Comparative Analysis of Trifecta Outcomes in Robot-Assisted Partial Nephrectomy for cT1a Versus cT1b + Renal Tumours—a Single-Centre Study

Abstract

Robotic assistance is being increasingly utilised for nephron-sparing surgery for complex renal masses. We evaluated the outcomes of robot-assisted partial nephrectomy (RAPN) for cT1a versus cT1b + renal masses by a comparative analysis of trifecta outcomes between these two groups of patients. We utilised our prospectively maintained database to identify patients undergoing RAPN for cT1a (group 1, n = 41) and cT1b + (group 2, n = 37) renal masses from April 2016 to March 2020. The oncological and trifecta outcomes were analysed using appropriate statistical methods. Out of 78 patients, trifecta was achieved in 30 (38.4%) patients. There was no statistically significant difference in trifecta between cT1a and cT1b + tumours (p = 0.152). We found a statistically significant difference between the two groups in terms of RENAL scores (p = 0.0005), PADUA score (p = 0.0002), and robotic console time (133.8 ± 42.8 Vs 170 ± 54.8 min for cT1a versus cT1b + , respectively) (p = 0.002). On multivariate analysis, warm ischemia time (p = 0.069), blood loss (p = 0.345), UCS repair (p = 0.691) and GFR reduction (p = 0.152) were not statistically different. There was no statistically significant difference in intraoperative and post-operative complications (p = 0.9317) or length of hospital stay (p = 0.112). Although recurrences were observed in two patients (5.4%) of the cT1b group, there was no statistical difference in the recurrence-free survival at 12 and 24 months. Our study shows that RAPN can be safely done for cT1b + renal tumours. These findings reinforce the view that RAPN should be considered a viable option for cT1b + lesions whenever technically feasible.

Introduction

Robot-assisted partial nephrectomy (RAPN) is increasingly employed in the treatment of localised renal tumours. Partial nephrectomy (PN) is the standard of care for cT1a and selected T1b + renal masses when technically feasible [1]. Historically, cT1b + tumours (> 4 cm) were treated with radical nephrectomy. However, given the increased risk of chronic kidney disease (CKD) and the resulting cardiac morbidity and mortality, the adoption of partial nephrectomy has risen rapidly recently to treat even larger and complex renal masses [2]. RAPN is associated with several advantages compared to open and laparoscopic approaches and is now being performed more commonly when available [3, 4]. Increasing experience with RAPN has facilitated the excision of complex lesions while maintaining excellent outcomes [5]. However, tumour complexity can increase the technical challenge of partial nephrectomy, resulting in potentially higher rates of collecting system entry, prolonged warm ischemia time (WIT), and resultant renal injury [6]. There is a dearth of studies comparing the safety, oncological, and functional outcomes of RAPN in cT1a vs higher complexity renal tumours in the Indian population who tend to present a decade earlier and in a more advanced stage when compared to their western counterparts [7–9]. We conducted a retrospective comparative study involving Indian patients who underwent RAPN in our centre to evaluate the trifecta rates in the two groups. Definition of trifecta varies between different studies. In our study, trifecta is defined as warm ischemia time ≤ 20 min, negative surgical margins, and no post-operative complications (≥ grade 3 Clavien grade) [10].

Materials and Methods

Overall, 78 patients underwent RAPN between April 2016 and March 2020 by a single surgical team at our centre. To avoid the ‘learning curve effect’, we excluded the initial 23 patients who underwent RAPN in the initial 5 years after the fellowship training of the lead surgeon (from 2011 to March, 2016). The data was acquired from a prospectively maintained database and analysed. The clinical staging of renal masses was determined by radiographic measurement of maximal tumour dimension and patients were stratified based on their tumour size into two groups: group 1—cT1a (≤ 4 cm) (n = 41) and group 2—cT1b + (> 4 cm) (n = 37). RAPN was advised for tumours whenever feasible. However, we recommended radical nephrectomy for patients with diffuse encasement of renal pedicles by tumour, diffuse invasion of collecting system, adjacent organ involvement, significant hilar lymphadenopathy, and when only less than 20% of global nephron mass could be preserved by partial nephrectomy.

We compared the demographic and clinical characteristics of patients including age, gender, body mass index (BMI), baseline estimated glomerular filtration rate (eGFR, ml/min/1.73 m²), Eastern cooperative oncology group (ECOG) score, tumour size (cm), RENAL & PADUA (preoperative and dimension for anatomical classification) nephrometry scores, tumour laterality and location (polar, anterior/posterior, hilar).

Perioperative and functional parameters including console time, WIT, estimated blood loss (EBL), collecting system repair, intraoperative complications, length of stay, percentage change in GFR at 3 and 12 months via di-ethylene tri-amine penta-acetic acid scan (DTPA) scan, post-operative complications (Clavien-Dindo grade), and positive surgical margins (PSM) were also analysed. DTPA renogram was performed in view of its accuracy and its ability to report the differential renal function, before and after RAPN [11]. Trifecta rates were calculated from the above data for both groups. Intermediate oncological and functional outcomes at 1- and 2-year follow-up were assessed as well. The factors impacting perioperative and functional outcomes were analysed in multivariate regression analysis.

Statistical Analysis

The data was compiled and analysed using MS Excel (R) office 2019, GraphPad prism 8.4.2, and SPSS version 23.0. Descriptive statistics were presented in the form of proportions/percentages for categorical variables and median/interquartile range (along with mean and standard deviation wherever deemed necessary) for continuous data; Fisher exact/chi-square test was used for the comparison of proportions (categorical variables). Continuous variables were analysed using the Mann–Whitney test (independent group/unpaired data) and Wilcoxon sign rank test (for paired data).

Univariate and multivariate analyses (controlling for age, gender, BMI, laterality and surgery year) were done using the linear (for continuous dependent) and logistic regression (for bivariate dependent variables) models to estimate parameter difference between cT1a and cT1b + groups. Standardised beta covariates for continuous parameters (linear regression) and odds ratio for the categorical parameters (logistic regression) were estimated. Recurrence-free survival and overall survival were assessed using the Kaplan–Meier (KM) survival analysis and a comparison of these survival parameters between the two groups was done using the log-rank test (overall comparison) and Gehan-Beslow-Wilcoxon test. p value of < 0.05 was considered significant.

Results

Out of a total of 78 patients with median age of 58.5(43.2–65.7) years and overall male to female ratio of 2.9:1, 41 and 37 patients underwent RAPN for cT1a and cT1b + tumours respectively. The demographic profile including median age (p = 0.677) and BMI (p = 0.319) was similar for both groups.

Table 1 shows that the median size of the tumour 5 (4.5–5.4) cm vs 3.2 (2.6–3.7) cm (p = 0.0001), median RENAL score (8 vs 7) (p = 0.0005), and median PADUA score (9 vs 8) (p = 0.0002) were higher in cT1b + group. Among low RENAL score tumours, 80% belonged to cT1a stage. On the other hand, 77.78% of high RENAL score tumours were staged cT1b + in our study. Upper polar tumours were seen with higher frequency in the cT1b + group (48.6% vs 19.5%) (p = 0.006). All patients underwent resection of the renal masses and intraoperative frozen section was not performed. Table 2 shows statistically significant difference in console time between two groups 132 (111–161) min versus 180 (123.5–207.5) min (p = 0.002). Table 3 shows the association of cT1b + group of tumours with various continuous outcome measures assessed via a multivariate linear regression model. The standard beta covariates were calculated and a positive beta covariate highlights positive association between the two measures. cT1b + tumours were found to be associated with significantly higher duration of surgery (p = 0.022) and robotic console time (p = 0.002). None of the patients of either group had Clavien grade 3 complication or positive surgical margin. Similarly, none of our patients had to be readmitted for complications up to 30 days of surgery. One patient developed transient atrial fibrillation on post-operative day 4, in the cT1b + group. However, it resolved with medical treatment and did not require ICU admission. This patient was discharged on seventh post-operative day. Other surgical parameters such as selective arterial clamping, early unclamping, blood loss, warm ischemia time, and ureterocalyceal system (UCS) repair were not significantly different between the two groups.

Table1.

Demographic profile and preoperative parameters of patients involved in the study

Variables	Group 1	Group 2	p-value*
No. of patients	41	37
Age in years M (IQR)	56 (46–63)	62 (44–66)	0.677
Males: Females	29:12	29:8	0.443
BMI, M (IQR)	26.3 (24.4–28.7)	27.2 (24.4–30.3)	0.319
Clinical tumour size (mm), M (IQR)	32 (26–37)	50 (45–54)	< 0.001
Baseline GFR (DTPA) in % affected kidney, M (IQR)	50 (48–52.9)	48.5 (47–52)	0.156
Laterality Right Left	19 22	20 17	0.615
Polar location Upper (%) Mid (%) Lower (%)	8 (19.5) 15 (36.6) 18 (43.9)	18 (48.6) 9 (24.3) 10 (27.1)	0.006 0.244 0.123
RENAL score M (IQR)	7 (5–8)	8 (7–9)	0.005
RENAL score stratification#	Low complexity (n = 20) n (%)	Intermediate complexity (n = 41) n (%)	High complexity (n = 9) n (%)	p-value
cT1a stage	16 (80)	19 (46.34)	2 (22.22)	0.007
cT1b + stage	4 (20)	22 (53.66)	7 (77.78)
Console time in minutes (mean ± SD)	136.95 ± 52.84	155.54 ± 56.90	180.78 ± 39.15	0.049µ
Warm ischemia time in minutes (Mean ± SD)	17.07 ± 7.22	24.72 ± 11.40	31.49 ± 9.57	0.022µ
Estimated blood loss in ml (mean ± SD)	130.53 ± 107.93	153.66 ± 128.16	175.00 ± 116.50	0.033µ

n number of patients, M median, IQR interquartile range.

*Fisher exact test (categorical parameters), Mann–Whitney’s U test (continuous parameters).

^µKruskal-Wallis test.

^#RENAL score unavailable for eight patients, n = 70 patients.

Table 2.

Operative and post-operative parameters

Variable	Group 1	Group 2	p-value
Console time in minutes M (IQR)	132 (111–161)	180 (123.50–207.50)	0.002
Selective arterial clamp	0	2	0.802
Early unclamping	17	12	0.413
Warm ischemia time in minutes M (IQR)	21 (15–24)	24.5 (18.3–29)	0.095
UCS repair n (%)	33 (80.5)	28 (75.7)	0.609
Blood loss in ml M (IQR)	100 (87.5–162.5)	100 (100–200)	0.431
Blood transfusion n (%)	0(0)	0(0)	> 0.999
Length of hospital stay in days M (IQR)	2 (2–2)	2 (2–3)	0.112
Postoperative complications	0	1 (grade 2—Clavien-Dindo)	0.931
Trifecta achieved n (%)	18 (43.9)	12 (32.4)	0.301
Follow-up months M (IQR)	9.3 (3.7–15.1)	6.3 (3.2–13.6)	0.584

n number of patients, M median, IQR interquartile range.

None of the patients in both groups had intraoperative complications or positive surgical margins.

*Fisher exact test (categorical parameters), Mann–Whitney’s U test (continuous parameters).

Table 3.

Multivariate analysis to estimate the parameter difference between two groups of the study

Multivariate analysis (β covariates)—linear regression
Parameters	Β	p value
Postop versus preop creatinine difference	− 0.103	0.390
Postop versus preop GFR difference	− 0.220	0.152
Warm ischemia time	0.214	0.069
Duration of surgery (minutes)	0.296	0.022
Robotic console time	0.035	0.002
Multivariate analysis (exp (b)—odds ratios)—logistic regression
Variables	Exp (β)	p-value
Trifecta outcome	0.457	0.152
UCS repair	0.781	0.691
Early renal hilar unclamping	0.664	0.432

Trifecta was achieved in more patients in the cT1a group (18 vs 12) but this difference was not statistically significant (p = 0.3016). Table 3 shows the odds ratios associated with cT1b and above group in achieving the various categorical outcome measures via a multivariate logistic regression model. It was seen that the cT1b group had 55.3% lower odds of achieving trifecta outcomes compared to cT1a group. The results however were not significant statistically (p = 0.152). It was evident that the cT1b group had 21.9% lower odds of needing UCS repairs compared to cT1a group. The results however were not significant statistically (p = 0.691). The cT1b group had 33.6% lower odds of undergoing early unclamping compared to cT1a group. The results however were not significant statistically as well (p = 0.432). The most recent median follow-up is 8.7 (3–30.97) months with 9.3 months for the cT1a group and 6.9 months for the cT1b + group. On multivariate analysis, there was no significant difference in GFR (p = 0.152) at the most recent follow-up.

There were two instances of cancer recurrence in the study. The first patient who presented with multiple unresectable local recurrences in 6 months of surgery had Fuhrman grade-3 pT2 clear cell RCC with no sarcomatoid/rhabdoid component. The second patient had cT1b upper pole tumour with failed radiofrequency ablation. Although her preoperative imaging did not reveal any adrenal involvement by the upper pole renal tumour, there was intra-operative suspicion of contiguous adrenal involvement as the tissue planes were not discernable and a concomitant adrenalectomy was therefore performed. The tumour turned out to be Fuhrman grade-2 pT4 clear cell RCC involving the ipsilateral adrenal gland as well. She also presented after 6 months of surgery with metastatic disease. Both the patients were referred to medical oncology for further management. The recurrence-free survival for cT1b + at 6, 12, and 24 months was 100%, 94.6%, and 94.6% respectively, and for cT1a tumours it was 100% at 6, 12, and 24 months (p = 0.144) (Table 4). No mortality was reported in either group in the entire period of the study. Figure 1 explains the recurrence-free survival over 2-year follow-up (KM chart).

Table 4.

Survival parameters

Groups	Number of recurrences	6-month RFS	12-month RFS	24-month RFS	P-value (Mantel-Cox)	Gehan-Breslow-Wilcoxon test
cT1a	0	100%	100%	100%	0.144	0.145
cT1b +	2	100%	94.60%	94.60%

RFS recurrence-free survival.

Fig. 1.

Recurrence-free survival (KM chart) over 2-year follow-up

Discussion

In this study, we analysed our prospectively maintained database of patients undergoing RAPN and found that despite a longer operating time, patients with tumours > 4 cm in size achieve similar oncological and functional outcomes as compared to those having smaller lesions, with no increase in complication rates. Moreover, the rates of achievement of trifecta are also not significantly different between the two groups.

The median warm ischemia time of 21 min (cT1a) and 24.5 min (cT1b +) in our study compares favourably with the other studies published so far [12–19]. While it is expected that the WIT would increase with increasing tumour size and complexity, this change has not been statistically significant in our group of patients. This is likely attributable to the benefits afforded by a coordinated and consistent operating room assistance, our overall robotic experience, judicious case selection (especially for larger tumours), and frequent use of the early unclamping technique. The inner renorrhaphy was done using continuous sutures in the early unclamping technique. The vascular clamps were then released, after which the outer renorrhaphy was performed by the ‘sliding clip’ method.

None of our patients developed an intraoperative complication and there were no conversions to open surgery in any of the groups. Only one patient (2.7%) in group 2 had Clavien-Dindo grade 2 complications postoperatively. The complication rate of robotic partial nephrectomy can range from 2.2 to 26.6% as reported in the previous series [12–19]. The lower complication rate in our series may be attributable to proper selection of patients and the experience of the operating team. The absence of surgical complications in the cT1b plus group suggests that RAPN is safe in this subset of renal tumours. This is indeed a significant finding as it enables us to offer this minimally invasive approach to patients with larger/more complex tumours with its attendant advantages of a short hospital stay (median of 2 days in our study), minimal blood loss, almost eliminating the need for intraoperative transfusions (0% in our study), and a faster recovery with minimal post-operative pain.

We also did not encounter any positive surgical margins (PSMs) in our patient which augurs well for their long-term oncological outcomes. In a study by Bernhard et al. which included 809 patients who underwent nephron-sparing surgery, clinical tumour size of more than 4 cm, PSM, and bilaterality were risk factors for ipsilateral recurrence [20, 21]. The PSM rates reported in the previous series of robotic partial nephrectomy vary between 0 and 6.5% [12–19]. Pavan et al. analysed a cohort of 5465 patients who underwent minimally invasive partial nephrectomy and found no significant difference with regard to PSM rates in cT1b plus tumours compared to cT1a tumours [22]. Although the true measure of oncological success is evident only on long-term follow-up, in our study, none of the patients in group 1 (cT1a) had a recurrence as compared to two recurrences in group 2 (cT1b plus) during the follow-up period with no discernible statistical difference seen between the two groups. Overall, no mortality was reported in either of the groups during the study period. Previous studies comparing RAPN in cT1a vs cT1b renal tumours have reported no difference in recurrence at median follow-up of 22 months [15]and progression-free survival at 24 months [16].

The concept of ‘Trifecta’ has been used recently to define the outcomes of NSS. The idea of trifecta has, however, changed over a period of time. It was initially described as negative surgical margins, preserved renal function, and no urological complications. The duration of warm ischemia time varies between previous studies. In our study, the achievement of trifecta includes negative surgical margin, a warm ischemia time of < 20 min, and no post-operative Clavien-Dindo grade III/IV complications [10, 23]. The achievement of trifecta varies in different series and has been reported to be between 58 and 81% for cT1b plus tumours [10, 24–28].

In our study, trifecta was achieved in 43.9% in group 1 (cT1a) and 32.4% in group 2 (cT1b plus) patients. In multivariate analysis, there was a 55% decrease in the odds for trifecta achievement in group 2 compared to group 1, but it was not significantly different (p = 0.152). Although our rate of trifecta achievement was lower as compared to the previous series of 277 patients reported by Kim et al., it can be partly explained by the higher cut-off for warm ischemia time(< 25 min) employed in their study [10]. Moreover, on follow-up, our study did not find any difference in the functional outcomes (serum creatinine level, change in GFR) between the two groups.

We recognise that our study is not without limitations. Firstly, our study is subject to selection bias because of its retrospective nature, wherein the most complex larger tumours would likely not have been taken up for RAPN. Future studies in this population with matched-pair analysis can address the inherent problem of selection bias associated with our retrospective study. However, we agree with the well-accepted fact that a minimally invasive approach like RAPN, and indeed NSS itself, should be applied judiciously rather than indiscriminately. Secondly, the relatively small number of patients and their limited follow-up period may not have allowed us to capture the full impact of perioperative differences between the two groups on the long-term oncological and functional outcomes. Larger and longer studies are indeed needed to clarify this aspect. Thirdly, our data represents results from a single robotic surgical team with an overall experience of almost 1000 robotic uro-oncological cases. Hence, these outcomes may not truly reflect the ‘real-world’ situation regarding the application of RAPN for larger tumours, especially at centres where this program is still at its nascent stages. However, we believe that our study provides further impetus to applying RAPN in complex and larger tumours and will likely increase the confidence in its safety, efficacy, and feasibility.

Conclusion

As per our study, selected patients with renal tumours > 4 cm (cT1b +) can be offered robot-assisted partial nephrectomy without compromising safety, oncological efficacy, and functional outcomes when compared to patients with smaller tumours. Patients with > 4 cm tumours and/or high nephrometry scores must be counselled regarding the prolonged console times and its implications appropriately. Not only the tumour size, but also the nephrometry scores are equally paramount in counselling patients about complexity and the risk of complications. With the widespread acceptance of robotic surgery and the implementation of newer technological platforms, RAPN is likely to be increasingly used in the management of larger/complex renal tumours. Studies with larger sample size and longer duration of follow-up are needed further to ascertain the evolving role of RAPN in these situations.

Declarations

Ethical Approval

As the study is retrospective, ethical committee approval was not required.

Prior Publication

Nil.

Permissions

Nil.

Conflict of Interest

The authors declare no competing interests.