Thermal Ablation Alone vs Thermal Ablation Combined with Transarterial Chemoembolization for Patients with Small (<3cm) Hepatocellular Carcinoma

Abstract

Purpose:

Thermal ablation (TA) and transarterial chemoembolization (TACE) may be used alone or in combination (TACE+TA) for the treatment of hepatocellular carcinoma (HCC). The aim of our study was to compare the time to tumor progression (TTP) and overall survival (OS) for patients who received TA alone or TACE+TA for HCC tumors under 3cm.

Materials and Methods:

This HIPAA-compliant IRB-approved retrospective analysis included 85 therapy-naïve patients from 2010–2018 (63 males, 22 females, mean age 62.4±8.5 years) who underwent either TA alone (n=64) or TA in combination with drug-eluting beads (DEB)-TACE (n=18) or Lipiodol-TACE (n=3) for locoregional therapy of early stage HCC with maximum tumor diameter under 3cm. Kaplan-Meier analysis was performed using the log-rank test to assess TTP and OS.

Results:

All TA and TACE+TA treatments included were technically successful. TTP was 23.0 months in the TA group and 22.0 months in the TACE+TA group. There was no statistically significant difference in TTP (p=0.64). Median OS was 69.7 months in the TA group and 64.6 months in the TACE+TA group. There was no statistically significant difference in OS (p=0.14). The treatment cohorts had differences in AFP levels (p=0.03) and BCLC stage (p=0.047). Complication rates between patient groups were similar (p=0.61).

Conclusion:

For patients with HCC under 3cm, TA alone and TACE+TA have similar outcomes in terms of TTP and OS, suggesting that TACE+TA may not be needed for these tumors unless warranted by tumor location or other technical consideration.

1. INTRODUCTION

Minimally-invasive image-guided thermal ablation (TA) techniques such as radiofrequency ablation (RFA) and microwave ablation (MWA) are curative therapies used in patients with early-stage hepatocellular carcinoma (HCC) who are not undergoing resection or transplantation [1]. Transarterial chemoembolization (TACE) is a targeted, regional therapy generally used in patients with larger, intermediate-stage HCC. TA and TACE when used alone have both been proven effective as individual therapies with Level 1A evidence [2–4], but the combination of TACE+TA also has the potential to benefit certain patients. Various clinical trials have shown the benefit of combining TACE with RFA for patients with intermediate- and large-sized HCC (3.1 – 5.0 cm and >5 cm, respectively), with the combination reducing the number of treatments needed to achieve technical success and lowering the rates of local tumor progression [5].

In addition to utilizing combination TACE+TA for intermediate- and large-sized tumors, practitioners from many institutions will also perform TACE+TA for small-sized HCC [6, 7]. However, it is unclear whether combination therapy benefits this subset of patients. The number of studies investigating TACE+TA for HCC <3cm is limited, and current guidelines from both the European Association for the Study of the Liver and the American Association for the Study of Liver Diseases indicate that more research is necessary [8–10]. Therefore, the purpose of our study was to retrospectively compare the time to tumor progression (TTP) and overall survival (OS) for patients who received TA alone or TACE+TA for HCCs <3cm.

2. MATERIALS AND METHODS

2.1. Patients

In this HIPAA-compliant IRB-approved retrospective study, medical records from 85 patients from 2010–2018 were collected who underwent either TA alone (n=64) or either drug-eluting beads (DEB)-TACE (n=18) or Lipiodol-TACE (n=3) treated consecutively with TA for locoregional therapy of early stage HCC. Included patients had a maximum single tumor diameter <3cm. TA modalities included MWA (n=35) and RFA (n=50). For the entire patient cohort, baseline tumor sizes ranged from 1.0–2.9cm, with a mean of 2.0±0.5cm. Tumors were considered subcapsular if the margin of the tumor abutted the capsule of the liver on CT or MRI. Tumors were considered next to vital or limiting structures if they were located within 1cm of the heart, gallbladder, central bile duct, IVC, hepatic artery or portal vein.

Sixty-three male and 22 female patients were included in the analysis. Ages ranged from 46 to 82 years, with mean age 62.4±8.5 years. Patients had Child-Pugh classifications of A (n=65) and B (n=20); BCLC stages were 0 (n=36) and A (n=49). HCC etiologies included hepatitis B virus (HBV) alone (n=2), hepatitis C virus (HCV) alone (n=38), ethanol (EtOH) alone (n=13), EtOH and HCV (n=15), non-alcoholic steatohepatitis (NASH, n=6), or other (n=11). The average total bilirubin (1.3±1.2 mmol/l), albumin (3.7±0.6 g/l), alanine aminotransferase (55.7±47 u/l), aspartate aminotransferase (72.3±53 u/l), alkaline phosphatase (118±57 u/l), gamma-glutamyl transferase (129±216 u/l), and alpha-fetoprotein (105±633 ng/ml) levels were recorded. Other baseline clinical characteristics are included in Table 1.

Table 1.

Baseline Patient Characteristics

	Total	TA	TACE + TA	p (Corrected)
	(n=85)	(n=64)	(n=21)	α = 0.0028
Sex, N (%)				0.37
Male	63 (74.1)	49 (76.6)	14 (66.7)
Female	22 (25.8)	15 (23.4)	7 (33.3)
Age				0.25 (1.0)
Years	62.4 ± 8.5	61.7 ± 8.9	64.2 ± 7.2
Transplant/Resection, n (%)				0.25
Yes	7 (8.2)	4 (6.3)	3 (14.3)
No	78 (91.8)	60 (93.7)	18 (85.7)
Recurrence, n (%)				0.66
Yes	52 (61.2)	40 (62.5)	12 (57.1)
No	33 (38.8)	24 (37.5)	9 (42.9)
Eitiology, n (%)				0.42
HBV	2 (2.4)	1 (1.6)	1 (4.8)
HCV	38 (44.7)	28 (43.8)	10 (47.6)
EtOH	13 (15.3)	11 (17.1)	2 (9.5)
EtOH and HCV	15 (17.6)	13 (20.3)	2 (9.5)
NASH	6 (7.1)	3 (4.7)	3 (14.3)
Other	11 (12.9)	8 (12.5)	3 (14.3)
Child-Pugh Score , n (%)				0.97
A	65 (76.4)	49 (76.6)	16 (76.2)
B	20 (23.5)	15 (23.4)	5 (23.8)
BCLC Stage, n (%)				0.047
0	36 (42.4)	31 (48.4)	5 (23.8)
A	49 (57.6)	33 (51.6)	16 (76.2)
Complication, n (%)				0.63
Yes	6 (7.1)	4 (6.2)	2 (9.5)
No	79 (92.9)	60 (93.8)	19 (90.5)
Number of Tumors, n (%)				0.21
1	64 (75.3)	50 (78.1)	14 (66.7)
2	17 (20.0)	11 (17.2)	6 (28.6)
3	4 (4.7)	3 (4.7)	1 (4.8)
Tumor Characteristics
Subcapsular Location, n (%)	22 (25.9)	15 (23.4)	7 (33.3)	0.37
Proximity to Limiting Structure, n (%)	13 (15.3)	7 (10.9)	6 (28.6)	0.051
Diameter of Largest Tumor (cm)	2.0 ± 0.5	1.9 ± 0.5	2.2 ± 0.5	0.01 (0.14)
Cumulative Tumor Diameter (cm)	2.4 ± 1.1	2.2 ± 1.1	2.7 ± 1.0	0.01 (0.09)
Clinical Characteristics
total bilirubin (mmol/l)	1.3 ± 1.2	1.4 ± 1.3	1.1 ± 0.6	0.46 (1.0)
albumin (g/l)	3.7 ± 0.6	3.7 ± 0.6	3.8 ± 0.5	0.66 (1.0)
ALT (u/l)	55.7 ± 47	54.7 ± 52	58.4 ± 30	0.08 (1.0)
AST (u/l)	72.3 ± 53	70.8 ± 52	76.7 ± 58	0.33 (1.0)
ALP (u/l)	118 ± 57	125 ± 62	96.3 ± 31	0.07 (1.0)
GGT (u/l)	129 ± 216	144 ± 245	83.1 ± 63	0.37 (1.0)
AFP (ng/ml)	105 ± 634	27.6 ± 63	323 ± 1200	0.001 (0.03)
creatinine (mg/dl)	0.90 ± 0.3	0.9 ± 0.3	0.88 ± 0.2	0.48 (1.0)
sodium (mmol/l)	138 ± 3.1	139 ± 3.0	137 ± 3.4	0.12 (1.0)
INR	1.1 ± 0.2	1.1 ± 0.2	1.2 ± 0.2	0.18 (1.0)
PT	12.3 ± 1.7	12.2 ± 1.7	12.6 ± 1.6	0.16 (1.0)
Platelets (x1000)	118 ± 83	125 ± 90	96.8 ± 53	0.41 (1.0)
Neutrophil (%)	63.1 ± 14	62 ± 13	66.6 ± 16	0.19 (1.0)
Lymphocyte (%)	24.3 ± 12	25 ± 11	22.1 ± 13	0.32 (1.0)

Baseline clinical characteristics of patient cohort. HCV, chronic hepatitis C; EtOH, alcoholic liver disease; HBV, chronic hepatitis B; NASH, non-alcoholic steatohepatitis; BCLC, Barcelona Clinic Liver Cancer; ALT, alanine aminotransferase; AST, aspartate aminotransferase; ALP, alkaline phosphatase; GGT gamma-glutamyl transferase; AFP, alpha-fetoprotein; INR, international normalized ratio; PT, prothrombin time. P values under 0.05 were considered significant after correction. Categorical values between groups were compared using Pearson’s Chi-squared or Fisher-Freeman-Halton where appropriate. Continuous values were compared between groups using a Students t-test or Mann-Whitney U where appropriate, with Bonferroni correction applied.

Patients were included if they had three or fewer tumors under 3cm prior to receiving TA or TACE+TA. HCC was diagnosed either by Organ Procurement and Transplantation Network (OPTN) imaging criteria or histopathological proof. Included patients did not have any evidence of residual tumor during their first post-procedure follow-up. At least one year of total follow-up time was required for inclusion in the study.

2.2. TA and TACE+TA Protocols

TA modalities included MWA and RFA performed under ultrasound and/or CT guidance. These procedures were performed by multiple fellowship-trained interventional radiologists with up to 21 years of experience. Either LeVeen (Boston Scientific, Marlborough, MA) or Coviden Cool-Tip (Medtronic, Dublin, Ireland) probes were used for RFA. Neuwave (Ethicon, Somerville, NJ) probes were used for MWA. For patients undergoing ablation alone, patients received either RFA (n=35) or MWA (n=29). A subset of RFA patients (n=13 out of n=35 total RFA patients) underwent laparoscopic RFA if this approach was recommended by the tumor board, and these procedures were performed by a surgical oncologist with over 20 years of experience. For TA procedures, the appropriate ablation margins were determined intraprocedurally by the attending physician. On US, an appropriate ring of hypodensity surrounding the ablated lesion was seen. Alternatively, either contrast enhanced CT or comparison to pre-treatment CT using appropriate anatomic landmarks was used to demonstrate appropriate margins.

For TACE+TA patients, TA occurred within 24 hours after the TACE procedure. Patients were either admitted overnight for a TA procedure the next day, or had TA performed immediately following TACE. TACE procedures were performed by multiple fellowship-trained interventional radiologists with up to 21 years of experience. TACE was performed using either drug-eluting LC beads (n=18, Boston Scientific, Marlborough, MA) or lipiodol (n=3, Guerbet, Villepinte, France). The total volume of embolization agent used, as well as the specific ratio of chemotherapy to embolization agent, was left to the clinical discretion of the interventional radiologist. For patients receiving TACE+TA, ablation modalities included RFA (n=15) and MWA (n=6). For TACE+TA procedures, the appropriate ablation margins were determined intraprocedurally by the attending physician. On US, an appropriate ring of hypodensity surrounding the ablated lesion was seen. Alternatively, contrast enhanced CT or comparison to pre-treatment CT using the appropriate anatomic landmarks was used to demonstrate appropriate margins.

For both TA and TACE+TA groups, ablation was performed until appropriate margins were achieved. In most cases, given the small tumor sizes, a single probe was used and repositioned rather than using additional probes. However, in those cases where an attending physician believed that an additional probe was necessary, ablation was performed in similar fashion until the appropriate margin was achieved.

Radiologic response was determined by follow-up imaging one month after the procedure. Imaging modalities included multiphasic CT or MR imaging. Image studies were examined for evidence of recurrent or residual tumor. Patients had follow-up office visits with the physician who performed the procedure and were evaluated for transplant eligibility. Patients were censored from TTP evaluation if they received a transplant. Figure 1 demonstrates MR imaging from a patient undergoing TA alone. Figure 2 demonstrates MR imaging from a patient undergoing TACE+TA.

Fig 1.

Pre-procedure (A), intra-procedure (C, D), and post-procedure (D) imaging of 53-year-old male who underwent US-guided RFA for HCC secondary to chronic Hepatitis C infection. Prior to treatment, a 2cm lesion with arterial enhancement (A), washout, pseduocapsule, and hepatobiliary phase hypointensity on MRI was noted at the borders of segment 7 and 8, consistent with OPTN 5B HCC. The patient was selected to undergo US guided RFA. The lesion was first visualized (B), biopsied, and then ablated (C) using a 15-gauge 4.0 cm LeVeen radiofrequency ablation device, first at a maximum of 130 watts at 4 minutes and 25 seconds before roll-off was achieved and then at 110 wats for 8 minutes and 49 seconds before roll-off was achieved. The patient underwent the procedure without complication. One month after the procedure, follow-up MRI was performed which did not demonstrate evidence of residual tumor at the ablation site in arterial phase (D) or portal venous phase. There was a small region of arterial enhancement lateral to the ablation cavity on arterial phase imaging (D), however this was non-specific and likely related to post-biopsy hyperemia. After 6 years of regular imaging follow-up this patient did not have recurrence of HCC.

Fig 2.

Pre-procedure (A), intra-procedure (B, C, D), and post-procedure (E) imaging for 60-year-old female who underwent TACE with subsequent CT-guided RFA for HCC secondary to chronic Hepatitis C infection. Prior to treatment, a 2.1cm lesion with arterial enhancement (A), washout, pseduocapsule, and hepatobiliary phase hypointensity on MRI was noted at segment 8, consistent with OPTN 5B HCC. Angiography was performed and a hypervascular mass was noted in the right hepatic lobe (B). A sub-branch of the right hepatic artery was treated with a 25mg of doxorubicin mixed with lipiodol. Prior to stasis, 100–300 micron Embospheres were added to the mixture of lipiodol and doxorubicin. Stasis was confirmed with post-embolization angiogram (C). Immediately after the TACE procedure, the patient was admitted to the hospital and CT-guided RFA was performed the next day with 2 cycles at 2 stations. A 15-gauge 3.5cm LeVeen RFA probe was used under CT guidance (D), treated at 140 watts for 7 minutes and 17 seconds until roll-off was achieved. A second ablation followed at 98 watts for 2 minutes and 16 seconds. A second application was then performed, first at 130 watts at 6 minutes and 35 seconds until roll-off, followed at 91 watts for 1 minute and 55 seconds. After the procedure a thin subcapsular hematoma was noted, but otherwise proceeded without complication. Follow-up MRI at 2 months did not reveal evidence of residual disease at the ablation site in either arterial (E) or portal venous phase. The patient underwent liver transplantation 19 months after the procedure and was censored from analysis. The patient did not show any evidence of tumor recurrence before this time.

To assess for the presence of procedural complications, procedure notes, discharge summaries, and follow-up notes were analyzed up to 30 days after the TA or TACE+TA or procedure. Patients received either CT or MRI approximately one month after procedures were completed, and then further imaging follow-up schedules were determined on a per-patient basis at the clinical discretion of the attending provider.

2.3. Statistical Analysis

Baseline clinical characteristics were reported using descriptive statistics. The primary endpoint analyzed was time to tumor progression (TTP) and overall survival (OS) after TA or TACE+TA therapy. TTP and OS were analyzed using the Kaplan-Meier method with log-rank test. For TTP analysis, patients were censored at time of death, liver transplant, or loss to follow-up. For OS analysis, patients were censored at the time of last follow-up if information regarding their survival status was unavailable. Between the TA and TACE+TA groups, clinical characteristics which could be represented as continuous variables were compared using Student’s t-test or Mann-Whitney U test. These results were then adjusted using Bonferroni correction to account for the number of continuous variables compared. Clinical characteristics represented as categorical variables were evaluated using Pearson’s Chi-square analysis or the Fisher-Freeman-Halton test. P-values under 0.05 divided by the total number of variables analyzed (n=18, α = 0.0028) were considered significant for continuous variables. For categorical variables, p-values under 0.05 were considered significant. Statistical analysis was performed using Python 3.7 with the SciPy library as well as SPSS Version 26. Kaplan-Meier curves were generated using GraphPad Prism 8.2.1.

3. RESULTS

All patients demonstrated absence of recurrent or residual tumor one month post-procedure as evidenced by multiphasic CT or MR imaging. At the time of analysis, 29 patients (34.1%) died and 56 (65.9%) were still alive. The median OS for the cohort was 69.7 months with a median follow-up time of 51.9 months. In the TA group, 17 (26.6%) patients died at the time of analysis and the median OS was 69.7 months with a median follow-up time of 41.2 months. In the TACE+TA group, 12 (57.1%) patients died at the time of analysis and the median OS was 64.6 months with a median follow-up time of 67.0 months. There was no significant difference in OS between groups with the log-rank test returning a p-value of 0.14 (Figure 3).

Figure 3.

Kaplan-Meier curve illustrating overall survival after TA and TA+TACE. Log-rank test returned a p-value of 0.14.

For the entire patient cohort, the median TTP was 23.4 months. At the time of analysis, 52 (61.2%) patients within the entire cohort experienced tumor recurrence. In the TA alone group, the median TTP was 23.0 months, with 40 (62.5%) patients having recurred at the time of analysis. In the TACE+TA group, the median TTP was 22.0 months with 12 (57.1%) patients recurring at the time of analysis. There was no significant difference in TTP between TA and TACE+TA groups (p=0.64, Figure 4). In the TA group, 4 patients (6%) were ultimately transplanted or resected. In the TACE+TA group, 3 patients (14%) were ultimately transplanted or resected. There was no statistically significant difference between transplant or resection rates in the two patient groups (p=0.25).

Figure 4.

Kaplan-Meier curve illustrating time to tumor recurrence after TA and TA+TACE. Log-rank test returned a p-value of 0.64. Patients were censored at time of death or liver transplant.

For the entire patient cohort, baseline tumor sizes of the single largest tumor ranged from 1.0–2.9cm, with a mean of 2.0±0.5cm. With respect to the number of tumors, n=64 (75.3%) had one tumor, n=17 (20.0%) had two tumors, and n=4 (4.7%) had three tumors. In the TA group, baseline sizes of the largest single tumor ranged from 1.0–2.9cm, with a mean of 1.9±0.5cm. When considering the cumulative diameter of multiple tumors, the sum of the diameters ranged from 1.0–7.2cm, with a mean of 2.2±1.1. In the TACE+TA group, baseline tumor sizes of the single largest tumor ranged from 1.1 to 2.9 cm, with a mean of 2.2±0.5. When considering the cumulative diameter of multiple tumors, the sum of the diameters ranged from 1.5–5.5cm, with a mean of 2.7±1.0. Before Bonferroni correction, there was a significant difference between TA and TACE+TA for the diameter of the largest single tumor (p=0.01). However, after Bonferroni correction this value became insignificant (p=0.14). There was also a significant difference between TA and TACE+TA for the cumulative tumor diameter before correction (p=0.01) which later disappeared (p=0.09). There was no significant difference between TA and TACE+TA with respect to the number of tumors (p=0.21).

For the entire patient cohort, tumors were in a subcapsular location in 22 (25.9%) patients. Tumors were located next to limiting structures in 13 (15.3%) patients. In the TA group, 15 (23.4%) patients had tumors in a subcapsular location and 7 (10.9%) had tumors located next to limiting structures. In the TACE+TA group, n=7 (33.3%) of patients had tumors in a subcapsular location and n=6 (28.6%) had tumors located next to a limiting structure. There was no significant difference between the presence of tumors in subcapsular locations (p=0.37) or tumors located next to limiting structures (p=0.051).

In the TA group, Child-Pugh scores were A (n=49, 76.6%) and B (n=15, 23.4%). BCLC stages were 0 (n=31, 48.3%) and A (n=33, 51.6%). In the TACE+TA group, CP scores were A (n=16, 76.2%), and B (n=5, 23.8%). BCLC stages were 0 (n=5, 23.8%) and A (n=16, 76.2%). There was no statistically significant difference between groups for CP scores (p=0.97). However, BCLC stage differed significantly between the TA and TACE+TA group (p=0.047). There was a statistically significant difference in AFP between the TA and TA+TACE groups (p=0.001) which did not disappear after correction (p=0.03).

In the TA group, 4 (6.2%) patients experienced a procedural complication. These complications included hemorrhage (n=2), pneumothorax (n=1), and hemoperitoneum (n=1). Intraprocedural probe repositioning and re-ablation was required n=6 patients (9.3%). In the TACE+TA group, 2 (9.5%) patients experienced a procedural complication. These complications included an access site hematoma (n=1) and a subcapsular hematoma (n=1). Intraprocedural probe repositioning and re-ablation was required in n=1 patient (4.7%). There was no statistically significant difference in complication rate between the TA and TACE+TA groups (p=0.63).

4. DISCUSSION

This study retrospectively compared OS and TTP in patients who received either TA or TACE+TA for early-stage HCC <3cm. Our findings imply that tumors <3cm do not seem to require TACE in addition to TA. Combination therapy is an emerging treatment option for patients with HCC, but the majority of studies investigating TACE+TA concern larger tumors [11, 12]. A meta-analysis of seven randomized control trials by Lu et al. indicated that HCC survival rates at 1, 3, and 5 years were substantially improved with TACE+TA vs TA alone, without a significant difference in major complications [13]. However, this effect was only noted for tumors >3cm. The study found no difference in survival rates for patients with HCC <3cm, indicating that TACE may not be important for the success of TA for smaller tumors. The usefulness of TACE+TA has been explored in a more limited capacity for patients with HCC <3cm. Shibata et al. randomized 93 patients with HCC <3cm to receive either TACE+TA or TA alone and found no statistically significant differences in LTP, OS, local PFS, or event-free survival rates [14]. However, it has been noted that this study may be biased towards lesions <2cm, potentially underrepresenting tumors 2–3cm which are more likely to recur [15]. Kim et al. examined TACE+TA vs TA alone for 314 patients specifically with tumors 2–3cm in diameter and found that combination treatment offered superior local tumor control, but no statistically significant improvement in survival [16]. Kim et al. performed their study at a Korean institution with HBV as the primary disease etiology. However, our patient cohort is primarily composed of patients with HCV and EtOH-related HCC, which may explain observed differences in local tumor control. Peng at al. conducted a randomized control trial of 189 patients and found that TACE+TA offered better OS and PFS in patients with tumors under 7cm [17].

Despite the inconclusive evidence, certain institutions will still perform TACE+TA for HCC <3cm, as retrospective studies by Ginsberg et al. and Biederman et al. indicate [6, 7]. Ginsberg et al. performed a retrospective analysis comparing radiofrequency and microwave ablation within the context of TACE+TA combination therapy. The median tumor size in their cohort was 3.1cm, meaning approximately half of their included patients had tumors <3cm. OS for the entire patient cohort was 39.0 months. This result cannot easily be compared with our data, as Ginsberg et al. did not report OS by tumor size, and patients with BCLC Stages A, B, and C were included. Biederman et al. identified 80 patients who received TACE+TA specifically for HCC <3cm at their institution over the course of five years. The authors cite studies which show improved effectiveness and OS with TACE+TA for tumors of various sizes as a rationale for using combination therapy [11, 17]. However, the authors of this study concede that the marginal benefit of combining TACE with TA is unknown for patients with small HCC and advocate for additional research which investigates this specific question. Their analysis indicated that TACE+TA was not significantly different from radiation segmentectomy in terms of TTP or OS. The studies by Ginsberg et al. and Biederman et al. indicate that TACE+TA is frequently performed for patients with HCC <3cm, despite unclear evidence supporting its use.

Our data suggests that TACE+TA is not warranted in patients with small HCC, as the choice of treatment did not lead to differences in OS (69.7 months vs 64.6 months, p = 0.14) or TTP (23 months vs 22 months, p = 0.64). We acknowledge that specific concerns regarding tumor visualization or proximity of tumor to vital or limiting anatomic structures may still warrant combination therapy in select patients with small tumors [18–21]. However, in our patient cohort we did not find that patients receiving TACE+TA had a significantly higher proportion of tumors located in a subcapsular location (p=0.37) or next to a vital structure (p=0.051). AFP levels were significantly higher in the TACE+TA group (p=0.03). The TACE+TA group also had a greater proportion of BCLC A patients (p=0.047). These factors may have contributed to the decision to offer combination therapy, as higher AFP levels [22] and more advanced BCLC stage [23] are associated with poorer prognosis. In our patient cohort, however, we did not appreciate differences between in TTP or OS between these groups. We suspect that a strong outlier with an AFP level of 5,760 ng/mL in the TACE+TA group accounts for the difference in AFP level between the TA and TACE+TA cohorts. The distribution of patients between BCLC stages reaches only borderline significant difference, so it is possible that we would not be able to observe differences in outcome with the available patient sample.

TACE+TA may not be a cost-effective treatment option for patients with HCC <3cm, as similar clinical outcomes can be achieved with TA alone. Performing TACE+TA requires additional procedure room time, instrumentation, and medication. While a detailed cost-effectiveness analysis is outside the scope of this paper, further research may offer additional insight.

There are several limitations to our study. The retrospective, single-institution nature of our analysis limits the number of patients available to power our study. It also risks sampling bias. In addition, it is possible that there were some differences in tumor size between our TA and TACE+TA groups. The initial p-value of 0.01 for the diameter of largest tumor and cumulative tumor diameter disappeared after Bonferroni correction. However, Bonferroni correction tends to be conservative and may increase the likelihood of making a type 2 error [24]. Finally, there is a borderline difference in proximity of tumors to limiting anatomic structures between the TA and TACE+TA groups. In the TA group, 7 (10.9%) patients had tumors located next to limiting structures, while in the TACE+TA group, 6 (28.6%) patients did (p=0.051). This suggests that patients who received TACE+TA may have often received this combination of therapy due to the location of the lesion, potentially biasing our results. Since lipiodol used during cTACE is radiopaque, deposition of lipiodol within a tumor can allow for improved visualization during subsequent TA. Proper visualization becomes important if a tumor is next to a limiting structure which may be sensitive to thermal damage.

In conclusion, this study found no difference between TA and TACE+TA with respect to TTP or OS, implying that TACE+TA may not be necessary unless tumor location or difficulty in tumor visualization necessitates the use of TACE to ensure completeness of treatment.

Highlights.

• For hepatocellular carcinoma <3cm, the combination of transarterial chemoembolization combined with thermal ablation does not improve overall survival when compared with thermal ablation alone.

• For hepatocellular carcinoma <3cm, the combination of transarterial chemoembolization combined with thermal ablation does not improve time to tumor progression when compared with thermal ablation alone.

• Small hepatocellular carcinoma in most situations can be treated with thermal ablation alone, except when the location of the tumor precludes this approach or additional visualization is needed.