Stereotactic and Robotic Minimally Invasive Thermal Ablation of Malignant Liver Tumors: A Systematic Review and Meta-Analysis

Background Stereotactic navigation techniques aim to enhance treatment precision and safety in minimally invasive thermal ablation of liver tumors. We qualitatively reviewed and quantitatively summarized the available literature on procedural and clinical outcomes after stereotactic navigated ablation of malignant liver tumors. Methods A systematic literature search was performed on procedural and clinical outcomes when using stereotactic or robotic navigation for laparoscopic or percutaneous thermal ablation. The online databases Medline, Embase, and Cochrane Library were searched. Endpoints included targeting accuracy, procedural efficiency, and treatment efficacy outcomes. Meta-analysis including subgroup analyses was performed. Results Thirty-four studies (two randomized controlled trials, three prospective cohort studies, 29 case series) were qualitatively analyzed, and 22 studies were included for meta-analysis. Weighted average lateral targeting error was 3.7 mm (CI 3.2, 4.2), with all four comparative studies showing enhanced targeting accuracy compared to free-hand targeting. Weighted average overall complications, major complications, and mortality were 11.4% (6.7, 16.1), 3.4% (2.1, 5.1), and 0.8% (0.5, 1.3). Pooled estimates of primary technique efficacy were 94% (89, 97) if assessed at 1–6 weeks and 90% (87, 93) if assessed at 6–12 weeks post ablation, with remaining between-study heterogeneity. Primary technique efficacy was significantly enhanced in stereotactic vs. free-hand targeting, with odds ratio (OR) of 1.9 (1.2, 3.2) (n = 6 studies). Conclusions Advances in stereotactic navigation technologies allow highly precise and safe tumor targeting, leading to enhanced primary treatment efficacy. The use of varying definitions and terminology of safety and efficacy limits comparability among studies, highlighting the crucial need for further standardization of follow-up definitions.


INTRODUCTION
Thermal ablation therapy is a validated curative-intended treatment option for malignant liver tumors, mainly hepatocellular carcinoma (HCC) and liver metastases from colorectal cancer (CRLM) (1,2). Following encouraging oncological outcome results for small tumors in the setting of limited disease, thermal ablation has been introduced into international treatment guidelines for both HCC and CRLM (3,4).
The key advantage of ablative treatments is their tissue-sparing nature, allowing preservation of a maximum of functioning liver tissue and favouring combination with a minimally invasive treatment access to further reduce procedure-related morbidity. The main challenge in achieving treatment success in such minimally invasive environments is the accurate and safe positioning of ablation probes to acquire adequate ablation volumes with full tumor coverage (5,6). The use of (contrastenhanced) ultrasound (US) imaging allows dynamic intraoperative tumor visualization and instrument guidance both in surgical and interventional radiology settings (7)(8)(9). For lesions remaining invisible due to small size, deep central location, obstructing gas/ ribs, or changes in liver parenchyma (10), computed tomography (CT) or magnetic resonance (MR) guidance enhances visibility but introduces constraints due to radiation exposure (11) or procedurerelated complexity. Especially for tumors requiring complex targeting trajectories, such as in the liver dome or caudal lobe, a safe and efficient targeting is often precluded (12,13).
To improve tumor accessibility, targeting accuracy, and treatment safety, stereotactic navigation systems have been introduced for use in minimally invasive surgery. These aim to enhance treatment precision by integrating computer assistance with imaging data, allowing stereotactic guidance of surgical instruments (14). Over the last two decades, increasing clinical experience with commercially available navigation devices has been reported, including several summary articles highlighting their main advantages (15)(16)(17)(18)(19)(20)(21). However, no systematic review of the available literature on the clinical application of this technology exists, and the true impact on procedural and clinical outcomes remains unknown. The aim of this study was to critically review and quantitatively summarize the available literature on targeting accuracy, procedural efficiency, and treatment efficacy when using stereotactic or robotic navigation technology for thermal ablation of malignant liver tumors in a minimally invasive setting.

Search Strategy
This study was conducted following the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines (22). The PRISMA checklist is available as Supplementary File 1. A systematic literature review was performed on May 20, 2020, searching the online databases Medline, Embase, and the Cochrane Central Register of Controlled Trials for all available full-text articles. Search terms were organized according to the population, intervention, comparison, and outcomes (PICO) criteria, without including comparison or outcome criteria to keep the search as broad as possible. Given the expected paucity of published literature, no limitation of publication dates was applied. The complete search strategy is shown in Table 1.

Study Selection
Eligible studies included original articles on adults, reporting on procedural or clinical outcomes when using stereotactic or robotic guidance for targeting and thermal ablation of malignant liver tumors, and using a minimally invasive (laparoscopic or percutaneous) approach. Stereotactic or robotic guidance was defined as the utilization of a tracked ablation probe or aiming device for targeting and subsequent therapeutic thermal ablation. Included studies reported on one or more of the following outcomes related to stereotactic thermal ablation of liver tumors: i) targeting accuracy, ii) procedural efficiency and safety, and iii) treatment efficacy. Excluded were i) review articles, ii) conference abstracts, iii) studies focusing on image fusion without the use of a tracked ablation probe or aiming device for tumor targeting, iv) studies analyzing efficacy of combined transarterial chemoembolization (TACE) and stereotactic ablation of liver tumors, v) case reports defined as <10 patients treated with navigated/robotic ablation, and vi) full texts written in languages other than English.
Two authors (PT and JE) independently determined eligibility for each citation by sequential review of titles, abstracts, and full texts using the predefined criteria. In case of disagreements, consensus was reached by group discussion. Reference lists from included studies were reviewed for additional citations not identified by the original search. Covidence (23) and Mendeley were used for screening and reference citation management.

Data Extraction and Risk of Bias Assessment
All data extracted for systematic review and meta-analysis were entered into a data spreadsheet available as [Supplementary File 2]. Main outcomes were i) targeting accuracy, defined as targeting errors resulting after stereotactic ablation probe positioning; ii) procedural efficiency and safety, including duration of the overall procedure and of stereotactic ablation probe positioning, numbers of probe readjustments, radiation exposure reported as dose length product (DLP), hospital length of stay (LOS), and clinical complications; and iii) treatment efficacy, including rates of technical success, primary and secondary technique efficacy, and local tumor progression (LTP). Definitions of treatment efficacy were summarized using the standardized terminology and reporting criteria for image-guided tumor ablation (24). The authors' detailed descriptive definitions of all endpoints are available in the [Supplementary File 2].
For quantitative analyses, data from studies with overlapping patient populations were obtained from the most relevant publications. These were chosen by prioritizing studies with the largest sample size and then by studies published most recently. Risk of bias in individual studies was evaluated for all outcomes according to the ROBINS-I tool (25), and studies with high risk of bias were excluded from meta-analysis. Studies reporting on ablation for uncommon subpopulations of liver tumors (i.e., very large tumors) were considered high risk of selection bias toward safety and efficacy outcomes. Comparative observational studies reporting outcomes without matching of tumors with respect to targeting complexity (e.g., tumor location) were considered medium risk of bias due to confounding. No quantitative analyses were performed for efficiency outcomes (probe readjustments, procedure, and targeting durations/DLP) and LTP, since heterogeneity across studies in definition and assessment of these outcomes was deemed too important.

Data Synthesis and Meta-Analysis
Baseline characteristics were summarized, and weighted averages, rates, and odds ratios (ORs) were reported according to individual number of lesions or patients, as appropriate.
Pooled estimates were calculated based on a modified inverse variance method, applying a continuous/binary random effects model given the assumed heterogeneity of included studies (DerSimionian-Laird method) (26). To avoid skewing of the variance toward zero, an arcsine transformation was applied for outcomes with small reported proportions (major complications, mortality) (27). Subgroup meta-analysis was performed to account for known sources of heterogeneity due to differing definitions for complication rates and primary technique efficacy. Between-study heterogeneity was reported as I 2 statistic and chi-square test of homogeneity. Sensitivity analyses were performed by comparing results from subgroup analyses vs. overall results and randomeffects vs. fixed-effects analyses. OpenMetaAnalyst (28), R (R Core Team, 2019), and RStudio (RStudio Inc., USA) were used for metaanalysis and generation of graphics.

RESULTS
After screening a total of 1,412 articles, 93 original works published in English and reporting on a minimum of 10 patients treated with stereotactic or robotic thermal ablation for malignant liver tumors were included for full-text screening. After additional exclusion of 59 works, 34 articles were included for qualitative analysis and 22 for quantitative analysis ( Figure 1). All of the 34 included works were single-center studies, of which 26 were retrospective studies , three were prospective case series (55)(56)(57), three were prospective cohort studies (58)(59)(60), and two were randomized controlled trials (61,62). Two studies reported results using a laparoscopic treatment access (36,55); the other 32 reporting on thermal ablations using a percutaneous approach. Six studies applied robotic targeting using mechanical tracking (39-41, 44, 60, 62), seven an electromagnetic (EM)-tracked dynamic technique (34,43,51,55,57,59,61), and the other 21 an optically tracked stereotactic aiming device. Baseline study, patient and lesion characteristics, applied ablation and navigation techniques, and reported outcomes of the included works are shown in Table 2. All data extracted from included studies are available in the Supplementary File 2. Results from meta-analyses including sensitivity analyses and selection bias assessment are available in Supplementary File 3.
The published literature on stereotactic or robotic guidance for thermal ablation of liver tumors increased continuously since the first clinical series in 2011. This was the case regarding all reported endpoints and most prominently for safety and treatment efficacy, as illustrated in Figure 2.

Targeting Accuracy
Of the 34 included studies, 10 reported on targeting accuracy, of which nine reported Euclidean, lateral, or angular targeting errors after stereotactic ablation probe positioning. The remaining study reported a 95.6% targeting accuracy, defined as the center of the ablated zone being located within a 5-mm range from the preoperatively defined ideal target point, as assessed on a 24-h CT/MRI scan (43). Pooled estimates for Euclidean, lateral, and angular targeting errors were 5.3 (95% CI 3.9, 6.7), 3.7 (3.0, 4.4), and 2.4 (1.7, 3.1) mm, respectively, with significant between-study heterogeneity ( Table 3). Summary estimates for lateral targeting errors are displayed in Figure 3.
Three studies investigated factors influencing targeting accuracy, of which two studies performed univariable between- group analyses and one group used multivariable linear regression. Mauri et al. (43) showed no influence on "correct targeting" by tumor entity, lesion characteristics, and applied guidance and ablation techniques. Widmann et al. (56) found larger lateral errors at the ablation probe tip in lesions located in "subphrenic plus fat" as opposed to "clear parenchymal" positions and in Segment II vs. Segment IV. Tinguely et al. (29) reported statistically significant higher lateral targeting errors with raising targeting trajectory lengths (0.2 mm per additional cm) and when targeting tumors in cirrhotic livers (by 0.7 mm) in a multivariable model, with no influence of challenging lesion locations or complex targeting trajectories on accuracy.

Procedural Efficiency and Safety
Eighteen studies reported on procedural efficiency, and all included works reported on safety related to stereotactic targeting for thermal ablation. The need for readjustment of ablation probes due to insufficient accuracy was reported varyingly across studies. Four studies reported mean numbers of ablation probe readjustments of n = 0 (62), 0.8 (40) (64), two applied the definitions proposed by Ahmed et al. (24), one applied the CIRSE classification, and 11 studies used other definitions or did not further specify. The overall complication rate ranged between 0% and 57.9%, the pooled estimate being 11.4% (CI 6.7, 16.1; I 2 87.9%, p < 0.01) in 16 studies included for meta-analysis. Major complications ranged between 0% and 20.5%, the 20.5% rate being reported in a study of patients undergoing stereotactic radiofrequency ablation (RFA) for very large (≥8 cm) tumors (49). The overall pooled estimate for major complication rate was 2.4% (CI 1.4, 3.6; I 2 20.7%, p = 0.198), which was lower in the subgroup applying the Clavien-Dindo classification (2.0%; CI 0.7, 4.0) vs.    General Tube discon   Only accuracies of initial probe placements are shown (before eventual readjustments).°D efined as "first-pass control": aiming at approximately 1 cm from the tumor. Not included in weighted average.  (30)], reported rates for technical success ranged from 90.2% to 100% (18 studies), for primary technique efficacy from 80.5% to 100% (27 studies) and for secondary technique efficacy from 90.2% to 100% efficacy (13 studies), and for LTP from 0% to 54% (21 studies). Quantitative analysis of primary technique efficacy (i.e., complete tumor ablation at the first follow-up imaging) according to time points of the first follow-up is summarized in Figure 3. Primary technique efficacy rates were reported to be higher in studies performing a first follow-up after 1-6 weeks (pooled estimate 93.6%; CI 88.9, 97.1) than in studies assessing primary technique efficacy at 6-12 weeks (pooled estimate 90.1%; CI 87.2, 92.7). Despite subgroup analysis, a statistically significant between-study heterogeneity remained in the former group ( Figure 6).
Six studies presented analyses on factors influencing treatment efficacy when using stereotactic tumor targeting, four of which included multivariable regression analyses. Tinguely et al. (29) showed a statistically significant influence of tumor size (</>3 cm) and targeting accuracy (</>5 mm) on LTP, with no influence by more complex intrahepatic tumor locations, in a multivariable model accounting for clustering by using Generalized Estimating Equation (GEE). Schaible et al. (41) reported tumor size (</>3 cm) next to the type of targeting approach (stereotactic vs. free-hand) to be a significant predictor of primary technique efficacy in a similar logistic GEE model. Lachenmayer et al. (33) found vessel proximity and tumor size (</>3 cm) as independent predictors for LTP. Hirooka et al. (59) reported the free-hand as opposed to the stereotactic approach to be significantly associated with "local residual recurrence" in Cox regression analysis. In univariable between-group comparisons, Widmann et al. (47) reported differences in "technique effectiveness" for lesion size </>5 cm and hollow viscera vicinity and no differences for tumor entity and lesion location. Bale et al. (45) found differences in "local recurrence" rates for lesions in proximity to vessels, bile ducts, and hollow organs.

Comparative Studies
Nine studies compared stereotactic vs. "free-hand" ablation for varying endpoints, including two randomized controlled studies (61,62) and three prospective cohort studies (59, 60) of which one used matched-pair analysis (58). Main study characteristics and reported results for targeting accuracy, procedural efficiency, and safety and treatment efficacy are summarized in Table 4.
Targeting accuracy was shown to be significantly enhanced when using stereotactic targeting in three out of four studies [one of them after manual adaptation of the ablation probe (39)]. The randomized controlled trial by Heerink at al (62). confirmed  Durations for overall procedures and for ablation probe positionings were reported variably across studies. Zhang et al. (61) showed a significant reduction in the number of CT scans used for interventions (7 vs. 10), in CT fluoroscopy time, and in total DLP when using EM-guided as opposed to free-hand targeting. Four other studies showed a reduction, and two studies showed an increase in total DLP in the stereotactic vs. free-hand cohort using ( Table 2). In the RCT of Heerink et al. (62), the number of navigational CT scans tended to be lower in stereotactic vs. free-hand procedures (5 vs. 7), and the increase in DLP was presumably due to the larger scan field necessary to include the optical reference fiducials.
With respect to treatment efficacy, Zhang et al. (34) showed a higher complete ablation rate in the first session (3 days after ablation) when using stereotactic EM guidance but equal primary technique efficacy at 1 month. All six studies included for meta-analysis assessed primary technique efficacy at a first follow-up imaging between 1 and 6 weeks. Stereotactic and robotic guidance led to enhanced primary technique efficacy rates as opposed to free-hand targeting, with a pooled OR of 1.94 (CI 1.18, 3.19) (Figure 7). Sensitivity analyses comparing random-to fixed-effects analysis showed comparable results,    File 3). Four propensity score matched analyses compared stereotactic thermal ablation for different subtypes of patients or lesions. Stereotactic RFA yielded similar safety and treatment efficacy outcomes when comparing CT "invisible" tumors ablated with MR image fusion to CT-visible tumors (30), octogenarians to a younger study population (52), lesions in a subphrenic location to non-dome locations (31), and HCC in a subcardiac position to a non-subcardiac location (32).

DISCUSSION
This work is a summary of the currently published knowledge on procedural and clinical benefits of applying minimally invasive stereotactic and robotic navigation technology for thermal ablation of malignant liver tumors. To our knowledge, this is the first systematic review and meta-analysis on this topic.
Stereotactic navigation technologies for use in liver surgery aim to enhance precision when performing tissue-sparing treatments such as "atypical" resections not following standard anatomical landmarks or locally targeted interventions such as thermal ablations. For the latter case specifically, the aims are to i) ensure adequate oncological results in a minimally invasive setting by precise ablation probe positioning and ii) enhance safety by defining trajectories that avoid injury to critical anatomical structures. Stereotactic navigation systems enable quantification of the accuracy with which ablation probes are positioned by measuring the error between the planned "optimal" and the final "real" probe position. This facilitates the intraoperative decision regarding an eventual probe readjustment and, importantly, allows a reproducible and comparable evaluation of targeting success. As summarized herein, pooled estimates for targeting accuracy in over 650 targeted lesions ranged from 2.4 to 5.3 mm, with relatively small confidence intervals and with all comparative studies showing reduced targeting errors as compared to free-hand techniques ( Table 4).
The pooled estimates for major complication rates for different definitions ranged between 1.8% and 5.2%, as expected when applying tissue-sparing treatments in a minimally invasive setting. Importantly, the number of probe readjustments was significantly reduced in three comparative studies, including two randomized controlled trials. This decrease in liver punctures until reaching adequate probe positions toward n = 1 represents an important virtue of stereotactic thermal ablation, leading to enhanced safety (60,62). It further allows to safely target a multitude of tumors in one treatment session (average number of tumor targeted in included studies up to 6; Table 2), keeping oncological indications for ablation treatment in mind. The high efficiency in tumor targeting did not unanimously translate into reduced overall procedure durations, where confidence intervals were wide and conflicting results were reported in comparison to free-hand targeting. An increased complexity in hardware setup and procedural workflows has been highlighted as potential drawbacks when using such novel technology (40). One way to optimize procedural efficiency when applying such novel technology is to invest in initial training of staff including anesthesia to create smooth and standardized clinical workflows in multidisciplinary teams. The latter allow relatively steep learning curves (36) and a reduction of inter-operator variance as a factor influencing outcomes after thermal ablation (47). The present study confirms enhanced reliability and reproducibility of outcomes as being one of the key advantages of using stereotactic ablation techniques (50).
Overall reported treatment efficacy rates were encouraging, leading to increased early local tumor control when using stereotactic navigation as opposed to manual guidance as shown in the latest available comparative studies (Figure 7). Of importance is the inconsistency in reported definitions describing technique efficacy and LTP ( Figure 5). While standardized terminology and reporting of outcomes criteria after image-guided ablation have been proposed (24), definitions allow wide ranges of authordependent interpretations. Especially the definition of "secondary treatment efficacy" allows for differing and unspecified numbers of retreatments over an undefined follow-up period to be included. Consequently, detectable tumor at the ablation site might only be defined as "local tumor progression" rates at 6 or 12 months after the initial ablation. This renders a straightforward comparison of reported results impossible. Despite subgroup meta-analysis according to time point of follow-up, subgroups remained highly heterogeneous due to variations in included patients, lesions, and applied types of stereotactic ablative treatments. This highlights the need for further standardization of follow-up definitions toward more clearly defined ranges of follow-up time points and more precise terminology. Using novel segmentation technology, a quantitative volumetric assessment of ablation margins will in the future allow a quantified distinction between "incomplete ablation" and true "tumour recurrence/progression at the ablation site" (65). Integrating such precise assessment of treatment success into refined follow-up terminology could contribute to the generation of more reproducible and comparable outcomes after stereotactic thermal ablation.
Varying types of navigation modalities were applied in the included studies, such as optically tracked aiming devices or EM-tracked ablation probes combined with US to CT/MR image fusion. An upcoming group of navigation systems uses robotic assistance devices, relying on mechanical tracking for automatic orientation of the robotic arm but still requiring manual insertion of ablation probes. These comprise important but only initial steps toward automation and standardization of stereotactic ablation on a larger scale. Future generations of robotic devices aim to perform automated probe positioning according to defined trajectories (66) and generate individualized ablation volumes for specified tumor configurations (67). The development of automated and dynamic patient tracking will further allow to address current challenges related to motion artifacts and facilitate ablation under sedation. No clear benefit regarding targeting accuracy when using current robotic vs. non-robotic approaches can be confirmed to date.
A potential limitation of the present work involves a potential bias in interpretation of qualitative analyses due to overlapping cohorts (Figure 1), which was not feasible to address, since studies overlapped varyingly for different outcomes assessed. Another important issue is the heterogeneity in tumor entities and applied ablation systems in the included studies, since variability in the response to applied ablation energies across tumor types and devices is known (68,69). In this study, we deliberately focused on targeting accuracy, procedural efficiency, and primary technique efficacy, which were not expected to differ significantly between tumor types, since they are primarily related to procedural technique rather than oncological aspects. Contrarily, many tumor-specific determinants affect ablation site recurrences and long-term tumor control, such as satellite nodules detectable on pathology but not imaging, underlying liver disease, mutational status and location of primary tumor, and different types and time points of chemotherapy regimens (70,71). Lastly, combination therapies with other interventional treatments complicate comparisons of long-term oncological outcomes. A variability in the calculation of morbidity and mortality rates (assessment per patient vs. per intervention vs. per lesion, 30-vs. 90-day period) remained. In some studies, it was further unclear if the positioned ablation probes, for which targeting accuracies were reported, were the ones applied for subsequent thermal ablation. This might affect conclusions regarding a relationship of targeting accuracy with resulting treatment efficacy. While accurate tumor targeting is the initial step for successful ablation, an independent correlation with early ablation site recurrence has been described (29).
As technology evolves, it can be expected that stereotactic and robotic interventions will increasingly become an integral part of the multimodality management of malignant liver tumors. Due to the increasing expertise in specialized centers, treatment indications for stereotactic thermal ablation are increasingly expanded. The present summary shows the encouraging results in all reported outcomes including for patients with tumors in difficult intrahepatic locations [centrally, liver dome (31), segment 1 (53)], very large tumors (49), and "vanishing" or CT-invisible tumors using image fusion (72). In practice, minimally invasive stereotactic ablation approaches are beneficial especially in these situations, where they allow a curative intended treatment when conventional image guidance techniques preclude an efficient and safe targeting. Key challenges remaining to be addressed when using current navigation technology are potential inaccuracies due to motion artifacts (73) and concerns regarding cost effectiveness (74).

CONCLUSIONS
Advances in stereotactic navigation technology allow highly precise, safe, and efficient minimally invasive ablation of malignant liver tumors, potentially leading to enhanced early treatment efficacy compared to traditional guidance techniques. Heterogeneity in terminology and time points in follow-up assessment limits comparability of safety and treatment efficacy among studies, highlighting the crucial need for further standardization and guidelines for follow-up definitions.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

AUTHOR CONTRIBUTIONS
PT and IP were involved in study concept and design, data acquisition, data analysis and interpretation, statistical analysis, and manuscript preparation and editing. SR and JE were involved in study concept and design, data analysis and interpretation, and manuscript editing and review. SW, KJ, DC, and JF were involved in study concept and design and manuscript editing and review. All authors contributed to the article and approved the submitted version.