Efficiency and safety of hepatic arterial infusion chemotherapy (HAIC) combined with anti‐PD1 therapy versus HAIC monotherapy for advanced hepatocellular carcinoma: A multicenter propensity score matching analysis

Abstract Purpose To investigate the clinical efficacy and safety of combination therapy of hepatic arterial infusion chemotherapy (HAIC) and anti‐programmed cell death protein‐1 (PD‐1) therapy in the treatment of advanced hepatocellular carcinoma (HCC). Methods In this retrospective clinical research, from March 2018 to December 2019, 1158 HCC patients categorized as BCLC stage C were reviewed for eligibility. We utilized propensity score matching (PSM) to mitigate initial disparities between the groups. The evaluation of the best tumor response was conducted in accordance with mRECIST 1.1 criteria. The difference in survival outcomes including overall survival (OS), progression‐free survival (PFS), and objective response rate (ORR) between groups were compared. Results Following the eligibility review, 453 patients underwent a combined treatment of HAIC with PD1 inhibitors (HAIC‐PD1 group), while 221 patients received HAIC monotherapy (HAIC group) met the inclusion criteria and were finally enrolled in this study. In the entire cohort, the HAIC‐PD1 group exhibited significantly prolonged overall survival (median overall survival: 40.4 months vs. 9.7 months, p < 0.001) and progression‐free survival (median progression‐free survival: 22.1 months vs. 5.8 months, p < 0.001). By propensity score, patients were matched according to baseline differences, resulting in all 442 patients in group HAIC‐PD1 (n = 221) and group HAIC (n = 221). After PSM adjustment, as well, the survival of the HAIC‐PD1 group was still distinctly longer than the HAIC group (median overall survival time, 40.4 months vs 9.7 months, p < 0.001; median progression‐free survival, 22.1 months vs 5.7 months, p < 0.001). Univariate and multivariable analysis demonstrated that AFP level, metastasis, and therapeutic schedule were independent predictive factors for overall survival. Conclusion The combination therapy of HAIC and PD1 inhibitors successfully extended OS to advanced HCC patients and could be a better choice than HAIC monotherapy.


| INTRODUCTION
Hepatocellular carcinoma (HCC), the prevailing histological variant of liver malignancy, ranked as the sixth neoplasm incidence, is identified as the third leading mortality attributable to cancer worldwide. 1,2From an epidemiological standpoint, there are many risk factors for primary liver malignancy, among which hepatitis B virus infection and alcoholic hepatitis are the main factors in Asia and Western countries respectively. 3,4Unfortunately, due to the highly insidious nature of HCC, a majority of cases have progressed to an advanced stage upon diagnosis and precluded the opportunity for curative local intervention.
][7][8] Moreover, traditional chemotherapy has a wide range of actions, unstable efficacy, serious systemic adverse reactions, and limited selectivity of individualized treatment options, which make it infrequently enrolled in medical practice.][11] In this setting, hepatic artery infusion chemotherapy (HAIC) is regarded as a potential substitute chemotherapy strategy.During the HAIC procedure, chemotherapeutic agents are administered directly into the tumor blood supply artery trunk through the pre-implanted microcatheter within the liver, so as to avoid the first-pass effect and constitute a superior systemic chemotherapy characteristic with elevated regional drug concentration and minimal systemic adverse effects.Despite undiminished international controversy, HAIC has been endorsed as a viable therapeutic regimen for advanced HCC by the Japan Society of Hepatology (JSH). 12It is noteworthy that advanced HCC patients are more commonly present with poor hepatic function.In this patient population, HAIC appears to demonstrate sufficient tolerance and promising therapeutic effects. 13In a phase III randomized trial, the HAIC-FOLFOX regimen demonstrated a median overall survival (OS) of 13.9 months, which surpassed the median OS of 8.0 months observed in the sorafenib group.The implementation of this treatment approach realized superior management of advanced HCC. 14 In recent times, the emergence of immunotherapy has garnered significant attention as a promising therapeutic modality for various solid tumors.For chronic inflammation-induced immune surveillance and dysregulation of the immune environment, programmed death-1 (PD-1) inhibitors can restrain tumor immune evasion by blocking the PD-1/PD-L1 pathway, increasing the intracorporal anti-tumor immunoreaction. 158][19] To date, the combined immunotherapeutic approach for HCC remains in the investigational phase and necessitates additional monitoring and scrutiny.
finally enrolled in this study.In the entire cohort, the HAIC-PD1 group exhibited significantly prolonged overall survival (median overall survival: 40.4 months vs. 9.7 months, p < 0.001) and progression-free survival (median progression-free survival: 22.1 months vs. 5.8 months, p < 0.001).By propensity score, patients were matched according to baseline differences, resulting in all 442 patients in group HAIC-PD1 (n = 221) and group HAIC (n = 221).After PSM adjustment, as well, the survival of the HAIC-PD1 group was still distinctly longer than the HAIC group (median overall survival time, 40.4 months vs 9.7 months, p < 0.001; median progression-free survival, 22.1 months vs 5.7 months, p < 0.001).Univariate and multivariable analysis demonstrated that AFP level, metastasis, and therapeutic schedule were independent predictive factors for overall survival.

Conclusion:
The combination therapy of HAIC and PD1 inhibitors successfully extended OS to advanced HCC patients and could be a better choice than HAIC monotherapy.

K E Y W O R D S
hepatic arterial infusion chemotherapy, hepatocellular carcinoma, immune checkpoint inhibitors, propensity score Here, we undertook such a retrospective study to expand upon the existing evidence base regarding the efficacy of combined HAIC and anti-PD1 treatment for advanced HCC patients.Specifically, we sought to examine the potential prognostic benefits of this therapeutic approach across various subgroups of HCC patients, with the ultimate goal of furnishing clinicians with an objective reference point for clinical decision-making.

| Clinical information
The Institutional Review Committees at each study center provided ethical approval for this retrospective multicenter study.The present investigation adhered to the principles outlined in the Declaration of Helsinki.
The eligibility of 1158 patients diagnosed with advanced HCC, who underwent treatment with either a combination of HAIC and PD1 inhibitors or HAIC monotherapy, was evaluated through the review of clinical data from four hospitals between March 2018 and December 2019.
The inclusion criteria for this study encompassed the following parameters: (1) diagnosis of HCC confirmed through histopathological biopsy or precise radiological findings in line with the American Association for the Study of Liver Disease (AASLD) guidelines. 20,21

| Treatment procedure
All HAIC treatments were performed using digital subtraction angiography by veteran interventional surgeons according to a standard protocol.Under local anesthesia, the modified Seldinger method was performed to puncture the femoral artery and sequentially inserted the 5F vascular sheath.Then, the 5-Fr Yashiro catheter (Terumo, Tokyo, Japan) was respectively inserted into the celiac trunk and superior mesenteric artery for angiography.After the tumor supply artery was clarified by radiography, the 2.7 F microcatheter (ASAHI, Tokyo, Japan) was coaxially inserted and super-selected to the tumor blood supply branch artery, the microcatheter head end position was determined by micro-ductal angiography.After the microcatheter pathway was constructed, a modified FOLFOX6 protocol, which included the administration of oxaliplatin at a dosage of 85 mg/m 2 via intravenous drip over a period of 2 h on Day 1, calcium folinate at a dosage of 400 mg/m 2 via intravenous drip over a period of 2 h on Day 1, and fluorouracil at a dosage of 400 mg/m 2 via intravenous injection on Day 1, followed by a continuous infusion of 2400 mg/m 2 over 46 h, was delivered through a microcatheter.The combined treatment group continued to receive anti-PD1 treatment within 2 days after the HAIC.The category and dosage of PD-1 inhibitors used in the HAIC-PD1 group were exhibited in Table S1.HAIC produce was repeatedly operated before each cycle.HAIC procedure was repeated every 3 weeks until tumor progression.Based on the judgment of adverse events or special causes, chemotherapy drugs (mainly oxaliplatin) and anti-PD1 drugs can be reduced or discontinued in advance as appropriate.Review of the enhancement CT every 6 weeks.According to mRECIST criteria to assess response to treatment.The follow-up deadline is December 2022.

| Tumor response assessment
The primary endpoints of the study were OS and progression-free survival (PFS).We compared these endpoints between the HAIC-PD1 and HAIC groups.OS was defined as the time from the start of treatment until death from any cause or the most recent follow-up.PFS was defined as the interval from the initiation of tumor treatment to disease progression, as evaluated based on the modified Response Evaluation Criteria in Solid Tumors 22 (mRECIST), or death from any cause.The efficacy of HAIC was assessed using the mRECIST criteria, which classified responses as complete response (CR), partial response (PR), stable disease (SD), or progressive disease (PD).Two experienced radiologists independently conducted all radiographic evaluations of tumor response.The secondary endpoints included the objective response rate (ORR) which represented the proportion of patients who achieved a tumor response categorized as CR or PR, and the disease control rate (DCR) which was defined as the percentage of patients who achieved a tumor response categorized as CR, PR, or SD.Adverse events (AEs) were assessed according to the Common Terminology Criteria for Adverse Events version 5.0. 23

| Propensity score matching analysis
We employed the 1:1 propensity score matching (PSM) method to mitigate the impact of confounding factors and balance the differences between the groups.The matching tolerance was adjusted to 0.02.The variables considered in the PSM analysis included age, gender, ECOG score, comorbidities, HBV infection, ascites, tumor size, number of tumors, vascular invasion, metastasis, ALBI grade, and AFP level.

| Statistical analysis
The analysis for this study was conducted using R software (Rstudio version 4.2.2).Continuous variables following a normal distribution were presented as mean ± standard deviation and analyzed using the Student's t-test.For variables that did not follow a normal distribution, the median was used and analyzed using the Mann-Whitney U-test.Categorical variables were reported as percentages and analyzed using either the chi-squared test or Fisher's exact test.Kaplan-Meier curves were used to estimate the OS and PFS between the groups in both the overall and matched cohorts, employing the Kaplan-Meier method.The Cox proportional hazards regression model was performed to identify independent risk factors influencing OS and PFS.Factors with a p < 0.10 in the univariate analysis were included in the multivariate analysis.A two-tailed p < 0.05 was considered statistically significant.

| Study population
The baseline characteristics of the patients enrolled in this study are presented in Table 1.A total of 674 patients with advanced HCC met the inclusion criteria and were included in the study.Among them, 453 patients received a combination of HAIC and PD1 inhibitors, while the remaining 221 patients received HAIC monotherapy.The patient selection process is illustrated in Figure 1.The majority of the matched cases were male and had a hepatitis B virus infection.Among them, 386 HBV patients (255 [56.3%] patients in the HAIC group vs. 131 [59.3%] patients in the HAIC-PD1 group, accounting for 59.3%, p = 0.462) received continuous antiviral therapy.The average maximum diameter of solid tumors was 11.0 cm, and 265 (59.7%) patients had more than three intrahepatic tumors.Additionally, there were 531 (78.8%) cases with vascular invasion and 354 (52.8%) cases with extrahepatic metastasis.After PSM, a total of 442 patients remained in the study, with 221 patients in the HAIC-PD1 group and 211 patients in the HAIC group.Before PSM adjustment, a higher percentage of patients in the HAIC alone group had vascular invasion (76.6% vs. 83.3%,p = 0.047) and poor liver function reserve (55.2% vs. 46.2%,p = 0.028) were observed than in the HAIC-PD1 group.However, all significant differences between the groups were eliminated after matching.

| Univariate and multivariate analysis
Univariate and multivariate analyses were conducted to identify predictors of OS and PFS, and the findings are presented in Table 2.In the univariate analysis, significant covariates associated with OS included alpha-fetoprotein (AFP), the presence of metastasis, and therapeutic schedule.Similarly, HBV infection, AFP, metastasis, and therapeutic schedule were found to be correlated with PFS.Furthermore, the multivariable Cox regression analysis revealed that AFP, metastasis, and therapeutic schedule were significant prognostic factors for OS.Additionally, HBV infection, metastasis, and therapeutic schedule were identified as significant prognostic factors for PFS.

| Subgroup analysis
Forest plots (Figure 3) were plotted to illustrate the comparison between subgroups.When analyzing the matched cohort, it was observed that, with the exception of the subgroup without HBV infection, the combination of HAIC and PD1 exhibited enhanced overall survival benefits and more effective tumor progression control in all other subgroups in comparison to the HAIC group.This finding implies that PD1 inhibitors may have even more promising efficacy in the treatment of advanced HCC in patients with HBV infection.

| Safety
The occurrence of adverse reactions between the HAIC-PD1 combination group and the HAIC monotherapy group revealed no significant difference (Table 4).It is worth noting that all adverse reactions were manageable, and symptomatic treatment led to significant relief.Importantly, no treatment-related deaths attributed to drug toxicity were reported during the treatment period.The most frequently observed adverse events in the HAIC-PD1 group included elevated transaminase levels, abdominal pain, and neutrophil reduction.Conversely, the HAIC group exhibited elevated transaminase levels, abdominal pain, and fatigue as the most common adverse events.

| DISCUSSION
From the perspective of genetics and epigenetics, HCC is a malignancy characterized by marked intratumoral heterogeneity, the difficulty of capturing which to a certain degree determines the sensitivity and plasticity of tumor cells to different treatment modalities.In comparison, treatment strategies in compliance with tumor burden and liver function appear to be currently suboptimal.In the absence of precise therapies based on intratumoral heterogeneity profiling, immunotherapy-based combination therapy represents a promising direction in augmenting tumor response. 24As one of the few available local treatment options, HAIC is also of great significance in managing the progression and prolonging survival of unresectable HCC patients.In our retrospective study, we utilized PSM to eliminate group differences, enrolled a larger number of cases across multiple centers, and conducted long-term follow-ups.Eventually, irrespective of whether in the entire cohort or in the PSM cohort, consistent conclusions for discrepant OS and PFS between groups were acquired.In the matched cohort of this study, a notable mOS of 40.4 months and advantageous ORR and DCR of 42.5% and 91.9% were obtained in the HAIC-PD1 group, respectively.
In the era of immunotherapy, anti-PD1 treatment plays a crucial role in the treatment of solid tumors.However, monotherapy with immune checkpoint inhibitors has shown low response rates for tumor regression.For example, the ORR of nivolumab, pembrolizumab, and cemiplimab are 20%, 17%, and 14.7%, respectively, with limited OS benefits. 15,25,26The main reason for this is that blocking the PD1/PD-L1 axis alone is insufficient to induce effective anti-tumor activity. 27In recent years, significant achievements have been made in combination therapies involving immunotherapy.According to the results reported from the IMbrave150 trial, the combination of atezolizumab and bevacizumab achieved an encouraging median survival time of 19.2 months, significantly surpassing the 13.4 months of the sorafenib group, particularly for the subset in China reaching a breakthrough 24.0 months. 28Another randomized trial reported a median OS of 18.7 months and a response rate of 24.0% based on the dual immunotherapy regimen of tremelimumab combined with durvalumab, which outperformed monotherapy immunotherapy. 29Subsequently, atezolizumab plus bevacizumab and tremelimumab plus durvalumab have been recommended as first-line treatments.Additionally, the combination of anti-PD1 treatment with local therapies has shown potential.XIn et al. reported that in the treatment of advanced HCC, the combination of HAIC with atezolizumab and bevacizumab achieved a remarkable ORR of 67.3% based on the mRE-CIST criteria and 44.2% based on the RECIST1.1 criteria. 30 retrospective study conducted by Mei et al. showed that adding anti-PD1 treatment to HAIC combined with lenvatinib extended the mOS of patients with advanced HCC by 7 months.31 Moreover, significant survival benefits from the combination of anti-PD1 treatment with TACE have been confirmed by multiple studies.32,33 Currently, combined immunotherapy of atezolizumab plus bevacizumab or durvalumab plus tremelimumab was significantly superior to standard sorafenib in prognosis according to these two clinical trials.Indeed, TKIs sequential therapy has been more effective in improving F I G U R E 3 Forestplot based on overall survival (A) and progression-free survival (B) of each subgroup.
prognosis than single TKIs.However, Kudo pointed out that the favorable prognosis achieved with TKIs sequential therapy can only be attained in the presence of long-term sustained better liver function, and even slight impairment of liver function can also severely impact treatment efficacy. 34,35Moreover, the use of TKIs in systemic therapy is an important factor contributing to elevated transaminase levels, which seem contradictory.In contrast, anti-PD1 therapy has shown better tolerability with regards to liver function in patients.In this study, a considerable proportion of patients with ALBI Grade 2 were included, and satisfactory outcomes were ultimately achieved in terms of OS, PFS and ORR.On this issue, the combination of HAIC and anti-PD1 treatment appears to be equally sensitive in patients with relatively poor liver function, a viewpoint also supported by previous research.On the other hand, patients with BCLC stage C represent a subset with poorer liver function, with most of them having a high burden of intrahepatic tumors and portal vein involvement, leading to the development of portal hypertension and worsening of liver function.Therefore, for the population at high risk of liver function deterioration in BCLC stage C, the combination of HAIC and anti-PD1 treatment appears to be a more effective choice.
The results suggest a benefit correlated with the combined application of HAIC and PD-1 blockades, which may be attributable to the synergistic effect between these two treatment regimens.7][38][39] Moreover, the administration of chemotherapy augments the antigenicity of tumors by eliciting ICD, thereby diminishing "off-target" immunosuppression within the tumor microenvironment.This, in turn, intensifies the immune response, regulates tumor progression and metastasis, and closely correlates with the notable response rate observed in the HAIC-PD1 group in this study.These findings underscore the profound impact of PD-1 inhibitors on the systemic immune microenvironment, as well as the synergistic interactions arising from the concurrent utilization of catheter-based chemotherapy and immunotherapy, which exert a pivotal role in restraining disease progression. 40n relation to subgroup analyses, notably, the integration of PD1 inhibitors with HAIC as compared to HAIC monotherapy exhibited significantly improved OS and PFS outcomes, irrespective of tumor size, tumor number, or the presence of vascular invasion.These findings suggest that HAIC-PD1 may offer a superior therapeutic approach for advanced HCC, regardless of the extent of tumor burden.Furthermore, the amalgamation of HAIC and PD1 demonstrates not only a therapeutic impact on intrahepatic lesions but also confers a survival benefit on patients with advanced HCC characterized by extrahepatic metastases.Besides, the HAIC-PD1 regime failed to attain superior survival outcomes compared to the HAIC monotherapy group only within the non-hepatitis B-related HCC subgroups.Possible underlying factors are as follows, on the one hand, prior researches have demonstrated that upregulation of regulatory factors such as PD-1, CTLA-4, and TIM-3 serves as a significant mechanism of immune dysfunction induced by chronic hepatitis B virus infection. 41Blockade of immune checkpoints may endeavor to enhance immune control through addressing these molecular pathways.On the other hand, checkpoint inhibitors and other immune-modulating agents can help to curtail T cell exhaustion and even promote their reinvigoration. 42Given the existing research, the risk of HBV reactivation remains rarely defined in cancer patients receiving immune checkpoint inhibitors, further in-depth research is needed to clarify the underlying mechanisms involved.
In the present study, the utilization of multivariate analysis revealed that AFP and metastasis emerged as common and statistically significant risk factors linked to OS and PFS.Furthermore, the examination of tumor tissue morphology demonstrated a correlation between heightened levels of HCC tumor markers and the extent of lesion invasion, encompassing vascular invasion and metastasis. 43Hitherto, several Phase III studies acknowledged persistent high AFP levels as a relevant factor for unsatisfactory survival and lesion progression. 44,45s well, this study provides evidence that HAIC-PD1 is an independent risk factor affecting OS and PFS in advanced-staged HCC patients.These findings support the positive effect of HAIC-PD1 in late-stage HCC and suggest that it may represent a promising therapeutic strategy for improving clinical outcomes.In terms of drug-related toxicity, the combined application of HAIC and PD-1 inhibitors is safe and well-tolerated.The HAIC-PD1 group did not demonstrate a significantly higher incidence of adverse events.The occurrence of Grade 3-4 adverse events was similar between the groups.Most adverse events were relieved with symptomatic treatment and did not worsen further.
In view of the current limitations of the administration of HAIC and targeted-immunotherapy, the synergistic effect of the combination of multiple treatment options can help remedy this deficiency.A number of current studies on combination therapy have been carried out one after another, aiming to explore individualized treatment options that can provide more survival benefits for different patients with tolerable side effects.The conclusions of similar studies related to HAIC in different centers seem to be contradictory, which may also be attributed to different HAIC or different dosing regimens of systemic therapy.Therefore, in the follow-up research, efforts should be made to standardize the therapeutic schedule and eliminate the differences caused by different treatment methods.
There is necessity to acknowledge that the present study has certain limitations due to its retrospective nature, which inherently introduces unavoidable selection bias.While we employed PSM to minimize between-group differences, it is possible that endogenous differences may still exist.Thus, there is a need for additional prospective randomized controlled trials to enhance the level of evidence and further validate our findings.Furthermore, it is crucial to highlight that our study was conducted in the specific milieu of China, a nation characterized by a substantial prevalence of hepatitis B virus.This circumstance warrants careful consideration, as it has the potential to impose restrictions on the extent to which our findings can be extrapolated to diverse populations.In addition, it is worth considering that different anti-PD1 drugs may yield varying research outcomes, thus necessitating a more rigorous experimental design to enhance methodological rigor.Lastly, this was a multicenter study, the HAIC process may have variations among different centers, which may affect the outcomes.
In conclusion, this study provides suggestion that the concomitant use of HAIC and PD-1 inhibitors offers advantageous survival benefits for patients with advanced HCC, while also being well-tolerated.Importantly, the combined application of HAIC and anti-PD-1 therapy demonstrates a more substantial survival advantage in patients with latestage HCC compared to HAIC monotherapy.

F I G U R E 1
Flowchart of the patient's selection process for this study.

F I G U R E 2
The Kaplan-Meier survival curves by Log-rank test for the HAIC-PD1 group and the HAIC group with or without propensity score matching (PSM) adjustment.(A) The Kaplan-Meier curves comparing the overall survival between the HAIC-PD1 group and the HAIC group, without adjusting for overall survival; (B) The Kaplan-Meier curves comparing the overall survival between the HAIC-PD1 group and the HAIC group, without adjusting for progression-free survival (PFS); (C).Comparison of PSM-adjusted overall survival between the HAIC-PD1 group and HAIC groups; (D) Comparison of PSM-adjusted PFS between the HAIC-PD1 group and HAIC groups.
. Based on mRECIST criteria, in the HAIC-PD1 group, tumor responses consist of 0.9% of CR, 23.1% of PR, 60.6% of patients with SD, and 8.1% of PD.In the HAIC alone group, including 0.4% of CR, 21.6% of PR, 60.7% of SD, and 17.2% of patients with Baseline characteristics of the study patients before and after PSM.
T A B L E 1 Univariate and multivariate analyses of predictors of survival after treatment.
Best tumor response before and after PSM.
Treatment-related adverse events.
T A B L E 4