Key words ablation procedures - MR-imaging - radiofrequency (RF) ablation
Introduction
Thermal ablation represents a locoregional treatment option for hepatocellular carcinoma
(HCC). Over the last decades this treatment option has gained popularity and has been
incorporated into various therapeutic regimens [1 ]. The Barcelona Clinic Liver Cancer (BCLC) guideline recommends thermal ablation
treatment for patients with HCC in the very early stage (BCLC 0) and early stage (BCLC
A). BCLC 0 corresponds to patients with a single HCC ≤ 2 cm. BCLC A includes patients
with a single HCC > 2 cm in size or up to 3 nodules ≤ 3 cm in size. Patients at these
two stages can potentially be cured with thermal ablation treatment [2 ]. Moreover, this treatment also makes it possible to limit disease progression in
patients awaiting transplantation [3 ]. Thermal ablation leads to the destruction of the tumor cells due to local hyperthermia
(heating tissue to at least 60 °C) [4 ]. The most commonly used thermal ablation modalities are radiofrequency (RFA) and
microwave ablation (MWA) [5 ]
[6 ]. The MWA technique represents a promising alternative to the well-established RFA
method. In this context, previous studies reported up to 45 % lower local tumor progression
(LTP) rates after MWA treatment compared to RFA [7 ]
[8 ]. Other advantages include larger ablation volumes, less interference from the heat-sink
effect, and a shorter duration of ablation [7 ]
[8 ].
Materials and Methods
Study Design
This prospective randomized clinical trial was approved by the local ethics committee
and written informed consent was provided by all patients in accordance with Health
Insurance Portability and Accountability Act (HIPAA) guidelines. The inclusion criteria
were as follows: (a) HCC diagnosed by histological and/or radiological examination;
(b) one thermal ablation treatment with MWA or RFA planned; (c) age over 18; (d) general
condition that allows magnetic resonance imaging (MRI) examination; (e) MRI examination
with 1.5 tesla or 3.0 tesla; (f) single lesion smaller than 5 cm; (g) up to 3 lesions,
each smaller than 3 cm; (h) no extrahepatic manifestation or vascular invasion. The
following patients were excluded: (a) general MRI contraindications; (b) pregnant
and breastfeeding women; (c) secondary carcinoma; (d) MRI contrast agent allergy;
(e) contraindication to thermal ablation treatment. All patients were randomly assigned
(1:1) to either the MWA group or the RFA group (25: 25) using a permuted block design
to ensure equal group sizes.
Patient Characteristics
Overall, 50 patients who met the inclusion criteria (38 males and 12 females, mean
age: 62.9 ± 10.5 years) were recruited for this study. The mean initial tumor diameter
was 23.3 ± 8.2 mm (range: 10.0–39.1 mm) in the MWA group and 20.7 ± 9.4 mm (range:
8.0–41.0 mm) in the RFA group. The mean total energy applied was 52.47 ± 19.92 kJ
(range: 14.4–98.4 kJ) in the MWA group and 194.24 ± 57.04 kJ (range: 84.0–348.0 kJ)
in the RFA group (p < 0.01). The distribution of tumor grading (G1, G2, G3) was comparable between the
MWA and RFA groups, recorded as (4/19/2) and (3/20/2), respectively (p > 0.05). Prior to study enrollment, 10 patients had received at least one thermal
ablation treatment and 9 patients had undergone partial hepatectomy (MWA/RFA, p = 0.46). 44 patients showed liver cirrhosis (MWA/RFA, p = 0.19) due to viral hepatitis in 32/44 cases (hepatitis C: 29, hepatitis B: 3),
non-alcoholic steatohepatitis (NASH) in 3/44 cases, and alcohol-related liver disease
(ARLD) in 9/44 cases. A total of 30 patients had undergone TACE therapy before ablation
(MWA = 17, RFA = 13, p = 0.38). Of the 50 patients, 28 were classified as BCLC A, with 15/28 belonging to
the MWA group and 13/28 to the RFA group. The mean initial tumor diameter in the MWA
subgroup was 28.6 ± 5.8 mm (range: 20.4–39.1 mm), while it was 28.2 ± 6.6 mm (range:
21.2–41.0 mm) in the RFA subgroup. There were no significant differences in these
parameters between the MWA and RFA groups (p > 0.05). Further characteristics of both groups were summarized in [Table 1 ].
Table 1
Patient characteristics.
Patient characteristics
RFA group
MWA group
Number of patients
25
25
Male/female
19/6
19/6
Age (years)*
62.7 ± 10.8
63.2 ± 10.3
HCC segment:
2
2/3
3
4
4/8
5
5/6
6
7
7/8
8
0
2
1
5
2
2
4
5
1
1
2
1
2
1
4
0
5
1
3
4
1
3
BCLC 0
12
10
BCLC A
13
15
Tumor grading (G1 / G2/G3)
3/20/2
4/19/2
HCV
15
14
HBV
2
1
NASH
1
2
Hepatic cirrhosis
24
20
Splenomegaly
7
4
Esophageal varices
2
3
Perihepatic ascites
13
2
Diabetes
3
2
Mean total energy applied (kJ)
194.24
52.47
Previous treatment
TACE
Partial hepatectomy
TACE and/or partial hepatectomy
Earlier ablation
13
6
15
4
17
3
18
6
Note: HCV = hepatitis C virus; HBV = hepatitis B virus; NASH = nonalcohol steatohepatitis;
TACE = transarterial chemoembolization; BCLC = Barcelona Clinic Liver Cancer
Measurements
MRI images were evaluated before ablation, 24 h after ablation, and subsequently after
a 3-, 6-, 9-, and 12-month follow-up. Data on LTP, overall survival (OS), disease-free
survival (DFS), and intrahepatic distant recurrence (IDR) were evaluated beyond the
first year by further regular follow-ups. Tumor volume and ablation area were calculated
by manual segmentation on each axial MRI slice. The surface area was multiplied by
the corresponding slice thickness to determine the volume. Measurements of the maximum
diameter, ADC value, and B50-SI were performed on axial slices. The ADC value and
B50-SI were measured in the tumor, as well as in the surrounding normal-appearing
liver tissue. The regions of interest (ROIs) were placed according to the small solid
sample method [9 ]. ROIs were placed freehand in the homogeneous marginal third of the tumor and the
surrounding homogeneous liver parenchyma to calculate tumor/normal-liver-parenchyma
ratios. The synchronization of the images allowed transfer of the ROIs from the B50
image to the exact same position of the corresponding ADC image. This method allowed
for precise measuring in the ADC image, even in cases where the HCC was difficult
to localize [10 ]. To reduce recall bias, measurements in the subsequent post-ablative MRI images
were taken at similar sections and locations as in the pre-ablative baseline image.
All measurements were performed by a trained investigator and supervised by a radiologist
with over 20 years of experience in interventional radiology.
Pre-Ablation Assessment and Ablation Procedure
Previous MRI images were studied by the interventional radiologist before treatment
in each patient to assess the anatomical position, size, and volume of the index tumor.
During the ablation treatment, clinical parameters, including blood pressure, electrocardiography,
and pulse oximetry measurements, were monitored. All ablations were performed under
CT planning and guidance (SOMATOM Definition AS, Siemens) [11 ]. A power of 200 W was applied for RFA. The MWA was conducted in three steps with
rising output powers (45–60 W, 65–80 W, and 85–100 W). Towards the end of treatment,
the puncture site of the inserted electrode was coagulated during retraction to prevent
tumor seeding or possible bleeding. The patients were subsequently observed and monitored
for the next 12 hours in the hospital. In the event of a deterioration of the patient's
condition or vital signs, a control CT was initiated to identify potential adverse
events (AE). Patients without symptoms and with normal vital signs were discharged.
Ablation Evaluation and Endpoints
Technical success was achieved when the ablation was conducted according to the recommended
manufacturer’s protocol and the ablation zone fully encompassed the index tumor on
the procedural CT examination [12 ]. During treatment, the extent of the ablation zone was checked by analyzing hypodense
tissue changes and small bubbles in the tumor area. Case examples are shown in [Fig. 1 ], [2 ]. Complete ablation was defined as a non-enhanced ablation zone completely covering
the index HCC on MRI. Technique efficacy has been accomplished when complete ablation
was present on the 24 h postprocedural, contrast-enhanced MRI examination [12 ]. AEs were categorized according to the Society of Interventional Radiology (SIR)
classification as mild, moderate, or severe [13 ]. Disease recurrences were monitored and assessed through regular follow-up appointments,
utilizing contrast-enhanced MRI. LTP was defined as the occurrence of a lesion connected
to the ablation zone. Time to local tumor progression (TLTP) was defined as the period
from the ablation date to LTP. The occurrence of new intrahepatic lesions with no
connection to the ablated area was described as IDR. DFS was the tumor-free period
starting from the ablation date. OS was calculated from the ablation date to the last
follow-up or death date.
Fig. 1 74-year-old male patient with HCC (arrow) in liver segment IVb with a diameter of
3.2 cm prior to RFA (a ). Axial CT image during the procedure with the radiofrequency applicator placed in
the HCC region (b ). The ablation duration was 18 min at an output power of 200 W. Post-ablation T1-weighted
MRI shows the non-enhanced ablation zone completely covering the index HCC (c ). Contrast-enhanced T1-weighted MRI 18 months after ablation shows a homogeneous
hypointense ablation zone without signs of a residual tumor (d ).
Fig. 2 T2-weighted MRI of 77-year-old male patient with HCC (arrow) in liver segment VIII
(a ). Axial CT image during ablation demonstrates the MW antenna in place (b ). Ablation was performed with the following power settings: 45 W for 1 min, 65 W
for 12 min, and 80 W for 3 min. Post-ablation axial MRI shows the ablation zone with
central necrosis (c ). Contrast-enhanced T1-weighted MRI 18 months after ablation shows LTP as a hypervascular
lesion that has developed in the cranial part of the ablation zone (arrow) (d ).
Technical Features
All ablation treatments were done using a dual-ablation system (Amica, MMS Medicor
Medical Supplies GmbH, Germany). The microwave applicator contained a mini-choked
coaxial antenna, whereas the radiofrequency applicator was monopolar. RFA included
output frequencies of 450 kHz and an output power of 200 W. MWA used output frequencies
of 2450 MHz and had a maximum output power of 140 W. The dual-ablation system included
internal water cooling to avoid shaft overheating.
Imaging Protocol
The standard imaging protocol comprised the following sequences: unenhanced and contrast-enhanced
T1- and T2-weighted MRI scans with a 1.5-T system or 3-T system with a 5-mm transverse
section thickness. The applied sequences included diffusion-transverse, EP-2D-Diff
(b50, b400, b800) HASTE, in- and opposed phase, TSE, FLASH-2D and contrast-enhanced
FLASH 2D dynamic phase.
Statistical Analysis
Statistical analysis was performed using Bias (Bias for Windows, version 11.06; Germany).
The measured values were represented as mean, standard deviation (SD), range, and
median. Categorical data were shown as counts and percentages. Baseline characteristics
were assessed with chi-square or fisher exact test, where appropriate. The Shapiro-Wilk
test was used to assess the normality assumptions of the collected data. Quantitative
data was then analyzed with Student’s t-tests or Mann-Whitney-U-tests. To compare
survival and recurrence data between both treatment groups, the Kaplan-Meier method
and log-rank tests were applied. The multivariate Cox proportional hazard regression
model was used to analyze the significance of parameters influencing DFS and OS. The
results were expressed as relative hazard = exp(coef) with 95 % confidence interval.
For all analyses, a p < 0.05 was considered significant.
Results
The ablation treatments of both methods had a technical success and a technique efficacy
rate of 100 % (50/50, p = 1.00). Neither moderate nor severe AEs were documented. Mild AEs occurred in a
total of six cases (5/6 minimal pleural effusions; 1/6 minimal pneumothorax), with
no significant difference between the two groups (p > 0.05). There were no significant differences in pre-ablation tumor volumes and
diameter among both groups (volume: p = 0.11, diameters: p = 0.20). The temporal progress of the ablation diameter is described in [Table 2 ]. Corresponding boxplots of the volume progress are displayed in [Fig. 3a ]. The study showed significantly larger ablation volumes for MWA (66.5 ± 35.8 cm3 ) compared to RFA (29.2 ± 22.2 cm3 ) on 24 h postprocedural imaging (p < 0.01). Further analysis showed that the index tumor diameter in the RFA group had
a significant influence on the OS (p = 0.04, rel. hazard: 1.07). In the MWA group, however, the influence was not significant
(p = 0.74, rel. hazard: 1.02).
Table 2
Diameter comparisons of HCC lesions and ablation zones between the MWA and RFA group.
Diameter (mm)
Preablation
24 hours
3 months
6 months
9 months
12 months
MWA
Mean
SD
Median
Range
23.3
8.2
22.5
10.0–39.1
49.7
9.9
51.9
26.4–74.2
42.4
7.3
42.7
28.3–58.4
37.1
8.7
36.1
21.8–58.8
36.8
7.1
36.8
25.9–53.5
34.8
8.6
34.8
15.5–51.7
RFA
Mean
SD
Median
Range
20.7
9.5
21.2
8.0–41.0
38.5
9.6
37.4
22.0–65.0
33.3
10.5
28.0
13.5–56.0
29.1
10.6
25.4
10.9–53.8
26.1
8.9
23.7
10.0–40.8
25.0
9.2
22.0
9.2–39.5
Fig. 3 The boxplots (MWA = black, RFA = grey) show the measured volume of the index HCC
(pre-ablation) and the subsequent ablation volume (a ) and the ADC ratios (lesion/normal liver parenchyma) from both groups (MWA = black;
RFA = grey) over time (b ) during follow-up.
The measured ADC value and B50-SI are presented in [Table 3 ], [4 ]. There were significant differences between MWA and RFA 24 h post-ablation imaging
(ADC ratio: p = 0.01; B50-SI ratio: p = 0.01). The mean ADC ratio ([Fig. 3b ]) shows an increase from pre-ablation to 12 months post-ablation for both ablation
methods (MWA: 1.27 to 1.41; RFA: 1.11 to 1.42). Conversely, the mean B50-ratio shows
a decrease (MWA: 1.99 to 1.6; RFA: 1.51 to 1.39). The ADC change from pre- to post-ablation
had no correlation to the LTP (p = 0.49) or OS (p = 0.43).
Table 3
ADC ratio progress and comparison between the MWA and RFA groups.
ADC value ratio
Pre-ablation
24 hours
3 months
6 months
9 months
12 months
MWA
Mean
SD
Median
Range
1.28
0.44
1.16
0.8–2.4
1.52
0.48
1.38
1.0–2.8
1.37
0.43
1.84
0.7–2.5
1.48
0.29
1.39
1.0–2.2
1.36
0.36
1.30
0.9–1.9
1.41
0.41
1.31
0.9–2.3
RFA
Mean
SD
Median
Range
1.11
0.42
1.07
0.4–2.2
1.23
0.35
1.15
0.8–2.3
1.41
0.28
1.49
0.7–1.7
1.38
0.30
1.29
0.9–2.0
1.33
0.35
1.42
0.9–1.9
1.42
0.38
1.47
0.7–1.9
Note: The ratio was calculated from the ADC value of the HCC (on pre-ablation image)
or the ablation zone (on post-ablation images) and normal liver parenchyma.
Table 4
B-SI ratio progress and comparison between the MWA and RFA groups.
B-value ratio
Pre-ablation
24 hours
3 months
6 months
9 months
12 months
MWA
Mean
SD
Median
Range
1.99
1.16
1.60
0.9–5.8
2.05
1.19
1.57
0.8–5.9
1.74
0.82
1.48
0.7–3.7
1.73
0.90
1.44
1.0–4.9
1.75
0.78
1.54
1.0–3.9
1.60
0.77
1.22
0.9–3.5
RFA
Mean
SD
Median
Range
1.51
0.54
1.38
0.6–2.8
1.31
0.48
1.18
0.3–2.3
1.44
0.34
1.42
0.8–2.0
1.28
0.44
1.18
0.6–2.3
1.23
0.32
1.29
0.7–1.8
1.39
0.54
1.32
0.9–2.7
Note: The ratio was calculated from the B-value of the HCC (on pre-ablation image)
or the ablation zone (on post-ablation images) and normal appearing liver parenchyma.
LTP within 2 years of the treatment date was 4/25 (16 %) for RFA and 1/25 (4 %) for
MWA (p = 0.056). The mean TLTP in the RFA group was 5.7 ± 4.5 months and the local progression
in the MWA group occurred after 17.8 months (p = 0.06, Log-rank test).
IDR within 2 years occurred in the RFA group in 14/25 (56 %) and in the MWA group
in 8/25 (32 %). For the DFS, the Kaplan-Meier method revealed a significant advantage
in the MWA group (p = 0.01) ([Fig. 4a ]). The OS rate showed no significant differences between the two procedures (p ≥ 0.14) ([Fig. 4b ]). The mean follow-up time after the ablation was 42.7 and 34.2 months for the MWA
group and the RFA group, respectively (p = 0.15). The 1, 2, and 3-year post-ablation OS rates were 100 %, 80 %, 72 % for MWA
and 72 %, 64 %, 60 % for RFA, respectively (p ≥ 0.14). In the subgroup of 28 BCLC Stage A patients, the overall survival was statistically
significant favoring MWA over RFA (p = 0.037, Log-rank test) ([Fig. 5a ]). The relative hazard ratio for MWA compared to RFA was 0.302 (95 % CI: 0.0974 to
0.9363). Additionally, in this subgroup, disease-free survival was also significantly
better for MWA (p < 0.01, Log-rank test), with a relative hazard ratio of 0.232 (95 % CI: 0.0832 to
0.6442) ([Fig. 5b ].).
Fig. 4 Kaplan-Meier curves show the disease-free survival data (MWA = black; RFA = grey)
(a ), the OS in months with numbers at risk shown below (MWA = black; RFA = grey) (b ), and the Kaplan-Meier curves of MWA and RFA with corresponding 95 % confidence intervals
(p = 0.95) (c ) (d ).
Fig. 5 Kaplan-Meier curves show the OS (a ) and DFS (b ) data of the BCLC A subgroup in months (MWA = black; RFA = grey) with numbers at
risk shown below.
Discussion
The ablation volumes of MWA-treated lesions were significantly larger than those treated
with RFA. This result could be explained by the lower heat-sink effect and a different
mechanism of energy deposition using MWA [5 ]
[14 ]. However, the rapid enlargement of the ablation zone in MWA was also reported as
a potential disadvantage, as the risk for damaging adjacent structures increased [15 ]. In this context, there were no significant differences between the ablation methods.
This underlines the assumption that MWA can be considered as safe as RFA [16 ]
[17 ]. Confirming the results of Ding et al., less IDR was registered in the MWA group
[18 ]. The lower occurrence of IDR might result from the significantly larger ablation
volumes created by MWA [19 ]. Therefore, neoplastic cells located in the surrounding area of the HCC lesion were
more likely to be ablated by MWA than RFA. In this study, DFS was significantly higher
for MWA compared to RFA, which is similar to the findings of Liu et al. [20 ]. It also strengthens the results of a meta-analysis by Facciorusso et al. [19 ]. Moreover, MWA showed favorable results for larger tumor sizes (≥ 2.5 cm) regarding
the LTP rate [8 ]
[21 ]. Glassberg et al. has stated that larger ablation zones of MWA might destroy neoplastic
cells more effectively and could have an impact on the LTP [8 ]. For smaller sized tumors (< 2.5 cm), the meta-analysis showed no significant difference
[8 ]. The mean tumor diameter in this study was 2.2 ± 0.9 cm and although the number
of LTP was lower in the MWA group, the difference was not statistically significant
(p = 0.06). This is consistent with the findings of Vietti Violi et al., who reported
no significant difference in local tumor progression or overall survival between the
two techniques when treating lesions that were 4 cm or smaller, over a 2-year follow-up
period [22 ]. The OS during the follow-up period was slightly longer in the MWA group, but similar
to other studies. No statistically significant differences were observed [8 ]
[21 ]
[22 ]
[23 ]. Liu et al. showed a better OS for MWA during a 5-year follow-up period [20 ]. In the subgroup of 28 BCLC Stage A patients, MWA showed statistically significant
benefits in both OS and DFS over RFA (p < 0.05, Log-rank test). These findings offer additional insight, given the limited
number of comparable subgroup analyses for BCLC Stage A patients in the current literature.
In general, the size of the index HCC was an important factor for the patient’s outcome
[24 ]. A meta-analysis by Casadei Gardini et al. demonstrated that HCC lesions smaller
than 3 cm showed no significant differences for OS rates [25 ]. In this context, this study showed that the OS might be influenced by the tumor
size in the RFA group (p = 0.04) but not in the MWA group (p = 0.74). Furthermore, ADC values in both groups were assessed for tissue characterization.
Studies examining the diffusion characteristics of tumors showed that necrotic areas
tend to have higher ADC values [26 ]
[27 ]. Other studies investigating TACE treatment indicated that ADC parameters might
be useful for the assessment of an early therapeutic response [28 ]
[29 ]
[30 ]. An increasing ADC signal correlated with successful TACE treatment and influenced
OS and DFS. In this study, both ablation techniques resulted in an increase in ADC
values (p < 0.05). Unlike the mentioned TACE-related studies, the prediction analysis revealed
no correlation between ADC changes, LTP, and OS.
The results of this study must be interpreted with caution and several limitations
merit consideration. The sample size of 25 patients per ablation study group was relatively
small. The assessment did not consider preexisting conditions of the patients, segmental
location of the HCC or its proximity to large vessels as the subgroups were too small
for statistical analysis. In a study conducted by Abe et al. [31 ], it was found that lesions located in liver segment 8 may be associated with a worse
prognosis.
In conclusion, both techniques are comparably effective in treating small- to medium-sized
HCC patients. No moderate or severe AEs were registered, and there were no statistically
significant differences in terms of LTP and OS between the two groups. While MWA demonstrated
a tendency towards better local tumor progression (LTP), it also showed a trend of
being superior to RFA regarding DFS with a lower rate of IDR. In a limited subgroup
of 28 BCLC A patients, a cautious yet statistically significant advantage for MWA
was observed in terms of both overall and disease-free survival (p < 0.05). However, due to the small sample size of both the general (n = 50) and sub-group populations (n = 28), these findings should be considered as preliminary indicators rather than conclusive
evidence. Further studies with larger sample sizes are needed to confirm these results.
Clinical relevance
The clinical relevance of this study lies in its indication that both MWA and RFA
are safe and effective for treating small- to medium-sized HCC, with MWA showing a
suggestive trend towards better LTP and DFS, thus warranting further investigation
in larger studies.
