Percutaneous Radiofrequency Ablation with or without Chemolipiodolization for Early-stage Hepatocellular Carcinoma: A Propensity Score-matched Analysis

Simple Summary: Percutaneous radiofrequency ablation (PRFA) is a useful treatment option for early-stage hepatocellular carcinoma. Additionally, chemolipiodolization (CL) is also reported to be useful in preventing recurrence. However, treatment outcomes of PRFA with or without CL for early-stage hepatocellular carcinoma have remained unclear thus far. In this study, we used propensity score-matched analysis to determine prognostic factors for overall survival after PRFA with or without CL. Our results showed that there was no significant difference between the two groups, indicating that PRFA with CL has no prognostic effects than PRFA without CL for early-stage hepatocellular carcinoma. Abstract: We aimed to compare prognostic factors for overall survival (OS) following percutaneous radiofrequency ablation (PRFA) with or without chemolipiodolization (CL) for early-stage hepatocellular carcinoma (HCC) using propensity score-matched analysis. We enrolled 221 patients with early-stage HCC who received PRFA with (n = 76) or without (n = 145) CL in Saga Central Hospital between April 2004 and December 2020. No significant difference was observed in OS between PRFA with and without CL cohorts (median survival time [MST]: 5.4 vs. 4.5 years; p = 0.0806). To reduce confounding effects, 108 patients were selected using propensity score-matched analysis (n = 54 for each treatment). No significant difference was observed in OS between the cohorts (MST: 3.6 vs. 4.0 years; p = 0.5474). After stratification according to tumor size, no significant difference was observed in OS for patients with tumor size ≥20 mm between PRFA with and without CL cohorts (MST: 3.4 vs. 3.5 years; p = 0.8236). PRFA with CL was not a significant prognostic factor in both univariate and multivariate analyses (p = 0.5477 and 0.9600, respectively). Our findings suggest that PRFA with CL does not demonstrate longer prognostic effects than PRFA without CL in early-stage HCC, regardless of tumor size. shows the results of univariate and multivariate analyses of OS obtained using propensity score-matched analysis (n = 108). Univariate analyses of OS revealed four variables as significant prognostic factors: Child–Pugh class (p = 0.0012), ALBI score (p <

Percutaneous RFA (PRFA) is a simple and minimally invasive treatment option for early-stage HCC [7]. Moreover, it is expected that combination therapy with PRFA and transarterial chemoembolization (TACE) can improve survival in patients with earlystage HCC [7]. In a prospective randomized trial, RFA combined with TACE was superior over RFA alone in improving the survival of patients with HCC [8]. Theoretically, it is possible to diminish the cooling effect of blood and consequently expand the area of ablation by performing TACE before PRFA. Recently, there has been accumulating evidence regarding the efficacy of PRFA with TACE for treating early-stage HCC [9][10][11].
Chemolipiodolization (CL) is performed before embolization in TACE as standard, which is a common practice in the treatment of HCC [12][13][14][15]. However, several studies have reported that CL plays a major role and embolization does not improve the survival [16,17], whereas other studies have drawn contrary conclusions [18]. CL is also performed before hepatic resection, and preoperative CL of the entire liver has been reported to be effective in reducing the incidence of postoperative recurrence and prolonging survival in patients with resectable HCC [19]. Thus, the evaluation of CL for HCC remains controversial.
CL for HCC is considered an advantage, in that the effect on the body is smaller than that in TACE because no embolization is performed. However, treatment outcomes of PRFA with or without CL for early-stage HCC remain unclear. Therefore, in this study, we aimed to determine the prognostic effects of PRFA with or without CL and the associated overall survival (OS) duration in early-stage HCC. In view of this, to reduce confounding effects, we performed propensity score-matched analysis.

Ethical Approval
The study was approved by the Ethics Committee of Saga Central Hospital (No. 21003) and was conducted according to the guidelines of the 1975 Declaration of Helsinki. more PRFA; therefore, we enrolled only consecutive patients who received PRFA at the first treatment.

Patients Receiving CL
Among the enrolled patients who received PRFA, some patients received CL without embolization the day before PRFA. CL was performed according to the same procedure as that described in previous studies [15,16,19]. After conventional visceral angiography, CL was performed by introducing an angiographic catheter into the feeding artery of HCC using Seldinger's technique. Angiographic survey of abdominal vessels such as the superior mesenteric artery and common hepatic vessels was performed to assess the arterial blood supply to the liver. Epirubicin was mixed water-soluble contrast medium and sterile water for injection. The mixture was thoroughly mixed with lipiodol. The injection was stopped either at the point of near stasis within the feeding artery of HCC or after the entire amount of the agent was administered.

Treatment Outcome
The treatment outcome was OS, defined as the time from the initiation of PRFA with or without CL to the date of death or the patient's last follow-up, whichever occurred first.

Statistical Analysis
The following baseline patient characteristics were analyzed using descriptive statistical methods: age, tumor size, albumin level, total bilirubin level, ALBI score, prothrombin time, AFP, and DCP. The data of these variables were calculated using the t-test, and the data of variables such as sex, etiology, Child-Pugh class, and BCLC stage were calculated using the chi-square test. Results are expressed as mean ± standard deviation (SD) and median (range) or n (%). Survival curves were constructed using the Kaplan-Meier analysis with the log-rank test. A p-value of <0.05 was considered to indicate statistical significance. JMP software (SAS Institute, Inc., Cary, NC, USA), version 15, was used for all statistical analyses.

Patient Characteristics
Overall, 501 patients were treated for early-stage HCC with PRFA, of whom 280 who received second or more PRFA were excluded; therefore, 221 consecutive patients who received PRFA at the first treatment were enrolled. Table 1 shows the characteristics of the 221 consecutive patients who were diagnosed with early-stage HCC and received either PRFA with (n = 76) or without (n = 145) CL. A higher proportion of patients tested positive for the large tumor size (p < 0.0001) in the PRFA with CL cohort, whereas a higher proportion of patients had BCLC stage 0 (p < 0.0001) in the PRFA without CL cohort. Age, sex, etiology, Child-Pugh class, albumin level, total bilirubin level, ALBI score, prothrombin time, AFP level, and DCP level were equivalent between the PRFA with and without CL cohorts. Results are expressed as the mean ± standard deviation and the median (range) or n (%).

Propensity Score-Matched Analysis
To reduce confounding effects, we performed propensity score-matched analysis to match patients treated with PRFA with CL (n = 76) with those treated with PRFA without CL (n = 145) [24,25]. The following 12 variables related to the prognosis of earlystage HCC were considered at the start of the follow-up: age, sex, etiology, Child-Pugh class, tumor size, BCLC stage, albumin level, total bilirubin level, ALBI score, prothrombin time, AFP level, and DCP level. The propensity scores (mean ± SD) of the patients treated with PRFA with or without CL were 0.4036 ± 2.3596 and -2.2509 ± 4.5160, respectively. We used these propensity scores to conduct one-to-one nearest neighbor matching within a caliper of 0.20, as previous studies have shown this SD percentage of the logit of the propensity score to be generally suitable as a caliper for propensity score-matched analysis [26]. Based on the results of the propensity scorematched analysis, 108 patients were selected (PRFA with CL, n = 54; PRFA without CL, n = 54). After the propensity score-matched analysis, the propensity scores (mean ± SD) of patients treated with PRFA with or without CL were -0.3854 ± 0.9998 and -0.4535 ± 0.9328, respectively. Table 2 shows the characteristics of 108 patients who were diagnosed with early-stage HCC and underwent PRFA with CL (n = 54) or PRFA without CL (n = 54) assessed using propensity score-matched analysis. No significant differences were observed for any variables between the PRFA with and without CL cohorts using propensity scorematched analysis. Results are expressed as the mean ± standard deviation and the median (range) or n (%).   Figure 3 shows the results of the Kaplan-Meier analysis of OS assessed with the log-rank test between the PRFA with and without CL cohorts in patients with a tumor size of ≥20 mm using propensity score-matched analysis. For patients with a tumor size of ≥20 mm, the MST was 3.5 years in the PRFA with CL cohort (red line; n = 26) and 3.4 years in the PRFA without CL cohort (blue line; n = 24) (p = 0.8236). The OS did not differ significantly between the PRFA with and without CL cohorts in patients with a tumor size of ≥20 mm after propensity score-matched analysis.  0.0001), AFP level (p = 0.0032), and DCP level (p = 0.0119). Multivariate analyses of OS identified two variables as independent significant prognostic factors: Child-Pugh class (p = 0.0245) and DCP level (p = 0.0442). PRFA with CL was not a significant prognostic factor in both univariate and multivariate analyses in propensity score-matched analysis. Abbreviations: OS = overall survival, HR = hazard ratio, CI = confidence interval, HCV = hepatitis C virus, BCLC = Barcelona Clinic Liver Cancer, ALBI = albumin-bilirubin, AFP = alpha-fetoprotein, DCP = des-gamma-carboxy prothrombin, CL = chemolipiodolization, PRFA = percutaneous radiofrequency ablation.

Univariate and Multivariate Analyses of OS After Propensity Score-Matched Analysis
All cutoff values were set for the median.

Discussion
In this study, we observed that the OS did not differ significantly between the PRFA with and without CL cohorts among the enrolled patients ( Figure 1). However, in the PRFA without CL cohort, tumor factors, especially tumor size, were significantly better (Table 1). To reduce confounding effects, we performed propensity score-matched analysis to match patients treated with PRFA with and without CL (Table 2). Again, the OS did not differ significantly between the PRFA with and without CL cohorts after propensity score-matched analysis (Figure 2). Similarly, the OS did not differ significantly between the PRFA with and without CL cohorts in patients with a tumor size of ≥20 mm after propensity score-matched analysis ( Figure 3). Furthermore, our results suggest that PRFA with CL does not demonstrate longer prognostic effects than PRFA without CL in early-stage HCC, regardless of the tumor size. Additionally, in both univariate and Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 4 October 2021 doi:10.20944/preprints202110.0038.v1 multivariate analyses, PRFA with CL was not a significant prognostic factor after propensity score-matched analysis (Table 3). There are several treatment strategies for early-stage HCC, such as PRFA and surgical resection. PRFA is an effective therapy for early-stage HCC wherein the tumor size is <3 cm in diameter and the number of tumors is <3 [7]. On the contrary, CL is an effective treatment option for unresectable HCC [27]. PRFA with TACE is reported as an effective treatment option for HCC [28]. PRFA combined with TACE was superior over PRFA alone in improving the survival of patients with HCC of <7 cm [8]. Percutaneous microwave coagulonecrotic therapy with TACE can also be used to effectively treat an HCC measuring >2 cm but <3 cm [29]. A meta-analysis indicated that PRFA combined with TACE can demonstrate higher tumor response rates and improvement in survival [30]. On the contrary, in a propensity score-matched study, embolization in TACE combined with PRFA could not improve the survival of patients with HCC within the Milan criteria [15]. However, only few studies have reported improved OS by PRFA with CL. Our study identified no significant value of adding CL to PRFA.
The reason why there was no significant difference in the OS of patients who underwent PRFA with CL is the induction of the cooling effect, although a previous study showed prolongation of survival in patients undergoing PRFA with TACE [8]. Obstruction of the hepatic artery is effective in reducing the cooling effect because the hepatic artery provides the main blood supply to HCC; therefore, PRFA is sufficient to expand the area of necrosis after hepatic artery obstruction [31]. Because CL is a procedure that does not use embolic substances, it is not good at reducing the cooling effect of the hepatic blood flow on PRFA, attributing to the diminishing of the cauterization effect. The removal of embolization in the combination therapy with TACE and RFA is not important for improving the survival rate of patients with HCC [15]. In our study, there was no significant difference between PRFA with CL and PRFA without CL; therefore, this may suggest the importance of embolization with PRFA.
In summary, PRFA in combination with CL could not improve OS in patients with early-stage HCC. The reason is that CL does not involve the use of embolic substances, which does not reduce the cooling effect. In addition, CL increases the risk of complications such as bleeding and infections, including biloma and skin abscess, and may increase medical expenses.
The present study had some limitations. First, our study had a single-center retrospective design with a relatively small sample size (n=221) for patients with earlystage HCC. Second, the treatment (PRFA with or without CL) was selected at the discretion of the chief physician, and patients were not randomized to the treatment. This resulted in a selection bias for patients with early-stage HCC. Third, therapeutic effects and adverse events in all cases could not be evaluated. Fourth, no further investigations were conducted after secondary treatment. Therefore, to overcome these limitaitons, a multicenter prospective study with a larger patient population should be conducted in the future.

Conclusions
The present study showed that PRFA with CL does not demonstrate longer prognostic effects than PRFA without CL for early-stage HCC. Our results suggest that PRFA with CL should not be employed for patients with early-stage HCC, regardless of the tumor size.

Funding: Not applicable
Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Ethics Committee of Saga Central Hospital (protocol code 21003 approved on 21 June 2021).
Informed Consent Statement: Informed consent was obtained from all patients regarding treatment. However, the consent regarding the study was waived because of the retrospective study design.
Data Availability Statement: The data that support the findings of this study are available from the corresponding author, M.Nakano, on reasonable request.