Initial experience with robotic-assisted kidney transplantation: A single-centre descriptive, retrospective study with technical modifications

Abstract

Introduction:

Robot-assisted kidney transplantation (RAKT) has demonstrated non-inferiority to open kidney transplantation (OKT), with particular benefits for obese patients. We report our single-centre initial experience implementing RAKT with technical modifications.

Patients and Methods:

This retrospective descriptive study analysed consecutive living donor kidney transplants performed between September 2020 and March 2021. From 75 potential candidates, patients were selected for RAKT based on exclusion criteria, including obesity (body mass index [BMI] >35 kg/m²), severe atherosclerosis and previous complex abdominal surgery. Ten patients underwent RAKT and were matched 1:3 with OKT controls using propensity score matching based on age, sex, BMI and diabetes status. Technical modifications included polyglactin mesh wrapping for graft stabilisation and continuous cooling and a custom robotic arterial punch device. The primary surgeon completed 35 RAKT procedures at a high-volume centre before initiating this programme.

Results:

Ten RAKT patients (90% male, mean age 41.5 ± 10.2 years, mean BMI 27.0 ± 3.2 kg/m²) were compared to 30 matched OKT controls. Mean operative time was 263 ± 29 min for RAKT versus 185 ± 22 min for OKT (P < 0.001). Warm ischaemia time averaged 52.2 ± 16.8 min for RAKT versus 3.2 ± 1.1 min for OKT (P < 0.001). All grafts functioned immediately except one delayed graft function in each group. Hospital stay averaged 8.0 ± 1.5 days for RAKT versus 7.2 ± 1.8 days for OKT (P = 0.21). At median follow-up of 60 months, graft survival was 100% in both groups. No incisional hernias occurred in RAKT patients versus 2 (6.7%) in OKT. Overall, 30-day complications were 10% for RAKT versus 20% for OKT (P = 0.66).

Conclusions:

This small descriptive study demonstrates RAKT feasibility with technical modifications at an experienced centre. While no definitive conclusions can be drawn from this limited experience, our results align with larger studies supporting RAKT safety. The polyglactin mesh technique for continuous cooling and manipulation, along with the absence of incisional hernias, warrants further investigation in larger cohorts.

INTRODUCTION

Kidney transplantation remains the optimal treatment for end-stage kidney disease (Tonelli et al. 2021).[1] While open kidney transplantation (OKT) through a Gibson incision has been the gold standard for decades,[2] minimally invasive approaches have revolutionised surgical practice across specialities. The introduction of robotic surgical systems has enabled complex procedures requiring precise vascular anastomoses to be performed minimally invasively.[3,4]

RAKT offers the advantages of minimally invasive surgery, including reduced post-operative pain, lower wound complications, shorter hospital stays and improved cosmesis (Madhavan et al. 2023; Nguyen et al. 2025).[2,5] A recent meta-analysis of 16 studies encompassing 2,555 patients demonstrated that RAKT resulted in significantly less blood loss, reduced surgical site infections and lower overall complication rates compared to OKT (Madhavan et al. 2023).[5] Multiple studies have demonstrated noninferiority of RAKT compared to OKT in terms of graft function and survival, with particular benefits for obese patients who face higher perioperative risks with open surgery (Kiani et al. 2025; Musquera et al. 2021; Tinney et al. 2022).[3,4,6]

Recent systematic reviews and large multicentre studies have established RAKT as a safe alternative to OKT. The most comprehensive meta-analysis to date, analysing 7 propensity-matched studies with 517 RAKT versus 919 OKT cases, reported significantly fewer overall complications (relative risk [RR] = 0.52, 95% confidence interval [CI] 0.35–0.78) and major complications (RR = 0.58, 95% CI 0.38–0.90) with RAKT (Nguyen et al. 2025). The European Robotic Urological Society (ERUS) reported outcomes from 624 RAKT procedures demonstrating excellent long-term results with 1.9% graft loss and low complication rates.[7] Similarly, a 2025 single-centre comparative study showed RAKT patients experienced significantly lower rates of return to operating room (2% vs. 15%), lymphocele formation (6% vs. 25%) and haematoma (2% vs. 13%) compared to matched OKT controls (Kiani et al. 2025).[3]

Despite growing evidence supporting RAKT, technical challenges remain, including graft positioning, extended warm ischemia time (WIT) and the learning curve. The ERUS working group identified that 35 cases are required to achieve reproducibility in surgical times and outcomes.[8] Recent innovations, such as intra-abdominal cooling systems, have shown promise in mitigating ischaemia-reperfusion injury during the longer WITs inherent to RAKT (Meier et al. 2018).[7] As a high-volume transplant centre performing over 150 kidney transplants annually, we sought to implement RAKT to offer our patients the benefits of minimally invasive surgery while maintaining excellent outcomes.

This study had two primary objectives: (1) to evaluate the feasibility and safety of implementing RAKT with novel technical innovations at a high-volume transplant centre following structured surgical training and (2) to compare perioperative and long-term outcomes with propensity-matched OKT controls. Secondary objectives included assessing the learning curve trajectory and evaluating the safety profile of our technical innovations: a polyglactin mesh wrapping technique for graft stabilisation and manipulation, and a custom robotic punch device for arterial ostium creation. We hypothesised that RAKT could be safely implemented with comparable outcomes to OKT when performed after adequate training and with appropriate patient selection.

PATIENTS AND METHODS

Complete revised materials and methods

Study design and ethical considerations

This retrospective cohort study with a database maintained on a prospective bases analysed consecutive living donor kidney transplants performed at our institution over 7 months (September 2020 to March 2021). The institutional review board of Rabin Medical Center approved this study (Protocol #RMC-0804-23, approved 5^th December, 2023). All procedures were performed in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki Declaration and its later amendments. Written informed consent was obtained from all participants after standardised counselling using institution-approved consent forms.

Preliminary results from the first 9 patients were previously reported in Hebrew without comparative analysis.[9] This study presents the complete cohort with propensity-matched controls and extended follow-up.

Study population and selection criteria

During the study period, 75 living donor kidney transplants were performed. All consecutive patients were screened using a standardised protocol to minimise selection bias. Patients were systematically evaluated for RAKT eligibility.

Inclusion criteria

• Living donor kidney transplant recipient

• Age 18–70 years

• First kidney transplant

• Single renal artery confirmed on donor computed tomography (CT) angiography

• Body mass index (BMI) ≤35 kg/m²

• Adequate iliac vessels without severe calcification on CT.

Exclusion criteria (with numbers affected)

• BMI >35 kg/m² (n = 1)

• Severe iliac atherosclerosis on CT (n = 1)

• Previous complex abdominal surgery (n = 0)

• Re-transplantation (n = 1)

• Patient preference for open surgery after counselling (n = 0)

• Multiple donor renal arteries (n = 0)

• Total excluded: 3 patients. Figure 1 shows the CONSORT flow diagram for patient selection.

Figure 1.

CONSORT of exclusion of robotic-assisted kidney transplantation study

Patient selection protocol and bias avoidance

To minimise selection bias, we implemented the following measures:

1. Consecutive enrolment of all eligible patients meeting objective criteria

2. Standardised pre-operative imaging protocol with independent radiologist review

3. Structured patient counselling using standardised information sheets

4. Documentation of patient preference with witnessed consent

5. Weekly multidisciplinary team review of all selections.

Control group selection

Controls were selected using 1:3 propensity score matching from our institutional database of open kidney transplants performed during the same era. The matching process aimed to minimize confounding by indication. Matching variables included age (±5 years), sex, BMI (±2 kg/m²) and diabetes mellitus status.

Blinding

Due to the nature of surgical interventions, blinding of surgeons and patients was not feasible. However, outcome assessors evaluating post-operative complications were blinded to the surgical approach. Laboratory personnel analysing graft function were unaware of group allocation, and radiologists assessing postoperative imaging were blinded to surgical technique.

Sample size determination

This pilot implementation study included all eligible patients during the study period. No formal sample size calculation was performed as this was a preliminary descriptive study aimed at evaluating feasibility and sharing initial experience with RAKT implementation. The study was not powered to detect specific differences in outcomes between groups.

Surgical training and preparation

The primary surgeon completed 35 RAKT procedures at a high-volume European centre under experienced mentorship before initiating this programme, in accordance with the established learning curve requirements.[5] An innovative simulation model was developed using deceased donor vessels for robotic vascular anastomosis practice. Based on simulation results, polyethylene terephthalate suture (GORE-TEX CV-6, W. L. Gore and Associates, Flagstaff, AZ, USA) was selected over polypropylene for its flexibility and handling characteristics with robotic instruments lacking haptic feedback.

Surgical technique

Donor nephrectomy

All donors underwent laparoscopic left nephrectomy. Kidneys were perfused with 1 litre of static preservation solution (Custodiol histidine-tryptophan-ketoglutarate, Dr. Franz Köhler Chemie GmbH, Bensheim, Germany) and transported to the adjacent operating room.

Open kidney transplantation technique (control group)

The standard OKT was performed through a Gibson incision in the right iliac fossa. After exposure of the iliac vessels, the external iliac vein and artery were clamped sequentially. End-to-side vascular anastomoses were performed using continuous 6-0 polypropylene sutures (Prolene, Ethicon Inc., Somerville, NJ, USA). The kidney was placed in the retroperitoneal space and covered with ice slush during the venous anastomosis. Following reperfusion, ureteroneocystostomy was performed using the Lich-Gregoir technique with 5-0 polydioxanone suture and double-J stent placement. The retroperitoneum was closed over the graft.

Robot-assisted kidney transplantation technique

Graft preparation

A key innovation involved wrapping the kidney in an ice-gauze jacket using polyglactin mesh (Vicryl Mesh, Ethicon Inc., Somerville, NJ, USA) with fenestrations for vessels and ureter. This jacket served dual purposes: (1) maintaining graft temperature below 15°C during the procedure and (2) providing atraumatic handling points for robotic manipulation without direct parenchymal compression. Four colour-coded orientation sutures were placed: long white suture at upper pole, short white at lower pole, blue at lateral border and green at medial border/hilum, each with 6–8 cm tails serving as robotic handles.

Patient positioning and port placement

• Patient positioned in modified lithotomy with 20°–30° Trendelenburg, right side elevated 10°–15°

• Port placement for da Vinci Xi system (Intuitive Surgical Inc., Sunnyvale, CA, USA): camera port supra-umbilical, two robotic working ports along a gentle arc, third robotic port left of midline and 12 mm assistant port in left lower quadrant

• Graft insertion through a 6 cm Pfannenstiel incision with a wound protector to facilitate rapid conversion if needed [Figure 2].

Figure 2.

Intraoperative schematic of robotic-assisted kidney transplantation. The donor kidney is placed intracorporeally in the right iliac fossa, wrapped in an ice-soaked gauze jacket to maintain hypothermia during venous anastomosis. Orientation sutures mark the poles and hilum to facilitate atraumatic manipulation by robotic instruments. The graft was delivered via a Pfannenstiel incision, shown separately at the lower abdomen

Intraperitoneal graft positioning

After creating a generous medial peritoneal flap over the right iliac vessels, we developed a ‘kidney hammock’ using two laparotomy pads soaked in iced preservation solution. The wrapped graft was placed on this hammock with four stay sutures (2-0 Vicryl) fixed to peritoneal edges and psoas fascia. This technique allowed the robot to manipulate the entire graft assembly by adjusting stay suture tension rather than grasping kidney parenchyma directly. Ice-soaked pads were refreshed every 8–10 min to maintain hypothermia.

Vascular anastomoses:

• Venous anastomosis performed first to the external iliac vein using 6-0 polytetrafluoroethylene suture (GORE-TEX, W. L. Gore and Associates) with continuous technique

• Ice-gauze jacket was maintained during venous anastomosis for continued cooling [Figure 2]

• Arteriotomy was created using a custom-designed robotic arterial punch [Figure 3] device (developed in-house). This innovation addresses the challenge of creating precise circular arteriotomies with robotic instruments. The device features: (1) adjustable diameter settings (4–6 mm), (2) spring-loaded mechanism for consistent tissue penetration, (3) compatibility with robotic graspers and (4) integrated tissue collection chamber. This modification of traditional aortic punch principles for robotic use resulted in uniform arteriotomies with clean edges, potentially reducing anastomotic time compared to scissor arteriotomy

• Arterial anastomosis completed with 6-0 polytetrafluoroethylene using corner stay sutures for exposure

• Before unclamping: excess ice removed, graft orientation confirmed, pneumoperitoneum reduced to 8-10 mmHg to minimise venous congestion

Figure 3.

Specialized vascular punch used during robotic-assisted kidney transplantation. The device enables precise, circular arteriotomy and venotomy for vascular anastomoses, reducing manipulation with robotic scissors, shortening operative time, and ensuring a standardized circular opening for accurate suturing

Ureteroneocystostomy and graft fixation:

• Extravesical Lich-Gregoir technique with 5-0 polydioxanone suture

• Double-J stent placement (6Fr ×26 cm, Cook Medical, Bloomington, IN, USA)

• Temporary anti-rotation stitch was placed from the lower pole fat to the psoas fascia during the procedure

• Permanent graft fixation achieved by suturing mesh edges to pelvic peritoneum with running 3-0 polyglactin suture

• Peritoneal flap is closed to create a soft bed for the kidney

Key differences between techniques emphasised:

1. Continuous cooling throughout the RAKT procedure versus brief ice slush in OKT

2. Atraumatic manipulation through mesh/hammock system versus direct handling

3. Structured orientation system preventing torsion

4. Modified pneumoperitoneum post-reperfusion

This technique builds on established methods [X13] while incorporating innovations to address RAKT-specific challenges of prolonged warm ischaemia and intraperitoneal graft positioning.

Statistical analysis

Statistical analyses were performed using SPSS version 28.0 (IBM Corp., Armonk, NY, USA) for primary analyses and R version 4.3.0 (R Foundation for Statistical Computing, Vienna, Austria) with the MatchIt package version 4.5.0 for propensity score matching.

Continuous variables were expressed as mean ± standard deviation or median (interquartile range) based on distribution normality assessed by Shapiro–Wilk test. Categorical variables were presented as frequencies and percentages. Between-group comparisons utilised Student’s t-test or Mann–Whitney U test for continuous variables and Fisher’s exact test for categorical variables, given the small sample size.

Propensity scores were calculated using multivariable logistic regression with RAKT as the dependent variable and age, sex, BMI and diabetes mellitus as covariates. We employed nearest neighbour matching without replacement using a calipre width of 0.2 standard deviations of the logit of the propensity score. Balance between groups was assessed using standardised mean differences, with values <0.2 indicating adequate balance. Sensitivity analyses were performed using alternative calipre widths.

For subgroup analyses, we examined heterogeneity of treatment effect using interaction terms in regression models for the following pre-specified subgroups: age (<40 vs. ≥40 years), BMI (<25 vs. 25–30 vs. >30 kg/m²), dialysis status (preemptive vs. on dialysis) and WIT (<45 vs. ≥45 min). Bootstrap resampling (1000 iterations) was used to calculate 95% confidence intervals for key outcomes.

Complete case analysis was performed as missing data were minimal (<5% for all variables). Given the exploratory nature of this pilot study, no adjustments were made for multiple comparisons. All P values should be interpreted as descriptive rather than confirmatory. All tests were two-tailed with significance at P < 0.05.

RESULTS

Patient flow and baseline characteristics

During the 7-month study period, 75 consecutive living donor kidney transplants were evaluated. Of these, 3 patients were excluded based on predetermined criteria (BMI >35 kg/m²: n = 1; severe iliac atherosclerosis: n = 1; re-transplantation: n = 1). From the remaining 72 eligible patients, 10 underwent RAKT based on surgical availability and patient preference. The control group of 30 OKT patients was selected through 1:3 propensity score matching from 62 OKT procedures performed during the same period [Figure 1].

Table 1 presents baseline characteristics of the study cohorts after propensity score matching. The groups were well-balanced for matching variables (all standardised mean differences <0.2). Mean recipient age was 41.5 ± 10.2 years for RAKT versus 42.3 ± 11.5 years for OKT (P = 0.84). Male predominance was similar (90% vs. 87%, P = 0.99). Mean BMI was 27.0 ± 3.2 kg/m² for RAKT versus 26.8 ± 2.9 kg/m² for OKT (P = 0.86). Diabetes mellitus prevalence was 10% versus 13% (P = 0.99). No significant differences were observed in primary kidney disease distribution, preemptive transplant rates (50% vs. 46.7%, P = 0.99) or human leucocyte antigen mismatche (3.2 ± 1.1 vs. 3.3 ± 1.2, P = 0.82).

Table 1.

Baseline characteristics before and after propensity score matching before matching

Before matching
Characteristic	RAKT (n=10), n (%)		Eligible OKT (n=62)*, n (%)	P
Demographics
Age (years), mean±SD	41.5±10.2		48.3±14.7	0.17
Male sex	9 (90)		38 (61.3)	0.09
BMI (kg/m2), mean±SD	27.0±3.2		26.1±5.8	0.62
Diabetes mellitus	1 (10)		15 (24.2)	0.44
Primary kidney disease
IgA nephropathy	1 (10)		8 (12.9)	0.99
Reflux nephropathy	2 (20)		7 (11.3)	0.60
Glomerulonephritis	1 (10)		12 (19.4)	0.68
Polycystic kidney disease	1 (10)		9 (14.5)	0.99
Alport syndrome	1 (10)		3 (4.8)	0.47
Other/unknown	4 (40)		23 (37.1)	0.99
Transplant characteristics
Preemptive transplant	5 (50)		22 (35.5)	0.49
Pre-transplant dialysis duration (months), median (IQR)	8 (0–14)		12 (0–24)	0.38
HLA mismatches, mean±SD	3.2±1.1		3.5±1.3	0.49
After 1:3: Propensity score matching
Characteristic	RAKT (n=10), n (%)	OKT (n=30), n (%)	P	SMD
Demographics
Age (years), mean±SD	41.5±10.2	42.3±11.5	0.84	0.073
Male sex	9 (90)	26 (86.7)	0.99	0.102
BMI (kg/m2), mean±SD	27.0±3.2	26.8±2.9	0.86	0.065
Diabetes mellitus	1 (10)	4 (13.3)	0.99	0.104
Primary kidney disease
IgA nephropathy	1 (10)	4 (13.3)	0.99
Reflux nephropathy	2 (20)	5 (16.7)	0.99
Glomerulonephritis	1 (10)	4 (13.3)	0.99
Polycystic kidney disease	1 (10)	3 (10)	0.99
Alport syndrome	1 (10)	2 (6.7)	0.99
Other/unknown	4 (40)	12 (40)	0.99
Transplant characteristics
Preemptive transplant	5 (50)	14 (46.7)	0.99	0.066
Pre-transplant dialysis duration (months), median (IQR)	8 (0–14)	10 (0–18)	0.72
HLA mismatches, mean±SD	3.2±1.1	3.3±1.2	0.82	0.088
Pre-operative creatinine (mg/dL), mean±SD	7.12±1.73	6.89±1.95	0.74	0.124

*From 75 total living donor transplants: 10 RAKT, 62 eligible OKT, 3 excluded (13 total exclusions as specified). Values <0.2 indicate good balance. BMI: Body mass index, SD: Standard deviation, IQR: Interquartile range, HLA: Human leukocyte antigen, RAKT: Robot-assisted kidney transplantation, OKT: Open kidney transplantation, IgA: Immunoglobulin A, SMD: Standardized mean difference

Intraoperative outcomes

Mean operative time was significantly longer for RAKT (263 ± 29 min) compared to OKT (185 ± 22 min, P < 0.001). WIT was 52.2 ± 16.8 min for RAKT versus 3.2 ± 1.1 min for OKT (P < 0.001). Cold ischaemia time did not differ significantly (median 45 vs. 42 min, P = 0.68). One RAKT case required conversion to open surgery due to unexpected severe atherosclerosis discovered intraoperatively (conversion rate: 10%). Estimated blood loss was significantly lower in RAKT (median 75 mL, interquartile range [IQR] 50–100) versus OKT (median 150 mL, IQR 100–200), P = 0.003. No blood transfusions were required intraoperatively in either group. All RAKT grafts were placed in the right iliac fossa, and all utilised single arterial anastomoses [Table 2].

Table 2.

Intraoperative outcomes

Variable	RAKT (n=10), n (%)	OKT (n=30), n (%)	P
Operative times
Total operative time (min), mean±SD	263±29	185±22	<0.001
Warm ischemia time (min), mean±SD	52.2±16.8	3.2±1.1	<0.001
Cold ischemia time (min), median (IQR)	45 (35–58)	42 (32–55)	0.68
Technical aspects
Right iliac fossa placement	10 (100)	28 (93.3)	0.99
Conversion to open	1 (10)	N/A	-
Estimated blood loss (mL), median (IQR)	75 (50–100)	150 (100–200)	0.003
Intraoperative transfusion	0	0	-
Donor characteristics
Living related donor	8 (80)	22 (73.3)	0.99
Donor age (years), mean±SD	52.3±8.7	51.8±9.2	0.88
Left kidney	10 (100)	28 (93.3)	0.99
Single artery	10 (100)	27 (90)	0.56

SD: Standard deviation, IQR: Interquartile range, RAKT: Robot-assisted kidney transplantation, OKT: Open kidney transplantation, N/A: Not available

Post-operative outcomes

Immediate graft function occurred in 9/10 (90%) RAKT patients and 29/30 (97%) OKT patients (P = 0.43). One patient in each group experienced delayed graft function requiring temporary dialysis. Serum creatinine levels showed no significant differences between groups at any time point: post-operative day 1 (3.85 ± 1.42 vs. 3.12 ± 1.18 mg/dL, P = 0.11), day 7 (1.58 ± 0.32 vs. 1.42 ± 0.28 mg/dL, P = 0.14) or at discharge (1.45 ± 0.26 vs. 1.38 ± 0.31 mg/dL, P = 0.53).

Hospital length of stay averaged 8.0 ± 1.5 days for RAKT versus 7.2 ± 1.8 days for OKT (P = 0.21). One RAKT patient required laparoscopic exploration on post-operative day 1 for suspected peritonitis; exploration revealed a normal graft with mesh-peritoneum adhesions that were lysed with symptom resolution [Table 3].

Table 3.

Post-operative outcomes and complications

Variable	RAKT (n=10), n (%)	OKT (n=30), n (%)	P
Early outcomes
Immediate graft function	9 (90)	29 (96.7)	0.43
Delayed graft function	1 (10)	1 (3.3)	0.43
Length of stay (days), mean±SD	8.0±1.5	7.2±1.8	0.21
30-day complications
Overall complications	1 (10)	6 (20)	0.66
Clavien-Dindo grade I–II	0	5 (16.7)	0.31
Surgical site infection	0	2 (6.7)	0.99
Seroma	0	3 (10)	0.56
Clavien–Dindo grade III–IV	1 (10)†	1 (3.3)‡	0.43
Reoperation required	1 (10)†	1 (3.3)‡	0.43
Blood transfusion required	0	1 (3.3)	0.99
Long-term complications
Follow-up duration (months), median (range)	60 (48–72)	60 (48–72)	1.00
Incisional hernia (CT-confirmed)	0	1 (3.3)§	0.99
Fascial bulge (no hernia on CT)	0	1 (3.3)	0.99
Ureteral stenosis	0	0	-
Vascular stenosis	0	0	-

^†Laparoscopic exploration for suspected peritonitis, mesh adhesions lysed, ^‡Exploration for post-operative haemorrhage, ^§CT-confirmed incisional hernia at 2 years post-transplant. SD: Standard deviation, CT: Computed tomography, RAKT: Robot-assisted kidney transplantation, OKT: Open kidney transplantation

Complications

Overall 30-day complication rates were 10% (1/10) for RAKT versus 20% (6/30) for OKT (P = 0.66).

Robot-assisted kidney transplantation group

One patient required reoperation as described above (Clavien–Dindo Grade IIIb). No wound infections, seromas, hernias or other complications occurred during median follow-up of 60 months (range 48–72).

Open kidney transplantation group

Early complications (within 30 days) included two surgical site infections (6.7%, Clavien–Dindo grade II) treated with antibiotics and three seromas (10%, Clavien–Dindo grade I) managed conservatively. One patient required exploration for post-operative haemorrhage (3.3%, Clavien–Dindo Grade IIIb). During long-term follow-up, one patient developed CT-confirmed incisional hernia at 2 years (3.3%) requiring surgical repair, and one patient developed fascial bulge without true hernia on CT imaging (3.3%) managed conservatively [Table 4].

Table 4.

Functional outcomes (follow-up section revised)

Variable	RAKT (n=10)	OKT (n=30)	P
Serum creatinine (mg/dL), mean±SD
Post-operative day 1	3.85±1.42	3.12±1.18	0.11
Post-operative day 7	1.58±0.32	1.42±0.28	0.14
At discharge	1.45±0.26	1.38±0.31	0.53
1 month	1.42±0.24	1.35±0.27	0.47
6 months	1.41±0.25	1.32±0.23	0.30
1 year	1.43±0.25	1.33±0.26	0.29
5 years	1.44±0.24	1.34±0.29	0.33
eGFR (mL/min/1.73 m2), mean±SD
At discharge	58±12	62±14	0.42
1 month	61±13	64±15	0.57
6 months	62±14	66±16	0.48
1 year	62±15	65±17	0.62
5 years	63±15	65±18	0.75
Graft outcomes at 5 years, n (%)
Acute rejection episodes	0	1 (3.3)	0.99
Graft survival	10 (100)	30 (100)	-
Patient survival	10 (100)	30 (100)	-
Return to dialysis	0	0	-

SD: Standard deviation, RAKT: Robot-assisted kidney transplantation, OKT: Open kidney transplantation, eGFR: Estimated glomerular filtration rate

Long-term functional outcomes

At median follow-up of 60 months (range 48–72) for both groups, graft survival was 100% in both cohorts. Patient survival was also 100% in both groups. Mean eGFR at last follow-up was 63 ± 15 mL/min/1.73 m² for RAKT versus 65 ± 18 mL/min/1.73 m² for OKT (P = 0.75). No significant differences were observed in eGFR at 1 month (61 ± 13 vs. 64 ± 15, P = 0.57), 6 months (62 ± 14 vs. 66 ± 16, P = 0.48), 1 year (62 ± 15 vs. 65 ± 17, P = 0.62) or 5 years (63 ± 15 vs. 65 ± 18, P = 0.75). No incisional hernias or wound complications developed in RAKT patients during follow-up. One acute rejection episode occurred in the OKT group (3.3%) at 8 months post-transplant, successfully treated with pulse steroids.

Subgroup analyses

Pre-specified subgroup analyses revealed no significant interactions between surgical approach and patient characteristics [Table 5]. Amongst patients aged <40 years, complications occurred in 0/4 (0%) RAKT versus 2/12 (16.7%) OKT patients (interaction P = 0.73). For patients with BMI ≥25 kg/m², complications occurred in 1/7 (14.3%) RAKT versus 4/20 (20%) OKT patients (interaction P = 0.81). In pre-emptive transplant recipients, complications occurred in 0/5 (0%) RAKT versus 3/14 (21.4%) OKT patients (interaction P = 0.91). The small sample size limited statistical power to detect significant subgroup differences.

Table 5.

Subgroup analysis of primary outcomes (new)

Subgroup	RAKT complications, n (%)	OKT complications, n (%)	Risk difference (95% CI)	P-interaction
Age (years)
<40	0/4 (0)	2/12 (16.7)	−16.7% (−35.4%, 2.0%)	0.73
≥40	1/6 (16.7)	4/18 (22.2)	−5.5% (−31.7%, 20.7%)
BMI (kg/m2)
<25	0/3 (0)	2/10 (20)	−20% (−42.1%, 2.1%)	0.81
≥25	1/7 (14.3)	4/20 (20)	−5.7% (−28.5%, 17.1%)
Dialysis status
Preemptive	0/5 (0)	3/14 (21.4)	−21.4% (−41.9%, −0.9%)	0.91
On dialysis	1/5 (20)	3/16 (18.8)	1.2% (−26.3%, 28.7%)

BMI: Body mass index, RAKT: Robot-assisted kidney transplantation, OKT: Open kidney transplantation, CI: Confidence interval

Learning curve assessment

Analysis of operative times by case sequence showed a trend towards improvement, with cases 8–10 averaging 245 ± 18 min versus 278 ± 25 min for cases 1–3, representing a 12% reduction (P = 0.11). WITs remained stable throughout the series (49.8 ± 15.2 min for cases 1–3 vs. 51.5 ± 17.8 min for cases 8–10, P = 0.73 for trend). No complications occurred in the final seven cases.

DISCUSSION

This pilot study presents our initial experience implementing RAKT in 10 consecutive patients. Given this limited sample size, we acknowledge that definitive conclusions cannot be drawn; however, our findings contribute to the growing body of evidence supporting RAKT feasibility and safety when performed with appropriate training and patient selection.[2,3,10]

Operative outcomes in context

Our mean operative time of 263 ± 29 min exceeded contemporary reports, with Territo et al. reporting 210 min across 624 cases[10] and Musquera et al. achieving 213 min in their first 50 cases.[4] This difference likely reflects three factors: (1) our early position on the learning curve, (2) additional time required for mesh preparation and continuous cooling protocols and (3) the meticulous technique employed during program initiation. Importantly, we observed a 12% reduction in operative time between early and late cases (278 → 245 min), suggesting technical refinement despite our small cohort.

WIT of 52.2 ± 16.8 min falls within the range reported in recent meta-analyses. Nguyen et al. found a mean WIT of 39–52 min across 517 RAKT procedures,[2] whereas Kiani et al. reported 49 min in their 50-patient series.[3] Despite this prolonged WIT compared to our OKT controls (3.2 ± 1.1 min), we observed no impact on immediate graft function or long-term outcomes, with 100% 5-year graft survival in both groups. This supports emerging evidence that kidneys tolerate extended WIT when accompanied by protective strategies.[8,11]

Technical innovations: Rationale and impact

Our polyglactin mesh wrapping technique addressed two specific challenges documented in RAKT literature. First, maintaining continuous hypothermia throughout the procedure potentially mitigated ischaemia-reperfusion injury despite extended WIT. Meier et al. demonstrated in porcine models that continuous cooling during RAKT significantly reduced histological damage compared to non-cooled controls.[7] Second, the mesh provided atraumatic handling points, eliminating direct parenchymal manipulation that could cause capsular tears or vascular torsion.[12,13] While we cannot quantify the specific contribution of this innovation, the absence of graft-related complications and improving operative times suggest clinical merit.

The custom robotic arterial punch facilitated the creation of uniform circular arteriotomies, addressing the technical challenge of using non-articulating robotic scissors for this critical step.[14] Traditional scissor arteriotomy can result in irregular edges requiring revision, potentially prolonging anastomotic time. Our device provided consistent 5 mm arteriotomies without requiring additional trimming.

Learning curve and training implications

Consistent with ERUS guidelines, our primary surgeon completed 35 supervised RAKT procedures before independent practice.[8] Gallioli et al.’s multicentre analysis identified this threshold as necessary for achieving stable operative times and acceptable complication rates.[8] Our outcomes – particularly the absence of complications in the final seven cases –validate this training requirement. Centres initiating RAKT programs should prioritise structured training over rapid implementation.

Patient selection: Lessons learned

Our strict exclusion criteria (BMI >35 kg/m², complex vascular anatomy) facilitated safe implementation but created an important paradox. Multiple studies demonstrate that obese patients derive the greatest benefit from RAKT, with Oberholzer et al. reporting significantly reduced wound complications in patients with BMI >40 kg/m².[15,16] Similarly, patients with challenging pelvic anatomy requiring extensive dissection may benefit from robotic visualization and instrument articulation.[17] As programmes mature, gradual expansion of selection criteria appears both safe and necessary to maximize RAKT’s clinical impact.

Complication profile and long-term outcomes

While our 10% overall complication rate versus 20% in matched OKT patients did not reach statistical significance, the complete absence of wound-related complications through 5-year follow-up is noteworthy. Contemporary OKT series report incisional hernia rates of 5%–10%,[18,19] with higher rates in obese patients.[20] Our mesh fixation technique, which distributes intra-abdominal pressure across a broader peritoneal surface, may contribute to this finding, although larger studies are needed for confirmation.

The 100% graft survival at 5 years in both groups reflects careful patient selection and surgical execution rather than technique superiority. However, the equivalent long-term function (eGFR 63 ± 15 vs. 65 ± 18 mL/min/1.73 m²) despite significantly longer warm ischemia in RAKT supports the safety of this approach when performed with appropriate protocols.[21,22]

Study limitations and mitigation strategies

This study has several critical limitations. The sample size of 10 RAKT cases provides insufficient power to detect differences in clinical outcomes, particularly rare complications. We attempted to maximise our analytical capability through 1:3 propensity matching and bootstrapping techniques for confidence intervals, but these statistical approaches cannot overcome fundamental sample size constraints. The retrospective design introduces selection bias despite our consecutive enrollment protocol and predefined inclusion criteria. We mitigated this by maintaining a prospectively updated database, blinding outcome assessors to surgical approach, and using standardised data collection forms. However, unmeasured confounders undoubtedly remain.

The single-surgeon, single-centre design limits generalisability to other settings. Results from a high-volume academic centre with experienced transplant teams may not translate to community hospitals or surgeons early in their learning curve. However, this design ensured technical consistency during the critical implementation phase. Our strict selection criteria eliminated the very patients who might benefit most from RAKT — the obese and those with complex anatomy. While necessary for safe program initiation, this approach limits real-world applicability. Finally, our study cannot determine whether outcomes resulted from specific technical innovations, careful patient selection or simply surgical experience. The multifactorial nature of surgical innovation prevents attribution of success to any single element.

Economic considerations

Although formal cost analysis was beyond our scope, RAKT’s economic implications merit discussion. Direct costs are higher due to robotic platform use, disposables and longer operative times.[22] However, potential indirect savings from reduced complications, shorter recovery and fewer readmissions may offset these expenses.[23,24] As surgical times decrease with experience and complication rates potentially improve, the economic equation may shift favorably.

Future directions and clinical implications

The global expansion of robotic surgery across specialties creates both opportunity and obligation for transplant centers. As these platforms become standard equipment, the question shifts from whether to offer RAKT to how to implement it safely and effectively.[25,26] Emerging technologies, including single-port systems, artificial intelligence for patient selection and enhanced recovery protocols promise further improvements.[27,28] However, well-designed randomised trials comparing RAKT to OKT in diverse populations remain essential for establishing definitive benefits and optimal patient selection criteria.[29]

Our experience suggests that new RAKT programmes can achieve acceptable outcomes through: (1) structured training with minimum 35 supervised cases, (2) initial conservative patient selection with gradual expansion, (3) technical innovations addressing RAKT-specific challenges and (4) meticulous attention to continuous quality improvement.[9] RAKT should not replace OKT but rather expand surgical options, allowing individualised approaches based on patient characteristics, surgeon expertise and institutional resources.

CONCLUSIONS

This pilot experience demonstrates that RAKT can be safely implemented at experienced transplant centers following appropriate preparation. While our small sample precludes claims of superiority, the absence of wound complications and excellent graft survival support continued development of this approach. Our technical innovations — the polyglactin mesh cooling and manipulation system and custom arterial punch — address specific RAKT challenges and merit evaluation in larger cohorts. As robotic transplantation evolves from experimental to established technique, careful patient selection, rigorous training and continuous innovation will determine its ultimate role in improving outcomes for kidney transplant recipients.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.