Efficacy and safety of HAIC alone vs. HAIC combined with lenvatinib for treatment of advanced hepatocellular carcinoma

To investigate efficacy and safety of hepatic arterial infusion chemotherapy combined with lenvatinib (HAIC-Len) and HAIC alone for the treatment of advanced hepatocellular carcinoma (Ad-HCC). Totally 349 patients with Ad-HCC participated in the research from February 2018 to October 2020. On the basis of propensity score matching (PSM), 132 and 110 cases were assigned to the HAIC group and the HAIC-Len group, respectively, with a ratio of 1:1. Progression-free survival (PFS), overall survival (OS), and complications were compared between two groups. The Kaplan–Meier method and log-rank test were utilized to estimate cumulative OS and PFS. Additionally, uni- and multi-variate Cox regression models were employed to identify significant independent factors. The median follow-up period in this study was set to be 20.8 months. Following PSM, the one-, two- and three-year cumulative OS rates in the HAIC-Len and HAIC groups were 63.6%, 12.1%, and 3.0%, and 47.2%, 11.8%, and 2.7%, respectively, with a significant difference (P < 0.001). The first-three-year cumulative incidence rates PFS in the HAIC-Len and the HAIC groups were 15.2%, 1.5%, and ND, and 11.8%, 4.5%, and 3.6%, respectively, with no significant difference detected (P = 0.092). BMI (HR 0.709. 95% CI 0.549, 0.915. P = 0.008) and AST (HR 1.005. 95% CI 1.003, 1.007. P < 0.001) represented independent prognostic factors for OS. Additionally, the two groups exhibited no significant difference in the incidence rates of adverse events. HAIC-Len significantly improved survival outcomes of patients with Ad-HCC and demonstrated acceptable toxicity compared to HAIC alone.


Introduction
Hepatocellular carcinoma (HCC) is a top cause of cancerrelated mortality, while over 50% of the global cases are reported in China [1,2]. Although surgical resection (SR) is currently the primary curative treatment for this disease, approximately 50-70% of patients lose the opportunity for SR as they are already in the advanced stage when diagnosed [3][4][5][6]. The median overall survival (OS) period of patients with advanced HCC (Ad-HCC) is only 4-8 months. To date, treatments by using multi-targeted tyrosine kinase inhibitor (TKIs), including lenvatinib, sorafenib, and atezolizumab plus bevacizumab (A + T), have been applied as the firstline therapy under multiple international guidelines [4]. Meanwhile, these treatments have been increasingly used for management of Ad-HCC in routine clinical practice, with confirmed survival benefits [5,7,8]. However, the curative effect of the TKI treatment remains unsatisfactory, even though the overall response rate (ORR) of A + T has reached Fang Long, Shali Chen and Ruidong Li have contributed equally to this work. 33% [9]. Therefore, a safe yet effective treatment is urgently needed for Ad-HCC.
Hepatic arterial infusion chemotherapy (HAIC) is currently a radical treatment method for Ad-HCC in a variety of countries, especially in Japan, where HAIC is regarded as the first-line treatment of this disease [10][11][12]. mFOLFOX6 (oxaliplatin plus fluorouracil and leucovorin) is a classic combined anticancer method with widely recognized systematic effectiveness for management of Ad-HCC. It has been confirmed that HAIC using the mFOLFOX6 regimen exhibits good safety and efficacy for the treatment of Ad-HCC, as it achieved superior survival outcomes compared to sorafenib [12]. Meanwhile, a randomized phase III trial of HAIC for Ad-HCC was conducted, and the results suggested that HAIC of infusion fluorouracil, leucovorin, and oxaliplatin contributed to longer OS compared with sorafenib in Ad-HCC [13]. Additionally, a randomized clinical trial indicated that the combination of HAIC with sorafenib resulted in increased OS of HCC patients with portal vein invasion (PVI) [14]. On the other hand, lenvatinib, which is a novel drug for TKIs, has been applied to Ad-HCC, with improved OS in Asian populations compared with sorafenib [15].
To date, limited data for comparing HAIC plus lenvatinib (HAIC-Len) and HAIC alone for the treatment of AD-HCC have been reported. In the present study, the survival outcomes and AEs of the two therapies were compared with each other to evaluate the efficacy and safety.

Study design
From February 2018 to October 2020, 879 cases with Ad-HCC subsequently underwent initial HAIC in three local hospitals by using a modified FOLFOX6 regimen were enrolled. HCC was diagnosed on the basis of standard guidelines in this field [16,17]. Some suspected cases were confirmed via needle biopsy. The inclusion criteria: (a) aged 18 to 75 years old, (b) Eastern Cooperative Oncology Group (ECOG) score < 2, (c) liver function of Child-Pugh class A, and (d) advanced HCC with PVI, but no extrahepatic metastases. The exclusion criteria: (a) history of alternative treatments before HAIC, Len, or a combination of these, (b) suffering from other malignancies, (c) liver function of Child-Pugh class C or B, (d) missing image data, and (e) follow-up for less than 6 months. 349 cases were initially involved and filtered on the basis of the patient recruitment pathway (Fig. 1). Among them, 217 cases who received only the HAIC treatment were assigned to the HAIC group and 132 cases who received HAIC combined with the lenvatinib treatment were assigned to the HAIC-Len group. Procedures and protocols applied in this study were approved by the Research Ethics Committees of Jinan University and Chinese PLA General Hospital. According to the retrospective nature, no informed consent was needed for this study.

Procedures of HAIC
Digital subtraction angiography (DSA) was used for all procedures. An artery sheath catheter was inserted into the femoral artery by using a modified Seldinger technique. After that, the feeding hepatic artery (the proper hepatic artery or its branche which covers all tumor areas to the maximum extent) was assessed by inserting a 5-Fr catheter (Terumo) into both the celiac trunk and the superior mesenteric artery.
Before HAIC administration, angiography was performed to evaluate the tumor vascularity and to identify any extrahepatic collateral vessels that might cause drug leakage or toxicity. If necessary, embolization of these vessels was performed with metallic coils or gelatin sponge particles. A micro-catheter (Terumo; 2.7-Fr) was inserted into the feeding artery. Additionally, the chemo-drugs were administered via the microcatheter in the HAI. A modified FOLFOX6 regimen involving leucovorin (dose, 200 mg/m 2 within 2 h on Day one), oxaliplatin (dose, 85 mg/m 2 within 2 h on Day one), and fluorouracil (dose, 400 mg/m 2 in bolus, and then a continuous infusion of 2,400 mg/m 2 within 46 h) was administered. We removed the sheath and the catheter when HAIC was completed. Treatment repetitions were generally performed every 21 days for four to six cycles, unless cases were intolerant of intrahepatic lesion progression or toxicity.

Protocol of treatment by using lenvatinib
Oral lenvatinib (Lenvimafi) was applied to the subjects. The initial dose was designed based on liver function and body weight. Specifically, patients with weight > 60 kg and Child-Pugh A classification started from12 mg once daily; patients with weight < 60 kg and the same liver function started from 8 mg once daily. A reduction in dosage or interruption of treatment was implemented when AEs were detected. Lenvatinib was administered unless patients were intolerant of radiological tumor progression or AEs.

Assessment of clinical outcomes
All subjects were assessed on June 31, 2021, which was their last follow-up date. Dynamic CT or MRI were employed to assess treatment responses according to the modified Response Evaluation Criteria in Solid Tumors (mRECIST). Herein, three endpoints of the two groups were investigated and compared. The primary end points included OS and PFS. OS was defined as the period from the date corresponding to initial treatment to the date of death or end of followup. PFS was regarded as the duration from the first treatment measure to tumor progression. The second endpoints included the disease control rate (DCR) and ORR. The third endpoint was patient safety, which was evaluated based on the symptoms and clinical laboratory tests. Additionally, the incidence rates and severity of adverse events were reported and assessed accordingly [19].

Statistical analysis
Continuous variables are denoted in a pattern of mean ± SD, and classified variables were denoted in the form of frequencies and percentages. A two-sample t-test and the Wilcoxon rank-sum test were utilized to analyze continuous variables if the normality was verified or not. The χ 2 test was used to analyze categorical variables. The Cohen's kappa statistics were used to obtain consistency in the treatment response evaluation. Estimation of OS and PFS was achieved by using the KM method and analyzed by using the log-rank test. We used PSM approach based on 1:1 (caliper of 0.02) to adjust possible imbalances in the treatment. Additionally, uni-and multi-variate Cox regression models had been used to determine independent prognostic factors (IPFs). Finally, statistical analyses were achieved by using R software. All significance tests were two-sided, with P < 0.05 denoted statistical significance.

Target population
Stratification of pre-and post-match cohorts in terms of baseline characteristics was executed by following the therapeutic schedule, as shown in Table 1. The median followup in the HAIC group and the HAIC-Len group was 21.2 (5.4-52.7 months) and 20.6 months (7.9-50.7 months), respectively. Significant differences were observed in standardized mean differences for the pre-match cohort (P < 0.05) in terms of Body Mass Index (BMI). After the PSM analysis based on a ratio of 1:1, 242 subjects were assigned to either the HAIC (n = 132) or HAIC-Len groups (n = 110), and all variables achieved adequate balance. Death rates of 28.8% (38/132) and 86.4% (95/110) were observed in the HAIC-LEN and HAIC groups, respectively, which constituted no significant differences (SSD) (P < 0.001). Progression event rates of 93.2% (123/132) and 93.6% (103/110) were observed in the HAIC-LEN and HAIC groups, with no SSDs detected (P = 0.887).

Radiological responses
The radiological response rates (RRR) between both groups were measured before and after PSM, and the results are shown in Table 2. Before PSM, DCR and ORR in the HAIC group and in the HAIC-Len group were 39.6% and 75.5%, and 44.0% and 72.0%, respectively, with no SSDs. Following PSM, the results were similar. Measurement of the ALBI scores of the two groups were executed all the way from baseline to posttreatment. The changes of ALBI scores were from − 2.60 to − 2.34 in in the HAIC group (P = 0.379) and

Discussion
Treatment strategies for Ad-HCC have been greatly supplemented by the emergence of lenvatinib, with most patients obtaining survival benefits. HAIC using the modified FOL-FOX6 regimen for Ad-HCC has been confirmed as a superior method to sorafenib in terms of safety and efficacy. Shi et al. reported that compared with using sorafenib alone, PVI led to an increase of the OS by 87.5% or 6.24 months by using sorafenib combined with HAIC and FOLFOX agents [14]. Therefore, HAIC combined with TKIs can generate favorable outcomes. Previous HAIC-related studies focused on HAIC combined with sorafenib as the first-line treatment for Ad-HCC [20][21][22]. However, a randomized phase III study revealed that the efficacy of lenvatinib was comparable to that of sorafenib for the OS of untreated patients with Ad-HCC [23]. Meanwhile, investigations using Cox regression models indicated that lenvatinib was more effective for the Asia-Pacific population [23,24]. The results of this study were in accordance with the tendency that lenvatinib combined with HAIC is superior to sorafenib combined with HAIC. However, this conclusion needs to be confirmed by future studies. The arterial infusion technique can also increase the survival rates of patients treated by HAIC. In the subsequent cycle of HAIC, the microcatheter tip position needed to be changed frequently (median = 2.5), and the newly developed hepatic-gastroduodenal collateral artery needed to be embolized (46.2% patients). Hence, repetitive catheterization and digital subtraction angiography applied before each HAIC session exhibited higher reliability in concentrating and avoiding exposure of other organs to anticancer drugs, compared with the implanted port catheter system. Indeed, a II randomized trial using a repetitive catheterization procedure, which can compare the outcomes of sorafenib combined with cisplatin-based HAIC and sorafenib alone, delivered positive results, while the other trial mentioned above delivered negative results [25].
Herein, the response rate was higher in the HAIC-Len group than in the HAIC group. One possible reason for this result may be that lenvatinib has multiple targets, which can trigger cell apoptosis, as well as increase in the sensitivity to FOLFOX agents [26]. In addition, the synergistic effect of 5-fluorouracil and oxaliplatin partly relies on the suppression (mediated by 5-FU) of drug transporters, whereas the activities of multidrug resistance-associated transporters can be inhibited by lenvatinib [27,28]. Lenvatinib can assist in vessel normalization in the case of HCC, as well as drug delivery [29].
Notably, a tenacious continuation of HAIC procedures may have led to deterioration of hepatic function in this study. Significant hepatic function deterioration was detected in both groups based on the assessment of changes in ALBI scores. However, the degrees of deterioration in the two groups were similar. In the uni-and multi-variate analyses, IPFs for Ad-HCC after PSM were assessed. The results showed that BMI and AST were IPFs for the OS, similar to conclusions presented in previous studies. It has been demonstrated that an appropriate BMI suggests good nutrition and a superior prognosis chance in anti-tumor treatments [30]. Additionally, the increase in AST is mainly distributed in the plasma and mitochondria of hepatocytes, indicating organelle-level damage to hepatocytes, severe liver damage, and poor prognosis [31]. Therefore, an increased AST level was a negative factor for Ad-HCC.
In this study, the categories and incidence rate of grade 3-4 AEs (7.3%) in the HAIC group were similar to those from previous HAIC studies [12,14]. The incidence rate of grade 3-4 AEs in the HAIC-Len group was consistent with that in the HAIC group, further confirming the safety and efficacy of HAIC-Len as a therapeutic approach for Ad-HCC. Additionally, no deaths associated with the treatment were observed in this study. In the HAIC-LEN group, the lenvatinib-related AEs, including hand-foot skin reactions and hypertension, were similar to those reported elsewhere. HAIC-related AEs, including liver dysfunction, myelosuppression, and gastrointestinal toxicity, showed lower incidence rates than those in previous studies [12,14]. This phenomenon might be attributed to the effective complication management and preventive measures (e.g., smearing ointment and regular monitoring of blood pressure) before treatment. No potential synergistic toxicity had been detected, suggesting that both treatments were clinically safe and feasible.
Nevertheless, this study exhibits a few limitations. First of all, this study was literally retrospective. Despite the application of PSM, post-matching sample size reduction may have influences on the results. Therefore, conclusions present in this study require verification in prospective randomized controlled trials. Second, this study only involved patients in China, while the hepatitis B virus serves as the predominant etiology of HCC in this country. Therefore, the applicability of the results to other countries, where the etiology of HCC is dominated by hepatitis C virus, has yet to be assessed. Finally, subsequent treatments may produce confounding factors. For instance, various cases changed the TKI drugs upon the presence of resistance to lenvatinib. As a result, it is challenging to assess the efficacy of these treatments, which may have influences on the OS of the cases.

Conclusions
Compared to HAIC alone, HAIC-Len led to dramatically improved OS in cases with Ad-HCC with acceptable treatment-related toxicity.
Funding The authors have not received any financial assistance.