, 18, 27], operative time [13,14,15,16, 18, 23, 27], positive surgical margins, [27] and change in eGFR [14, 16, 18, 27], in relation to R.E.N.A.L. and different complexity grades. On the other hand, other studies reported significant difference in the rate of some outcome parameters according to R.E.N.A.L. including the rate of total complications [18, 23], ischemia time [14, 15, 17, 18, 23, 27], the total operative time [17], estimated blood loss [14,15,16,17,18,19,20,21,22,23], length of hospital stay [14,15,16,17,18,19,20,21,22,23], renal function impairment [15] and the possibility of conversion to RN [17, 27].

As there was no statistically significant difference in nearly all of the outcome parameters according to different complexity levels using R.E.N.A.L., we tried to modify R.E.N.A.L. hoping that it will better predict the postoperative outcome and complications. We incorporated both the hilar position and RPS. Additionally, a significant limitation to the original R.E.N.A.L was giving an equal strength to its different components although some items may affect the complexity of the surgery more than others. Consequently, we modified the scoring of some items with giving more strength to E and N. These modifications were supported by what reported in the previous literature. Tomaszewski and colleagues evaluated the renal pelvic anatomy as an independent predictor of urine leak in 255 patients undergoing open PN. Despite the presence of more lesions with higher R.E.N.A.L. scores in patients with an extrarenal pelvis, the presence of an intrarenal pelvis was highly independently associated with risk of urine leak and rate of secondary intervention [20]. This was confirmed in other studies [29]. In the study reported by Liu et al. [18], only nearness to the collecting system “N”, out of the five components of R.E.N.A.L., was significantly associated with both incidence of complications and postoperative hemorrhage. Furthermore, Bruner et al. [30] reported that “E” score out of other components of R.E.N.A.L. was a significant predictor of postoperative urine leak. Tomaszewski et al. [29] reported that “E” score had a strong independent association with urine leak while the overall NS and its other components were not associated with urine leak.

The application of these modifications was associated with marked improvement in the ability of the complexity level to predict different outcome parameters. This was confirmed by analysis of different predicting factors for occurrence of complications which revealed MNS, complexity level according to MNS and hilar position as the only significant predictors. However, original R.E.N.A.L or its individual components (R, E, N, L) were found to be non-significant.

Our study has some limitations. The most important one is the relatively small sample size. Another limiting factor was the use of warm ischemia in some patients and zero ischemia in others. Moreover, we did not evaluate the split renal function. Furthermore, we had different pathologies including benign lesions. However, our newly modified score was associated with better prediction of the outcome of PN including complications even in the presence of a small sample size. The comparison of MNS versus R.E.N.A.L. added to the strength of the study. The prospective evaluation is another strong point. The short period of recruitment of patients with use of the same surgical technique by the same surgeon helped to abolish their effects on outcome. These finding will be better confirmed in a larger studies using the same ischemia protocol. Additionally, a longer follow-up may help to detect the effect of positive surgical margin on the rate of recurrence free survival.

5 Conclusions

The newly modified nephrometry score was associated with better prediction of the outcome of PN when compared to R.E.N.A.L. nephrometry score. Thus, it can be used for better prediction of the outcome following partial nephrectomy.