Efficacy and safety of transcatheter arterial chemoembolization combined with microwave ablation for hepatic hemangiomas (> 5 cm)

Abstract

BACKGROUND

Hepatic hemangiomas represent the most prevalent benign liver tumors. Surgical management of large symptomatic hepatic hemangiomas remains controversial and there is an increasing interest in minimally invasive techniques, such as transcatheter arterial chemoembolization (TACE) and microwave ablation (MWA).

AIM

To evaluate the efficacy and safety of TACE combined with MWA for large hepatic hemangiomas.

METHODS

This retrospective cohort study was conducted at Peking Union Medical College Hospital between January 2015 and January 2024. Eighty-two patients with hepatic hemangiomas > 5 cm were divided into two groups: Observation (TACE + MWA, n = 50) and control (TACE, n = 32). Tumor diameter and treatment outcomes were evaluated at baseline, 12 months, and > 3 years. Appropriate statistical tests were chosen based on the type and distribution of the data.

RESULTS

At baseline, the median tumor diameter was 8.3 (range: 5.0-19.2) cm in the observation group and 8.5 (range: 5.0-20.0) cm in the control group. The median follow-up duration was 44.6 (95% confidence interval: 36.7-52.5) months. At 12 months post-treatment, the observation group demonstrated a higher tumor reduction ratio compared to the control group (50.98% vs 23.28%, respectively; P < 0.001). The objective response rate was 93.94% in the observation group, which was significantly higher than that in the control group (33.33%) (P < 0.001). No recurrence occurred in the observation group, while one case occurred in the control group. Notably, no cases of hemoglobinuria or acute kidney injury were reported in the observation group.

CONCLUSION

Combination treatment enhances tumor shrinkage, promotes long-term tumor control, and reduces the complications associated with MWA, thereby presenting a promising alternative to surgical resection.

Key Words: Hepatic hemangioma; Transcatheter arterial chemoembolization; Microwave ablation; Interventional treatment; Safety; Liver

Core Tip: This study compares two treatment options for large liver hemangiomas, a type of benign tumor. We found that combining transcatheter arterial chemoembolization and microwave ablation was more effective at shrinking tumors and preventing recurrence compared to transcatheter arterial chemoembolization alone. Patients with large liver hemangiomas experienced fewer complications on the combined treatment which may be a safer alternative to surgery.

INTRODUCTION

Hepatic hemangiomas are the most prevalent benign liver tumors, with an incidence rate of approximately 2.5% to 5.0%, and are more frequently observed in women aged 30-50 years[1,2]. Although most asymptomatic small hemangiomas (diameter < 5 cm) do not require intervention, larger tumors (diameter > 5 cm) or those presenting with clinical symptoms, such as persistent abdominal pain, gastrointestinal discomfort, or psychological distress, often necessitate therapeutic management[3,4]. Research indicates that approximately 40% of patients with hepatic hemangiomas seek treatment for psychological anxiety[5,6].

According to the 2016 European Association for the Study of the Liver (EASL) guidelines[1], surgical resection is the preferred treatment for symptomatic and large hemangiomas. However, as outlined in the EASL and Associazione Italiana Studio del Fegato guidelines[7], surgery is applicable only to a subset of patients. The surgical resection of large symptomatic hepatic hemangiomas remains controversial due to the associated invasive risks[1,7]. This has led to increased interest in minimally invasive interventional treatment, such as transcatheter arterial chemoembolization (TACE) and microwave ablation (MWA)[3,6,8-10].

TACE induces tumor ischemia and shrinkage by occluding the supplying arteries, while MWA generates heat through high-frequency electromagnetic waves, resulting in thermal coagulative necrosis of the tumor. While effective, TACE may be limited by its reduced efficacy in poorly vascularized tumors and less significant tumor shrinkage, whereas MWA carries a risk of hemolysis-related complications[9,11-13]. A combined TACE and MWA strategy has been proposed to mitigate these respective limitations, potentially enhancing therapeutic outcomes while reducing MWA-related complications[4,14]. However, research on this combined treatment strategy remains insufficient, particularly concerning its safety and efficacy compared with TACE alone, and there is a notable lack of long-term follow-up and systematic evaluations of efficacy. This retrospective cohort analysis aimed to systematically compare the safety and efficacy of TACE combined with MWA with those of TACE alone for the treatment of hepatic hemangiomas. The clinical outcomes of the combined treatment strategy were evaluated in terms of tumor reduction, symptom relief, alleviation of psychological anxiety, and overall quality of life.

MATERIALS AND METHODS

This retrospective cohort study was conducted at Peking Union Medical College Hospital between January 2015 and January 2024 and included 82 patients with hepatic hemangiomas. The patients were divided into two groups based on the treatment modality: The observation group (n = 50), which received a combination of TACE and MWA, and the control group (n = 32), which received TACE alone. The study adhered to strict inclusion and exclusion criteria to ensure both scientific rigor and reproducibility of the results.

All enrolled patients underwent a comprehensive pretreatment evaluation conducted by a multidisciplinary team (MDT) that included hepatobiliary surgeons, interventional radiologists, and anesthesiologists. During this evaluation, surgical candidacy was assessed for each patient, and surgical resection was considered in all cases. However, for most patients, surgical resection was either declined due to personal preferences or was deemed contraindicated due to medical comorbidities. Given the symptomatic burden and ineligibility or reluctance to undergo surgical resection, the MDT collectively determined that a minimally invasive interventional approach consisting of TACE combined with MWA was a clinically appropriate and justifiable alternative. This decision was based on the expected safety profile and efficacy of the combined treatment, which was tailored to address both the tumor characteristics and the patient’s overall clinical condition. The MDT’s decision-making process was guided by a thorough consideration of individual patient factors, including the risk-benefit profile of surgical resection vs the proposed interventional treatment.

Inclusion criteria

The inclusion criteria were as follows: (1) Tumor diameter: The longest diameter of the hepatic hemangioma was > 5 cm, confirmed by contrast-enhanced magnetic resonance imaging (MRI) or computed tomography (CT); (2) Clinical symptoms: Patients exhibited persistent abdominal pain, gastrointestinal discomfort, or other symptoms related to hepatic hemangiomas; (3) Tumor growth trend: Imaging follow-up indicated tumor growth exceeding 1 cm in the past 2 years; and (4) Patients demonstrated psychological anxiety and were undergoing interventional therapy.

Exclusion criteria

The exclusion criteria were as follows: (1) Comorbid malignant tumors: Presence of other primary or metastatic malignant tumors; (2) Severe systemic diseases: Patients with severe dysfunction of the liver, kidneys, heart, or lungs; (3) Coagulation disorders: Patients with a severe bleeding tendency (platelet count < 50 × 10⁹/L, prothrombin time > 18 seconds, or international normalized ratio > 1.5) were excluded; and (4) Other contraindications: Patients with vascular abnormalities, such as the Kasabach-Merritt phenomenon, characterized by thrombocytopenia and consumptive coagulopathy.

Diagnosis

The diagnosis of hepatic hemangioma was primarily confirmed through contrast-enhanced CT and/or MRI and was characterized by peripheral nodular enhancement in the arterial phase and centripetal filling in the venous phase. In cases with atypical tumor morphology, a pathological biopsy was performed to confirm the diagnosis of hepatic hemangioma.

Treatment procedure

All patients were treated by four interventional radiologists, each with more than 15 years of experience, using standardized techniques for TACE and MWA. The team also included anesthesiologists and nurses to ensure smooth execution of the procedures and patient safety.

The control group underwent TACE using the ionic contrast agent, diatrizoate meglumine and diatrizoate sodium (Urografin^® 76%, 370 mgI/mL; Bayer AG, Berlin, Germany, H10970164), guided by a Philips Allura DSA system (Allura Xper FD20). The Seldinger technique was employed to puncture the right femoral artery, followed by selective catheterization of the common hepatic artery, during which hepatic arteriography was performed. Upon confirming the location of the lesion, the catheter was advanced into the blood vessel of the tumor, and the embolizing agent was slowly administered under fluoroscopic guidance. The embolizing agent used was a bleomycin-super-liquid iodized oil emulsion consisting of 15000-30000 IU of bleomycin (H20123357) mixed with 5-20 mL of super-liquid iodized oil (Guerbet, France; H20050307). The exact volume of the emulsion was determined based on the tumor volume, which ranged from 1 mL to 10 mL (1:2 to 1:1 ratio). Embolization ceased when enhancement was observed in the small portal vein surrounding the lesion, or when reflux from the supplying artery occurred. Hepatic arteriography confirmed occlusion of the feeding vessels, thereby completing TACE.

In the observation group, MWA (water-cooled MWA system; Yigao Medical, China) was performed following TACE. The ablation protocol was standardized across all patients, with careful consideration of the tumor characteristics and patient safety. The patients received intravenous anesthesia, followed by the insertion of an endotracheal tube or laryngeal mask airway, with mechanical ventilation employed to control respiration, thereby ensuring patient stability during the ablation process. CT guidance was used to identify the precise location of the lesion, followed by meticulous planning of the puncture path to minimize damage to surrounding vital organs. To mitigate the risk of bleeding from the hemangioma, the ablation probe was typically directed through the normal liver tissue to reach the target lesion. The output power of MWA was generally set between 30 W and 80 W, with the ablation time determined according to the tumor volume. Ablation sites were carefully planned to optimize tumor necrosis. At a typical output power of 50 W, an ablation needle was inserted to deliver energy, and the ablation zone was designed to adequately cover the tumor. For hemangiomas with tumor diameter > 5 cm, a single point of ablation was focused on, with a target range of approximately 3.0-3.5 cm along the anteroposterior axis and a diameter of approximately 2 cm from the needle tip. The ablation zone was determined based on the manufacturer’s technical specifications and the clinical experience of the interventional radiologist. For giant (> 10 cm) or strategically located tumors, a two-circle overlapping ablation technique was used to ensure complete tumor coverage within the ablation zone. Throughout the procedure, the microwave probe was maintained at least 1 cm away from critical structures such as the bile ducts and inferior vena cava. Thermal shielding technology was utilized as necessary to protect these vital structures. MWA commenced at the edge of the tumor where the thermal sink effect was relatively minimal, thereby facilitating better bleeding control. Finally, needle tract ablation was performed by withdrawing the ablation electrode to further decrease the risk of postoperative bleeding. Following general anesthesia for ablation, the patients underwent 12 hours of electrocardiogram monitoring, 6 hours of fasting, and hydration with 1000-2000 mL of normal saline, adjusted according to the patient’s specific condition.

Postoperative evaluation and follow-up

In this study, dynamic monitoring of the liver and kidney function was conducted before treatment, 1-week post-treatment, and 1-month post-treatment. We focused on key biochemical markers, including alanine aminotransferase (ALT), aspartate aminotransferase, total bilirubin (TBIL), direct bilirubin (DBIL), albumin (Alb), and creatinine. Postoperative monitoring included the white blood cell (WBC) count, neutrophil count, platelet count, and urinalysis to diagnose and monitor potential hemoglobinuria and acute kidney injury (AKI). Special attention was paid to complications related to postoperative hemolysis, with prompt detection and management of changes in hemoglobin and red blood cell levels. Beyond surveillance for hemolysis, these comprehensive hematologic assessments were also systematically performed to monitor potential bone marrow suppression, a recognized-though typically mild-systemic effect of bleomycin chemotherapy[15,16].

During imaging follow-up, enhanced MRI scans were performed at 1 month, 6 months, and 12 months after treatment to assess the longest tumor diameter and shrinkage of hepatic hemangiomas. Currently, the Response Evaluation Criteria in Solid Tumors (RECIST)[17] are not commonly used to assess hepatic hemangioma. This study employed RECIST to analyze the imaging results, thereby ensuring the objectivity and accuracy of treatment evaluations. For patients presenting with multiple hepatic hemangiomas, the longest lesion meeting all inclusion criteria at baseline was designated as the index lesion. All efficacy outcomes, including tumor diameter changes, tumor reduction ratio, and response assessment by RECIST 1.1 criteria, were primarily based on the measurements and evaluations of this pre-defined index lesion. This study employed the Barthel Index to quantify changes in patients’ daily living abilities by utilizing statistical analysis to compare admission and discharge scores and evaluate improvements in postoperative self-care capabilities. Health-related quality of life and psychological status were assessed using widely validated scales, including the generalized anxiety disorder-7, Patient Health Questionnaire-9, Perceived Stress Scale, and 36-Item Short Form Health Survey[18-20], along with an in-house-designed psychological questionnaire. These psychological assessments were completed by patients as self-reported questionnaires, administered during routine follow-up outpatient visits. The clinical research team members responsible for questionnaire administration and data collection were not blinded to the treatment allocation (i.e., TACE combined with MWA or TACE alone). Special attention was paid to patients undergoing treatment for anxiety and psychological stress, with follow-ups of their postoperative anxiety levels and alleviation of clinical symptoms.

Study endpoints

The primary endpoints of this study were safety, assessed using the Clavien-Dindo classification[21] of ablation-related complications, and imaging, evaluated using the RECIST. Secondary endpoints included the clinical response, which encompassed the evaluation of postoperative symptom improvement, such as pain relief, restoration of digestive function, enhancement of psychological status, and the recurrence rate of residual tumors. The follow-up period extended from January 2015 to January 2025, providing long-term data to support treatment.

Statistical analysis

The sample size for this retrospective cohort study was determined by the complete enumeration of all eligible patients with hepatic hemangiomas (diameter > 5 cm) who met the predefined inclusion and exclusion criteria and underwent either TACE combined with MWA (observation group, n = 50) or TACE alone (control group, n = 32) at Peking the Union Medical College Hospital between January 2015 and January 2024, and for whom comprehensive clinical outcome data were available. All data were analyzed using SPSS software (version 26.0; IBM Corp., Armonk, NY, United States). The statistical review of the study was performed by a biomedical statistician. Appropriate statistical methods were chosen based on the type and distribution of the data, with statistical significance set at P < 0.05. For some of the key metrics, the risk ratios and their 95% confidence intervals (CIs) were calculated and reported to assess the effect sizes and clinical relevance of the differences between groups.

To evaluate the statistical power of the current study to detect differences in the primary efficacy endpoint, post-hoc power analysis was performed for the 12-month objective response rate (ORR). This analysis, based on the observed ORRs of 93.94% in the observation group (n = 50) and 33.33% in the control group (n = 32), was conducted using G*Power software (version 3.1.9.7; Faul, Erdfelder, Lang, and Buchner, 2007; Heinrich Heine University Düsseldorf, Düsseldorf, Germany). The analysis was set for a two-sided test with a significance level (α) of 0.05. The resulting statistical power (1-β) was 0.9999995, indicating that the study was amply powered to detect the observed between-group difference.

Ethical statement and informed consent

This study was reviewed and approved by the Ethics Committee of Peking Union Medical College Hospital (approval No. 1-24PJ1959). All study participants, or their legal guardians, provided informed written consent prior to study enrollment. Data collection and analysis strictly adhered to the ethical guidelines for medical research involving humans, as outlined in the Declaration of Helsinki, ensuring full protection of patient privacy.

RESULTS

Baseline characteristics

This study included 82 patients who were categorized into an observation (n = 50) or control (n = 32) group according to the type of intervention. No statistically significant differences were observed between the groups in terms of age, sex, baseline tumor diameter, disease duration, lesion distribution, and count, comorbidities, preoperative anxiety status, or other critical demographic and clinical indicators (all P > 0.05). This indicated that the baseline characteristics of the participants were comparable (Table 1).

Table 1 Baseline characteristics, n (%).

Characteristics1	All (n = 82)	Control group (n = 32)	Observation group (n = 50)	P value2
Age (years)	46 (40-54)	47 (41-55)	44 (39-52)	0.122
Longest tumor diameter (cm)	8.4 (5-20)	8.5 (5-20)	8.3 (5-19.2)	0.860
Disease duration (years)	6 (1-13)	7 (2-12)	5 (1-14)	0.668
Sex				0.527
Female	65 (79.27)	27 (84.38)	38 (76.00)
Male	17 (20.73)	5 (15.62)	12 (24.00)
Measurable lesions3				0.993
Single	73 (89.02)	29 (90.62)	44 (88.00)
Multiple	9 (10.98)	3 (9.38)	6 (12.00)
Lesion distribution				0.909
Right lobe	59 (71.95)	24 (75.00)	35 (70.00)
Left lobe	10 (12.20)	3 (9.38)	7 (14.00)
Junction of the right and left lobes	11 (13.41)	4 (12.50)	7 (14.00)
Caudate lobe	2 (2.44)	1 (3.12)	1 (2.00)
Lesion classification4				0.624
Huge	63 (76.83)	26 (81.25)	37 (74.00)
Giant	19 (23.17)	6 (18.75)	13 (26.00)
Symptoms				0.099
Asymptomatic	52 (63.41)	16 (50.00)	36 (72.00)
Abdominal discomfort	27 (32.93)	15 (46.88)	12 (24.00)
Gastrointestinal symptoms	3 (3.66)	1 (3.12)	2 (4.00)
Comorbidities				0.473
No comorbidities	67 (81.71)	25 (78.12)	42 (84.00)
Hypertension	10 (12.20)	5 (15.62)	5 (10.00)
Hyperlipidemia	2 (2.44)	0 (0.00)	2 (4.00)
Diabetes mellitus	2 (2.44)	1 (3.12)	1 (2.00)
SLE	1 (1.22)	1 (3.12)	0 (0.00)
Anxiety status				1.0
No anxiety	75 (91.46)	29 (90.62)	46 (92.00)
Mild anxiety	7 (8.54)	3 (9.38)	4 (8.00)
Reason for treatment				0.128
Enlargement of hemangioma	52 (63.41)	19 (59.38)	33 (66.00)
Symptomatic hemangioma	18 (21.95)	6 (18.75)	12 (24.00)
Anxiety	2 (2.44)	0 (0.00)	2 (4.00)
≥ 2 reasons	10 (12.20)	7 (21.88)	3 (6.00)

¹Continuous variables are expressed as mean ± SD for normally distributed data or median (interquartile range) for non-normally distributed data based on the Shapiro-Wilk test for normality. Categorical variables, including sex and lesion classification, are presented as frequencies (percentages).

²P value was calculated using the independent samples t-test for normally distributed data and the Mann-Whitney U test for non-normally distributed data. The χ² test or Fisher’s exact test was used for categorical variables, and Fisher’s exact test was applied when the expected frequencies were < 5.

³In this study, measurable lesions were defined as lesions with a longest diameter > 5 cm, reflecting a clinically significant disease burden. For patients with multiple hepatic hemangioma lesions (n = 9), we predefined the longest diameter lesion as the “index lesion” for this patient. Each of these 9 patients contained 2 to 3 hemangiomas larger than 5 cm in diameter. Patients with multiple lesions were represented by the index lesion in statistical analyses.

⁴Huge was defined as hemangiomas whose longest tumor diameter ranges from 5 to 10 cm. Giant was defined as hepatic hemangiomas whose longest tumor diameter is > 10 cm.

SLE: Systemic lupus erythematosus.

Ablation parameters of the observation group

As shown in Table 2, most patients in the observation group (n = 50) underwent a single ablation (47/50, 94.0%), while three patients (6.0%) required two ablation sessions due to the tumor diameter or clinical strategy. On average, surgeries in 66% of the patients required one probe, while 34% used two probes per ablation, with an average of 5.78 ± 3.00 ablation sites. Regarding the energy parameters, 50 W was the most common average output power (34/50, 68.0%) used in the cases, followed by 15 cases (30.0%) using 60 W, and only one case (2.0%) employing 40 W. The average duration of the ablation procedure was 27.88 ± 14.11 minutes, suggesting that the procedure duration may vary based on the tumor diameter and tissue characteristics. Furthermore, based on postoperative responses and clinical observations, the embolization to ablation interval was categorized into three groups: < 1 week (32.0%), 1-3 weeks (20.0%), and > 3 weeks (48.0%). In the observation group, ablation was generally performed within the appropriate time window following the initial embolization to enhance local control in lesions with residual or potential recurrence risk.

Table 2 Ablation parameters of the observation group, n (%).

Ablation parameters1	Observation group (n = 50)
Number of ablation procedures
1	47 (94.00)
2	3 (6.00)
Number of ablation probes
1	33 (66.00)
2	17 (34.00)
Average ablation sites	5.78 ± 3.00
Ablation power (W)2
40	1 (2.00)
50	34 (68.00)
60	15 (30.00)
Ablation duration (minute), mean ± SD	27.88 ± 14.11
Embolization to ablation interval3
< 1 week	16 (32.00)
1-3 weeks	10 (20.00)
> 3 weeks	24 (48.00)

¹Continuous variables are expressed as mean ± SD if they are normally distributed or expressed as median ± interquartile range if they are non-normally distributed; categorical variables are expressed as frequency (percentage).

²Ablation power (W) represents the average power output of the ablation device during the procedure. The power settings were adjusted according to the lesion volume and tissue properties.

³The embolization to ablation interval refers to the time interval between embolization and ablation procedures, categorized as < 1 week, 1-3 weeks, and > 3 weeks. Categorization was based on postoperative patient responses and clinical observations.

Postprocedural adverse events

Perioperative adverse events were classified according to the Clavien-Dindo grading system, ranging from mild (grade I) to severe complications (grade IV), as detailed in Table 3. There were no significant differences in the occurrence of common adverse events, including postprocedural pain, fever, nausea, and vomiting, between the control and observation groups (P > 0.05). The incidence of mild perihepatic hemorrhage was significantly higher in the observation group than in the control group (P = 0.019). Similarly, the incidence of grade II liver injury was significantly higher in the observation group than in the control group [36.00% (18/50) vs 6.25% (2/32), P = 0.005]; further analysis revealed that the risk of grade II liver injury in the observation group was 5.76 times higher than that in the control group (risk ratios = 5.76; 95%CI: 1.43-23.17). Notably, seven patients (14.00%) in the observation group developed jaundice, although this was not significantly different from the control group (P = 0.071). However, all complications were classified as Clavien-Dindo grade I or II and were effectively managed with conservative treatment. These conditions resolved within 1 week postoperatively without the need for further invasive interventions, underscoring the overall safety and controllability of the combined treatment approach. Furthermore, no cases of hemoglobinuria or AKI were detected in this study, suggesting that the combined treatment maintained a favorable safety profile, effectively mitigating the risk of severe renal complications or hemoglobinuria associated with MWA. Notably, detailed hematologic monitoring (Table 4) revealed that the key parameters, including white blood cell counts, neutrophil counts, and platelet counts, remained stable post-procedure. This stability, coupled with the absence of anemia, neutropenia, or thrombocytopenia requiring intervention (Table 3), indicated minimal discernible impact on bone marrow function from the administered bleomycin regimen in this cohort. Less common complications such as bile fistulas or liver abscesses were not observed, indicating that the incidence of such complications was low within the study population.

Table 3 Summary of postprocedural adverse events.

Adverse events1	Clavien-Dindo grade	All (n = 82)	Control group (n = 32)	Observation group (n = 50)	P value2
Postprocedural pain	I	20 (24.39)	8 (25.00)	12 (24.00)	1.0
Fever	I	5 (6.10)	1 (3.12)	4 (8.00)	0.669
Nausea and vomiting	I	17 (20.73)	6 (18.75)	11 (22.00)	0.94
Mild perihepatic hemorrhage	I	10 (12.20)	0 (0.00)	10 (20.00)	0.019a
Anemia3	I	0 (0.00)	0 (0.00)	0 (0.00)	1.0
Neutropenia3	I	0 (0.00)	0 (0.00)	0 (0.00)	1.0
Thrombocytopenia3	I	0 (0.00)	0 (0.00)	0 (0.00)	1.0
Hemoglobinuria	I	0 (0.00)	0 (0.00)	0 (0.00)	1.0
Hepatic injury	II	20 (24.39)	2 (6.25)	18 (36.00)	0.005a
Jaundice	II	7 (8.54)	0 (0.00)	7 (14.00)	0.071
Biliary fistula	II	0 (0.00)	0 (0.00)	0 (0.00)	1.0
Liver abscess	II	0 (0.00)	0 (0.00)	0 (0.00)	1.0
Acute kidney injury	IV	0 (0.00)	0 (0.00)	0 (0.00)	1.0

^aP < 0.05; P values < 0.05 were considered statistically significant. Risk ratio for hepatic injury in the observation group compared to the control group was 5.76 (95% confidence interval: 1.43-23.17).

¹Adverse events are listed in ascending order according to the Clavien-Dindo classification, from minor (grade I) to more severe complications (grade IV).

²P value was calculated using the χ² test to compare the distribution of adverse events between the groups.

³Hematologic parameters were monitored considering potential bleomycin-induced bone marrow suppression. Anemia: Hemoglobin < 10.0 g/dL. Neutropenia: Absolute neutrophil count < 1.5 × 10⁹/L. Thrombocytopenia: Platelet count < 75 × 10⁹/L.

Table 4 Comparison of pre- and post-treatment laboratory parameters, mean ± SD.

Characteristics	All (n = 82)	Control group (n = 32)	Observation group (n = 50)	P value1
ALT (U/L)
Pre-treatment	25.34 ± 22.85	16.25 ± 11.10	31.16 ± 26.39	< 0.001
Postoperative within 1 week	74.55 ± 122.15	30.62 ± 38.36	102.66 ± 147.19	< 0.001
Postoperative at 1 month	23.39 ± 26.06	18.84 ± 7.13	26.30 ± 32.68	0.227
TBIL (μmol/L)
Pre-treatment	12.81 ± 5.43	11.20 ± 4.53	13.85 ± 5.74	0.034
Postoperative within 1 week	18.83 ± 10.52	14.71 ± 7.53	21.46 ± 11.35	< 0.001
Postoperative at 1 month	11.86 ± 4.05	11.20 ± 3.78	12.28 ± 4.20	0.248
DBIL (μmol/L)
Pre-treatment	3.88 ± 1.68	3.44 ± 1.65	4.16 ± 1.66	0.038
Postoperative within 1 week	4.66 ± 2.36	3.59 ± 1.21	5.35 ± 2.65	< 0.001
Postoperative at 1 month	3.65 ± 1.48	3.38 ± 1.46	3.83 ± 1.47	0.166
Albumin (g/L)
Pre-treatment	43.85 ± 3.94	44.06 ± 5.10	43.72 ± 3.02	0.33
Postoperative within 1 week	40.95 ± 3.08	41.72 ± 3.31	40.46 ± 2.84	0.09
Postoperative at 1 month	43.98 ± 2.57	43.81 ± 3.34	44.08 ± 1.96	0.759
WBC (× 109/L)
Pre-treatment	5.81 ± 1.52	5.77 ± 1.48	5.83 ± 1.56	0.844
Postoperative within 1 week	8.47 ± 3.08	8.86 ± 3.77	8.21 ± 2.56	0.794
Postoperative at 1 month	5.69 ± 1.11	5.69 ± 0.95	5.69 ± 1.21	0.555
Neutrophils (× 109/L)
Pre-treatment	3.78 ± 1.38	3.53 ± 1.07	3.94 ± 1.53	0.274
Postoperative within 1 week	7.02 ± 2.98	7.34 ± 3.67	6.81 ± 2.46	0.761
Postoperative at 1 month	3.59 ± 1.04	3.54 ± 0.83	3.63 ± 1.16	0.775
Platelets (× 109/L)
Pre-treatment	218.45 ± 56.88	220.47 ± 65.79	217.16 ± 51.05	0.799
Postoperative within 1 week	198.87 ± 60.03	201.38 ± 80.27	197.26 ± 43.29	0.868
Postoperative at 1 month	225.18 ± 53.08	228.53 ± 63.39	223.04 ± 45.87	0.816
Hemoglobin (g/L)
Pre-treatment	134.45 ± 17.30	134.97 ± 16.50	134.12 ± 17.96	0.83
Postoperative within 1 week	128.63 ± 20.91	130.38 ± 15.33	127.51 ± 23.90	0.936
Postoperative at 1 month	132.59 ± 13.43	132.88 ± 14.73	132.40 ± 12.67	0.659
Creatinine (μmol/L)
Pre-treatment	63.68 ± 13.16	62.09 ± 13.94	64.70 ± 12.67	0.232
Postoperative within 1 week	59.72 ± 12.02	59.36 ± 11.12	59.95 ± 12.68	0.985
Postoperative at 1 month	63.57 ± 9.05	62.25 ± 9.02	64.42 ± 9.06	0.175

¹P value was calculated using the χ² test to compare the distribution of adverse events between the groups.

Continuous variables are expressed as mean ± SD if they are normally distributed or expressed as median ± interquartile range if they are non-normally distributed; categorical variables are expressed as frequency (percentage). ALT: Alanine transaminase; DBIL: Direct bilirubin; TBIL: Total bilirubin; WBC: White blood cell.

Pre- and post-treatment laboratory parameters

Table 4 presents the dynamic monitoring results of key laboratory parameters, including the levels of ALT, TBIL, DBIL, Alb, WBCs, neutrophils, platelets, hemoglobin, and creatinine, for both groups at multiple time points (pre-treatment, 1-week post-treatment, and 1-month post-treatment). Overall, ALT and TBIL levels were significantly higher in the observation group at certain time points (including 1-week post-treatment) than in the control group (P < 0.001). This finding suggests that the combined intervention had a more pronounced short-term effect on liver cell injury. Follow-up assessments 1 month post-treatment indicated that the differences in ALT and TBIL between the two groups had gradually diminished and were no longer significantly different (P > 0.05). Additionally, no significant differences were observed in the Alb, WBC, platelet, or hemoglobin levels at any of the time points (all P > 0.05).

Post-treatment tumor response and follow-up outcomes

The median follow-up time for this study was 44.6 months (95%CI: 36.7-52.5 months), with a quartile range spanning from 21.3 months to 82.5 months. As indicated in Table 5, no statistically significant difference was observed in the longest tumor diameter between the two groups at the 1-month follow-up (P = 0.70). However, at the 6- and 12-month follow-ups, the observation group exhibited a significantly smaller tumor diameter than the control group (P < 0.05). The tumor reduction ratio at 12 months postoperatively was significantly greater in the observation group (50.98%, interquartile range: 37.5-61.8) than in the control group (23.28%, interquartile range: 17.5-47.27); P < 0.001), as illustrated in Figure 1. RECIST evaluation revealed a higher ORR in the observation group than in the control group (93.94% vs 33.33%). In the long-term follow-up period of 1-3 years, and in some cases extending beyond 3 years, the observation group consistently demonstrated a superior complete response and ORR. These outcomes are summarized in Figure 1. Notably, two patients in the control group experienced tumor recurrence during the extended follow-up, whereas no recurrence was observed in the observation group.

Figure 1 This figure compares the transcatheter arterial chemoembolization group and the transcatheter arterial chemoembolization + microwave ablation group at 12 months and > 3 years. The top charts display the complete response and objective response rate, illustrating significantly higher values for transcatheter arterial chemoembolization (TACE) + microwave ablation (MWA) at 12 months (P < 0.001) and a trend toward better outcomes at > 3 years (P = 0.059). The lower charts show the tumor diameter at 12 months: TACE + MWA achieved significantly greater tumor reduction (P < 0.05), whereas at > 3 years, the difference was not statistically significant (P = 0.29). These findings highlight the enhanced early tumor control and suggest a sustained long-term clinical benefit using the TACE + MWA combination. CR: Complete response; MWA: Microwave ablation; ORR: Objective response rate; TACE: Transcatheter arterial chemoembolization.

Table 5 Post-treatment tumor response and follow-up outcomes.

Variable	All (n = 82)	Control group (n = 32)	Observation group (n = 50)	P value1
Longest tumor diameter at 1 month	7.0 (5.2-8.2)	6.2 (5.9-9.0)	7.0 (4.95-7.95)	0.70
Longest tumor diameter at 6 months	4.4 (3.2-7.0)	7.5 (4.6-8.9)	4.1 (3.1-6.0)	< 0.05a
Longest tumor diameter at 12 months	4.05 (2.8-6.6)	6.6 (3.5-7.1)	3.7 (2.6-5.2)	< 0.05a
Tumor reduction ratio (%)	46.16 (26.48-58.73)	23.28 (17.5-47.27)	50.98 (37.5-61.8)	< 0.001c
RECIST				< 0.001c
CR	4/51 (7.84%)	1/18 (5.56%)	3/33 (9.09%)
PR	33/51 (64.71%)	5/18 (27.78%)	28/33 (84.85%)
SD	14/51 (27.45%)	12/18 (66.67%)	2/33 (6.06%)
PD	0	0	0
ORR2	37/51 (72.55%)	6/18 (33.33%)	31/33 (93.94%)
Longest tumor diameter at 1-3 years (n = 38)3	3.2 (0.5-5.8)	4.5 (3.5-7.9)	2.8 (0-4)	< 0.05a
RECIST				< 0.01b
CR	7/38 (18.42%)	1/12 (8.33%)	6/26 (23.08%)
PR	24/38 (63.16%)	5/12 (41.67%)	19/26 (73.08%)
SD	6/38 (15.79%)	5/12 (41.67%)	1/26 (3.85%)
PD	1/38 (2.63%)	1/12 (8.33%)	0
ORR2	31/38 (81.58%)	6/12 (50%)	25/26 (96.15%)
Longest tumor diameter at > 3 years (n = 29)3	3.6 (2-5.5)	4.6 (2.05-7.95)	3 (0-4)	0.29
RECIST				0.059
CR	5/29 (17.24%)	1/9 (11.11%)	4/20 (20.00%)
PR	21/29 (72.41%)	5/9 (55.56%)	16/20 (80.00%)
SD	2/29 (6.90%)	2/9 (22.22%)	0
PD	1/29 (3.45%)	1/9 (11.11%)	0
ORR2	26/29 (89.66%)	6/9 (66.67%)	20/20 (100.00%)

^aP < 0.05.

^bP < 0.01.

^cP < 0.001.

¹P values were calculated using the χ² test or independent samples t-test, as appropriate. P values < 0.05 were considered statistically significant.

²The objective response rate is defined as the sum of the complete response and partial response.

³At 1-3 years: n = 38 (control group: n = 12; observation group: n = 26); at > 3 years: n = 29 (control group: n = 9; observation group: n = 20). Postoperative follow-up data were primarily collected through enhanced magnetic resonance imaging, while other data were obtained through telephone follow-ups or outpatient visits.

CR: Complete response; ORR: Objective response rate; PD: Progressive disease; PR: Partial response; RECIST: Response Evaluation Criteria in Solid Tumors; SD: Stable disease.

Figure 2 depicts the imaging results of a 48-year-old female with a 14-cm hepatic hemangioma, showing the tumor’s appearance on pre-treatment CT (Figure 2A and B), intraoperative angiographic embolization (Figure 2C and D) and post-embolization CT (Figure 2E), application of MWA (Figure 2F), and marked tumor reduction on the 40-month follow-up MRI (Figure 2G and H), consistent with a complete response.

Figure 2 A 48-year-old female patient with giant hemangiomas. A postoperative follow-up conducted 40 months later showed a significant reduction in tumor diameter. According to the Response Evaluation Criteria in Solid Tumors, the tumor was classified as partial response. The patient currently exhibits no signs of recurrence. This case highlights the potential of combined transcatheter arterial chemoembolization (TACE) and microwave ablation in achieving significant, long-term tumor control and maintaining recurrence-free status in patients with giant hepatic hemangiomas. A and B: Pre-treatment contrast-enhanced computed tomography images taken during the arterial and portal phases, illustrating the longest tumor dimensions of 14 cm × 10 cm; C and D: Intraoperative angiography performed during TACE. This was followed by selective catheterization of the feeding artery and the administration of embolic agents for TACE treatment; E: A post-TACE non-contrast computed tomography scan demonstrating the accumulation of embolic agents at the periphery of the tumor; F: The microwave ablation treatment applied to the hepatic hemangioma; G and H: Contrast-enhanced magnetic resonance images reveal the ablation zone of the treated tumor.

Comparison of hospitalization evaluation indicators

Table 6 presents the activities of daily living scores of both groups at admission and discharge. At admission, the mean ± SD activities of daily living scores were 81.65 ± 21.36 overall, 82.50 ± 20.52 in the control group, and 81.00 ± 20.52 in the observation group (P = 0.849). At discharge, the corresponding scores were 96.71 ± 6.99, 96.41 ± 7.32, and 96.90 ± 6.84, respectively (P = 0.975). No statistically significant differences were observed between the two groups at any time point (all P > 0.05).

Table 6 Comparison of hospitalization evaluation indicators.

Variable1	All (n = 82)	Control group (n = 32)	Observation group (n = 50)	P value
Admission	81.65 ± 21.36	82.50 ± 20.52	81.10 ± 22.07	0.849
Discharge	96.71 ± 6.99	96.41 ± 7.32	96.90 ± 6.84	0.975

¹Activities of daily living scores are expressed as mean ± SD.

DISCUSSION

In recent years, the continuous advancement of interventional therapy technologies has led to minimally invasive treatments for hepatic hemangiomas, which have gradually emerged as viable alternatives to traditional surgical procedures[1,16]. Transcatheter arterial embolization and TACE have been employed to shrink tumors and alleviate symptoms by obstructing the blood supply. However, these treatments often require repetition because of the potential for vascular recanalization, which carries the risk of recurrence[10,14]. In the field of thermal ablation therapies, radiofrequency ablation (RFA) and MWA are widely used for the treatment of hepatic hemangiomas. According to Chinese experts, including Gao et al[4], RFA induces tumor necrosis through thermal coagulative necrosis. Although RFA is minimally invasive and facilitates rapid recovery, it is susceptible to thermal sink effects in highly vascular lesions, which can result in severe hemolysis and associated complications. A multicenter retrospective analysis by Wu et al[22] reported an 85.91% incidence of hemoglobinuria following RFA. In contrast, MWA employs dielectric heating principles, enabling it to achieve higher temperatures in a shorter timeframe, thus creating a larger and more uniform necrotic area while being less influenced by the cooling effects of surrounding blood flow[23-25]. However, due to the abundant blood supply and complex vascular structure of hepatic hemangiomas, MWA still encounters thermal sink effects caused by blood flow, leading to hemoglobinuria, hemolytic jaundice, and AKI, with incidence rates ranging from 3.8% to 48.6%[9,11-13]. Consequently, identifying both safe and effective minimally invasive treatment methods has become a pivotal research focus.

In this study, no instances of hemoglobinuria or AKI were observed in the patients who underwent TACE combined with MWA. This result notably diverges from those of previous studies in which hemolysis-related complications with RFA or MWA were reported. These hemolytic complications primarily stem from extensive erythrocyte destruction induced by thermal ablation in highly vascularized hepatic lesions, causing the release of hemoglobin and other cellular contents into the circulation, and ultimately impairing renal function through tubular injury and inflammation-driven processes[9,26]. The absence of such complications in our study is primarily attributable to the pre-ablation embolization strategy. Initial TACE markedly reduced the tumor blood volume and limited erythrocyte exposure and fragmentation during subsequent MWA. By decreasing intratumorally blood flow, TACE enhanced heat delivery efficiency and minimized hemolysis, aligning with the strategic embolization approach reported by Kacała et al[14] and Cai et al[11]. Furthermore, Fei and Hongsong[6] conducted a meta-analysis, revealing that shorter ablation durations were associated with reduced hemolysis rates, supporting our approach of targeted, abbreviated MWA sessions enabled by pre-ablation embolization. Our finding suggests that the “embolization followed by ablation” strategy not only enhances the efficacy of ablation but also mitigates the risk of hemolysis, a common complication associated with a rich blood supply[1,4,14,25]. These results underscore the superior safety profile of the TACE + MWA combination therapy compared to MWA alone.

The median follow-up time for this study was 44.6 months (95%CI: 36.7-52.5 months), with a quartile range spanning from 21.3-82.5 months, ensuring the long-term effectiveness of the treatment and the significance of the follow-up data. This comprehensive analysis better reflects the evolving trends in treatment effectiveness and recurrence risk, particularly in the management of giant hemangiomas (> 10 cm), enabling a more thorough evaluation of treatment comprehensiveness and durability. The study data indicated that at 12 months post-treatment, the median longest tumor diameter in the observation group decreased to 3.7 cm (interquartile range: 2.6-5.2 cm), which was significantly lower than the median of the control group at 6.6 cm (interquartile range: 3.5-7.1 cm; P < 0.05). The ORR in the observation group was 93.94% compared to 33.33% in the control group. During the long-term follow-up (> 3 years), the complete response and ORR in the observation group increased to 20.00% and 100.00%, respectively, both of which were significantly higher than those in the control group (11.11% and 66.67%, respectively; P = 0.059). As illustrated in Figure 1, the combined therapy achieved markedly better outcomes. These results demonstrate that the combined treatment strategy offers distinct advantages in terms of inhibiting tumor growth and reducing recurrence risk, effectively addressing the limitations of single treatments in managing hemolytic complications and the necessity for repeated procedures in large/giant hemangiomas. Guided by MDT evaluation, sequential TACE + MWA offered a key minimally invasive alternative for patients unsuitable for or declining surgery. This combined approach is particularly advantageous for those with giant hemangiomas (> 10 cm), pronounced symptoms, high expectations for cytoreduction, or lesions challenging for standalone MWA due to size or critical location (e.g., major vascular proximity, high heat-sink risk), where TACE may enhance subsequent MWA efficacy and safety[6,9,12,14].

All ablation treatments were conducted under general anesthesia to ensure a painless experience and minimize physiological stress responses. General anesthesia enhances safety and operability when addressing larger or more complex lesions, thereby optimizing treatment outcomes[4,6,27]. Furthermore, the combined approach underscores the importance of a well-planned needle insertion strategy. As indicated in Table 2, the observation group required an average of 5.78 ± 3.00 ablation sites, with ablation primarily performed at a power setting of 50-60 W. This not only reduced the ablation time but also ensured a more extensive necrotic area. The ischemia and inflammatory edema induced by TACE augmented conduction[4,14]. Although the observation group experienced a higher incidence of mild perihepatic hemorrhage (20.00%) and hepatic injury (36.00%) at 1-week post-treatment than the control group (0.00% and 6.25%, respectively), these adverse events were classified as Clavien-Dindo grade I or II and resolved rapidly with conservative management and short-term symptomatic treatment, with the majority improving within 1 week postoperatively. These findings underscore the overall safety and controllability of the combined therapeutic approach and reinforce its clinical viability without the need for further invasive interventions. One month post-treatment, ALT, TBIL, DBIL, and other indicators returned to levels comparable to those in the control group (P > 0.05), suggesting that liver function impairment was largely self-limiting. Stringent monitoring and timely interventions can effectively mitigate the risk of further deterioration[22,23,28]. Beyond mitigating MWA-induced hemolysis, a comprehensive safety assessment of the TACE approach also considers the systemic toxicities of bleomycin. While bleomycin’s primary dose-limiting concern is pulmonary toxicity (not observed in our cohort, likely due to patient selection and localized intra-arterial delivery), its potential for bone marrow suppression is generally characterized as mild and infrequent with regional administration at modest doses[14,15]. Our study’s findings strongly corroborate this favorable hematologic safety profile. Comprehensive hematologic follow-up (Table 4) demonstrated no clinically significant deviations in the white blood cell counts, neutrophil counts, platelet levels, or hemoglobin concentrations post-procedure, and no anemia requiring intervention was documented (Table 3). These observations align with recent literature on localized bleomycin therapy for hepatic hemangiomas[14,16]. The targeted intra-arterial administration, specific bleomycin dosage used for this benign condition, and likely preserved baseline marrow function in our patients are factors potentially contributing to this excellent hematologic tolerance. Nevertheless, systematic hematologic monitoring remains a cornerstone of prudent management when utilizing bleomycin, ensuring patient safety even when systemic risks appear low.

In terms of improving clinical symptoms, interventional therapy has obvious advantages in the treatment of hepatic hemangiomas. Preoperative data indicated that 32.93% (27/82) of patients experienced abdominal discomfort, with 46.88% (17/32) in the control group and 24.00% (12/50) in the observation group. Gastrointestinal symptoms occurred in 3.66% of the patients, with one case (3.12%) in the control group and two cases (4.00%) in the observation group. This distribution aligns with prior studies, such as the study by Kong et al[9], in which the proportion of patients treated with RFA or MWA seeking treatment for abdominal pain or discomfort ranged from 7.56% to 11.63%. After treatment, 97.56% (80/82) of patients were asymptomatic, with 96.88% (31/32) in the control group and 98.00% (49/50) in the observation group. Only two patients (2.44%) reported persistent symptoms. The overall symptom improvement rate was approximately 93%; the control group improved from 50.00% to 3.12% (93.76% improvement), and the observation group improved from 28.00% to 2.00% (92.86% improvement). Shi et al[12] demonstrated that the complete symptom relief rate after MWA was 88.6%, compared to 69.2% in the transcatheter arterial embolization group. Consequently, interventional therapy for hepatic hemangiomas not only effectively alleviates symptoms but may also mitigate patients’ anxiety. This study evaluated changes in the psychological state and clinical symptoms of patients with hepatic hemangiomas before and after treatment, using widely validated scales, including the generalized anxiety disorder-7 and 36-Item Short Form Health Survey, to quantify changes in anxiety, depression, stress, and overall quality of life, with particular emphasis on anxiety relief and clinical symptom improvement. Among the 82 patients, nine (11.0%) exhibited anxiety prior to treatment. Following intervention, improvement was observed in six of these patients (66.7%), and only three (33.3%) continued to display anxiety. This finding is consistent with the treatment strategies outlined in the EASL guidelines[1] and corroborates the results of a multicenter study conducted by Tang et al[5] on the clinical symptoms and psychological status of patients with hepatic hemangiomas.

Short-term follow-up (within 12 months) revealed no recurrence in either group. However, during the 1- to 3-year follow-up period, one case of recurrence was observed in the control group, whereas no recurrence was observed in the observation group. The long-term follow-up imaging (Figure 2) confirms the durability of tumor control achieved with combined therapy. This trend indicates that embolization followed by ablation may help reduce the risk of long-term recurrence, particularly in patients with giant hemangiomas. Gao et al[4] highlighted that TACE for hepatic hemangiomas > 10 cm carries a risk of recurrence, while combined ablation reduces postoperative residue and enhances long-term tumor control. Some patients were evaluated as having a partial response or stable disease after treatment, as their tumors remained active. Additionally, some patients did not receive secondary treatment because of the decision to continue observation. Follow-up remains a reasonable strategy[9,14,29].

Although originally developed for malignancies, the RECIST standard provides valuable utility in assessing the treatment response of benign hepatic hemangiomas, particularly after interventional therapies[17]. RECIST measures changes in tumor diameter, specifically the reduction in the longest diameter, offering a standardized, reproducible method for evaluating therapeutic outcomes. However, RECIST has limitations in the context of benign tumors, such as hepatic hemangiomas, where treatment goals often focus on symptom relief and tumor growth inhibition rather than complete tumor eradication. Beyond RECIST measurements, volumetric assessment offers a potentially more precise method for quantifying the changes in the tumor burden, particularly for hepatic hemangiomas with irregular morphology, where three-dimensional measurements can more accurately reflect therapeutic efficacy than single-diameter assessments[14,16]. Furthermore, while the modified RECIST criteria[30] are currently validated primarily for assessing the response in hepatocellular carcinoma, their core principle of evaluating viable tumor tissue based on arterial phase enhancement could hold considerable relevance for appraising the treatment outcomes in highly vascular benign lesions such as hepatic hemangiomas following embolization and/or ablation. The modified RECIST criteria might provide a more comprehensive and accurate reflection of the true clinical value of TACE combined with MWA in this setting.

This study has several important limitations. First, this was a single-center retrospective cohort study with a limited sample size (n = 82). Although post-hoc power analysis for the primary efficacy endpoint (12-month ORR) demonstrated exceptionally high statistical power (> 0.999) for detecting the observed significant differences, the sample size might still pose limitations for the statistical power of analyses concerning certain secondary endpoints or within specific subgroups. Second, the efficacy assessment of the multiple hepatic hemangiomas focused primarily on the response of the index lesion. This may have incompletely captured precise changes in all secondary lesions and is an inherent challenge when retrospectively analyzing multiple lesions. Future studies may consider more detailed individualized follow-up of all treated lesions in patients with multiple lesions. Third, loss to follow-up was observed, particularly in patients with follow-up periods > 3 years. In addition, considerations of economic impact and healthcare accessibility are also paramount in clinical decision-making. Although this retrospective study did not include a formal health economic evaluation, future research should prioritize direct comparative studies evaluating TACE combined with MWA against surgical approaches, including laparoscopic resection. The lack of consensus on the optimal procedural protocols for combining TACE and MWA indicates that the results of this study may not fully represent the best possible clinical outcomes. Variability in the application of the combined treatment could affect the consistency of the results across different patient populations. Future multicenter, prospective randomized controlled trials are essential to clarify the ideal combination strategies, including determination of the optimal interval between TACE and MWA, ideal MWA power settings, and types of embolic agents that should be used.

CONCLUSION

In conclusion, TACE combined with MWA for the treatment of hepatic hemangiomas > 5 cm demonstrates superior efficacy compared to TACE alone, as evidenced by a higher tumor reduction ratio, complete response, and long-term local control. This combination therapy also effectively mitigates the risk of hemolytic complications potentially associated with MWA monotherapy, exhibiting an excellent safety and efficacy profile that offers an improved long-term prognosis for patients. Nevertheless, prospective, multicenter randomized controlled trials are warranted to definitively confirm these encouraging outcomes and refine patient selection criteria for this combined therapeutic strategy.