A Randomized, Single-Blind Clinical Trial Comparing Robotic-Assisted Fluoroscopic-Guided with Ultrasound-Guided Renal Access for Percutaneous Nephrolithotomy
This article is commented on by the following:
Abstract
Purpose:
We conducted a randomized, single-blind clinical trial comparing the surgical outcomes of robotic-assisted fluoroscopic-guided (RAF group) and ultrasound-guided (US group) renal access in mini-percutaneous nephrolithotomy (PCNL).
Materials and Methods:
We recruited patients who underwent mini-PCNL with ureteroscopic assistance for large renal stones between January 2020 and May 2021. Block randomization was performed using online software. Automated needle target with x-ray was used for fluoroscopic-guided renal access in the RAF group. PCNL was performed by residents using a pneumatic lithotripsy system with 16.5Fr/17.5Fr tracts. The primary outcome was single puncture success, and the secondary outcomes were stone-free rate, complication rate, parameters measured during renal access and fluoroscopy time.
Results:
In total, 71 patients (35 in US group, 36 in RAF group) were enrolled. No difference was seen in the single puncture success rate between the US and RAF groups (34.3% and 50.0%, p=0.2). In 14.3% cases in the US group vs no cases in the RAF group, the resident was unable to obtain access due to difficult targeting (p=0.025). The mean number of needle punctures was significantly fewer, and the median duration of needle puncture was shorter in the RAF group (1.83 vs 2.51 times, p=0.025; 5.5 vs 8.0 minutes, p=0.049, respectively). The stone-free rate at 3 months after surgery was 83.3% and 70.6% in the RAF and US groups, respectively (p=0.26). Multivariate analysis revealed that RAF guidance reduced the mean number of needle punctures by 0.73 times (p=0.021).
Conclusions:
RAF renal access in mini-PCNL may have further potential applications in this field.
| AI |
artificial intelligence |
| ANT-X |
automated needle targeting with x-ray |
| CT |
computerized tomography |
| KUB |
kidney ureter bladder |
| PCNL |
percutaneous nephrolithotomy |
| RAF |
robotic-assisted fluoroscopic |
| SFR |
stone-free rate |
| URS |
ureteroscopy |
| US |
ultrasound |
Percutaneous renal access is the most crucial step during percutaneous nephrolithotomy (PCNL); therefore, most endourologists approach needle puncture and dilation with maximum precaution.1 The CROES PCNL Study Group reported that PCNL had a 15% overall complication rate, including nearly 5% complications with a Clavien-Dindo grade of III or worse.2 Clinically significant complications such as massive bleeding requiring blood transfusion and/or embolization, thoracic and bowel injuries, renal pelvis perforation and even sepsis are possible consequences of inadequate renal access during PCNL.3 Therefore, researchers have been seeking ways to improve techniques and training.
There are 2 major renal access methods: fluoroscopic guidance and ultrasound (US) guidance with or without fluoroscopic image assistance.4 These 2 techniques are thought to be comparable in terms of postoperative outcomes, including stone-free status and complication rate; however, this finding can only be applied when procedures are performed by experts. A recent systematic review concluded that the learning curve for PCNL was 60 cases for gaining renal access and 105 cases for achieving acceptable stone-free rates (SFRs).5 While some new modalities such as endoscopic guidance,6,7 computerized tomography (CT) guidance,8 real-time virtual sonography assistance9 and electromagnetic guidance10 have been introduced in clinical practice, renal access is still challenging for beginners or under specific conditions. Given the complexity of 3D image structure as gauged from intraoperative 2D images provided to surgeons, pre- or intraoperative planning using immersive virtual reality helps us understand the structure and facilitate a safe and effective procedure. However, these techniques still require expertise in targeting the calyx at the time of renal puncture, and subjectivity may cause uncertain or inaccurate access, which results in the development of complications and residual fragments.
In our previous research, we introduced an artificial intelligence (AI) puncture trajectory finder that when integrated with robotic arm movement, can provide robotic-assisted fluoroscopic (RAF)-guided renal puncture in a phantom model, a porcine model and in humans.11,12 This AI platform, automated needle targeting with x-ray (ANT-X: NDR Medical Technology, Singapore), is thought to be beneficial for residents or novices for achieving acceptable surgical outcomes. To examine the efficacy and safety of RAF-guided renal access during PCNL, we conducted a randomized controlled trial comparing RAF- with US-guided PCNL at an academic center.
Materials and Methods
Study Design and Participants
This single-center, open-label, randomized clinical trial compared RAF-guided and US-guided mini-PCNL with ureteroscopy (URS) assistance. The study was approved by the Nagoya City University Certified Research Review Board (No. 2019A002) and conducted at Nagoya City University Hospital between October 2019 and September 2021. It was conducted in accordance with the Declaration of Helsinki and the Clinical Trials Act. This study was also registered in the Japan Registry of Clinical Trials (No. jRCTs042190074). All patients provided written informed consent for participation.
The eligibility criteria were as follows: age 16–80 years, and detection of >15 mm renal and/or ureteral stones requiring PCNL. Pregnant or possibly pregnant women, patients with active pyelonephritis or those undergoing anti-platelet/coagulation therapy within 1 week before surgery, those for whom general/lumbar anesthesia was considered difficult by an anesthesiologist and those receiving best supportive care due to terminal carcinoma were excluded.
Randomization and Outcomes
Patients who visited or were referred to our outpatient clinic were evaluated for study eligibility and recruited by individual physicians. After consent procedures, block randomization was performed using online software stratifying for age (≤30, 31–40, 41–50, 51–60, 61–70, ≥71 years), sex, laterality, total stone burden (≤20, 21–30, 31–40, 41–50, ≥50), hydronephrosis grade (none, mild, moderate, severe, atrophic) and presence of staghorn stones. Each participant was assigned to the US-guided or RAF-guided group by an independent investigator. The surgical team was informed of the allocation a few weeks before the surgery; however, the patients were blinded.
The primary outcome was single puncture success, and the secondary outcomes were SFR, overall complication rate, complication (Clavien-Dindo classification) rates, parameters during renal access and fluoroscopy duration. The successful renal access was defined as the renal access in which a guidewire was placed into the renal collecting system with urine backflow. Stone-free status was defined as follows: no residual fragments larger than 4 mm detected in kidney ureter bladder (KUB) radiography at 1 month after surgery, and no residual fragments larger than 2 mm detected by CT 3 months after surgery.
The study flowchart is summarized in Figure 1.
Figure 1. Study flowchart. Asterisk indicates block randomization, which was performed with adjustment for age, sex, laterality, total stone burden, hydronephrosis grade and presence of staghorn stones.
Intervention and Surgical Technique
All PCNLs were performed with URS assistance with or without laser lithotripsy under general anesthesia. We utilized a pneumatic lithotripsy system (Swiss LithoClast® Master J, Electro Medical Systems, Nyon, Switzerland) through a mini-percutaneous nephrolithotomy (MIP, Karl Storz, Tuttlingen, Germany) with a 16.5Fr/17.5Fr, 21Fr/22Fr or 24Fr operating sheath depending on stone volume and complexity. For percutaneous renal puncture, a 24-gauge trocar needle (Hanaco Medical Co., Ltd., Saitama, Japan) was used; thereafter, the percutaneous tract into the renal collecting system was made by a 1-step dilation tool.
For RAF-guidance, ANT-X (see supplementary Figure, https://www.jurology.com) was utilized to gain renal access. The procedure was performed with the patient in the prone position. After fluoroscopic visualization of the renal collecting system using contrast medium, the robot arm was mounted on the patient’s back, and the needle tip was marked on the skin above the desired calyx. The puncture trajectory alignment was implemented by applying the bull’s-eye technique by AI calculation based on the location between markers and a needle tip point. The surgeon inserted a needle into the target calyx by confirming the depth through a tilted fluoroscopic view (Fig. 2). In all cases, the procedure from the ANT-X setup through the needle insertion was performed by a single surgeon, who was a novice in fluoroscopic-guided percutaneous access. The remaining procedures, including tract creation and fragmentation, were performed by a main PCNL surgeon in each case.
Figure 2. RAF percutaneous renal access. A, fluoroscopic image of an upper calyceal stone. B, a needle insertion point on the patient’s skin is marked under the fluoroscopic confirmation of renal collecting system. C and D, the software monitor view connected to the device and C-arm. The robotic arm is mounted on the patient’s back and keeps the target calyx in the inner yellow circle. Three marker balls are detected for calibration, then a needle is set in the holder before alignment (C). After the alignment, the needle trajectory is determined as a bull’s-eye position and automatically positioned by the robotic arm (D). E, the external view of the device with needle after the alignment. F, the needle is inserted from the targeted upper calyx. The C-arm is tilted to visualize the depth. G, the wire placement into the collecting system is confirmed by both fluoroscopy and the ureteroscopic view (inset).
For US guidance, ARIETTA 65® (Hitachi, Tokyo, Japan) with 6-1 or 5-1 MHz biopsy transducers (FUJIFILM Healthcare Co., Tokyo, Japan) was used for renal access. The procedure was performed with the patient in either the prone or Galdakao modified supine Valdivia position, based on surgeon’s decision considering stone and anatomical features. After needle insertion into the renal collecting system, the percutaneous tract was created mainly by fluoroscopic guidance with or without URS assistance. The primary surgeon who performed the first attempt at renal puncture in each case was a resident.
We fragmented the stones using a pneumatic lithotripsy probe (power 100%, frequency 12 Hz), and fragments were removed from the percutaneous tract by continuous backflow of percutaneous and ureteroscopic irrigation. We placed a 4.8Fr ureteral stent in all cases and 12Fr nephrostomy tubes in case of residual fragments or risk of postoperative bleeding/infection.
Statistical Analysis
The null hypothesis was that the RAF group had a superior single needle puncture over the US group. This was based on the results of the prior phantom study, showing 95% and 70% single needle puncture rate in the RAF and US groups, respectively.13 The sample size was calculated with a 0.05 type 1 error, 0.8 power, and a 10% superior margin considering differences between the phantom study and the real clinical setting. Therefore, a minimum sample size was set for 35 per group.
All statistical analyses were performed using EZR software (R Foundation for Statistical Computing, Vienna, Austria),14 and significance was set at p <0.05. Categorical variables, such as single puncture success as a primary outcome, and SFR, complications and other perioperative parameters as secondary outcomes, were compared using chi-square and Fisher’s exact tests. Continuous variables, such as duration of each surgical step, including fluoroscopy and laboratory data changes, were compared using 2-sample t-tests and the Mann-Whitney U test, depending on the distribution pattern for secondary and other outcome evaluation. Since the sample size of this randomized controlled trial was small, there were potential confounding variables unequally distributed between the 2 arms. Therefore, as the subgroup analyses, logistic and linear regression was performed to mitigate potential inequality of covariates between the groups. In particular, the SFR and overall complication were assessed with adjusting covariates assumed to impact these outcomes during the surgical planning, using a logistic regression model at the study beginning. The needle puncture and percutaneous access duration, as well as the number of punctures, were assessed with covariates that estimated potential unequal distribution, using a linear regression model.
Results
In total, 71 patients (35 in US group, 36 in RAF group) were enrolled. The mean age at surgery was 59.7 and 55.9 years in the US and RAF groups, respectively. Approximately 40% patients were female, and approximately one-third of stones were located in the ureteropelvic junction. Twenty-six percent and 36% of patients had staghorn stones in the US and RAF groups, respectively. Other preoperative factors in the 2 groups were comparable (Table 1).
| US-Guided | RAF-Guided | |
| No. pts | 35 | 36 |
| Mean yrs age (SD) | 59.7 (11.5) | 55.9 (13.9) |
| No. female sex (%) | 14 (40.0) | 14 (38.9) |
| No. lt side (%) | 17 (48.6) | 17 (47.2) |
| Mean kg/m2 BMI (SD) | 24.12 (4.39) | 25.02 (4.93) |
| No. main stone location (%): | ||
| Upper calyx | 8 (22.9) | 2 (5.6) |
| Middle calyx | 7 (20.0) | 12 (33.3) |
| Lower calyx | 6 (17.1) | 10 (27.8) |
| UPJ | 12 (34.3) | 12 (33.3) |
| Proximal | 1 (2.9) | 0 (0.0) |
| Distal | 1 (2.9) | 0 (0.0) |
| No. staghorn (%): | ||
| Partial | 6 (17.1) | 11 (30.6) |
| Complete | 3 (8.6) | 2 (5.6) |
| No. bacteriuria (%) | 14 (40.0) | 12 (33.3) |
| Median cm3 stone vol (IQR) | 6.58 (3.47, 14.48) | 6.58 (4.01, 15.67) |
| Median HU stone density (IQR) | 1,283 (1,081, 1,481) | 1,367 (1,181, 1,500) |
| No. hydronephrosis (%): | 11 (31.4) | 11 (30.5) |
| Mild | 1 (2.9) | 2 (5.6) |
| Moderate | ||
| Severe | 3 (8.6) | 1 (2.8) |
| No. pre-stenting (%) | 6 (17.1) | 4 (11.1) |
Primary Outcome and Puncture Details
The single puncture success rate was 34.3% and 50.0% in the US and RAF groups, respectively (p=0.2). However, the mean number of needle punctures was significantly smaller in the RAF group (1.82 times vs 2.51 times, p=0.025). In the US group, the resident was unable to obtain access for 14.3% cases; this was not seen in the RAF group (p=0.025, Table 2).
| US-Guided | RAF-Guided | p Value | |
| No. pts | 35 | 36 | |
| No. single puncture success (%) | 12 (34.3) | 18 (50.0) | 0.2 |
| No. needle punctures (%) | 2.51 (1.50) | 1.83 (0.94) | 0.025 |
| No. cases resident was unable to obtain access (%) | 5 (14.3) | 0 (0.0) | 0.025 |
| Median mins needle puncture duration (IQR)* | 8.00 (4.50, 14.50) | 5.50 (3.00, 10.00) | 0.049 |
| Median mins percutaneous access duration (IQR)† | 17.0 (12.0, 26.5) | 14.0 (8.8, 19.0) | 0.11 |
| Median seconds fluoroscopic duration (IQR) | 594 (447, 778) | 660 (534, 924) | 0.18 |
| No. stone-free status 1 mo after surgery (%)‡ | 31 (88.6) | 34 (97.1) | 0.4 |
| No. stone-free status 3 mos after surgery (%)§ | 24 (70.6) | 30 (83.3) | 0.3 |
| No. overall complications (%) | 6 (17.1) | 10 (27.8) | 0.4 |
| No. complication within 1 mo after surgery (%):ǁ | 0.9 | ||
| I | 2 (5.7) | 2 (5.6) | |
| II | 3 (8.6) | 5 (13.9) | |
| III | 1 (2.9) | 2 (5.6) | |
| No. complication between 1 and 3 mos after surgery (%):ǁ | 0.17 | ||
| I | 3 (8.8) | 0 (0.0) | |
| II | 1 (2.9) | 1 (2.8) |
Secondary Outcomes
The median needle puncture duration was also significantly shorter in the RAF group (5.5 minutes vs 8.0 minutes, p=0.049), whereas no differences were seen in percutaneous access and fluoroscopic duration between the 2 groups. SFR, confirmed by KUB radiographs, was 88.6% and 97.1% at 1 month after surgery, and 70.6% and 83.3% at 3 months after surgery in the US and RAF groups, respectively (p=0.356 and p=0.26, respectively). No between-group differences were observed in the overall complication rate, complications within 1 month after surgery, or complications between 1 and 3 months after surgery (Table 2).
Table 3 summarizes the intra- and postoperative parameters. PCNL was performed predominantly by residents in both groups (74.3% and 55.6% in the US and RAF groups, respectively; p=0.3). The median number of prior PCNL cases that lead surgeons had experienced was 14 and 12 cases in the US and RAF groups, respectively (p=0.2). The groups showed similar trends in the selection of renal calyces for percutaneous access. Patients’ positioning was different between the US and RAF groups (prone in 60.0% and 94.4%, respectively; p=0.001). There were no significant differences in other intraoperative parameters. There were no significant differences in specific complications, including serum leukocyte increase, decrease in hemoglobin, and estimated glomerular filtration rate at 1 day and at 3 months after surgery.
| US-Guided | RAF-Guided | p Value | |
| No. pts | 35 | 36 | |
| No. main PCNL surgeon (%): | 0.3 | ||
| Resident | 26 (74.3) | 20 (55.6) | |
| Fellow | 4 (11.4) | 7 (19.4) | |
| Attending | 5 (14.3) | 9 (25.0) | |
| Median prior PCNL cases lead surgeons had experienced (IQR) | 14 (10, 58) | 12 (10, 18) | 0.2 |
| No. puncture calyx (%): | 0.8 | ||
| Upper | 14 (40.0) | 11 (30.6) | |
| Middle | 10 (28.6) | 13 (36.1) | |
| Lower | 11 (31.4) | 12 (33.3) | |
| No. Fr tract size (%): | 0.4 | ||
| 17.5 | 32 (91.4) | 31 (88.7) | |
| 22 | 1 (2.9) | 5 (11.2) | |
| 24 | 2 (5.7) | 0 (0.0) | |
| No. positioning (%): | 0.001 | ||
| Prone | 21 (60.0) | 34 (94.4) | |
| Supine | 14 (40.0) | 2 (5.6) | |
| No. tubeless (%) | 29 (82.9) | 29 (80.6) | >0.9 |
| Median mins device setup duration (IQR) | 4.00 (2.00, 5.00) | 4.50 (3.00, 7.00) | 0.5 |
| Median mins fragmentation duration (IQR) | 52.0 (33.0, 84.0) | 61.0 (35.8, 94.0) | 0.8 |
| Median mins surgical duration (IQR) | 103 (75, 141) | 112 (82, 149) | 0.5 |
| No. fever (%)* | 3 (8.6) | 4 (11.1) | >0.9 |
| No. fragment obstruction (%) | 3 (8.6) | 0 (0.0) | 0.12 |
| No. renal hemorrhage (%) | 0 (0.0) | 3 (8.3) | 0.2 |
| No. urinary damage (%) | 7 (20.0) | 4 (11.1) | 0.3 |
| No. organ injury (%) | 1 (2.9) | 1 (2.8) | >0.9 |
| No. sepsis (%) | 1 (2.9) | 1 (2.8) | >0.9 |
| Median serum WBC×103/μl increase 1 day after surgery (IQR) | 3.30 (1.90, 5.50) | 3.30 (1.27, 4.72) | 0.4 |
| Median gm/l serum Hb decrease 1 day after surgery (IQR) | 0.60 (0.25, 1.50) | 0.90 (0.28, 1.60) | 0.6 |
| Median ml/min/1.73 m2 serum eGFR decrease 1 day after surgery (IQR) | −0.60 (−3.70, 4.85) | 2.15 (−3.73, 8.23) | 0.4 |
| Median serum WBC×103/μl increase 3 mos after surgery (IQR) | −0.40 (−1.00, 0.40) | −0.40 (−0.95, 0.55) | 0.8 |
| Median gm/l serum Hb decrease 3 mos after surgery (IQR) | 0.00 (−0.50, 0.60) | −0.05 (−0.70, 0.30) | 0.5 |
| Median ml/min/1.73 m2 serum eGFR decrease 3 mos after surgery (IQR) | 1.30 (−3.20, 7.10) | 1.15 (−0.77, 3.85) | >0.9 |
Further causal evaluation with logistic regression analysis demonstrated that larger percutaneous tract size, defined as either 22Fr or 24Fr, which was larger than the size used in the majority of cases (17.5Fr), was associated with an increased odds ratio for a higher overall complication rate (OR=13.7, p=0.032). No association was found between SFR 3 months after surgery and factors including RAF-guided renal access (Table 4). Linear regression analysis revealed that only RAF-guided renal access was associated with a decreased number of renal punctures (estimate −0.73, p=0.021). Preoperative and intraoperative factors, including RAF-guided renal access, were not associated with needle puncture and percutaneous access duration (Table 5).
| Stone-Free Status 3 Mos after Surgery | Overall Complications | |||
| OR (95% CI) | p Value | OR (95% CI) | p Value | |
| Intercept | 0.57 (0.02–17.2) | 0.8 | 0.93 (0.02–34.8) | >0.9 |
| RAF guidance | 2.06 (0.59–7.14) | 0.3 | 1.62 (0.45–5.76) | 0.5 |
| Attending surgeon performing PCNL | 0.77 (0.20–2.92) | 0.7 | 1.59 (0.43–5.90) | 0.5 |
| Larger percutaneous tract size* | 1.51 (0.16–14.8) | 0.7 | 13.7 (1.3–151) | 0.032 |
| BMI | 1.08 (0.94–1.24) | 0.3 | 0.94 (0.81–1.08) | 0.4 |
| Stone vol | 0.98 (0.95–1.01) | 0.3 | 0.98 (0.94–1.02) | 0.3 |
| Needle Puncture Duration | Percutaneous Access Duration | No. Punctures | ||||
| Estimate (95% CI) | p Value | Estimate (95% CI) | p Value | Estimate (95% CI) | p Value | |
| Intercept | 1.27 (−9.07–11.6) | 0.8 | 13.2 (−0.58–26.9) | 0.06 | 1.51 (−0.17–3.19) | 0.078 |
| RAF guidance | −3.76 (−7.59–0.07) | 0.054 | −3.70 (−8.78–1.39) | 0.15 | −0.73 (−1.36–−0.11) | 0.021 |
| Attending surgeon performing PCNL | 2.75 (−1.45–6.95) | 0.2 | 1.59 (−3.99–7.17) | 0.6 | 0.07 (−0.61–0.75) | 0.8 |
| BMI | 0.32 (−0.09–0.72) | 0.12 | 0.07 (−0.47–0.6) | 0.8 | 0.03 (−0.04–0.09) | 0.4 |
| Stone vol | −0.03 (−0.12–0.06) | 0.5 | 0.11 (0–0.23) | 0.059 | 0.01 (−0.01–0.02) | 0.5 |
| Neg preop hydronephrosis | 2.89 (−0.86–6.64) | 0.13 | 4.93 (−0.06–9.91) | 0.053 | 0.39 (−0.22–1) | 0.2 |
Discussion
In this study, the RAF group demonstrated a smaller number of needle punctures and a shorter needle puncture duration, which can positively impact the patient’s outcome. In fact, renal access during PCNL is directly related to surgical outcomes, including complications.15 While no differences were found in postoperative complications between the US and RAF groups in our trial, Gorbachinsky et al reported that multiple renal accesses were associated with postoperative decrease in renal function in PCNL.16 Therefore, fewer needle punctures might be beneficial for postoperative renal function preservation. In addition, consistent with other large volume evidence,17 the current study revealed that cases with larger percutaneous tract size (22Fr and 24Fr sheaths) were associated with a higher complication risk. This finding indicated that RAF-guided percutaneous access did not independently influence the occurrence of complications; however, tract size did. Moreover, in 14.3% of US-guided punctures, a change of surgeon was needed owing to the difficulties associated with gaining access, whereas no surgeon change was required during RAF-guided renal access. Percutaneous access was used by most residents in this study, indicating that US-guided access was challenging for them. RAF-guided access was achievable by a novice surgeon. This may fill the gap between experts and trainees in the real-world trend of teaching facilities—wherein not all procedures are performed by experts, and similar outcomes between US- and fluoroscopy-guided procedures are expected.4
Robotic-assisted surgeries have been widely used by urologists and have become the standard approach for several subspecialties, with a surge of research interest.18 As in other fields, kidney stone surgeons may have a greater appreciation of robotic technologies, such as the Avicenna Roboflex (ELMED Medical Systems, Ankara, Turkey) for URS, which controls scope maneuver and supports lithotripsy from a distant console.19 Robotic percutaneous access to the kidney, introduced by John Hopkins University, enables the control of percutaneous needle insertion with a robot arm while surgeons monitor live fluoroscopic images.20 There are alternative image navigation tools that use fluoroscopic guidance, such as StealthStation™ (Medtronic, Minneapolis, Minnesota), Micromate™ (iSYS, Kitzbühel, Austria) and AcuBot (Georgetown Medical Center, Washington, D.C.).21 These devices provide more precise percutaneous access by holding needles and fine-tuning needle movement; however, needle trajectory toward the target has to be navigated by the surgeons themselves. In contrast, the current robotic technique, ANT-X, used in this study has a unique feature: automation of the needle puncture trajectory provided by the AI platform. Our study showed that this technique allows the novice surgeon to perform renal access safely and effectively with less needle puncture frequency and duration compared with a conventional US-guided procedure. Certainly, these robotic-assisted surgeries have the risk of device malfunction; therefore, backups and training for alternative methods should always be performed as a precaution. Indeed, in our study, 1 case was converted to US-guided PCNL due to a machine preparation error. In addition to this safety evaluation, the cost-effectiveness of robotic-assisted surgeries also needs to be evaluated.
Several other modalities support renal access during PCNL. Under US guidance, real-time virtual sonography provides the fusion of real-time US images and preoperative CT images that are input into the US machine software beforehand.9 The individual anatomical analysis is also achieved with 3D printing, which is mainly utilized for simulation. Ghazi et al reported the usefulness of patient-specific simulation using a 3D-printed hydrogel model for PCNL training.22 Moreover, technological innovation with 3D models has provided us with a cross-reality tool for the simulation and navigation of renal access during PCNL.23 Virtual reality,24 augmented reality using an iPad® (Apple Inc., Cupertino, California)25 and mixed-reality hologram,26 can reduce radiation exposure as well as renal puncture time or number of attempts. Electromagnetic-guided percutaneous access has promising features with direct navigation from the ureteroscope tip, which provides semiautomatic targeting with minimal radiation and a short learning curve.10,27 These modalities are examples that improve anatomical understanding and thus aid in quicker preoperative planning and navigation of the surgeon’s self-maneuver of the needle puncture. However, the technique for adjusting the angle toward the target calyx with 2D intraoperative images is still challenging for young or inexperienced urologists, and the image navigation tool requires extra time to prepare. Although our previous benchtop study showed that the device setup time was longer in RAF-guided puncture than that in US-guided puncture, the current clinical trial demonstrated no between-group difference in device setup time. We assumed that this comparable result was due to the support of our medical engineering staff, who prepared the device simultaneously.
One could overcome this technical difficulty in percutaneous needle targeting by the application of AI. Checcucci et al updated promising research utilizing AI for percutaneous needle puncture.23 Needle tracking with precise segmentation of US images could be achieved by deep learning architecture. It could also provide needle trajectory prediction during percutaneous puncture. Moreover, the robotic system to support US-guided percutaneous punctures by fine-tuning the needle position is available in proof of concept status.28 Combining these needle detection and trajectory calculations may be an ideal technology integration to achieve precise percutaneous access. RAF-guided renal access consists of real-time image and trajectory navigation. Since no cases lost the needle targeting into the targeted calyx by a single surgeon, this platform may be the most promising solution for future automatic percutaneous puncture.
This study had several limitations. Although we calculated the sample size based on the results of a previous phantom study, a multicenter study with a larger volume may be statistically more effective to obtain more sophisticated evidence. The comparison between US- and fluoroscopic-guided robotic techniques might reflect more than just the difference between freehand and RAF-guided percutaneous access. In addition, the cost estimation should be provided for comparing these 2 methods in order for RAF to be applied to our daily practice. Finally, the heterogeneity of surgical methods, such as percutaneous tract size, types of lead surgeon at the PCNL side and patients’ positioning, were potential confounding factors that masked the influence of robotic assistance on the percutaneous access outcome. However, we believe that these results reflect real-world practice in terms of the recent global trend whereby the PCNL technique has been shifting to US guidance or utilizing new navigation systems.
Conclusions
RAF-guided PCNL can be performed as safely and effectively as US-guided PCNL, even by novice surgeons, using fluoroscopic guidance. It may provide more accurate and faster percutaneous access, which could potentially reduce the perioperative complication rate. With their promising features, robotic-assisted platforms will be a beneficial innovation in percutaneous procedures.
Acknowledgments
We thank the operating room, radiology and medical engineering staff at Nagoya City University Hospital who supported the installation of the RAF-guided system as well as NDR Medical Technology for their help in ANT-X preparation. We would also like to thank Editage for English proofreading.
References
- 1. : New technology in ureteroscopy and percutaneous nephrolithotomy. Curr Opin Urol 2016; 26: 95. Google Scholar
- 2. : The clinical research office of the Endourological Society Percutaneous Nephrolithotomy Global Study: indications, complications, and outcomes in 5803 patients on behalf of the CROES PCNL study group. J Endourol 2011; 25: 11. Google Scholar
- 3. : Incidence, prevention, and management of complications following percutaneous nephrolitholapaxy. Eur Urol 2012; 61: 146. Google Scholar
- 4. : Ultrasound or fluoroscopy for percutaneous nephrolithotomy access, is there really a difference? A review of literature. J Endourol 2021; 35: 241. Google Scholar
- 5. : Learning curves in urolithiasis surgery: a systematic review. J Endourol 2018; 32: 1008. Google Scholar
- 6. : Endoscopically guided percutaneous renal access: “seeing is believing.” J Endourol 2006; 20: 451. Google Scholar
- 7. : Retrograde endoscopic-assisted percutaneous renal access: a novel “lasso” technique to achieve rapid secure access to the collecting system. J Endourol 2008; 22: 591. Google Scholar
- 8. : Computed tomography for percutaneous renal access. J Endourol 2009; 23: 1633. Google Scholar
- 9. : A new navigation system of renal puncture for endoscopic combined intrarenal surgery: real-time virtual sonography-guided renal access. Urology 2017; 109: 44. Google Scholar
- 10. : Ureteroscopy-assisted percutaneous kidney access made easy: first clinical experience with a novel navigation system using electromagnetic guidance (IDEAL stage 1). Eur Urol 2017; 72: 610. Google Scholar
- 11. : Robot-assisted fluoroscopy-guided renal puncture for endoscopic combined intrarenal surgery: a pilot single-centre clinical trial. BJU Int 2021; 127: 307. Google Scholar
- 12. : Automated needle targeting with x-ray (ANT-X)–robot-assisted device for percutaneous nephrolithotomy (PCNL) with its first successful use in human. J Endourol 2021; 35: e919. Google Scholar
- 13. : Robot-assisted fluoroscopy versus ultrasound-guided renal access for nephrolithotomy: a phantom model benchtop study. J Endourol 2019; 33: 987. Google Scholar
- 14. : Investigation of the freely available easy-to-use software ‘EZR’ for medical statistics. Bone Marrow Transpl 2013; 48: 452. Google Scholar
- 15. : Percutaneous nephrolithotomy: technique. World J Urol 2017; 35: 1361. Google Scholar
- 16. : Evaluation of renal function after percutaneous nephrolithotomy—does the number of percutaneous access tracts matter?J Urol 2016; 196: 131. Link, Google Scholar
- 17. : Mini percutaneous nephrolithotomy is a noninferior modality to standard percutaneous nephrolithotomy for the management of 20–40 mm renal calculi: a multicenter randomized controlled trial. Eur Urol 2021; 79: 114. Google Scholar
- 18. : Robotic surgery research in urology: a bibliometric analysis of field and top 100 articles. J Endourol 2019; 33: 389. Google Scholar
- 19. : A new robot for flexible ureteroscopy: development and early clinical results (IDEAL stage 1–2b). Eur Urol 2014; 66: 1092. Google Scholar
- 20. : Robotic percutaneous access to the kidney: comparison with standard manual access. J Endourol 2002; 16: 471. Google Scholar
- 21. : AcuBot: a robot for radiological interventions. IEEE Trans Robot Autom 2003; 19: 927. Google Scholar
- 22. : Validity of a patient-specific percutaneous nephrolithotomy (PCNL) simulated surgical rehearsal platform: impact on patient and surgical outcomes. World J Urol 2022; 40: 627. Google Scholar
- 23. : Percutaneous puncture during PCNL: new perspective for the future with virtual imaging guidance. World J Urol 2022; 40: 639. Google Scholar
- 24. : Pilot assessment of immersive virtual reality renal models as an educational and preoperative planning tool for percutaneous nephrolithotomy. J Endourol 2019; 33: 283. Google Scholar
- 25. : iPad-assisted percutaneous nephrolithotomy (PCNL): a matched pair analysis compared to standard PCNL. World J Urol 2020; 38: 447. Google Scholar
- 26. : Percutaneous kidney puncture with three-dimensional mixed-reality hologram guidance: from preoperative planning to intraoperative navigation. Eur Urol 2022; 81: 588. Google Scholar
- 27. : Electromagnetic guided percutaneous renal access outcomes among surgeons and trainees of different experience levels: a pilot study. Urology 2020; 136: 266. Google Scholar
- 28. : Ultrasound-guided needle insertion robotic system for percutaneous puncture. Int J Comput Assist Radiol Surg 2021; 16: 475. Google Scholar
Submitted January 21, 2022; accepted April 28, 2022; published May 13, 2022.
Support: This work was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology, Japan (Nos. 19H03791, 20K09544 and 20K21658), and grants from the Hori Science and Arts Foundation, Nitto Foundation and CASIO Science Promotion Foundation.
Conflict of Interest: Kazumi Taguchi and Yung Khan Tan report advisory roles for NDR Medical Technology. The remaining Authors have no conflicts of interest to declare.
Ethics Statement: This study was approved by the Nagoya City University Certified Research Review Board (No. 2019A002) and was conducted in accordance with the Declaration of Helsinki and the Clinical Trials Act. All patients provided written informed consent for participation.
Authors’ Contributions: Study concept and design: K. Taguchi, S. Hamamoto, H. Fukuta, Y. K. Tan, T. Yasui. Acquisition of data: K. Taguchi, T. Sugino, R. Unno, T. Kato, N. Kawai. Analysis and interpretation of data: A. Okada, T. Sugino, R. Unno, T. Kato, R. Ando. Drafting of the manuscript: K. Taguchi, S. Hamamoto, T. Sugino, R. Unno. Critical revision of the manuscript for important intellectual content: All authors. Statistical analysis: K. Taguchi, S. Hamamoto, T. Sugino, R. Unno, R. Ando. Obtaining funding: K. Taguchi, S. Hamamoto, T. Yasui. Administrative, technical, or material support: Y. K. Tan, N. Kawai. Supervision: T. Kato, H. Fukuta, Y. K. Tan, A. Okada, N. Kawai, T. Yasui. Other: None.
Japan Registry of Clinical Trials No.: jRCTs042190074.
Data Sharing: Data are available for bona fide researchers who request it from the authors.
