Physical Activity Improves Outcomes of Combined Lenvatinib Plus Anti-PD-1 Therapy In Patients With Unresectable Hepatocellular Carcinoma

Background: Adaptive resistance is one of the major hinderances for unresectable hepatocellular carcinoma (HCC) treated with lenvatinib and anti-programmed cell death protein 1 (anti-PD-1) therapy. Physical activity is known to have anti-cancer effects, including immunomodulatory actions, inuencing patients’ outcomes. This study investigated the hypothesis that physical activity synergizes with combined lenvatinib plus anti-PD-1 therapy to enhance ecacy in patients with unresectable HCC. Methods: The physical activity levels of patients with unresectable HCC treated with combination therapy using lenvatinib plus anti-PD-1 antibodies were recorded by questionnaire. Patients were categorized according to physical activity levels (active vs. sedentary). The primary outcome was overall survival (OS), analyzed using the Kaplan-Meier method with a log-rank test. Secondary outcomes included objective response rate (ORR) and progression-free survival (PFS). Factors associated with survival and ORR were analyzed using regression analyses. A subcutaneous syngeneic HCC model was generated in C57BL/6 mice. Mice were randomized to receive placebo, combined lenvatinib plus anti-PD-1 antibodies or combination therapy plus physical activity. Tumors were measured every 3 days and harvested for immunohistochemistry analysis at 20 mm maximum diameter. Results: Fifty-nine patients with unresectable HCC were categorized to active (n=28) or sedentary (n=31) groups. The active group had higher albumin and des-γ-carboxy prothrombin levels and lower hepatitis B virus load at baseline; other clinical and oncologic characteristics were comparable between the two groups. Patients in the active group had signicantly longer OS (HR = 0.220, 95% CI, 0.060-0.799) and PFS (HR = 0.158, 95% CI, 0.044-0.562) and higher ORR (OR = 4.571, 95% CI, 1.482-14.102) than patients in the sedentary group. Regular physical activity was independently associated with OS, PFS and ORR. The mouse model showed that physical activity signicantly suppressed tumor growth and prolonged survival of tumor-bearing mice. Furthermore, physical activity inhibited Treg cell inltration and immune checkpoint expression (including CTLA4, TIGIT and TIM3) induced by long-term combined lenvatinib plus anti-PD-1 therapy, improving ecacy.


Background
Primary liver cancer is the fth most commonly diagnosed cancer and second leading cause of cancerrelated death in China. [1] Hepatocellular carcinoma (HCC) represents 75%-85% of primary liver cancer and is a major global health problem. [2] The majority of patients with HCC are diagnosed at an advanced stage, resulting in limited treatment options and poor prognosis. [3] Systemic therapy, including sorafenib, lenvatinib, and immune check point blockade (ICB), should be considered in patients with unresectable HCC. However, none of these treatment options have achieved satisfactory e cacy when used as monotherapy in published trials. [4][5][6][7][8] Recently, a phase 1b trial testing the combination of lenvatinib and pembrolizumab has shown encouraging results, with a response rate of 36% reported. [9] Nevertheless, adaptive resistance is still one of the major hinderances for those patients.
Upregulation of alternative immune checkpoints is considered one mechanism underlying adaptive immune resistance. [10] When resistance to anti-programmed cell death protein 1 (anti-PD-1)/antiprogrammed cell death ligand 1 (anti-PD-L1) therapy develops, the proportion of T cells expressing alternative immune checkpoints, including cytotoxic T-lymphocyte-associated protein 4 (CTLA4) and Tcell immunoglobulin mucin-3 (TIM3), increases. [11,12] Regulatory T (Treg) cells play a pivotal role in both maintaining immune homeostasis and tumor immune escape. [13] Moreover, PD-1 blockade signi cantly enhances the in ltration of immunosuppressive Treg cells, contributing to adaptive resistance to immunotherapy. [14] It is reported that lenvatinib reduces Treg cell in ltration and activates immune pathways, resulting in reprograming of the tumor microenvironment, which may contribute to improvement of the e cacy of anti-PD-1 therapy. [15,16] Suppressing the in ltration of Treg cells and alternative immune checkpoints level is therefore an attractive strategy to improve the therapeutic e cacy of combined lenvatinib and anti-PD-1 antibody (Ab) therapy.
Regular physical activity is associated with a lower risk of developing several cancers, including HCC. [17,18] In addition, physical activity reduces the risk of total deaths and cancer-related deaths in both colon and breast cancer. [19][20][21] Physical activity may reduce tumor cell proliferation, suppress epithelialmesenchymal transition (EMT), and promote intra-tumoral perfusion/vascularization. [22][23][24] Moreover, physical activity elicits anticancer effects by reducing systemic in ammation and countering immunosenescence. [25,26] We hypothesize that physical activity may synergize with combined lenvatinib plus anti-PD-1 therapy through immunomodulatory effects to enhance the therapeutic e cacy of this treatment regimen.

Patients
Patients with unresectable HCC treated with combined lenvatinib plus anti-PD-1 antibody at Zhongshan Hospital from June 1, 2018 to September 30, 2020 were included in this retrospective study. Inclusion criteria were as follows: (1) patients met the clinical diagnostic criteria for HCC, [27] with or without pathological diagnosis; (2) HCC was unresectable and not suitable for transarterial chemoembolization (TACE); (3) treatment with lenvatinib combined with anti-PD-1 antibody was given as rst-line systemic therapy; and, (4) an Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1. Patients were excluded if they had other malignancies, received previous systemic treatment or had no tumor evaluation after initiation of the combined treatment regimen. This study was approved by the Hospital Research Ethics Committee. Informed consent was obtained from all patients.
Grouping and measurement of physical activity Patients took regular physical activity during combined lenvatinib plus anti-PD-1 treatment were assigned to the active group, otherwise they were assigned to the sedentary group. Physical activity was measured by questionnaire as previously described. [19,28,29] Patients were questioned by phone survey in June, 2021. Patients were asked the following questions: (1) Do you take any leisure time physical activity during pharmacotherapy? (2) What kind of physical activity do you usually take? (brisk-walk, swimming, ball games, Tai Chi, or other aerobic activity) (3) How many days per week do you take leisure time activity? (4) What is your total time (minutes) of leisure time physical activity per day? (5) On a scale of 1 to 10, please rate the intensity of your leisure time activity (perceived scale: 1-4 = low intensity, 5-6= moderate, 7-10 = vigorous). The criteria for regular physical activity were based on the American College of Sports Medicine guidelines. [30] Patients were considered to engage in regular physical activity if they met any of the following criteria: (1) no less than 5 d·wk −1 of moderate aerobic activity for ≥30 min·d −1 ; (2) no less than 3 d·wk −1 of vigorous aerobic activity for ≥30 min·d −1 ; (3) no less than 3-5 d·wk −1 of mixed intensity activity for ≥30 min·d −1 ; or (4) any of the above before or within 1 month after the initiation of combination therapy until 3 months before death or last follow-up.

Treatment outcomes
The primary treatment outcome was overall survival (OS). Secondary outcomes included objective response rate (ORR) and progression-free survival (PFS). All outcomes were assessed using magnetic resonance imaging and modi ed response evaluation criteria in solid tumors (mRECIST) criteria for hepatocellular carcinoma response assessment. [31] Cells and reagents Hepa1-6 cells, derived from the BW7756 tumor in a C57L mouse, were obtained from the American Type Culture Collection (ATCC) (Manassas, VA, USA). The cells were cultured in high-glucose Dulbecco's modi ed eagle medium (DMEM) (BasalMedia Technologies Co., Shanghai, China) supplemented with 10% fetal bovine serum (FBS) (Yeasen Biotechnology Co., Shanghai, China) and 1% penicillinstreptomycin (Yeasen Biotechnology Co., Shanghai, China) at 37°C under a 5% CO 2 atmosphere.

Subcutaneous syngeneic mouse model
A subcutaneous syngeneic HCC model was generated by subcutaneously injecting ~3*10 6 Hep1-6 cells in 100 µl PBS on the backs of 5-6-week-old female C57BL/6J mice (Charles River Laboratories) (n=21). Animals were weighed and tumor volume was assessed every three days. Once the maximum diameters reached 10 mm, animals were randomly assigned to receive combination therapy (lenvatinib plus anti-PD-1 Ab; n=7), combination therapy plus physical activity (n=7), or placebo (control, drug vehicle plus mouse isotype IgG; n=7).
Lenvatinib (dissolved in 3 mM HCl, 10 mg/kg) was administered daily by oral gavage. [32] Anti-PD-1 Ab (200 µg/mouse) was intraperitoneally administered every ve days. [33] Physical activity was facilitated by placing running wheels in cages; running distance was recorded by electromagnetic sensors in the combination therapy plus physical activity group. Mice were sacri ced and tumors harvested once tumor maximum diameter reached 20 mm. The study was performed in compliance with guidelines for the use of animals established by the institution ethical committee and the "Tumor induction in mice and rats IACUC Guideline" [34].

Immunohistochemical staining and analysis
Immunohistochemical staining was conducted as previously described [35,36]  Tumor viability, de ned as the proportion of tumors presenting viable cells in a sample (i.e. excluding necrotic regions or granulation tissue), was assessed on hematoxylin and eosin slides. All analyses were performed by an expert pathologist blinded to the treatment arms.

Statistical analysis
All statistical analyses were performed using SPSS (Version 22, Chicago, IL, USA). Continuous variables were compared with the Student's t-test (equal variances assumed) or Wilcoxon rank sum test (equal variances not assumed). Categorical variables were analyzed by Chi-squared test or Fisher exact test.
Kaplan-Meier survival analysis with the log-rank test was used to evaluate the associations between various interventions and survival. Univariate and multivariate logistic regression was used to determine factors affecting ORR. Multivariate Cox regression was used to determine variables associated with survival. For the analysis, a p-value <0.05 in a two-tailed test was considered statistically signi cant. Graphs were drawn with Graphpad Prism (Version 8.0.2, San Diego, CA, USA).

Results
Baseline characteristics of patients Fifty-nine patients were eligible for this study. Twenty-eight patients were assigned to the active group and 31 to the sedentary group. Up to June 30, 2021, the median follow-up duration was 13 months (range, 3-28 months). The baseline clinical characteristics are summarized in Table 1. Compared to patients in the sedentary group, patients in the active group had higher albumin and des-γ-carboxy prothrombin (DCP) levels and lower hepatitis B virus (HBV) viral load at baseline (p<0.05). Other clinical characteristics were comparable between the two groups.  Figure 1A). The 1-year PFS rate was 82.1% in the active group vs. 23.0% in the sedentary group (p<0.001, Figure 1B). The median PFS in the active group was 23 months, longer than that in the sedentary group (6 months, p<0.001

Physical activity improved the effect of combination therapy in syngeneic HCC mice
To elucidate why physical activity improved the outcomes of HCC patients treated with combination therapy, we established a subcutaneous syngeneic HCC mouse model. The average running distance per mouse per week in the combination therapy plus physical activity group is shown in Figure 2A. Compared to the control group, both combination therapy and combination therapy plus physical activity groups had lower tumor burden (p<0.05 and p<0.001, respectively; Figure 2B). Physical activity synergized with combination therapy to further improve its inhibitory effect on tumor growth (p<0.05; Figure 2B).
Both combination therapy and combination therapy plus physical activity groups showed prolonged survival (time to 20 mm in maximum diameter) of tumor-bearing mice compared to the control group, and mice in the combination therapy plus physical activity group had the longest lifespan (median survival of 15, 30 and 54 days in control, combination therapy and combination therapy plus physical activity groups, respectively, p<0.001; Figure 2C).
Finally, tumor viability was assessed to further understand the synergistic effect of physical activity and combination therapy. The combination therapy plus physical activity group had the lowest tumor viability (p<0.0001 vs. control group and p<0.01 vs. combination therapy group), followed by the combination therapy group (p<0.0001 vs. control group) ( Figure 2D).

Mechanism of the synergistic effect of physical activity and combination therapy
To elucidate why physical activity improved the e cacy of combination therapy, immunohistochemical staining of major immune cell populations and alternative immune checkpoints was performed. H-scores were generated to quantitatively analyze expression levels among the three groups. Both combination therapy and combination therapy plus physical activity groups showed increased CD4+ T and Treg cell in ltrations and TIGIT expression compared to the control group (Supplementary gure 1 & Figure 3). The combination therapy plus physical activity group had elevated immune cell in ltration, including macrophage and CD8+ T cells, and increased CTLA4 and VISTA expression (Supplementary gure 1 & Figure 3). Compared to the combination therapy group, the combination therapy plus physical activity group tended to exhibit reduced in ltration of Treg cells and expression of several immune checkpoints, including TIGIT and TIM3.
Interestingly, we observed that mice receiving short-term (<15 days, n=3) and long-term (≥15 days, n=4) combination therapy showed distinct immunophenotypes. Short-term combination therapy had a similar immunophenotype to the control group, while long-term combination therapy exhibited an immunosuppressive phenotype, with increased in ltration of Treg cells and higher levels of alternative immune checkpoints, including CTLA4, TIGIT, TIM3 and VISTA ( Figure 3). However, the combination therapy plus physical activity group showed reduced in ltration of Treg cells and expression of CTLA4, TIGIT and TIM3 ( Figure 3).
Overall, we observed that physical activity inhibited the in ltration of immune-suppressive Treg cells and the adaptive upregulation of alternative immune checkpoints to enhance the therapeutic e cacy of combination therapy.

Discussion
In this study, we found that regular physical activity was associated with improved outcomes in patients with unresectable HCC receiving combined lenvatinib and anti-PD-1 therapy. Compared with the sedentary population, physically active patients had an approximately 80% lower risk of death and progression and four times greater likelihood of reaching objective response criteria.
Ample studies have shown that physically active lifestyles reduce the risk of multiple cancers, including HCC. [17,18,37] Following primary treatment, physical activity has been consistently shown to have positive effects on vigor and vitality, cardiorespiratory tness, quality of life, depression, anxiety, pain and fatigue; physical activity also reduces the overall risk of death and cancer-related death in cancer survivors. [19][20][21]38] During primary treatment with lenvatinib, physical activity might enable patients to receive longer-term treatment through improvement of their physical and psychological condition.
To elucidate the reason why physical activity synergizes with combination therapy, we established a subcutaneously implanted HCC mouse model. Since mice are natural runners, [39] running wheels were placed in cages of the combination therapy plus physical activity group to facilitate voluntary running.
The mouse model showed that physical activity improved the therapeutic e cacy of combination therapy with retarded tumor growth and prolonged survival, consistent with clinical ndings in humans.
We found that both combination therapy and combination therapy plus physical activity were associated with an immunosuppressive tumor microenvironment, which contradicts the ndings of previous studies.
[ 15,16] However, the treatment duration in the present study was much longer (median treatment duration of 30 days) than those reported in previous studies. [15,16] As short-term and long-term treatments showed distinct immunophenotypes, this may account for the contradictory ndings of the present study. It is suggested that the immunostimulatory effect of lenvatinib might decrease with prolonged treatment duration.
Our ndings suggest that physical activity enhances combined lenvatinib plus anti-PD-1 therapy by counteracting the immunosuppressive tumor microenvironment induced by long-term treatment. Compared to tumors treated with combination therapy, the addition of physical activity inhibited the in ltration of immunosuppressive Treg cells and the expression of several immune checkpoints, including CTLA4, TIGIT and TIM3, reprogramming the tumor immune microenvironment. Notably, the immunomodulatory function of physical activity during long-term treatment postponed tumor progression and prolonged OS.
In addition to immunomodulatory effects, physical activity has been associated with other anti-cancer mechanisms. It was reported that physical activity suppressed tumor growth by promoting p53-driven apoptosis. [40] Our previous study found that moderate swimming inhibited liver cancer progression through suppression of transforming growth factor-beta-induced epithelial-mesenchymal transition. [23] It was also reported that exercise promoted a shift towards a more "normalized" tumor microenvironment by improving intra-tumoral perfusion/vascularization. [41,42] Multiple mechanisms may therefore account for the anti-cancer effects of physical activity.
These ndings add to the evidence supporting physical activity as an important lifestyle intervention in patients with cancer. In recent years, multiple international organizations have published physical activity recommendations for patients living with and beyond cancer, including the American Cancer Society, [43] the American College of Sports Medicine, [44] Cancer Care Ontario, [45] and the Clinical Oncology Society of Australia, [46] and Exercise and Sports Science Australia. [47] The speci c value of physical activity in promoting an enhanced treatment response in patients receiving combination lenvatinib plus anti-PD-1 therapy should be explored further in this context.
Our study is not without limitations. Firstly, the retrospective design of the study may have introduced bias. A further prospective study should be designed, so as to control bias and quantify the physical activity level. Secondly, the duration of combination treatment might in uence the intra-tumor immunophenotype, which was not accounted for at the time of study conception. Further studies are needed to con rm the effect of duration of combination treatment on the tumor microenvironment.

Conclusions
The present study suggests that physical activity improved the therapeutic e cacy of long-term combined lenvatinib plus anti-PD-1 therapy in patients with HCC through reprograming the tumor microenvironment from an immunosuppressive to immunostimulatory phenotype. This study provides evidence for recommending physically active lifestyles to patients with unresectable HCC receiving combined lenvatinib and anti-PD-1 therapy. Representative images captured at 40X. * p<0.05, ** p<0.01, *** p<0.001, **** p<0.0001. Abbreviations: HCC, hepatocellular carcinoma.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.