Infiltration of T Cells and Programmed Cell Death Ligand 1-expressing Macrophages as a Potential Predictor of Lenvatinib Response in Hepatocellular Carcinoma

Article information

J Liver Cancer. 2020;20(2):128-134

Publication date (electronic) : 2020 September 30

doi : https://doi.org/10.17998/jlc.20.2.128

Pil Soo Sung^1,2, Sung Woo Cho¹, Jaejun Lee^1,3, Hyun Yang^1,3, Jeong Won Jang^1,2, Si Hyun Bae^1,3, Jong Young Choi^1,2, Seung Kew Yoon^1,3

¹The Catholic University Liver Research Center, College of Medicine, The Catholic University of Korea, Seoul, Korea

²Division of Gastroenterology and Hepatology, Department of Internal Medicine, Seoul St. Mary’s Hospital, College of Medicine, The Catholic University of Korea, Seoul, Korea

³Division of Gastroenterology and Hepatology, Department of Internal Medicine, Eunpyeong St. Mary’s Hospital, College of Medicine, The Catholic University of Korea, Seoul, Korea

Corresponding author: Pil Soo Sung, Division of Gastroenterology and Hepatology, Department of Internal Medicine, College of Medicine, The Catholic University of Korea, 222 Banpo-daero, Seocho-gu, Seoul 06591, Korea, Tel. +82-2258-2073, Fax. +82-2-3481-4025, E-mail; pssung@catholic.ac.kr

Received 2020 June 13; Revised 2020 July 27; Accepted 2020 July 30.

Abstract

INTRODUCTION

Liver cancer is the fourth most common cause of malignancy-associated mortality.1 Hepatocellular carcinoma (HCC) accounts for 80% of the primary malignant liver cancers.1 In 2007, sorafenib, a multi-target tyrosine kinase inhibitor, was approved for the treatment of unresectable HCC. Although sorafenib improves the overall survival of patients with unresectable HCC, the survival benefit is modest.2,3 Similar to sorafenib, lenvatinib is an oral multi-target tyrosine kinase inhibitor. Evidence from preclinical studies indicates that its potent antiangiogenic activity is associated with the inhibition of the fibroblast and vascular endothelial growth factor (VEGF) signaling pathways.4 A recent international multi-center clinical trial comparing lenvatinib with sorafenib as the first-line treatment for unresectable HCC found that the median overall survival of patients treated with lenvantinib was non-inferior to that of patients treated with sorafenib, although lenvatinib was superior to sorafenib in terms of objective response rate and progression-free survival.5 Consequently, lenvatinib was approved as the first-line therapy for unresectable HCC in countries including Korea, Japan, China, the EU, and the USA.6,7

HCC is recognized as an immune-tolerant malignancy.8 Overexpression of immune inhibitory molecules, such as programmed cell death protein-1 (PD-1), leads to the exhaustion of activated CD8+ T cells.1,9 Moreover, the infiltration of various immune cells, such as tumor-associated macrophages (TAMs), leads to an immunosuppressive microenvironment.10,11 However, only 14–18% of the patients who receive pembrolizumab or nivolumab monotherapy demonstrate objective tumor responses.12–14 Moreover, unlike that in other solid tumors, no significant association was identified between tumor cell programmed cell death ligand 1 (PD-L1) expression and anti-PD-1 responses in HCC, as reported by the Keynote-224 and CheckMate-040 studies.9,15,16

A previous report has demonstrated that lenvatinib requires the help of CD8+ T cells to exert antitumor activity.17 Moreover, tumor volume was significantly reduced through the use of lenvatinib in combination with the anti-PD-1 antibody in a mouse syngeneic Hepa1–6 HCC tumor model.17 These findings in mice have suggested that lenvatinib exerts immunomodulatory and antitumor effects. In this study, we evaluated the role of infiltrating immune cells in lenvatinib-treated unresectable HCC through the application of immunohistochemistry and multicolor flow cytometry of patient’s samples, which were obtained by biopsy.

METHODS

This study was conducted after approval from the Institutional Review Boards of Seoul St. Mary’s Hospital and Eunpyeong St. Mary’s Hospital (XC20RIDI0063).

RESULTS

1. Patient characteristics

The baseline clinical characteristics of the nine enrolled patients are shown in Table 1. The median patient age was 57 years, and the principal etiology of HCC was hepatitis B virus infection (67%). Extrahepatic metastases were reported in one patient, and portal vein tumor thrombosis was identified in four patients. The median alpha-fetoprotein (AFP) level was 2,031 ng/mL. All patients were classified as having Child-Pugh class A at the time of lenvatinib start. Prior to systemic therapy, most patients underwent local–regional therapies, such as trans-arterial chemoembolization or hepatic arterial infusion chemotherapy. The median lenvatinib treatment duration was 9 weeks. Patients with bodyweights of more than 60 kg received 12 mg of lenvatinib and those weighing less than 60 kg received 8 mg.

Table 1

Baseline characteristics of the enrolled patients

2. Objective responses to lenvatinib were associated with intra-tumoral/peri-tumoral infiltration of T cells or PD-L1-expressing macrophages

First, we compared the frequency of intra-tumoral or peri-tumoral T cells and macrophages between lenvatinib responders (CR+PR, n=4) and non-responders (SD+PD, n=5) using immunohistochemistry data (Fig. 1). Lenvatinib responders showed a significantly higher frequency of intra-tumoral (Fig. 1A) and peri-tumoral (Fig. 1B) T cells than lenvatinib non-responders. However, there were no significant differences in the frequency of intra-tumoral (Fig. 1C) and peri-tumoral (Fig. 1D) macrophages between lenvatinib responders and non-responders. Total PD-L1-expressing cells were represented by combined positive scores, and lenvatinib responders presented significantly higher numbers than non-responders.

Figure 1

Objective responses to lenvatinib are associated with intra-tumoral/peri-tumoral infiltration of T cells or programmed cell death ligand 1 (PD-L1)-expressing cells. (A) Frequency of intra-tumoral T cells in lenvatinib responders (n=4) and lenvatinib non-responders (n=5). (B) Frequency of peri-tumoral T cells in lenvatinib responders and lenvatinib non-responders. (C) Frequency of intra-tumoral macrophages in lenvatinib responders and lenvatinib non-responders. (D) Frequency of peri-tumoral macrophages in lenvatinib responders and lenvatinib non-responders. (E) Frequency of PD-L1-expressing cells in lenvatinib responders and lenvatinib non-responders. Bar graphs represent the mean±standard deviation. ^*P<0.05; ^†P<0.01.

3. Tumor-associated macrophages were the primary cells expressing PD-L1 in HCC

Next, we investigated which cell types expressed PD-L1 in the tissues of lenvatinib responders. Fig. 2A is representative of the immunohistochemistry of lenvatinib responders. This panel includes three distinct parts: T cell compartment (CD3+), macrophage compartment (CD68+), and tumor-cell compartment. Interestingly, PD-L1 was generally expressed only in the macrophage compartment and was rarely expressed in T cell or tumor-cell compartments (Fig. 2A). Flow cytometry using intra-tumoral liver biopsy specimens demonstrated that TAMs, which expressed high levels of CD206 and CD163, also expressed high levels of PD-L1 in lenvatinib responders. Moreover, PD-L1 expression showed a positive correlation with HLA-DR expression, thereby indicating that these TAMs might activate cognate effector T cells after PD-1/PD-L1 blockade (Fig. 2B). In contrast, lenvatinib non-responders showed less CD3+ T cell infiltration in both peri-tumoral and intra-tumoral regions. Moreover, CD68+ TAMs did not express PD-L1 in lenvatinib non-responders, although a considerable number of TAMs infiltrated (Fig. 2C).

Figure 2

Tumor-associated macrophages are the primary cells expressing programmed cell death ligand 1 (PD-L1) in hepatocellular carcinoma after lenvatinib treatment. (A) Immunohistochemistry of CD3, CD68, and PD-L1 in the biopsy specimen of a lenvatinib responder. (B) Flow cytometry analysis of PD-L1, Human Leukocyte Antigen-DR isotype, CD206, and CD163 expression in intra-tumoral macrophages. (C) Immunohistochemistry of CD3, CD68, and PD-L1 in the biopsy specimen of a lenvatinib non-responder.

DISCUSSION

Lenvatinib was recently reported to exert significant immunomodulatory activity in a Hepa1–6 mouse HCC model.17 Another recent study using single-cell analysis has demonstrated that lenvatinib reduces the population of TAMs.22 These findings suggest that lenvatinib not only potently inhibits angiogenesis but also exerts considerable immunomodulatory effects. Moreover, recent reports have demonstrated that the expression of PD-1 and other inhibitory checkpoints is upregulated by VEGF-A secreted by cancer and stromal cells and that immune regulatory function is disrupted when VEGF receptor signaling is targeted with VEGF receptor inhibitors.23,24 These studies suggest that lenvatinib can promote antitumor immunity by inhibiting VEGF signaling pathways, resulting in the activation of CD8+ T cells and a reduction in the TAM population.

Monocytic cells in the peripheral circulation are recruited to specific tumor microenvironments and become TAMs.25 In general, macrophages can be classified within two main groups: classically activated macrophages (M1) and alternatively activated macrophages (M2).25 In general, TAMs exhibit the M2 phenotype, and their infiltration is known to be associated with poor prognosis in HCC.9,11,26 However, the function and phenotypes of TAMs in human HCC tissues have not been precisely characterized thus far. In different cancer types, including lung cancer, conflicting results have been found regarding the correlation of TAM counts in surgical specimens and clinical prognoses.27 It appears that tumor-expressed PD-L1 directly inhibits T cells, but the significance of PD-L1 expression by TAMs for the regulation of the effector phase of tumor-specific T cells remains unclear.27 In HCC, PD-L1 expression is generally very weak in tumor cells, and the expression of PD-L1 in TAMs may justify tumor immunogenicity. Interestingly, TAMs in lenvatinib responders in our study seems not to be fully M2-polarized because they express high levels of HLA-DR, which is a crucial M1 marker. Therefore, it can be suggested that these tumors are immunogenic and can respond to treatments modulating antitumor immune responses.

Noteworthy, our study also suggests that treatment with lenvatinib plus anti-PD-1/PD-L1 blockade may be a promising combinatorial strategy for immune-based HCC therapy. In general, two different drugs with different mechanisms of T cell restoration have synergistic anti-tumor activity. Therefore, treatment with lenvatinib (anti-VEGF activity) + anti-PD-1/PD-L1 will probably result in better responses in HCC than treatment only with the single agent.

This study includes the following limitations: 1) A retrospective study design, 2) a small number of patients, 3) a short follow-up period, 4) a single-arm design.

In conclusion, our study demonstrates that tumor immunogenicity, reflected by T cell and PD-L1-positive macrophage infiltration, affects the response of unresectable HCC to lenvatinib. Future studies with larger numbers of patients may elucidate the precise function of TAMs and the immune-modulatory function of lenvatinib treatment in HCC.

Article information Continued