Improved Outcomes of Thermal Ablation for Colorectal Liver Metastases: A 10-Year Analysis from the Prospective Amsterdam CORE Registry (AmCORE)

Background To analyze long-term oncological outcomes of open and percutaneous thermal ablation in the treatment of patients with colorectal liver metastases (CRLM). Methods This assessment from a prospective, longitudinal tumor registry included 329 patients who underwent 541 procedures for 1350 CRLM from January 2010 to February 2021. Three cohorts were formed: 2010–2013 (129 procedures [53 percutaneous]), 2014–2017 (206 procedures [121 percutaneous]) and 2018–2021 (206 procedures [135 percutaneous]). Local tumor progression-free survival (LTPFS) and overall survival (OS) data were estimated using the Kaplan–Meier method. Potential confounding factors were analyzed with uni- and multivariable Cox regression analyses. Results LTPFS improved significantly over time for percutaneous ablations (2-year LTPFS 37.7% vs. 69.0% vs. 86.3%, respectively, P < .0001), while LTPFS for open ablations remained reasonably stable (2-year LTPFS 87.1% [2010–2013], vs. 92.7% [2014–2017] vs. 90.2% [2018–2021], P = .12). In the latter cohort (2018–2021), the open approach was no longer superior regarding LTPFS (P = .125). No differences between the three cohorts were found regarding OS (P = .088), length of hospital stay (open approach, P = .065; percutaneous approach, P = .054), and rate and severity of complications (P = .404). The rate and severity of complications favored the percutaneous approach in all three cohorts (P = .002). Conclusion Over the last 10 years efficacy of percutaneous ablations has improved remarkably for the treatment of CRLM. Oncological outcomes seem to have reached results following open ablation. Given its minimal invasive character and shorter length of hospital stay, whenever feasible, percutaneous procedures may be favored over an open approach. Supplementary Information The online version contains supplementary material available at 10.1007/s00270-022-03152-9.

, P = .12). In the latter cohort (2018)(2019)(2020)(2021), the open approach was no longer superior regarding LTPFS (P = .125). No differences between the three cohorts were found regarding OS (P = .088), length of hospital stay (open approach, P = .065; percutaneous approach, P = .054), and rate and severity of complications (P = .404). The rate and severity of complications favored the percutaneous approach in all three cohorts (P = .002). Conclusion Over the last 10 years efficacy of percutaneous ablations has improved remarkably for the treatment of CRLM. Oncological outcomes seem to have reached results following open ablation. Given its minimal invasive character and shorter length of hospital stay, whenever
Thermal ablation can be performed via an open, laparoscopic or percutaneous approach. Laparoscopic ablation is increasingly being performed due to its minimal invasive character compared to ablations via laparotomy, and local control rates are reported to be comparable between the two approaches. [11] However, laparoscopic ablation is technically more demanding and requires a fairly high level of expertise, which is presumably the reason that it is not yet widely embraced worldwide [4,12,13]. The percutaneous approach is mainly preferred in patients whose comorbid conditions preclude surgery, for centrally located tumors otherwise requiring a major resection (parenchyma-sparing), or in patients with regional or local tumor progression after prior local liver treatment [14][15][16][17][18]. This minimally invasive percutaneous approach is known for its favorable safety profile with low major complications rates (1.3%-2.4%) [14,19,20] .
As oncological outcomes of thermal liver ablation differ substantially among semi-recently published papers and evidence regarding the potential improvement over time, in terms of local control and time-to-local tumor progression, is lacking, this single-center Amsterdam Colorectal Liver Met Registry (AmCORE) based study aimed to analyze local disease control and survival outcomes following thermal ablation in patients treated for hepatic metastases from colorectal cancer over the last 10 years.

Material and Methods
Patients Data were sourced from a prospective, longitudinal tumor registry for patients with hepatic metastases from colorectal cancer. All patients were treated at the Amsterdam UMC, location Vrije Universiteit (Amsterdam, the Netherlands), a tertiary referral institution for hepatobiliary and gastrointestinal malignancies. Approval was granted from the affiliated Institutional Review Board (reference number 2021.0121).
Between January 2010 and February 2021, 449 consecutive patients with liver-only metastatic colorectal carcinoma underwent open or percutaneous thermal ablation with RFA or MWA (Fig. 1). One-hundred fifteen patients were excluded for having no available follow-up data at our institute. Although higher morbidity rates have never been reported after simultaneous liver ablation and bowel resection, partial hepatectomy plus colon surgery is known to be associated with a significant increased postoperative morbidity rates [40]. To overcome potential outcome interference, 15 patients were excluded having received simultaneous bowel resection. The remaining 329 patients underwent 541 procedures for 1350 liver metastases. Preprocedural treatment planning (e.g., angle of probe insertion) was performed prior to all procedures, and for percutaneous sessions, all needles/antennae were inserted under real-time computed tomography (CT) imaging. All patients had an Eastern Cooperative Oncology Group status of B 2. The diagnosis of CRLM was based on crosssectional imaging containing CT, magnetic resonance imaging (MRI) and [18F]-fluoro-2-deoxy-D-glucose ( 18 F-FDG) positron emission tomography (PET)-CT scans. Treatment planning was routinely discussed in a multidisciplinary tumor board. An open rather than a percutaneous approach was chosen in case of liver metastases needing concomitant partial hepatectomy or when a percutaneous approach was technically not feasible due to the position of the tumor (e.g., in close proximity to the stomach).
Although induction systemic therapy is not standard of care within the Netherlands, three patient categories did often receive induction systemic therapy first, namely: (A) patients with locally advanced primary (rectal) cancer, (B) patients with unresectable but potentially downstagable CRLM or with difficultly resectable disease if systemic therapy is likely to reduce procedural risk, and (C) patients with early metachronous disease. Chemotherapy regimen consisted of either capecitabine or irinotecan monotherapy, capecitabine and oxaliplatin (CAPOX), capecitabine ? irinotecan (CAPIRI), folinic acid ? 5-fluorouracil ? oxaliplatin (FOLFOX) or folinic acid ? 5fluorouracil ? irinotecan (FOLFIRI). Additional monoclonal antibodies (bevacizumab or panitumumab) were added in case of potentially downstagable disease. Conformal to national guidelines, no patients received adjuvant systemic therapy. [41].
The baseline characteristics of all enrolled patients are summarized in Table 1

Ablation Method
The vast majority of open and percutaneous ablations were performed by three interventional radiologists (BM, JV, MM) who have performed and/or supervised [ 100 imageguided tumor ablations. The staff in our department has been almost stable over the last ten years. Approximately one-third of the procedures were performed by two interventional radiologists at the same time. During approximately 60% of all ablation procedures, the senior interventional radiologist (MM) was present. The procedure and other study-related details are given in supplementary materials (Appendix 1).

Efficacy Evaluation and Follow-Up Strategy
Within the first two weeks after the initial procedure, a quality control contrast-enhanced CT scan was performed when there was a potential inadequate safety margin (0-5 mm) in combination with sub-optimal tumor conspicuity and needle visibility during the procedure [6]. This allowed for an early completion ablation procedure, if indicated. Follow-up should have consisted of at least one cross-sectional imaging modality study to reliable exclude or detect LTP. Regular follow-up consisted of [18]F-FDG- PET CT scans every 3 months after the initial ablation during the first year of follow-up and roughly every 6 months thereafter, according to national guidelines [41] and the standardization paper [2]. Additional MRI was only performed in case of uncertainty whether LTP was present. Follow-up imaging was reviewed by the interventional oncology team, certified diagnostic abdominal radiologists and nuclear physicians. If loco-regional disease recurrence was found on follow-up imaging, optimal retreatment was offered based on recommendations of the multidisciplinary team, depending on the extent of the disease in the liver, hepatic function, extrahepatic metastases and general condition of the patient.

Data Collection and Statistical Analysis
For the sake of oncological outcome analyses, the entire cohort was divided into three subgroups (2010-2013, 2014-2017 and 2018-2021). Standard demographic, clinical and surveillance data were retrieved from the   .010 a electronic database. Categorical variables are reported as frequencies (with or without percentage; %), whereas continuous variables are presented as median (IQR, interquartile range) or mean (± SD, standard deviation). Differences between the three subgroups in terms of baseline variables and outcomes were determined by using the Pearson Chi-square (v2) test for categorical variables ( a ) and the one-way ANOVA ( b ) for comparison of means between the three subgroups. Endpoint definitions were used along the consensus guidelines for the definition of time-to-event endpoints in image-guided tumor ablation by Puijk et al. [27] To study the primary endpoint, a time-to-event superiority analysis was used to analyze local tumor progression. LTP was defined as growth of tumor tissue at the initial treated tumor site [2,27]. Patients were followed until the first recorded evidence of LTP (event) or until the last followup exam for those alive without LTP. Local tumor progression-free survival (LTPFS) curves, per patient and per tumor, were estimated using the Kaplan-Meier method and compared between subgroups using the log-rank test. Death without LTP was considered a competing risk. Variables with P \ .05 in the univariable analysis were included in the final multivariable model. Hazard ratios (HR) and 95 percent confidence intervals (95% CI) were calculated. Using backward selection procedure, results of step-by-step removed variables were reported. Results are from last step before removal. Secondary endpoints were overall survival (OS) and safety. OS probability was estimated using the Kaplan-Meier method (time from the first ablation until the date of death or to the last follow-up visit or exam) for the entire cohort. Death during the index hospitalization or within 30 days after treatment was considered perioperative mortality. Safety in terms of complications was evaluated and reported using the standardized Common Terminology Criteria for Adverse Events (CTCAE) grading system, version 4.0 and 5.0. [2,27,42].
Statistical analyses were performed in consultation with an independent statistician (BLW) using SPSSÒ software, version 24.0 (IBMÒ, Armonk, New York, USA) [43] and the R software package, version 3.6.3 (R Foundation,

Fig. 2 Kaplan-Meier survival curves indicating local tumor progression-free survival (LTPFS) per treated tumor (A) and per patient (B) after all thermal ablation sessions. Numbers at risk correspond to the amount of tumors and number of patients respectively. Death without local tumor progression (LTP) is censored (competing risk)
Vienna, Austria) [44]. Statistical significance was established for P \ .05. All results were reported according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for reporting observational study data. [45].

Complications and Length of Hospital Stay
Grade 1-5 complication rate in the entire cohort was 20.5% (111/541 procedures; Table 2). The severity of complications did not change over time (P = .404). The rate and severity of complications favored the percutaneous approach in all three cohorts (2010-2013, P = .069; 2014-2017, P = .129; 2018-2021, P = .020). Sub-analysis of procedures were thermal ablation was used solely (in other words without simultaneous resection or irreversible electroporation in case of open procedures), revealed no difference in complication rate between the three time frames (P = .406).
Overall procedure-related mortality was 1.5% (5/329) in the entire cohort. One patient deceased 7 days after combined liver resection and ablation due to massive pulmonary embolism (30-day mortality 0.4%; n = 1/329). Five others died from postoperative complications between 30 and 90 days: one due to massive portal thrombosis and multi-organ failure 5 weeks after combined percutaneous ablation and irreversible electroporation, and three due to abdominal abscesses and cardiopulmonary failure 8-9 weeks after combined liver resection and open ablation.
For open ablations, the mean length of hospital stay did not significantly differ between the three time frames (mean 6.9 days [SD 5.9]; P = .065). Mean hospitalization after percutaneous procedures was 1.4 days (SD 2.6) with no differences between the three cohorts (P = .054).

Overall Survival
A total of 99 patients (30.1%) deceased during follow-up (Table 2). Of them, 93 died from disease progression. Survival probability after the first ablative treatment was 92.0%, 78.8%, 45.9% and 26.8% at 1, 3, 5 and 10 years, respectively (Fig. 5). For the entire cohort, the median OS after the first ablation procedure was 54. the 2014-217 cohort. The median OS for the latter cohort was not met. The median OS did not significantly improve over the last decade (P = .088), nor differed for patients treated by open or percutaneous ablation (P = .888).

Discussion
Over the past decades, thermal ablation has become the standard treatment option to eradicate small unresectable CRLM (B 3 cm) and a fair alternative for deepseated resectable CRLM that would otherwise require major hepatectomy [1,2,4]. Though advances in energy delivery in methods for precise probe placement and in ablation confirmation techniques have, often prematurely, been introduced as alleged improvements, our results underwrite technological progresses made over time. The improvement over time, in terms of LTPFS, especially for patients being treated with CT-guided percutaneous ablations, was the most remarkable finding in our study. OS did not significantly improve over the last 10 years. Whether this reflects an absent correlation between survival and local treatment failure, especially given the relative ease to repeat ablations, or the gradual acceptance to offer curative intent ablations to more complex cases with higher disease burden, remains unknown.
Results of this study compare well with OS and LTPFS data published in other recent series regarding thermal ablation of CRLM [1, 14, 35-39, 46, 47]. We have reached the point where the local tumor progression rate after percutaneous ablation has approached results following open ablation as well as following partial hepatectomy, as the most recent surgical series report R1/R2 rates varying from 12 to 46% [48][49][50][51][52]. Outcomes of this current cohort study are again underlining the necessity to conduct a randomized controlled trial comparing standard partial hepatectomy to its less invasive competitor thermal ablation for smaller-size resectable CRLM (B 3 cm). Although the phase III randomized LAVA trial (ISRCTN52040363) attempted to randomize high surgical risk CRLM patients to surgery or thermal ablation, recruitment feasibility was not established during the pilot stage, and therefore, the trial closed early without having gathered data regarding the primary endpoint two-year disease-free survival [53]. The interim results of the COLLISION trial (NCT03088150), presented at CIRSE 2021 and ECIO 2022, confirm thermal ablations' superior safety profile, shorter hospital stay, equal to superior local control and similar OS compared to partial hepatectomy; the final results are eagerly awaited [54,55]. Though a recent comparative analysis favored thermal ablation with regard to OS, LTPFS and eventual local control for small-size (B 3 cm), stereotactic body radiation therapy (SBRT) does challenge thermal ablation for intermediate-size (3-5 cm) CRLMs; the ongoing COLLISION-XL trial (NCT04081168) will hopefully provide clarity. [56].
Although speculative, the improvement over time, in terms of LTPFS, for patients being treated with percutaneous ablation should probably be contributed to (A) gained experience and (B) technological advancements made during the last decades. A multitude of minor improvements with regard to energy delivery spectrum, antenna and generator design (e.g., Thermosphere TM technology, multiple antennae systems or stereotactic navigation), anesthesia and breath-hold techniques, real-time image guidance (e.g., administration of intra-arterial contrast via an hepatic artery catheter) and the use of rigid and non-rigid image fusion and registration platforms allowing intraprocedural completion ablations seem to have led to this major quality improvement [6, 7, 22, 24-26, 31, 57-61].
Some limitations need to be addressed. The median follow-up period in the 2018-2021 cohort was sufficient (11.5 months), but inevitably lower compared to the earlier cohorts. This may have led to the situation where some patients in the latest cohort are still susceptible to developing LTP (immortality time bias), though this only applies to a small amount of tumors; as historically seen, the vast majority of LTPs are detected within the first 3-9 months following local treatment and a clear LTPFS plateau is reached after roughly 18 months follow-up (Fig. 2a) [9]. Reported study data were analyzed from prospectively kept records, and potential confounders were excluded by uniand multivariable analyses, which does not fully guarantee that residual confounding has been eliminated. The fact that periprocedural chemotherapy regimens and follow-up imaging protocols did not change over time decreases the likelihood for residual bias. The lack of a comparison between laparoscopic and open ablated tumors could be a potential limitation as in certain cases the laparoscopic approach might be superior to the open approach in terms of safety and length of hospital stay. Due to technological advancements in energy delivery and reduced procedure time, MWA was gradually favored over RFA, even though previously published data showed no significant difference in terms of local disease control [6,60,61]. Nonetheless, the ablation modality need to be addressed as potential confounder. In addition, the specific ablation devices used in this c Fig. 4 Kaplan-Meier survival curves indicating local tumor progression-free survival (LTPFS) per time frame and approach. A, B Analysis of open and percutaneous thermal ablation per treated tumor and per patient respectively, C and D patients treated with open ablation, analysis per treated tumor and per patient respectively, E and F patients treated with percutaneous ablation, analysis per treated tumor and per patient respectively. Numbers at risk correspond to either the amount of tumors or the number of patients. Overall comparison log-rank (Mantel-Cox) test is reported per graph. Death without local tumor progression (LTP) is censored (competing risk) study may render the comparative results as they do not necessarily represent all current day ablation systems. Although mutant RAS and BRAF status are known to be associated with LTP [47,62], these tumor characteristics were not routinely measured over the last decade, resulting in high rates of missing data. Furthermore, it should be noted that the national guideline recommendations not routinely offer neo-adjuvant or adjuvant chemotherapy for locally treatable disease, what differs from several other countries and regions, and hence, it may be challenging to compare our results with series where patients were routinely offered (neo-)adjuvant systemic therapy [41]. However, the national guideline recommendations did not change over time and were actually re-established following the recent publication of two clinical trials of which one showed no difference in OS for perioperative chemotherapy (EORTC 40983) [63] and one showed an inferior OS for adding adjuvant chemotherapy (JCOG 0603) [64] .
In conclusion, the efficacy of percutaneous ablations for CRLM in terms of local tumor progression-free survival has improved remarkably over the last 10 years and seems to have approached oncological outcomes following open ablations. Over the last decade, no differences were found regarding length of hospital stay, rate and severity of complications, and overall survival. Given its minimal invasive character and shorter length of hospital stay, whenever feasible, percutaneous procedures may be favored over an open approach.
Funding This study was not supported by any funding.

Declarations
Conflict of interest The authors declare that they have no conflict of interest.
Ethical Approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.
Informed Consent This study has obtained IRB approval from the Amsterdam UMC, location Vrije Universiteit (reference number 2021.0121) and the need for informed consent was waived.
Consent for Publication For this type of study, consent for publication is not required.

Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1007/s00270-022-03152-9. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons. org/licenses/by/4.0/.