Introduction
Hepatocellular carcinoma (HCC) is the most common hepatic malignancy, accounting for
approximately 75% of liver cancers globally, with poor prognosis.[1]
[2]
[3] Surgical resection and transplant remain the gold standard for HCC given the high
curative potential; for unresectable HCC, locoregional therapies (ablation, transarterial
chemoembolization [TACE], radioembolization) can be performed for curative intent,
palliation, or bridge to transplant.[2]
[4] Specifically, for patients with Barcelona Classification of Liver Cancer (BCLC)
0 or A, surgical resection or percutaneous ablation are preferred, depending on transplant
candidacy and lesion size.[2] Percutaneous ablation offers comparable overall survival to resection for lesions
≤ 3 cm while minimizing invasiveness and cost, with microwave ablation (MWA) favored
over radiofrequency ablation (RFA) for lesions ≤ 4 cm given its ability to achieve
better tumor necrosis.[2]
[5]
[6]
Although measures to reduce the risk of off-target ablation and injury have been described,
patients with unresectable HCC may be poor candidates for ablation if lesions are
in difficult locations to execute a percutaneous approach (e.g., caudate lobe or adjacent
to major vessels, the hilum, hepatic dome, biliary structures, heart, or bowel).[4]
[7]
[8]
[9]
[10]
[11]
[12]
The 2022 BCLC guidelines names transarterial radioembolization with yttrium-90 (Y90)
as an effective alternative to TACE and ablation in early- and intermediate-stage
HCC.[2]
[13]
[14]
[15] Furthermore, recent studies indicate Y90 radiation segmentectomy (Y90-RS) may have
comparable outcomes to MWA for early-stage HCC.[6]
[7]
[16] Herein, we compare the safety and efficacy of Y90-RS and MWA when performed with
curative intent for HCC lesions in locations otherwise suboptimal for MWA.
Materials and Methods
Patient Selection and Inclusion Criteria
Institutional Board Review approval was obtained for this single-center retrospective
study of patients who underwent Y90-RS or MWA (±prior TACE) for HCC with curative
intent between January 2014 and July 2019. Treatment decisions for HCC patients at
our institution were made with multidisciplinary consensus at a weekly tumor board
involving interventional radiology, hepatology, oncology, and transplant surgery.
Patients in both groups had lesions in suboptimal locations for percutaneous ablation,
defined as lesions within 5 mm of the liver dome, capsule, hilum, or gallbladder;
locations where needle placement or achieving sufficient margins may be difficult.[17]
[18] Those with lesions > 3 cm, vascular invasion of the tumor, extrahepatic metastases,
portal venous thrombosis, prior resection, or orthotopic liver transplant were excluded.
Y90 radioembolization was performed by two interventional radiologists with institutional
authorized user status. All other procedures were conducted by one of five interventional
radiologists with 1 to 15 years of experience. Y90-RS dosimetry and treatment protocol
using glass microspheres TheraSphere (Boston Scientific Corp., Marlborough, Massachusetts,
United States) have been previously detailed.[9]
[16]
[19]
[20] Dose was calculated using the Medical Internal Radiation Dosimetry dosing model
for perused volume. Whole lobar volume was used for dose calculation with a target
of 120 Gy and subsequently delivered selectively into the segment.[9] Dose to perfused volume was sufficiently higher (> 400 Gy). MWA probes (MicroThermX
Varian Medical Systems, Austin, Texas, United States) were placed percutaneously under
computed tomography (CT) guidance (±ultrasound [US]) to obtain a 5- to 10-mm circumferential
margin. Examples for treatment and follow-up imaging are provided for Y90-RS ([Fig. 1]) and MWA ([Fig. 2]).
Fig. 1 Example case of yttrium-90 (Y90)-radiation segmentectomy (RS). (A) Axial postcontrast T1-weighted image demonstrates a rounded, 1.8-cm arterial phase
enhancing lesion in segment 5/1 with washout (LR-5). (B) Transradial angiogram demonstrates an enhancing mass in segment 5/1. (C) Post-Y90-RS contrast-enhanced axial computed tomography (CT) demonstrating the treatment
zone of 2.9 cm diameter with no viable tumor.
Fig. 2 Example case of microwave ablation (MWA). (A) Pretreatment axial contrast-enhanced computed tomography (CT) demonstrates Liver
Imaging Reporting and Data System (LI-RADS) 5 lesion measuring 1.7 × 1.5 in segment
1 close to the hilum of the liver. (B) Intraprocedural image of MWA probe placement in segment 1 lesion. (C) Follow-up axial contrast-enhanced T1 magnetic resonance imaging (MRI) demonstrating
necrosis of the tumor and infarction of the left lobe of the liver.
Neither temperature monitoring nor hydrodissection were performed prior to MWA. The
tract was coagulated while removing the thermal probe, followed by CT to rule out
complications.
Treatment Response and Toxicity Assessment
Follow-up visits and imaging were performed 6 weeks postprocedure and at 3-month intervals
thereafter. Tumor response and progression was assessed by independent subspecialty
trained abdominal radiologists with > 10 years of experience on 3-month follow-up
imaging per the modified Response Evaluation Criteria in Solid Tumors algorithm.[21] Time to progression (TTP) was recorded for all patients during the study period.
Liver function tests and treatment complications were assessed at 1-month postprocedure.
Complications were characterized per the Common Terminology Criteria for Adverse Events
(CTCAE).[22] The threshold for minor laboratory events was ≤ grade 3.
Statistical Analysis
Fischer's exact test was performed for categorical variables, and Student's t-tests for nominal variables. The threshold of significance was p < 0.05. All statistical analyses were conducted using SPSS version 25 (IBM, Armonk,
New York, United States).
Results
Patient Demographics and Characteristics
Fifty patients with 53 lesions were included; 23 lesions in 20 patients received Y90-RS
and 30 lesions in 30 patients received MWA. Baseline demographics and disease characteristics
are displayed in [Table 1]. There were no significant differences in demographics between groups (p > 0.05). All patients were Eastern Cooperative Oncology Group performance status
0 to 1. Of the MWA cohort, 19 were Child-Pugh class A, 5 were B, and 6 were C and
the mean pretreatment laboratory values were as follows: Model for End-stage Liver
Disease sodium (MELD-Na) 12.7 ± 4.6, alpha-fetoprotein (AFP) 848 ± 3168.0, aspartate
aminotransferase (AST) 71.9 ± 49.1, alanine aminotransferase (ALT) 48.0 ± 32.4, and
total bilirubin 2.4 ± 2.7. Of the Y90-RS cohort, 15 were Child-Pugh class A, 4 were
B, and 1 was C and pretreatment laboratory values were as follows: MELD-Na 10.5 ± 3.3
(Y90-RS), AFP 762.2 ± 1793.8 (Y90), AST 50.3 ± 30.5 (Y90), ALT 30.1 ± 16.9 (Y90),
and total bilirubin 1.6 ± 1.1 (Y90). The MWA group had significantly higher pretreatment
transaminases (p < 0.05) and the Y90-RS group had significantly larger tumor diameter; mean tumor
diameter was 2.9 ± 1.0 in those treated with Y90-RS and 2.3 ± 0.9 for MWA (p < 0.05). Lesions were located adjacent to the following structures: dome (n = 22), capsule (n = 16), hilum (n = 9), and gallbladder (n = 6). There was no difference in the distribution of lesions ([Table 1], p > 0.05). Nineteen (63%) MWA patients were treated in combination with prior conventional
TACE.
Table 1
Patient demographics and clinical status before treatment initiation
|
Characteristic
|
All (n = 50)
|
MWA (n = 30), TACE/MWA (n = 19), MWA alone (n = 11)
|
Y90 (n = 20)
|
p-Value
|
|
Demographics
|
|
Sex
|
|
Male
|
31 (62)
|
18 (63)
|
13 (65)
|
0.13
|
|
Female
|
19 (38)
|
12 (27)
|
7 (35)
|
|
Age
|
|
Mean
|
64.5
|
62.5
|
67.4
|
0.1
|
|
Range
|
27–86
|
27–80
|
53–86
|
|
Disease (at treatment)
|
|
Child-Pugh
|
|
A
|
34 (68)
|
19 (63)
|
15 (76)
|
0.33
|
|
B
|
9 (18)
|
5 (17)
|
4 (19)
|
|
C
|
7 (14)
|
6 (20)
|
1 (5)
|
|
MELD-Na score
|
|
Mean (SD
|
11.8 (4.2)
|
12.7 (4.6)
|
10.5 (3.3)
|
0.09
|
|
Range
|
6.0–21.0
|
6.0–21.0
|
6.0–17.0
|
|
Serum AFP
|
|
Mean (SD)
|
807.8 (2864.0)
|
848.8 (3618.0)
|
762.2 (1793.8)
|
0.93
|
|
Range
|
1.4–16217.0
|
1.4–16217.0
|
3.0–6175.5
|
|
AST
|
|
Mean (SD)
|
64.2 (43.1)
|
71.9 (49.1)
|
50.3 (30.5)
|
0.03
|
|
Range
|
16.0–202.0
|
18.0–202.0
|
16.0–131.0
|
|
ALT
|
|
Mean (SD)
|
40.8 (28.3)
|
48.0 (32.4)
|
30.1 (16.9)
|
0.04
|
|
Range
|
7.4–157.0
|
16.0–157.0
|
7.4–62.0
|
|
Total bilirubin
|
|
Mean (SD)
|
2.0 (2.2)
|
2.4 (2.7)
|
1.6 (1.1)
|
0.25
|
|
Range
|
0.2–13.0
|
0.3–13.0
|
0.2–4.8
|
|
Lesions, n
|
53
|
30
|
23
|
|
Location, n
|
|
Capsule
|
16 (30)
|
11 (37)
|
5 (22)
|
0.23
|
|
Dome
|
22 (42)
|
10 (33)
|
12 (52)
|
|
Hilum
|
9 (17)
|
4 (13)
|
5 (22)
|
|
Gallbladder
|
6 (11)
|
5 (17)
|
1 (4)
|
|
Tumor diameter, cm
|
|
Mean (SD)
|
2.6 (0.9)
|
2.3 (0.9)
|
2.9 (1.0)
|
0.01
|
|
Range
|
1.0–5.0
|
1.0–4.5
|
1.2–5.0
|
Abbreviations: AFT, α-fetoprotein; ALT, alanine aminotransferase; AST, aspartate aminotransferase;
MELD-Na, Model for End-stage Liver Disease-sodium; MWA, microwave ablation; SD, standard
deviation; TACE, transarterial chemoembolization; Y90, yttrium-90.
Note: Parenthetical values are percentages unless otherwise indicated.
Posttreatment Outcomes
Posttreatment outcomes are displayed in [Table 2]. The mean follow-up was 17.6 months following Y90-RS, and 14.7 months following
MWA (p = 0.65). At 3-month follow-up, 22/23 (96%) of Y90-RS treated lesions achieved complete
response (CR) versus 23/30 (77%) of MWA treated lesions (p = 0.05). One (4%) Y90-RS patient demonstrated stable disease and 7 (23%) MWA patients
demonstrated partial response. During the study period, disease progression occurred
in 4 (17%) patients treated with Y90-RS versus 7 (23%) treated with MWA (p > 0.05). TTP was 23.5 months following Y90-RS and 6.7 months following MWA (p < 0.0001).
Table 2
Tumor response and complications
|
Characteristic
|
All (n = 53)
|
MWA (n = 30), TACE/MWA (n = 19), MWA alone (n = 11)
|
Y90 (n = 23)
|
p-Value
|
|
Response (mRECIST)
|
|
CR
|
45 (85)
|
23 (77)
|
22 (96)
|
0.05
|
|
PR
|
7 (13)
|
7 (23)
|
0 (0)
|
|
SD
|
1 (2)
|
0 (0)
|
1 (4)
|
|
PD
|
0 (0)
|
0 (0)
|
0 (0)
|
|
Total, n
|
10 (19)
|
10 (33)
|
0 (0)
|
|
Minor, n
|
3 (6)
|
3 (10)
|
0 (0)
|
|
Wound infection
|
1 (2)
|
1 (3)
|
0 (0)
|
|
Abdominal pain
|
1 (2)
|
1 (3)
|
0 (0)
|
|
Nausea
|
1 (2)
|
1 (3)
|
0 (0)
|
|
Major, n
|
7 (13)
|
7 (23)
|
0 (0)
|
|
Arterioportal fistula
|
1 (2)
|
1 (3)
|
0 (0)
|
|
Pneumothorax
|
1 (2)
|
1 (3)
|
0 (0)
|
|
Liver infarction
|
2 (4)
|
2 (7)
|
0 (0)
|
|
Capsular burn
|
1 (2)
|
1 (3)
|
0 (0)
|
|
Rectus sheath hematoma
|
1 (2)
|
1 (3)
|
0 (0)
|
|
Hepatic artery/portal vein injury
|
1 (2)
|
1 (3)
|
0 (0)
|
|
Time to progression
|
|
n
|
11 (20)
|
7 (23)
|
4 (17)
|
0.6
|
|
Mean TTP, mo
|
12.8
|
6.7
|
23.5
|
< 0.0001
|
|
Follow-up
|
|
Mean, mo
|
15.6
|
14.7
|
17.6
|
0.65
|
|
Range
|
1.0–53.0
|
1.0–53.0
|
1.0–47.0
|
Abbreviations: CR, complete response; mRECIST, modified Response Evaluation Criteria
in Solid Tumors; MWA, microwave ablation; PD, progressive disease; PR, partial response;
SD, stable disease; TACE, transarterial chemoembolization; TTP, time to progression;
Y90, yttrium-90.
Note: Parenthetical values are percentages unless otherwise indicated. Complications
were categorized by the CTCAE (Common Terminology Criteria for Adverse Events) with
minor events defined as grades 1–3 and major events as grades 4–5.
No treatment-related adverse events occurred following Y90-RS during the follow-up
period whereas 10 (33%) occurred following MWA, including 3 minor complications (wound
infection, abdominal pain, and nausea) and 7 major complications including arterioportal
fistula (n = 1), pneumothorax (n = 1), liver infarction (n = 2), capsular burn (n = 1), rectus sheath hematoma (n = 1), and hepatic vasculature injury (n = 1). Neither treatment groups experienced major posttreatment laboratory grade adverse
events at 1-month follow-up ([Table 3]).
Table 3
Posttreatment labs
|
Characteristic
|
All (n = 50)
|
MWA (n = 30), TACE/MWA (n = 19), MWA alone (n = 11)
|
Y90 (n = 20)
|
p-Value
|
|
AST
|
|
Mean (SD)
|
59.4 (35.2)
|
65.6 (37.6)
|
53.2 (31.5)
|
0.26
|
|
Range
|
21.0–163.0
|
24.0–163.0
|
21.0–135.0
|
|
ALT
|
|
Mean (SD)
|
33.7 (16.6)
|
40.2 (16.9)
|
26.3 (12.9)
|
0.01
|
|
Range
|
6.6–86.0
|
18.0–86.0
|
6.6–49.0
|
|
Total bilirubin
|
|
Mean (SD)
|
2.2 (3.1)
|
2.6 (3.9)
|
1.6 (1.3)
|
0.3
|
|
Range
|
0.3–18.4
|
0.3–18.4
|
0.4–5.3
|
Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; MWA,
microwave ablation; SD, standard deviation; TACE, transarterial chemoembolization;
Y90, yttrium-90.
Note: Parenthetical values are percentages unless otherwise indicated.
Discussion
RS delivers microspheres loaded with a predetermined, personalized tumor-absorbed
dose of Y90 into the segmental vessel, thereby selectively targeting the lesion while
minimizing radiation exposure to normal liver parenchyma.[11]
[23]
[24]
[25] Recent landmark single-arm studies have demonstrated the efficacy of Y90-RS for
unresectable early-stage HCC.[4]
[6]
[23] The RASER study described the first prospective cohort of very-early to early HCC
(mean 2.1 cm) treated with Y90-RS in which 90% of patients achieved CR after one treatment
with a median response duration of 635 days.[7] Target lesion progression had a cumulative incidence of 4% after 1 year and 12%
after 2 years and actuarial overall survival was 96% at 2 years. These findings were
similar to the LEGACY study of 162 patients with tumors measuring up to 8 cm (median
2.7 cm), which demonstrated a 2- and 3-year overall survival of 94.8 and 86.6%, respectively,
and a best objective response rate of 88.3%, consistent with CR rates from previous
retrospective analyses.[7]
[26]
[27]
[28] Our smaller retrospective cohort shows similar CR rates of 96% with Y90-RS and 77%
with MWA ± TACE. Furthermore, disease progression occurred after 23.5 months following
Y90-RS and 6.7 months following MWA ± TACE, with a longer TTP following Y90-RS.
There is a paucity of literature directly comparing Y90-RS to percutaneous ablation.[6]
[7]
[26] Arndt et al evaluated the outcomes of MWA versus Y90-RS in 68 treatment-naive unresectable
lesions < 4 cm. Response rates were comparable between treatment groups with 87.5%
(MWA) and 88.0% (Y90-RS) achieving CR at 3 months (p = 0.565) as was overall survival (RS 46.1 months, MWA 46.0 months) and adverse events.
All clinical and laboratory CTCAE toxicities were ≤ grade 3 with no significant difference
in laboratory toxicities, although there was a higher incidence of clinical toxicities,
that is, ascites (p = 0.048), fatigue (p = 0.028), and abdominal pain (p = 0.035) following Y90-RS. Notably, these differences were eliminated after propensity
score matching and progression-free survival was longer following Y90-RS (59.0 vs.
44.3 months; p = 0.021). The authors attributed this to Y90-RS achieving negative margins more frequently,
corroborating the results in our study.[6] Therefore, our difference in CR may be attributed to Y90-RS obtaining better negative
margins compared with MWA for lesions that are adjacent to critical structures.
Biederman et al performed a propensity score-matched study in which 121 locoregional
treatment-naive patients with lesions ≤ 3 cm were treated with Y90-RS or combined
TACE/MWA. No statistically significant differences were observed in procedure-related
complications (8.9% TACE/MWA vs. 4.9% Y90-RS; p = 0.46), yet major CTCAE clinical toxicities only occurred following TACE/MWA, including
pneumothorax (n = 3), TACE access site hematoma (n = 2), biloma (n = 1), and subcapsular hematoma (n = 1). The Y90-RS group experienced two access site hematomas during pretreatment
mapping angiography. Grade 3 to 4 elevations in bilirubin and AST occurred infrequently
and with similar incidence in both treatment groups. Overall and target lesion CR
was comparable regardless of propensity score matching.[16] However, in contrast to our study, lesions treated with MWA in this study were in
favorable locations.
Complete necrosis following thermal ablation is as high as 97.3% for tumors < 3 cm,
but local recurrence is reported to be frequent (up to 26%), particularly for those
in areas suboptimal for ablation as a 0.5-cm safety margin is required to prevent
local recurrence.[7]
[29]
[30]
[31] Gozzo et al investigated recurrence following percutaneous thermal ablation for
213 patients, where local recurrence occurred in 12.4% at 1 year and 19.7% over the
total follow-up period.[32] Lesions in suboptimal locations may be more prone to incomplete ablation margins.
In our study, 23% of the patients had disease recurrence after MWA, which was not
statistically different from that of Y90-RS (17%; p = 0.6), however, there are considerations with respect to assessing response rates.
The tumoricidal impact of MWA is mediated by hyperthermal damage resulting in coagulative
necrosis, which becomes quickly visible on imaging, whereas radioembolization causes
delayed hemorrhagic necrosis and edema, with persistent enhancement of the exposed
area. Therefore, follow-up imaging within 3 months may result in underestimation of
response, and larger tumors may even require > 6 months to fully appreciate tumor
necrosis.[10]
[20]
With respect to safety, our study found Y90-RS to be a comparable treatment strategy
compared with MWA for lesions adjacent to critical structures; there were no clinical
complications following Y90-RS and a rate of 33% following MWA. A review from 2019
assessed the clinical outcomes and toxicity of 155 cases of Y90-RS, in which only
two patients developed radiation-induced liver injury, the most commonly reported
side effect was fatigue, and all CTCAE events were ≤ grade 3.[33] Similarly, the LEGACY and RASER trials reported grade 3 events in 19.1 and 24.1%
of patients, respectively, with the most common being transient fatigue and leukopenia.[7]
[26] Transient laboratory toxicities were the only adverse events noted in our cohort
of Y90-RS patients. In contrast, our study showed a higher minor and major complication
rate for MWA, with 7 major complications including arterioportal fistula (n = 1), pneumothorax (n = 1), liver infarction (n = 2), capsular burn (n = 1), rectus sheath hematoma (n = 1), and hepatic vasculature injury (n = 1).
Reported complication rates following MWA are between 2.2 and 61.5%, with major complications
ranging from 2.6 to 4.6%.[34]
[35] Critical structures such as the gallbladder, bowels, hilum, pericardium, diaphragm,
and large vessels may complicate MWA due to their higher dielectric constants, rendering
them more susceptible to necrosis than healthy liver parenchyma.[36]
[37]
[38] The challenges presented by lesions within 1 cm of critical structures are twofold:
(1) critical structures are susceptible to mechanical damage or delayed complications
if ablated, and (2) the avoidance of damaging these structures can result in incomplete
ablation of the lesion.[39]
[40]
Puncture-related complications include pneumothorax, pleural effusion, hemothorax,
intraperitoneal bleeds, and tumor seeding.[41]
[42] Damage during probe placement can be mitigated by CT and US guidance or by reducing
motion artifact with decreased ventilated tidal volume or the Valsalva maneuver.[12]
[43] Cauterization of the intraparenchymal tract reduces the risk of seeding and direct
puncture of the tumor can be avoided by creating an overlapping ablation zone with
multiple probes.[12]
[44] The endovascular approach of Y90-RS offers a theoretical advantage over percutaneous
ablation by eliminating the risk of damage to surrounding structures or tract seeding.
Thermal-related complications include damage to gastrointestinal or biliary structures
(e.g., gallbladder perforation, bile duct stenosis, cholecystitis, biloma), liver
abscesses, or portal vein thromboses.[41]
[45] Additionally, large caliber vessels can alter the ablation zone away from the vessel,
termed heat sink effect, leading to incomplete tumor ablation, although MWA has been
shown to be less susceptible to this phenomenon than RFA.[45]
[46] Hydrodissection provides thermal protection by physically separating the lesion
from critical structures; it can also reduce heat sink effect by displacing large
vessels.[17]
[18] A recent study detailed 66 patients with subcapsular tumors and tumors near critical
structures who underwent hydrodissection-assisted MWA; although CR occurred in 91.4%,
3.0% had a major complication (two hepatic abscesses, one biliary injury).[17] Additional disadvantages of hydrodissection include fluid overload, diffusion from
the injection site, risk of peritoneal seeding, and perihepatic bleeding from forceful
dissection.[47]
[48]
Temperature monitoring can monitor for thermal damage to surrounding structures or
achievement of threshold temperature when placed adjacent to the tumor margin.[49] In a study of 89 patients with 96 lesions adjacent to the diaphragm, thermal monitoring
needles placed between the lesion and diaphragm allowed for temperature control between
50 and 60°C. There was no difference in complete ablation rate (94.8%) when compared
with control lesions, although the study group had statistically insignificant, yet
higher, incidence of shoulder pain, pleural effusion, nausea, and vomiting.[50] Although such adjunct techniques can mitigate MWA complications, they lead to increased
procedure time and complexity and are not without inherent risks.
Limitations of this study include its small size, single-center and retrospective
design, limited follow-up period, and inconsistent MWA pretreatment with TACE. In
addition, this study was not randomized and therefore there is a potential for bias
to have been introduced during case selection. A longer follow-up period (> 3 months)
may be beneficial to avoid underestimation of treatment response following Y90-RS.
Future studies necessitate a multi-institutional comparison of treatment-naive lesions
with a longer follow-up period. Additionally, comparison using radiology-pathology
correlation and an expanded characterization of toxicities will be prudent. The use
of TACE prior to MWA has been postulated to increase response rates, however, we did
not examine if prior TACE improved efficacy of MWA. Despite its limitations, our study
directly compared Y90-RS and MWA for lesions in particular locations, demonstrating
comparability in safety and efficacy between the treatment modalities.
Conclusion
Y90-RS is a suitable alternative to MWA as a first-line therapy for early-stage HCC,
particularly where regional anatomy presents challenges for percutaneous ablation.
Y90-RS offers both a favorable treatment response and an excellent safety profile,
supporting its use as an important curative therapy for early-stage HCC.
