Results and Lessons Learned on Robotic Assisted Kidney Transplantation.

INTRODUCTION: Nowadays, minimally invasive surgery in kidney transplantation is a reality thanks to robotic assistance. In this paper, we describe our experience, how we developed the robotic assisted Kidney transplantation (RAKT) technique, and analyze our results. Material and Methods. This is a retrospective study of all RAKTs performed at our center between July 2015 and March 2020. We describe the donor selection, surgical technique, and analyze the surgical results and complications. A comparison between the first 20 cases and the following ones is performed.
RESULTS: During the aforementioned period, 82 living donor RAKTs were performed. The mean age was 47.4 ± 13.4 and 50 (61%) were male. Mean body mass index was 25 ± 4.7 and preemptive in 63.7% of cases. Right kidneys and multiple arteries were seen in 14.6% and 12.2%, respectively. Mean operative and rewarming time was 197 ± 42 and 47 ± 9.6 minutes, respectively. Five cases required conversion to open surgery because of abnormal kidney vascularization. Two patients required embolization for subcapsular and hypogastric artery bleeding without repercussion. Three kidneys were lost, two of them due to acute rejection and one because venous thrombosis. Late complications requiring surgery included one kidney artery stenosis, one ureteral stenosis, two lymphoceles, and three hernia repairs. We noticed a significant reduction in time between the first 20 cases and the following ones from 248.25 ± 38.1 to 189.75 ± 25.3 (p < 0.05). With a mean follow-up time of 1.8 years (SD 1.3), the mean creatinine was 1.52 (SD 0.7) and RAKT graft survival was 98%.
CONCLUSIONS: The robotic approach is an attractive, minimally invasive method for kidney transplantation, yielding good results. Further studies are needed to consider it a standard approach.

1. Introduction

Kidney transplant (KT) is the treatment of choice for end stage renal disease (ESRD) because it offers better survival and quality of life compared with dialysis treatment [1]. The surgical technique for kidney transplantation has not changed significantly over the last decades, probably due to technical difficulties and the necessity of abdominal incision for graft introduction.

Nowadays, minimally invasive surgical techniques are preferred to open ones in order to reduce morbidity in many surgeries. In this way, the techniques of laparoscopy and, more recently, robotics have spread around the world. But the application of laparoscopy to kidney transplantation has not succeeded due to its difficulty and low reproducibility rate. Only a few centers were able to perform this technique safely [2, 3].

Thus, robotics has filled this gap and has permitted us to obtain the capability to perform intracorporeal vascular anastomosis assisted by the DaVinci® surgical system (Intuitive Surgical, Inc.,) safely and reproducibly. For this reason, during the last 5 years, this technique has been introduced in many centers around the world with promising results, making minimal invasive kidney transplantation a reality.

Our department has a wide experience in open kidney transplantation (OKT) [4] and surgical innovation, being the first Spanish center performing a deceased donor kidney transplant in 1965 by Gil-Vernet et al. Regarding minimally invasive techniques, our group started a laparoscopic living donor nephrectomy program in 2002 with posterior introduction of minimally invasive techniques in kidney living donor nephrectomy such as assisted transvaginal natural orifice transluminal endoscopic surgery (NOTES) [5] and laparoendoscopic single site (LESS) in 2009 [6].

Following our previous experimental work, we developed a laparoscopic animal study on kidney transplantation, without its translation to a human setting because of the difficulties and poor feasibility [7]. Our wide experience in OKT and robotic surgery encouraged us to start our program of robotic assisted kidney transplantation (RAKT) during the summer of 2015, being one of the three European pioneer centers for this technique and currently the European center with the highest number of cases [8].

The aim of this study is to describe our experience, explain how we developed the RAKT technique, and analyze our results.

2. Material and Methods

A retrospective review of a prospectively maintained database was performed on consecutive RAKT recipients performed between July 2015 and March 2020 to assess surgical results, complications, and functional outcomes.

A description of recipient and donor selection, as well as surgical technique and evolution, is carried out. A comparison between the first 20 cases and the following ones is included. The Institutional Ethics Committee of the Hospital Clínic of Barcelona approved the study and, due to the nature of retrospective data review, waived the need for informed consent from individual patients.

3. Recipient and Donor Selection

After a medical work up, donors and recipients were assessed by a nephrologist and urologist indicating the feasibility of the transplantation. A high-resolution angio Ct scan is required to assess donor kidney pedicle and recipient iliac region. Paired cases are reviewed by a multidisciplinary team (urologist, nephrologist, radiologist, transplant coordinator, anesthesiologist, immunologist, ethics committee, etc.).

At the beginning of the program, only left kidneys were accepted for RAKT. After surgical technique consolidation, right kidneys were accepted and even kidneys with multiple vessels.

RAKT was originally indicated only to first transplant recipients without any vascular calcifications. Over time, we extended the indications: second kidney transplantation, and currently, we accept recipients with small and nonconcentric external iliac vascular calcification.

4. Surgical Technique Development

Our RAKT technique follows the principle surgical technique described by Menon et al. with some minor changes over time [9, 10].

Historically, our group used Ringer's lactate to perfuse the living donor kidney because of the low warm ischemia time and low rewarming time in our series. After the first cases of RAKT, we noticed a slow creatinine normalization so we decided to use Celsior® to minimize cell damage.

After kidney extraction and perfusion, the graft is carefully prepared in bench surgery. It is very important to ligate all small vessels to avoid any bleeding after reperfusion. The kidney is wrapped in a gauze jacket with ice, making a small window for artery and vein exposure. The lower pole of the kidney is marked with a longer suture (Figure 1).

Figure 1

Wrapped Kidney before introduction.

The recipient is placed in the decubitus supine position with open legs and 30° of Trendelenburg. Six trocars are placed (four 8 mm robotic trocars, a 5 and 12 mm trocar, see Figure 2.

Figure 2

Trocar location.

Following the original technique, we initially introduced the kidney through a midline umbilicus incision using a GelPOINT® device (Applied Medical). In order not to undock the robot arms during kidney introduction, we decided to move to a Pfannenstiel incision. This location permits an easy open conversion if any problem occurs during surgery. Our previous experience with the NOTES-assisted approach in kidney surgery has permitted us to translate this technique to kidney transplantation, using the vagina as a natural orifice for graft insertion in selected cases. The transvaginal approach requires a wide vagina to permit kidney insertion without difficulties and avoiding kidney damage. In this approach, an Alexis® wound protector retraction is placed through the vagina. This device aids kidney introduction and avoids pneumoperitoneum leakage.

The first step of the surgery is vessel preparation for anastomosis. This technique requires wider vascular dissection compared to the open one. Before kidney insertion, it is important to create a peritoneal flap for kidney repositioning after transplantation. The bladder is also prepared for ureteral anastomosis, so an extravesical submuscular tunnel is performed.

The wrapped kidney is introduced through the elected incision, taking into account its orientation, to avoid any error. The kidney is then positioned medially to the vessels to permit comfortable vascular anastomoses.

In some cases, iliac vein transposition is required to reduce the distance between vessels, which is especially useful in right kidneys with a short vein. The iliac vein is clamped with bulldogs. The venotomy is performed using robotic Pots scissors and is then flushed with heparin solution. A 6/0 Gore-Tex CV-6 (W.L. Gore and Associates Inc., Flagstaff, AZ, USA) running suture is then performed with a knot in each vertex (18 cm). After completing the vein suture, a bulldog clamp is placed on the graft vein to check the anastomosis' tightness. Iliac vein clamps are then placed on the iliac artery. Arteriotomy is performed using the same Pots scissors, and the remaining blood is removed by flushing heparin solution. A 6/0 Gore-Tex running suture is then performed (15 cm). A new bulldog clamp is placed on the renal artery, and the iliac artery bulldogs are then removed.

After kidney revascularization, the gauze jacket is removed for kidney color inspection. To reduce the possible deleterious effect of pneumoperitoneum on kidney graft function, we reduce the pressure from 12 to 10 mmHg.

As a safety measure, from the fifty cases, we perform an intracorporeal ultrasound to confirm good renal flow. This is due to a failure in detecting abnormal perfusion problems in one of our first cases, in which a technical problem with vein anastomosis caused a bad vein drainage that ended with open reanastomosis and final transplantectomy after 24 hours.

The ureteroneocystostomy was performed according to the Lich-Gregoir technique with ureteral stent insertion. A two 4/0 V-Loc™ (Medtronic) is used for the anastomoses (video in the Supplementary Material for comprehension (available here)).

5. Statistics

For the purpose of the present study, a descriptive analysis was performed. Continuous and categorical variables were presented as mean (SD) and n (%), respectively. We used Fisher's exact test to compare differences between the first 20 patients and the following 20. Significance was p value < 0.05. Statistical analyses were performed with Microsoft SPSS-PC+, version 25.0 (SPSS, Chicago, IL, USA).

6. Results

During the aforementioned period, 82 RAKTs were performed at our institution. Donors and recipient's characteristics are described in Table 1. Donors were mostly females (75.6%). Minimally invasive surgeries for donor nephrectomy were performed in all donors but one 64 (78.1%) laparoscopic, 15 (18.3%) transvaginal assisted, 2 (2.4%) LESS, and one mini open technique. ABO incompatible couples represented 20.7% of cases.

Table 1

Donors and recipients characteristics (n = 82).

Variable	Results
Donor characteristics
Donor sex (male/female), n (%)	20 (24.4)/62 (75.6)
Donor age (mean, SD)	53.8 (10.4)
Donor BMI (mean, SD)	24.9 (3.1)
Donor side, (left/right), n (%)	70 (85.4)/12 (14.6)
Vascular anatomy, n (%)
Multiple arteries	10 (12.2)
Multiple vein	2 (2.4)
Recipient characteristics
Recipient sex (male/female), n (%)	50 (61)/32 (39)
Mean recipient age at surgery, yr (SD)	47.4 (13.4)
Mean BMI (SD)	25 (4.7)
Medical history, n (%)
Diabetes mellitus	3 (3.6)
Nephroangiosclerosis	12 (14)
Polycystic kidney disease	15 (18)
Immunological disease	5 (6)
Glomerulonephritis	22 (22.9)
Interstitial nephropathy	4 (4.8)
Others	21 (15.2)
Preemptive yes/no, n (%)	52 (63.4)/30 (36.6)
Relationship with the donor, n (%)
Parent	25 (30.9)
Brother/sister	17 (21)
Wife/husband	29 (35.8)
Others	10 (12.3)
ABOi, n (%)	17 (20.7)

The mean warm ischemia time was 3 ± 1.6 minutes. The right kidney was removed in 12 cases (14.6%). Ten cases had multiple arteries, which required end to side anastomoses to obtain one final vessel in 8 cases. In two patients, two independent arterial anastomoses were performed. A double vein was seen in two cases, a small vein was discarded in one case, and anastomosis in a pantaloon fashion was performed in the other one. In nineteen cases, the graft was introduced via an umbilical incision (23.3%), in 57, a Pfannenstiel incision (69.4%) was used and in 6 cases (7.3%), the kidney was introduced through the vagina.

The left iliac fossa was used in eleven cases (13.4%) due to previous KT in all but one with iliac right fossa bowel adhesion. A transposition of the external iliac vein was performed in 10 cases to permit easier vein anastomosis.

Five cases were converted to the open approach due to abnormal perfusion after reperfusion. One case ended with vein thrombosis and required transplantectomy 24 hours later; in two cases, vein reanastomosis was carried out because of abnormal vein drainage, due to a vein valve and a vein rotation, respectively. One transvaginal case with multiple arteries had bad perfusion requiring kidney reperfusion and reanastomosis. In the last converted case, the kidney recovered good perfusion spontaneously after opening the abdomen.

Total mean surgical and rewarming time was 197 minutes (SD 42) and 47 minutes (SD 9.6), respectively. All surgical data are described in Table 2. In Figure 3, we show surgical time over cases, converted cases, and mean operative time every 10 cases.

Table 2

Surgical data (n = 82).

Variable	Mean (SD)
Operative time (min)	197 (42)
Warm ischemia time (min)	3 (1.6)
Rewarming time (min)	47 (9.6)
Arterial anastomosis time (min)	17 (5.36)
Vein anastomosis time (min)	19 (4.8)
Ureterovesical anastomosis time (min)	20 (6.5)
Estimated blood loss (cm3)	130 (100)

Figure 3

Surgical time over cases, converted cases, and mean operative time every 10 cases.

We have analyzed operative time and rewarming time over cases; we have compared the first 20 cases with the following 20 until the end. We noticed a significant reduction in time between the first 20 cases and the following ones, from 248.25 ± 38.1 to 189.75 ± 25.3 (p < 0.05) (Figure 4).

Figure 4

Mean surgical and rewarming time every 20 cases.

Recipients received immunosuppression with induction treatment with either basiliximab (Simulect®, Novartis) or a lymphocyte depleting antibody (Thymoglobulin® or Grafalon®) according to immunologic risk. Moreover, all patients received tacrolimus and steroids and either mycophenolic acid (CellCept® or Myfortic®) or an mTOR inhibitor (rapamune® or Certican®). In the ABOi recipient, the desensitization protocol consisted of rituximab, plasma exchange or immunoadsorption, and immunoglobulins.

Early complications (30 days) included one graft loss due to vein thrombosis. Besides, this two bleeding were successfully treated with selective embolization of a subcapsular renal and hypogastric artery, respectively. Two kidneys were lost because of acute rejection in an ABO compatible and incompatible couple, respectively (Table 3).

Table 3

Early complications according the Clavien Dindo classification.

Complications	n (%)
Grade I
Wound hematoma	1 (1.2)
Ileus	1 (1.2)
Grade II
Transfusion	6 (7.2)
Grade IIIa
Embolization	2 (2.4)
Grade IIIb
Transplantectomy for vein thrombosis	1 (1.2)
Transplantectomy for acute rejection	2 (2.4)
Grade IV	0
Grade V	0

None of the recipients analyzed presented hematuria or wound infection. Postoperative creatinine evolution is shown in Figure 5. Mean hospital stay was 7.9 days (SD 2.5).

Figure 5

Mean creatinine post-op evolution.

Late complications included two lymphoceles that requires surgical marsupialization and three (3.6%) umbilical incision hernias and two ureteral stenoses developed over time, treated by balloon dilatation and anastomoses to the native ureter, respectively, and one arterial stenosis in a terminolateral anastomosis requiring angioplasty. One graft was lost due to chronic rejection at 2.5 years of transplantation. Late complications are described in Table 4.

Table 4

Late complications according the Clavien Dindo classification.

Complications	n (%)
Grade I
Lymphocele	1 (1.2)
Grade II
Rejection (humoral/cellular)	8 (9.6)
Grade IIIa
Ureteral stenosis	1 (1.2)
Angioplasty	1 (1.2)
Grade IIIb
Lymphocele	2 (2.4)
Hernia repair	3 (3.6)
Ureteral stenosis	1 (1.2)
Grade IV	0
Grade V	0

With a mean time of follow-up of 1.8 years (SD 1.3), the mean creatinine was 1.52 (SD 0.7), patient survival 100%, and RAKT graft survival was 95.12% (Figure 6).

Figure 6

Graft survival.

7. Discussion

Innovation is a challenge and requires some kind of risk that has to be minimal in a surgical setting. For this reason, a new surgical technique has to be tested in an animal model before human implantation. As an innovative group, we were one of the first groups using the NOTES-assisted or hybrid approach for vaginal kidney extraction with excellent results in the donor [5]. We also introduced LESS in kidney surgery after developing the technique in an animal model [11].

Kidney transplantation is a different scenario because it requires abdominal incision for graft introduction and vascular anastomoses. This surgery has not changed over time because of the difficulty with current tools. In this way, we tried to introduce laparoscopy into kidney transplant surgery, and to do so, we conducted a laparoscopic kidney transplant animal model. We included ten pigs in the study that were submitted to laparoscopic radical nephrectomy and then laparoscopic kidney transplantation. Despite surgeries being performed with 2D vision and the surgeon describing the procedure as extremely difficult, we managed to complete eight cases with good functional results after 24 hours [7]. Using the experimental model described by us, the first totally intracorporeal laparoscopic living donor kidney transplantation was performed with good functional results, but the conclusion was that the surgery was too demanding and it was not repeated [2].

Robotics has changed this scenario and a minimally invasive kidney transplant technique becomes a reality. As in any other innovative surgery, its implementation comes with potential harm during the learning curve. The group of Menon and Ahlawat used the IDEAL (the Innovation, Development, Exploration, Assessment, and Long-term study) Guidelines in the development of RAKT [12], proving its safety and efficacy when performed by surgeons experimented in robotics [9, 10].

Before starting our program, we visited the group of Dr. Alawhat in India to learn the technique, and then we planned our first case. This first case was selected carefully to avoid any problem: a young patient, preemptive with excellent iliac vessels.

As in all innovative surgeries, after the first cases, we adopted some surgical changes to facilitate and ameliorate the former technique. One of these changes has been the kidney introduction site. We started using periumbilical incision like others groups [13] but we moved to a Pfannenstiel approach or even transvaginal insertion in selected recipients. The Pfannenstiel incision permits a quicker intra-abdominal positioning of the kidney and a lateral prolongation of the incision in case of urgent conversion to an open approach. Another reason for this change is the proven higher rate of hernia in periumbilical incision compared to Pfannestiel [14]. This issue is especially important as the transplant population has a higher risk of wound complications related to immunosuppressive treatment [15]. In our series, three cases presented with incision hernia (3.6%); all of them from the umbilical incision, although this percentage is lower than other publications (9-16%) [16]. Our group has a wide experience with using natural orifices for organ removal [5, 17]. The use of the vagina as a channel for kidney insertion permits us to reduce incisions even further, as has been already described in obese female RAKT, with good results [18, 19].

In our series, right kidneys and multiple arteries were used in 15% and 12% of cases, respectively. In ten right kidney cases, a transposition of the iliac vein was carried out to facilitate anastomosis. This technique had previously been reported with good results in open cases [20]. Multiple arteries have also been used by other centers showing similar functional results [21]. An arterial stenosis of the terminolateral anastomosis performed in bench surgery was seen in a converted case during follow-up, requiring angioplasty with final good results (1.2%). Anyway, it is possible to anastomose two arteries with a reasonable increase of ischemia time.

One issue that concerns all transplant surgeons is the operative time, and the more important rewarming time that can have a deleterious effect on graft function. In our series, after the first 20 cases, operative time became competitive with the open approach (mean operative time in 197 ± 45 minutes with a rewarming time of 47± min). Similar results are described by Ahlawat et al. [22]. In this paper, they demonstrated improved skills after 20-25 cases. In the study recently published by the ERUS-RAKT group, the cutoff for cases was 35 [23].

The last aspect we would like to comment is the risk of conversion. It is very important to understand that in the case of any nonideal perfusion of the graft that cannot be managed with the robot, conversion to open surgery is mandatory to solve it. As described in Results, we converted five cases to open surgery. Similar results were found for the RAKT-European Robotic Urological Society (ERUS) group [2]. Despite the three early lost kidneys (one because of a surgical issue), functional results are good and similar to other series [16].

8. Conclusion

Surgical innovation is important and has to be carried out with warranties to reduce any potential complication to a minimum. The robotic approach is an attractive, minimally invasive method for kidney transplantation, with good results. Further studies are needed to consider it a standard approach.