CT-monitored minimal ablative margin control in single-session microwave ablation of liver tumors: an effective strategy for local tumor control

Percutaneous thermal ablation therapy is one of major treatment options both for primary and metastatic liver tumors, such as hepatocellular carcinoma (HCC) and colorectal liver metastasis (CRLM). Radiofrequency ablation (RFA) is the most widely studied technique, while microwave ablation (MWA) is increasingly used as it allows larger ablation zone and shorter ablation time [1]. In ablation therapy, local tumor control is the primary goal which may improve long-term outcomes of patients. Accumulated evidence support that, for effective local tumor control, not only complete destruction of visible tumor on imaging, but creating a sufficient ablative margin in all dimensions is critical [2‐5]. For now, widely accepted criteria for minimal ablative margin are ≥ 5 mm for HCC and ≥ 10 mm for CRLM, which have been reported to significantly lower local tumor progression (LTP) rates [6, 7].

Technical success of ablation treatment is routinely assessed on post-ablation contrast-enhanced CT (CECT) or MRI [8, 9], and repeat ablation session is performed based on the imaging results. However, considering the multi-step process from the initial MWA session to repeat treatment session, intra-procedural monitoring of treatment efficacy followed by additional ablation all within the same initial session has the potential to improve efficiency and patient convenience. In ablation therapy of liver tumors, B-mode ultrasound (US) is commonly used both for guiding and monitoring of the procedure. However, it has limitations as hyperechoic cloud of gas bubbles generated by ablation and the accompanying posterior acoustic shadow make accurate assessment difficult. In this regard, intra-procedural CECT can be a promising tool. With in-suite CECT with multi-planar reconstruction (MPR), 3D ablative margin status can be evaluated by comparing post-ablation images to pre-treatment images, which would not only help define tumors needing additional ablation, but also pinpoint exactly the margin where additional overlapping ablation may be required [10]. Until now, little is known about the added value of in-suite CECT imaging during liver tumor ablation in comparison to the conventional process which only includes post-ablation imaging.

This study aimed to investigate the usefulness of intra-procedural CECT-monitored minimal ablative margin control in MWA for successful ablation treatment of liver tumors.

This retrospective study was approved by the institutional review board, and met the requirements of the Declaration of Helsinki. The requirement of informed consent was waived.

A total of 334 liver tumors in 172 patients were finally included in the analysis which were treated in 222 independent MWA sessions (Fig. 1). Index liver tumors included 240 HCCs and 94 CRLMs, of which 73 were perivascular and 235 were subcapsular, and although the majority of tumors were less than 30 mm in diameter, there were 41 tumors exceeding 30 mm. Consistent with the general indications for MWA, the tumors included in this study were all less than 50 mm with one exception, whereas one was larger than 50 mm on a retrospective measure (i.e., 57 mm). Patient and tumor characteristics are described in Table 1. There were 5 minor complications, including 3 small peri-hepatic hematomas, one transient pneumothorax, and one sub-segmental infarct.

In local ablation treatment of liver tumors, intra-procedural imaging for margin monitoring could speed up the decision-making process of repeat ablation and thereby improve workflow. Our study demonstrated that intra-procedural CECT frequently directed within-session additional overlapping ablation (18.9% of tumors) by detecting suboptimal margins after MWA, ultimately yielding complete ablation coverage in 99.4% and sufficient minimal ablative margin (≥ 5 mm for HCC and ≥ 10 mm for CRLM) in 77.5% of tumors. This strategy provided good local tumor control outcomes, with an estimated 6-month, 1-year, and 2-year LTP rates of 3.2%, 7.5%, and 12.9%, respectively. Importantly, this outcome can be achieved in a single session, avoiding multi-session treatment courses.

Although other advanced imaging techniques such as real-time fusion imaging and contrast-enhanced US are being used for intra-procedural assessment of ablative margin [13], intra-procedural CT may have added advantages as it is known to be useful for guiding the ablation of tumors in challenging anatomical locations [14, 15]. In other words, the effectiveness of intra-procedural CT includes not only margin assessment, but also its ability to enable an accurate execution of additional ablation determined according to the assessment. Thus, in our study, the estimated frequency of tumors with sufficient margin at post-MWA MRI increased from 61.4 to 77.5%, quite high in comparison to prior studies for HCC [16] or CRLM [17], therefore supporting the concept of immediate CECT assessment during the procedure. According to our study results, tumors of larger size and subcapsular location showed higher frequency of insufficient ablative margin after an expected ablation completion, therefore requiring additional intra-session ablation. These results are in line with the general perception that larger tumor size and subcapsular location were factors that rendered percutaneous ablations technically more difficult. While MWA may have an advantage over RFA in treating larger tumors [18], full coverage of a large index tumor plus a sufficient 3D margin would still be challenging even with multiple probes. For subcapsular tumors, accurate tumor targeting can be difficult due to poor visibility by real-time US, and the probe location and energy application may need to be adjusted to prevent damage to adjacent structures [19]. Therefore, the use of intra-procedural CECT to confirm sufficient minimal ablative margin should be more actively considered for such technically challenging tumors.

In our study, local tumor control (i.e., 1-year estimated LTP rates of 7.5%) with single-session CECT-monitored MWA was comparable or superior to recent studies of MWA in liver malignancies [20‐22]. In addition, our study confirmed ablative margin status as the most important independent predictor for LTP which highlights the importance of achieving sufficient minimal margin [13]. While the concept is not new, it is challenging to widely adopt due to the inherent difficulty in accurately assessing the ablative margin status by imaging during the ablation itself. At the very least a CECT would provide the best ablation zone boundary definition, but the index tumor boundary is no longer visible within the ablation zone. Therefore, the conventional method of visual registration of pre- and post-treatment images can be challenging and time-intensive. Recently, a software-based registration method has become available and was used based on operator discretion in our study population. As recent studies have shown the value of software-based registration for more sensitive detection of suboptimal margins [5, 16, 23], the full effect of software-based registration methods needs to be further tested in a prospective study.

It is notable that patients who ultimately did reach sufficient minimal margin, based on immediate MRI assessment, had very low rates of LTP (1.0%, 2.1%, and 6.9% at 6 months, 1 year, and 2 years). This low rate of LTP is rare in the published literature on percutaneous ablation of liver tumors, with the exception of a recent report using the technique of “no-touch” RFA, which ensures circumferential margin ablation with 3 electrodes enclosing a tumor but does not violate the tumor itself during the procedure [24]. In their series of “no-touch” RFA, LTP was virtually non-existent; however, their study population only included HCC averaging 1.6 cm and no more than 2.5 cm, therefore a super-selected group. For larger tumors and for tumors in subcapsular and perivascular locations, such enclosure techniques are often not technically applicable [24]. On the contrary, in our series, we had larger tumors (21.6% [72/334] exceeding 25 mm; including 41 tumors exceeding 30 mm), large number of subcapsular (235/334) and perivascular (73/344) tumors, and substantial number of CRLM (94/344). Therefore, we believe that in the end, for all tumors regardless of size, location, and histology, the ultimate equalizer is the pursuit of adequate minimal margins irrespective of ablation device, relying on operator skill to optimally position the applicators, and intra-procedural assessment to ensure full execution of the intended plan. In addition to ablative margin status, larger tumor size and perivascular tumor location were independent factors of LTP which were well correlated with prior studies of ablation treatment for liver malignancies [22, 25‐28]. Of note, however, these features were not as strong an indicator of poor outcome as margin status, and the overall LTP rates in our study cohort despite including such tumors were very low, indicative of the power of a local tumor ablation strategy that primarily aims for true sufficient minimal ablation margins, while taking effective measures to compensate for larger tumor size and perivascular locations.

Our study had several limitations. First, as this study was of a retrospective design, the MWA protocol to determine additional ablation was not strictly controlled. Although there was general consensus between operators on a criterion of sufficient minimal ablative margin, i.e., ≥ 5 mm for HCC and ≥ 10 mm for CRLM, whether or not to perform additional ablation for each tumor was subjectively decided by considering margin status as well as other factors such as the risk of complication or expected efficacy of additional ablation. Prospective studies using detailed study protocols are needed to accurately validate the value of intra-procedural CECT. Second, in our study, post-MWA MRI was used as a reference standard for margin status, while intra-session additional ablation was determined based on CECT which may raise the concern about potential discrepancies for margin assessment depending on imaging modalities. However, our approach can be justified by the following reasons: CT can be applied quickly and conveniently for an intra-procedural use, and MRI is superior to CT in discriminating between index tumor and ablation margin [29]. Lastly, regarding treatment outcomes, only per-tumor results of LTP were analyzed in our study. Future studies with long-term follow-up data should reveal per-patient outcomes such as overall survival.

In conclusion, in MWA of liver tumors, achieving sufficient minimal ablation margins is highly predictive of local tumor control. Intra-procedural CECT monitoring of minimal ablative margin facilitates identification of potentially suboptimal minimal margins and guides immediate additional intra-session ablation, therefore optimizing local tumor control with a single-session treatment.