University of Groningen
CT-guided percutaneous interventions
Heerink, Wouter
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Heerink, W. (2019). CT-guided percutaneous interventions: Improving needle placement accuracy for lung and liver procedures. Rijksuniversiteit Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
O
ver the past 20 years, thermal ablation has emerged as a successful treatment method for hepatic malig-nancies (1–3). Radiofrequency ablation and microwave ablation (MWA) are currently recommended for the treat-ment of hepatocellular carcinoma (HCC) and colorectal liver metastases in patients who are unfit for surgery or in combination with surgery (4–6). The main drawback of percutaneous liver tumor ablation is local recurrence of disease, with reported ablation site recurrence rates rang-ing from 5.0% to 32.1% (2,7–9). Risk factors for recur-rences include larger tumor size, peritumoral vascularity, and insufficient ablation margin surrounding the tumor (10–12). The latter can be caused by inaccurate ment of the ablation antenna. Currently, antenna place-ment is mostly performed manually (freehand), with CT as a frequently used imaging modality for guidance.To improve antenna placement accuracy, various ro-botic needle guiding systems have been developed. They
generally offer higher positioning accuracy and/or improved procedure times compared with freehand needle placement (13–16). However, to our knowl-edge, none of these devices have been compared with freehand CT-guided needle placement in randomized controlled trials. We developed a robotic needle posi-tioning system that proved highly accurate in a phan-tom study (17).
The purpose of this study was to compare the ac-curacy of robotic-guided antenna placement versus free-hand antenna placement in study participants with liver tumors treated with percutaneous CT-guided MWA in a randomized controlled trial. We hypothesized that robotic MWA antenna placement increases accuracy. Results were stratified according to difficulty (in-plane vs out-of-plane antenna placement). The primary end point was the number of repositionings required to ob-tain an adequate antenna position.
Robotic versus Freehand Needle Positioning in
CT-guided Ablation of Liver Tumors:
A Randomized
Controlled Trial
Wouter J. Heerink, MSc • Simeon J. S. Ruiter, MSc • Jan Pieter Pennings, MD • Benno Lansdorp, PhD • Rozemarijn Vliegenthart, MD, PhD • Matthijs Oudkerk, MD, PhD • Koert P. de Jong, MD, PhD
From the Center for Medical Imaging–North East Netherlands (W.J.H., R.V., M.O., K.P.d.J.), Department of Radiology (W.J.H., J.P.P., R.V.), and Department of Hepato-Pancreato-biliary Surgery (S.J.S.R., K.P.d.J.), University of Groningen, University Medical Center Groningen, Hanzeplein 1, 9713 GZ Groningen, the Netherlands; and DEMCON Advanced Mechatronics, Enschede, the Netherlands (B.L.). Received July 20, 2018; revision requested September 4; final revision received November 5; accepted December 3. Address correspondence to W.J.H. (e-mail: w.j.heerink@umcg.nl ).
Supported by Samenwerkingsverband Noord-Nederland (grant T2017). Conflicts of interest are listed at the end of this article.
Radiology 2019; 00:1–7 • https://doi.org/10.1148/radiol.2018181698 • Content codes:
Purpose: To compare the accuracy of freehand versus robotic antenna placement in CT-guided microwave ablation (MWA) of liver
tumors.
Materials and Methods: This study was conducted as a prospective single-center nonblinded randomized controlled trial (Netherlands
Trial Registry, NTR6023). Eligible study participants had undergone clinically indicated CT-guided MWA of liver tumors and were able to receive a CT contrast agent. Randomization was performed per tumor after identification on contrast material–enhanced CT images. The primary outcome was the number of antenna repositionings, which was compared by using the Mann-Whitney U test. Secondary outcomes were lateral targeting error stratified by in-plane and out-of-plane targets and targeting time.
Results: Between February 14 and November 12, 2017, 31 participants with a mean age of 63 years (range, 25–88 years) were
in-cluded: 17 women (mean age, 57 years; range, 25–77 years) and 14 men (mean age, 70 years; range, 52–88 years). The freehand study arm consisted of 19 participants, while the robotic study arm consisted of 18 participants; six participants with multiple tu-mors were included in both arms. Forty-seven tutu-mors were assessed; five tutu-mors were excluded from the analysis because of techni-cal limitations. In the robotic arm, no antenna repositioning was required. In the freehand arm, a median of one repositioning was required (range, zero to seven repositionings; P , .001). For out-of-plane targets, lateral targeting error was 10.1 mm 6 4.0 and 5.9 mm 6 2.9 (P = .007) for freehand and robotic procedures, respectively, and for in-plane targets, lateral targeting error was 6.2 mm 6 2.7 and 7.7 mm 6 5.9, respectively (P = .51). Mean targeting time was 19 minutes (range, 8–55 minutes) and 36 minutes (range, 3–70 minutes; P = .001) for freehand and robotic procedures, respectively.
Conclusion: Robotic antenna guidance reduces the need for antenna repositioning in microwave ablation to accurately target liver
tu-mors and increases accuracy for out-of-plane targets. However, targeting time was greater with robotic guidance than with freehand targeting.
Robotic versus Freehand Needle Positioning in CT-guided Liver Tumor Ablation
2 radiology.rsna.org n Radiology: Volume 00: Number 0— 2019
Materials and Methods
Study DesignThis prospective single-center nonblinded randomized study was performed in the Netherlands and was approved by the institutional review board of the University Medical Center Groningen. The study protocol can be found in the Nether-lands Trial Registry (NTR6023). The study was financially aided by Samenwerkingsverband Noord-Nederland (grant T2017). DEMCON Advanced Mechatronics (Enschede, the Netherlands) provided equipment (the robotic device) for the study. Neither agency had a role in study design or data analysis. Only the authors who are not employed by DEMCON had control of inclusion of any data and in-formation that might present a conflict of interest. Written informed consent was obtained from all participants prior to the procedure.
Study Participants
Participants who were candidates for CT-guided MWA of the liver were eligible to participate if (a) their tumor size was less than 50 mm, (b) the number of tumors was three or fewer, and (c) they were 18 years of age or older. Participants were excluded if (a) their tumor was adjacent to biliary structures,
(b) they were unable to undergo general anesthesia, or (c) they
were unable to tolerate the CT contrast agent. All participants were discussed in a multidisciplinary tumor board meeting for CT-guided percutaneous liver ablation.
Randomization and Masking
Procedures were randomized per tumor by using block ran-domization, as described previously by Arifin (18). Results were revealed from sealed opaque envelopes after tumors were given unique identifying number by a researcher (W.J.H., with 4 years of experience) who was not involved with participant selec-tion or clinical decision making.
Procedures
All procedures were performed with the participant in gen-eral anesthesia and in a stable position on a vacuum mattress to eliminate patient movement. Procedures were performed with a 64–multidetector row CT system (Somatom Sensation 64; Siemens Medical Systems, Erlangen, Germany). Image acquisition and antenna manipulation were performed after controlled apnea. This guarantees maximum relaxation of the elastic recoil in the thorax and thus a reproducible position of the liver. A contrast material–enhanced CT study was per-formed for planning. Navigational CT images were acquired after each antenna manipulation. Tube voltage was 100 kVp, and quality reference tube current was 110–179 mAs. The planning scan was acquired prior to the admission of intra-venous contrast agent (90–110 mL of Iomeron 300; Bracco Imaging, Milan, Italy). Section thickness and section incre-ment were 2 mm. After identification, tumors were given an identifying number, and a target was selected inside the tumor on the contrast-enhanced CT image by using the in-tervention suite software on the scanner. These targets were stored in key images from which the exact target coordinates were extracted.
Robotic Approach
Robotic-guided procedures were performed with the Nee-dle Positioning System (NPS; DEMCON Advanced Me-chatronics), which consists of a robotic arm that slides on a rail parallel to the CT table (Fig 1). Bolting the rail to the CT table takes less than 5 minutes and is done at the start of the day. Patients can get on and off the table eas-ily with the system in place. After patient positioning, the planning CT scan is acquired and the entry point is marked on the skin. The skin is disinfected, and surgical drapes are placed. Subsequently, a disposable sterile cover is put over the robotic arm before it is positioned above the sterile skin. The needle guide itself attaches to the arm through a hole in the sterile cover. The needle guides are reusable and cleaned with autoclave sterilization. The robotic arm is positioned manu-ally on top of the entry point on the skin, using a disposable pointer on the needle guide. After the device is locked into place, a registration CT scan is acquired, on which the four CT fiducials situated inside the robotic arm are visualized. This is a non–contrast-enhanced scan with similar settings as the plan-ning CT scan. The registration scan, together with the selected target, are sent to the system, after which the device automat-ically orients its needle guide toward the target. Subsequently, the MWA antenna can be clipped into the needle guide and the insertion is performed manually to the depth specified by the system. After a control CT scan, the MWA antenna is unclipped from the needle guide and mechanical ventilation is restarted. A more detailed description of the system and the procedure can be found in Arnolli et al (17).
For the freehand procedures, use of the vacuum mat-tress and controlled apnea was similar. The microwave an-tenna was positioned manually, and CT images were used to verify the position and, if necessary, the antenna was reposi-tioned and checked again by using CT until the position was
Abbreviations
DLP = dose-length product, HCC = hepatocellular carcinoma, MWA = microwave ablation
Summary
For CT-guided microwave ablation of liver tumors, robotic antenna guidance offers significant advantages over freehand antenna ment in terms of the number of antenna repositionings and place-ment accuracy.
Implications for Patient Care
n For CT-guided microwave ablation of the liver, out-of-plane
targeting errors were 40% lower with robotic guidance than with freehand localization.
n For CT-guided microwave ablation, robotic antenna guidance
re-duced the need for antenna repositioning, which has the potential to improve ablation efficacy.
n Longer targeting times for robotic guidance than for freehand
placement suggests the need for further optimization of robotic guidance.
Outcomes
Primary outcome was the number of antenna repositioning attempts required to reach a position inside the tumor. This was verified during the procedure by a radiologist who did not perform the antenna placement (J.P.P.). Advancement of the antenna along the same path to reach the correct depth was not considered repositioning. The secondary outcomes were antenna placement accuracy, targeting time, number of CT scans acquired during the MWA procedure, dose-length product (DLP), number of incomplete ablations at the 1-week control contrast-enhanced CT study, and complications.
The accuracy of antenna positioning was defined in multiple ways. The Euclidian error was determined as the distance be-tween the antenna tip and the target, and the lateral error was determined as the shortest distance between the needle path and the target. The angles between target, skin entry point, and nee-dle tip were also determined (Fig 3). These measurements were determined automatically after selection of the skin entry point and the tip of the MWA antenna by radiologist J.P.P. both after initial antenna insertion as well as when the antenna’s position was deemed adequate.
The targeting time was defined as time from target selection until adequate antenna position. For the robotic approach, this included application of the sterile cover, the placement of the needle guide, acquisition of the fiducial CT scan, and data trans-fer. The robot installation at the start of the day, which took approximately 5 minutes, was not included because it was not repeated for each procedure. For solitary tumor ablations, total procedure time was determined as the time from the patient’s arrival in the CT room to the time of the patient’s departure from the CT room. The number of CT scans acquired to reach this position and the corresponding DLP were determined as well, excluding the contrast-enhanced CT scans. For the robotic procedures, this included the fiducial scan.
Complications were monitored and classified as minor or major according to the International Working Group on Image-Guided Tumor Ablation standards of terminology and reporting criteria (19). Minimal perihepatic fluid or blood collection was considered an expected side effect and was therefore not classi-fied as a complication.
To investigate the differences between in-plane and out-of-plane approaches, the outcomes of the freehand and robotic adequate for ablation. “Adequate” was defined as a position
within the tumor, with the potential exception of a position with close proximity to the liver capsule. If ablation cycles in multiple areas in the tumor were necessary, the reposition-ing of the ablation antenna was performed by hand. Hence, only the placement of the antenna toward the initial position differed between the two study groups. Subsequently, MWA was performed according to the protocol provided by the manufacturer.
For the first 29 procedures (in 26 participants), the Em-print MWA system (Medtronic, Minneapolis, Minn) was used. Because of a global recall during the study by Medtronic, a different ablation system (Amica, HS Medical, Rome, Italy) was used for the remainder of the study (13 procedures in eight participants). A contrast-enhanced CT examination was performed directly after the procedure. If the ablation zone was determined to be inadequate, additional ablation was performed. A follow-up contrast-enhanced CT examina-tion was performed 1 week after the ablaexamina-tion procedure to evaluate if the ablation zones completely covered the tumor and included a safety margin.
Data Analysis
The freehand and robotic groups were compared with regard to participant age, tumor type, tumor diameter, tumor depth, applied ablation energy, angulation (in-plane and out-of-plane procedures), and treatment type (first ablation and ablation of recurrence). The tumor diameter was measured as the long-axis diameter in the transverse plane. Tumor depth was determined as the Euclidian distance between the entry point on the skin and the selected target (Fig 2). Measurements were performed by a radiologist with 10 years of experience (J.P.P.). The proce-dures were categorized as in plane or out of plane on the basis of the angle of approach. If the microwave antenna was angled 5 or more degrees from the axial plane, the approach was con-sidered to be out of plane.
Figure 1: The robotic system moves on a rail parallel to the CT table
(1). The system is manually maneuvered into place and the device is
locked with a pushbutton (2). After automatic orientation, the micro-wave antenna is clipped in the needle guide (3) and can be inserted to the specified depth. Note that for clarity, this image is taken without the use of a sterile cover.
Figure 2: Schematic of accuracy measures. a = Angle error, Elateral
is the lateral distance between antenna tip and target, EEucl = Euclidian
Robotic versus Freehand Needle Positioning in CT-guided Liver Tumor Ablation
4 radiology.rsna.org n Radiology: Volume 00: Number 0— 2019
shows the inclusion flowchart. Five of the robotic approaches (in five participants, two of whom had more than one tumor) were excluded from analysis. In one instance, only three of the device fiducials were inside the scanner’s field of view. An-other approach proved to be technically impossible because of steep cranial angulation. A third robotic approach suffered from technical malfunctioning of the system, which was re-solved before the operators proceeded to the next target. The other two procedures were excluded because of antenna de-flection that was so substantial that it prevented liver penetra-tion. These procedures were finished by using the freehand approach and were replaced in our analysis by including five additional tumors in three additional participants and in one participant who had previously been included.
For the final analysis, 21 robotic approaches in 16 par-ticipants and 21 freehand approaches in 18 parpar-ticipants were included. Table 1 shows the participant characteris-tics. Twenty-one participants had a single tumor, nine had two tumors, and one had three tumors; six participants were included in both arms. Twenty-six colorectal liver metasta-ses were treated (in 17 participants), 12 HCCs were treated (in seven participants), and four tumors were benign liver nodules. Overall, mean participant age was 61 years (range, 25–88 years), mean tumor diameter was 23.1 mm 6 11.1, and mean depth from skin was 88.4 mm 6 29.1. For the ro-botic and freehand groups, respectively, tumor diameter was 21.2 mm 6 10.2 and 24.9 mm 6 11.9 (P = .31) and tumor depth was 92.2 mm 6 31.8 and 84.0 mm 6 26.1 (P = .33). For men, the mean age was 70 years (range, 52–88 years); for women, mean age was 57 years (range, 25–77 years). For the robotic and freehand arms, respectively, the mean par-ticipant ages were 60.1 years 6 15.2 and 62.8 years 6 15.2 (P = .52). Tumor characteristics for freehand and robotic procedures are presented in Table 2.
group were additionally compared, stratified by antenna angula-tion category.
Statistical Analysis
Sample size calculation.—An analysis of 50 consecutive
CT-guided freehand liver ablations performed at our depart-ment showed that the mean number of antenna manipula-tions required to reach an adequate ablation position was 2.1 6 1.3 (standard deviation). To enable us to demonstrate that the robotic approach will improve to a mean of 1.2 6 0.2 needle manipulations with a power of 0.8 and an a of .5, the number in both arms was calculated to be 21. Power calcula-tions were performed by using G-power 3.1.9.2 (Dusseldorf, Germany) (20).
All variables were checked for normality by using the Shapiro-Wilk test. The means and standard deviations were determined for continuous parametric variables (age, tumor diameter, tumor depth, all accuracy measures) and were compared by using the in-dependent samples t test. For categoric parameters (tumor type, angulation, treatment type, liver side, incomplete ablations), the x2 test was used. For nonparametric variables (number of
repo-sitionings, number of CT scans), the median and range were determined and were compared by using the Mann-Whitney U test. Statistical significance was set at P , .05. Statistical analyses were performed in SPSS 23 (IBM, Armonk, NY).
Results
Between February 7, 2017, and November 13, 2017, 47 tu-mors in 31 participants were included in our study. Figure 3
Table 1: Participant Characteristics
Characteristic Datum
No. of participants 31
Mean age (y) 63
Age range (y) 25–88
No. of men 14
Mean age (y) 70
Age range (y) 52–88
No. of women 17
Mean age (y) 57
Age range (y) 25–77
No. of tumors per participant
1 21 2 9 3 1 Robotic procedure 12 Freehand procedure 13 Both* 6
* Participants with multiple tumors could be randomized in both groups.
Figure 3: Study participant flowchart. * = In six study participants with multiple tumors, randomization on a per-tumor basis resulted in randomization into both groups. CE = contrast enhanced, n = number of tumors, NPS = Needle Placement System.
shows that robotic guidance is of value for CT-guided liver abla-tion with complex angulaabla-tions.
The number of CT scans was equal in both groups, but the DLP for the robotic group was significantly higher because of a greater scan length required to visualize all the fiducials. The ro-botic targeting times were, on average, 20 minutes longer than the freehand targeting times (18 minutes, P = .001). This can be attributed to the application of the sterile cover and to some technical issues with data transfer. It did not significantly affect the total procedure times for solitary tumors, which were 2 hours 36 minutes and 2 hours 40 minutes (P = .76) for freehand and robotic procedures, respectively. For shorter procedures without general anesthesia, this may be more substantial.
In the robotic group, three ablations were incomplete even though the lateral error was measured as less than 5 mm. In one participant, no contrast-enhanced scan could be acquired at the end of the procedure because of a malfunctioning CT scanner. The other two ablations were in the same participant, where tu-mor borders were difficult to visualize at contrast-enhanced CT. Hence, in these instances, a lack of ablation zone feedback likely played a role. This feedback is important, because complete ab-lation is also dependent on the creation of predictable abab-lation zones. The latter is affected by heat-sink and tumor type (10,21). The incomplete ablation in the freehand group occurred in a dif-ficult out-of-plane procedure where the antenna was repositioned seven times, resulting in a final lateral error of 15 mm.
Devices comparable to the Needle Placement System include MAXIO and ROBIO (Perfint Healthcare, Chennai, India), two large floor-mounted robotic devices with a reported accuracy of 6.5 mm (13); the CAS-One (CAScination, Bern, Switzerland), a table-mounted navigation system that uses visual fiducials re-quiring a line of sight, with a reported accuracy of 4.9 mm (15); and the iSYS1 (iSYS Medizintechnik, Kitzbühel, Austria), which The number of
reposi-tionings was zero (range, zero to zero) and one (range, zero to seven) for robotic and freehand procedures, respectively. For initial placement, the Euclidian error was smaller for robotic procedures—namely, 10.2 versus 22.2 mm (P , .001). All other error measures were also smaller for the robotic approach, with val-ues of less than half of those for the freehand approach (P , .005). For adequate antenna position, the Eu-clidian error of the robotic procedures was still smaller than that of the freehand procedures—namely, 9.3 versus 13.8 mm (P = .02). All technical outcomes can be found in Table 3.
Mean targeting time for freehand procedures was 19 minutes, as compared with 36 minutes for robotic procedures (P = .001). Total procedure times for solitary tumor ablations were 2 hours 40 minutes and 2 hours 36 minutes for robotic (n = 10) and freehand (n = 11) procedures, respectively (P = .76). Total abla-tion time and amount of energy applied until the post-MWA contrast-enhanced CT scan were acquired were approximately similar between groups. The numbers of incomplete ablations in the robotic and freehand groups at the follow-up CT were three and one (P = .34), respectively. There were two complications in this study. The first, which occurred in the freehand group, consisted of an abscess that required laparotomy with resection of the left liver lobe. The second, which occurred in the robotic group, consisted of a self-resolving pneumothorax.
Table 4 shows the technical outcomes, stratified for in-plane and out-of-plane procedures. Eleven of the 21 freehand proce-dures were out of plane, compared with 13 of the robotic pro-cedures. In the in-plane procedures, targeting times were 18 and 32 minutes (P = .004) for freehand and robotic procedures, re-spectively. In the out-of-plane procedures, targeting times were 19 and 37 minutes, respectively. For in-plane procedures, only the depth error was significantly better (10.1 vs 3.2 mm, P = .03) for the robotic group, but other error measures were similar. For the out-of-plane procedures, the robotic system outperformed freehand positioning on all error measures (P , .05 for all).
Discussion
The goal of our study was to compare antenna placement accuracy in robotic and freehand CT-guided liver tumor ablation. With ro-botic guidance, antenna repositioning was not needed in any of the study participants. However, in the freehand group, between zero and seven repositionings were needed. For out-of-plane pro-cedures, the lateral accuracy improved from 16.1 to 5.6 mm. This
Table 2: Tumor Characteristics
Characteristic Robotic Group Freehand Group P Value
No. of tumors 21 21
Tumor type
HCC 4 8 .42
Colorectal liver metastasis 14 12
Adenoma 2 1
Regenerative liver nodule 1
Liver side Left 1 7 .045 Right 20 14 Tumor diameter (mm)* 21.2 6 10.2 24.9 6 11.9 .31 Tumor depth (mm)* 92.9 6 31.8 84.0 6 26.1 .33 Angulation
No. of in-plane procedures 8 10 .76
No. of out-of-plane procedures 13 11
Treatment type
First ablation of tumor 19 18 .63
Recurrent ablation of tumor 2 3
Note.—Unless otherwise specified, data are numbers of tumors or procedures. HCC = hepatocellular carcinoma.
Robotic versus Freehand Needle Positioning in CT-guided Liver Tumor Ablation
6 radiology.rsna.org n Radiology: Volume 00: Number 0— 2019
study, Engstrand et al (23) analyzed the accuracy and procedural safety of the CAS-One for CT-guided percutaneous MWA of liver tumors. In 28 tumors, they reported a lateral accuracy of 4.0 mm, although no comparison with freehand positioning was made. None of the aforemen-tioned studies differentiated between in-plane and out-of-plane procedures.
There were some limi-tations to our study. Ran-domization was performed per tumor and tumors were considered independent, even though correlation be-tween tumors in one patient can occur. For example, ac-curacy can be affected by patient size or level of cir-rhosis. Additionally, five procedures could not be finished with the robotic device because of extreme antenna deflection, limited fiducial detection, or in-sufficient locking pressure. The problem with antenna deflection was most likely exacerbated by the flexible, blunt MWA antennae that were used initially. We did not encounter the same is-sue with the antennae that were used for the last 15 procedures. Finally, great lengths were taken in this study to eliminate motion, including the use of general anesthesia with suspended respiration and the use of vacuum mattress. Because in many centers CT-guided liver abla-tion is not performed with the patient in general anesthesia, this limits the generalizability of this study.
In a future study, we intend to use a coaxial needle with a rigid stylet that can be exchanged for the MWA antenna and prevent deflection from the intended path. The next version of the robotic system will work with three instead of four fiducials, increasing the range of potential suitable positions. Additionally, we are plan-ning to use robotic guidance with conscious sedation to determine whether providing patients with breathing instruction is sufficient to result in accurate antenna placement or if respiratory tracking systems would need to be used.
is most similar to the Needle Placement System in using CT fidu-cials, with a reported accuracy of 2.3 mm (22). Despite the rela-tively large number of (experimental) robotic systems, there are few randomized patient studies that assessed their functionality in real clinical environments. We have identified only two random-ized controlled trials that compared robotic with freehand needle placement, with only one trial performed in patients with liver tu-mors. In 2005, Patriciu et al (14) tested a robotic device (AcuBot) in a randomized study with only 14 patients. The AcuBot re-duced the number of needle repositionings and targeting time, but needle placement accuracy was not reported. To this date, the AcuBot is not on the market. In another nonrandomized clinical
Table 3: Technical Outcomes of Robotic and Freehand Procedures
Outcome Robotic Procedures Freehand Procedures P Value
No. of needle repositionings 0 (0–0) 1 (0–7) ,.001
First placement accuracy*
Euclidian error (mm) 10.2 6 5.2 22.2 6 11.3 ,.001
Lateral error (mm) 6.4 6 4.2 15.0 6 10.0 .001
Depth error (mm) 3.6 6 5.3 10.8 6 9.1 .003
Angle error (degrees) 4.5 6 3.2 10.2 6 5.8 ,.001
Adequate placement accuracy*
Euclidian error (mm) 9.3 6 4.2 13.8 6 6.4 .02
Lateral error (mm) 6.6 6 4.2 8.2 6 3.9 .20
Depth error (mm) 3.4 6 2.5 9.4 6 7.3 .004
Angle error (degrees) 4.6 6 3.1 6.1 6 3.8 .15
Targeting time (min) 36 (3–70) 19 (8–55) .001
Ablation time (min) 16 (5–26) 19 (2–85) .18
Total procedure time (h:min)*† 2:40 6 0:26 2:36 6 0:29 .76
Ablation energy (kJ) 72 (6–125) 92 (2–383) .10
No. of CT scans to achieve adequate MWA
antenna position 2 (1–7) 3 (1–7) .18
No. of navigational CT scans required 5 (2–11) 7 (2–16) .09
DLP per pre-MWA contrast-enhanced
CT study (mGy · cm) 149 (74–372) 142 (67–219) .17
DLP for navigational CT scans to adequate
position (mGy · cm) 307 (107–1584) 108 (28–443) ,.001
DLP for total navigational CT scans (mGy · cm) 557 (107–2806) 219 (54–866) .004
DLP per post-MWA contrast-enhanced
CT study (mGy · cm) 170 (61–361) 131 (62–334) .02
Appearance at follow-up contrast-enhanced
CT study after 1 week .34
Apparent complete ablation 17 20
Apparent incomplete ablation 3‡ 1
No contrast-enhanced CT performed§ 1 0
Hematoma 2 1
Complications
Abscess 1
Minor pneumothorax 1
Note.—Unless otherwise specified, data are medians, with ranges in parentheses. DLP = dose-length prod-uct, MWA = microwave ablation. Navigational CT scan = CT image acquired to verify antenna position. * Data are means 6 standard deviations.
† Total procedure time is reported for solitary tumor ablations only.
‡ Two of the three incomplete ablations occurred in one patient, in whom contrast-enhanced CT was not
performed directly after ablation because of a CT scanner malfunction.
§ In one case, no follow-up CT examination was performed because the lesion was poorly delineable at
To our knowledge, our study was the first substantial random-ized controlled trial to analyze robotic needle placement for abla-tion of liver tumors. We showed that robotic antenna guidance offers advantages over freehand antenna placement in terms of the number of antenna repositionings needed and placement accuracy. The advantage is most prominent for out-of-plane procedures. Author contributions: Guarantor of integrity of entire study, K.P.d.J.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; agrees to ensure any ques-tions related to the work are appropriately resolved, all authors; literature research, W.J.H., R.V.; clinical studies, W.J.H., S.J.S.R., J.P.P., M.O., K.P.d.J.; experimental studies, W.J.H., B.L., M.O.; statistical analysis, W.J.H., M.O., K.P.d.J.; and manu-script editing, all authors.
Disclosures of Conflicts of Interest: W.J.H. disclosed no relevant relation-ships. S.J.S.R. disclosed no relevant relationrelation-ships. J.P.P. disclosed no relevant rela-tionships. B.L. Activities related to the present article: is an employee of DEMCON Advanced Mechatronics; institution has received a European Grant. Activities not related to the present article: disclosed no relevant relationships. Other relation-ships: institution receives payment for patent(s) issued to DEMCON Advanced Mechatronics. R.V. disclosed no relevant relationships. M.O. disclosed no relevant relationships. K.P.d.J. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: disclosed no relevant rela-tionships. Other relationships: disclosed no relevant relarela-tionships.
References
1. Curley SA, Izzo F, Delrio P, et al. Radiofrequency ablation of unresectable primary and metastatic hepatic malignancies: results in 123 patients. Ann Surg 1999;230(1):1–8. 2. Tanis E, Nordlinger B, Mauer M, et al. Local recurrence rates after radiofrequency ablation
or resection of colorectal liver metastases. Analysis of the European Organisation for Re-search and Treatment of Cancer #40004 and #40983. Eur J Cancer 2014;50(5):912–919. 3. Seror O, N’Kontchou G, Nault JC, et al. Hepatocellular carcinoma within Milan
criteria: no-touch multibipolar radiofrequency ablation for treatment–long-term re-sults. Radiology 2016;280(2):611–621 [Published correction appears in Radiology 2016;280(3):981.].
4. European Association For The Study Of The Liver; European Organisation For Re-search And Treatment Of Cancer. EASL-EORTC clinical practice guidelines: man-agement of hepatocellular carcinoma. J Hepatol 2012;56(4):908–943.
5. Dervenis C, Xynos E, Sotiropoulos G, et al. Clinical practice guidelines for the management of metastatic colorectal cancer: a consensus statement of the Hellenic Society of Medical Oncologists (HeSMO). Ann Gastroenterol 2016;29(4):390–416. 6. Gillams A, Goldberg N, Ahmed M, et al. Thermal ablation of colorectal liver metas-tases: a position paper by an international panel of ablation experts, the Interventional Oncology Sans Frontières meeting 2013. Eur Radiol 2015;25(12):3438–3454.
7. Kao WY, Chiou YY, Hung HH, et al. Risk factors for long-term prognosis in he-patocellular carcinoma after radiofrequency ablation therapy: the clinical implica-tion of aspartate aminotransferase-platelet ratio index. Eur J Gastroenterol Hepatol 2011;23(6):528–536.
8. Lam VW, Ng KK, Chok KS, et al. Incomplete ablation after radiofrequency ablation of hepatocellular carcinoma: analysis of risk factors and prognostic factors. Ann Surg Oncol 2008;15(3):782–790.
9. Hof J, Wertenbroek MW, Peeters PM, Widder J, Sieders E, de Jong KP. Outcomes after resection and/or radiofrequency ablation for recurrence after treatment of colorectal liver metastases. Br J Surg 2016;103(8):1055–1062.
10. Lam VW, Ng KK, Chok KS, et al. Risk factors and prognostic factors of local re-currence after radiofrequency ablation of hepatocellular carcinoma. J Am Coll Surg 2008;207(1):20–29.
11. Lu DSK, Raman SS, Limanond P, et al. Influence of large peritumoral vessels on outcome of radiofrequency ablation of liver tumors. J Vasc Interv Radiol 2003;14(10):1267–1274. 12. Kim YS, Rhim H, Cho OK, Koh BH, Kim Y. Intrahepatic recurrence after
percu-taneous radiofrequency ablation of hepatocellular carcinoma: analysis of the pattern and risk factors. Eur J Radiol 2006;59(3):432–441.
13. Koethe Y, Xu S, Velusamy G, Wood BJ, Venkatesan AM. Accuracy and efficacy of percutaneous biopsy and ablation using robotic assistance under computed tomog-raphy guidance: a phantom study. Eur Radiol 2014;24(3):723–730.
14. Patriciu A, Awad M, Solomon SB, et al. Robotic assisted radio-frequency ablation of liver tumors: randomized patient study. Med Image Comput Comput Assist Interv 2005;8(Pt 2):526–533.
15. Wallach D, Toporek G, Weber S, Bale R, Widmann G. Comparison of freehand-navigated and aiming device-freehand-navigated targeting of liver lesions. Int J Med Robot 2014;10(1):35–43.
16. Stoianovici D, Cleary K, Patriciu A, et al. Acubot: a robot for radiological interven-tions. IEEE Trans Robot Autom 2003;19(5):927–930.
17. Arnolli MM, Buijze M, Franken M, de Jong KP, Brouwer DM, Broeders IAMJ. System for CT-guided needle placement in the thorax and abdomen: a design for clinical acceptability, applicability and usability. Int J Med Robot 2018;14(1):e1877. 18. Arifin WN. Random sampling and allocation using SPSS. Educ Med J 2012;4(1):
e129–e143.
19. Ahmed M, Solbiati L, Brace CL, et al. Image-guided tumor ablation: standardization of terminology and reporting criteria–a 10-year update. Radiology 2014;273(1):241–260. 20. Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible statistical power
analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods 2007;39(2):175–191.
21. Heerink WJ, Solouki AM, Vliegenthart R, et al. The relationship between applied energy and ablation zone volume in patients with hepatocellular carcinoma and colorectal liver metastasis. Eur Radiol 2018;28(8):3228–3236.
22. Kettenbach J, Kara L, Toporek G, Fuerst M, Kronreif G. A robotic needle-positioning and guidance system for CT-guided puncture: Ex vivo results. Minim Invasive Ther Allied Technol 2014;23(5):271–278.
23. Engstrand J, Toporek G, Harbut P, Jonas E, Nilsson H, Freedman J. Stereotactic CT-guided percutaneous microwave ablation of liver tumors with the use of high-frequency jet ventilation: an accuracy and procedural safety study. AJR Am J Roent-genol 2017;208(1):193–200.
Table 4: Technical Outcomes Stratified according to Angulation
Outcome
In-Plane Angulation Out-of-Plane Angulation
Freehand Robotic P Value Freehand Robotic P Value
No. of tumors 10 8 11 13
Depth from skin (mm) 85.6 6 27.5 89.7 6 33.8 .78 82.6 6 25.9 94.8 6 31.7 .32
No. of needle repositionings* 1 (0–5) 0 (0–0) .004 1 (0–7) 0 (0–0) .001
Targeting time (min)* 18 (4–25) 32 (3–43) .01 19 (7–55) 37 (12–84) .02
First placement accuracy
Euclidian error (mm) 20.1 6 3.6 10.4 6 4.8 .03 24.1 6 11.5 10.1 6 5.6 .002
Lateral error (mm) 13.9 6 9.7 7.7 6 5.9 .13 16.1 6 10.7 5.6 6 2.7 .009
Depth error (mm) 10.1 6 8.1 3.2 6 2.7 .03 11.4 6 10.2 3.9 6 6.5 .04
Angle error (degrees) 8.9 6 5.3 5.5 6 4.4 .16 11.3 6 6.3 3.6 6 2.0 .002
Adequate placement accuracy
Euclidian error (mm) 13.0 6 6.7 10.4 6 4.8 .36 14.5 6 6.4 9.2 6 4.0 .02
Lateral error (mm) 6.2 6 2.7 7.7 6 5.9 .51 10.1 6 4.0 5.9 6 2.9 .007
Depth error (mm) 10.1 6 8.2 3.2 6 2.7 .03 8.7 6 6.8 3.5 6 2.5 .03
Angle error (degrees) 4.4 6 1.9 5.5 6 4.4 .47 7.7 6 4.3 3.7 6 2.0 .006
Note.—Unless otherwise specified, data are means 6 standard deviations. * Data are medians, with ranges in parentheses.
