RFA Lowers IDO Activity and Slows Non-Ablated Lesion Growth in Multifocal Hepatic Carcinoma


 Background: The aim of this study was to determine whether radiofrequency ablation (RFA) could activate immunity and slow non-ablated lesion growth in multifocal hepatic carcinoma.Methods: We performed a retrospective study on patients with multifocal hepatic carcinoma and assessed the non-ablated lesion growth rate between patients who received RFA and those who did not. In self-controlled study, before and three weeks after RFA, blood samples were collected from patients who received RFA to allow for comparisons of the alpha-fetoprotein (AFP) level, indoleamine 2,3-dioxygenase (IDO) concentration as assessed by enzyme-linked immunosorbent assay (ELISA) and IDO activity as assessed by high-performance liquid chromatography (HPLC).Results: A total of 66 patients were included in the retrospective analysis: there were 46 (69.7%) patients in the treatment group and 20 (30.3%) patients in the control group. The mean growth rate of non-ablated tumors was 0.0291 ± 0.0965 mm/d in the treatment group and 0.0947 ± 0.0754 mm/d in the control group (P = 0.001). The mean concentrations of IDO before and after RFA were 15.57±4.06 ng/ml and 7.53±1.56 ng/ml in 45 eligible patients, and this difference was significant (P = 0.034). The mean IDO activity values were 29.7±22.03 and 25.25±1.75 before and after RFA, respectively (P = 0.031). AFP decreased significantly after RFA, but there was no significant correlation between the decrease in AFP and the decrease in IDO concentration and activity.Conclusions: RFA may induce abscopal effects in hepatic carcinoma patients, manifested by a decrease in IDO concentration and activity and a slowed growth rate in non-ablated lesions.


Background
Liver cancer ranks sixth in morbidity and fourth in mortality among all types of cancer [1]. Because most patients miss the opportunity for surgery when diagnosed, radiofrequency ablation (RFA) has become an effective supplementary therapy for multifocal hepatic carcinoma. Under the guidance of imaging methods such as computerized tomography and ultrasound, the electrode is directly placed in the tumor lesions. By delivering radiofrequency energy, the tumor and adjacent tissues are devitalized. In addition, previous studies have discovered that RFA could in uence the host immune microenvironment, such as increasing local dendritic cells (DC) in ltration and the expression of CD80 and CD86 [2], enhancing tumor-associated antigen-speci c T cell and natural killer cell responses, and increasing programmed cell death protein 1 (PD-1) expression in primary tumors [3,4]. Dromi et al. demonstrated that RFA can degrade non-ablated lesions in murine tumor models [5], and the treatment is designed to have an abscopal effect. The concept rst proposed by Mole in 1953 is that irradiation can induce non-irradiated or distant tumor lesions to shrink [6]. However, the incidence of abscopal effects is low, and these effects mostly occur in immunogenic tumors [7]. Presently, there are no reports about RFA inducing abscopal effects in humans.

Page 3/21
The topic of immune checkpoint inhibitors is increasingly popular in tumor immune escape research. The IDO family is a group of enzymes, including IDO1, IDO2 and TDO. IDO1 is the rate-limiting enzyme of the kynurenine (Kyn) pathway [8][9][10]. IDO1 mediates intracellular immune suppression through mechanisms involving three proteins: general control nonderepressible 2 (GCN2), mammalian target of rapamycin (m-TOR) and aromatic hydrocarbon receptor (AhR). These rst two mechanisms are mediated by activating L-Tryptophan (L-Trp) de ciency signals, such as GCN2/eukaryotic transcription initiation factor-2α (eIF-2α), and inhibiting the master signal of L-Trp adequacy, amino acid-sensing kinase 1 (GLK1)/m-TOR [11,12]. Kyn produced by L-Trp through IDO1 is the third mechanism and is an endogenous ligand of AhR.
The binding of Kyn and AhR results in the differentiation of immature CD4+ T cells into regulatory T cells and induces the expression of IDO1, further inhibiting the immune response of T cells [13] and decreasing the immunogenicity of DC [14]. Whether IDO activity changes could participate in the abscopal effects induced by RFA activation in multifocal hepatic carcinoma has not yet been studied. We investigated RFA-related abscopal effects and their mechanism. Methods 1. Methods

Retrospective analysis
We retrospectively analyzed 66 patients with hepatoma seen at Sun Yat-sen University Cancer Center from June 2013 to July 2016. In the treatment group, patients received RFA once between two imaging examinations; in the control group, patients did not receive treatment between the two examinations.

Enrollment criteria
The enrollment criteria for the retrospective analysis were as follows: (1) Patients had a measurable mass other than the ablated lesions.
(2) Patients had no other antitumor treatment between the two evaluations.
(3) Patient disease complied with the clinical or pathological diagnostic criteria for primary hepatoma.

Exclusion criteria
The exclusion criteria were as follows: (1) Patients had no measurable non-ablated lesions.
(2) Patient received anti-tumor treatment other than RFA between the two assessments.
(3) Patients were not reviewed in time.
(4) Patients were too weak to tolerate RFA. (5) Patients had a history of chronic autoimmune diseases (e.g., Addison's disease, multiple sclerosis, Hashimoto's thyroiditis, and hypophysitis etc.) (6) Patients had taken immunosuppressive drugs in the treatment period.

Tumor growth rate
We measured the diameter changes of the largest nonablated lesions and divided it by the days between the two reviews to obtain the growth rate of the tumors. An electronic balance was used to weigh 0.1361 g KH 2 PO 4 powder, which was mixed into 800 ml of ultrapure water on an electronic balance, and then the pH was adjusted to 4.0 with hydrochloric acid. Then, the volume was adjusted to 1L. The solution was mixed with pure methanol at a ratio of 3:1 for use in chromatography and treated with ultrasound for 15 minutes after ltration of the microporous membrane.
2.2.2 Kyn storage solutions (10 mmol/L) Kyn powder (0.0208 g) was weighed, dissolved in ultrapure water, transferred into a 10 mL volumetric bottle with a glass rod, diluted in ultrapure water to 10 mL, and mixed to obtain the standard storage solution of Kyn with a concentration of 10 mmol/L. The stock solution was aliquoted into 1.5 mL Eppendorf tubes and stored in a refrigerator at -20 °C.

Trp storage solutions (10 mmol/L)
Trp powder (0.0204 g) was weighed with an electronic balance, dissolved in a small amount of ultrapure water, transferred into a 10 mL volumetric bottle with glass rod, diluted in 10 mL of ultrapure water, and mixed to obtain the standard Trp storage solution with a 10 mmol/L concentration. The stock solution was aliquoted into 1.5 mL Eppendorf tubes and stored in a refrigerator at -20 °C.

Trichloroacetic acid (30% (m/v))
Trichloroacetic acid (1.5 g) was weighed, mixed with 50 mL of ultrapure water, dissolved and moved into a brown bottle for storage at room temperature.

The assay of IDO
We measured the activity and concentration of IDO before and after RFA in patients who were newly diagnosed with primary hepatoma in our center and underwent CT-guided RFA from May 2014 to June 2016. All patients were fully aware of the risk of the operation and signed informed consent before treatment. This study was approved by the Ethics Committee of Sun Yat-sen University Cancer Center.

Enrollment criteria
The principal eligibility criteria were as follows: (1) Con rmed diagnosis of primary hepatoma according to histopathological criteria or the criteria of the American Society of Hepatology for the diagnosis of hepatic carcinoma; (2) Eastern Cooperative Oncology Group (ECOG) performance status ≤ 1; (3) Hepatic function Child-Pugh Grading A or B; (4) International normalized ratio <1; (5) Platelets > 50,000/cm 3 ; (6) No history of hepatic encephalopathy; (7) Receipt and acknowledgement of informed consent.

Quantitative determination of IDO concentration
Venous blood samples (3 ml) were collected in dry test tubes before and after RFA and centrifuged at 3,000 rpm for 10 minutes in a high-speed centrifuge within 2 hours of collection. Then, serum was transferred into an EP tube with a pipette and stored at -80 °C.

Operational processes by ELISA
(1) The multi-welled plate was removed from the aluminum foil bag after 20 minutes at room temperature, and the sample size was calculated according to the number of standard and the number of blank control wells; (2) Standard wells, sample wells and blank wells were set up, and blank wells had nothing added.
Standard wells had different concentrations of the standards added in 50 μl; (3) The sample wells had 10 μl of sample added, and then 40 μl of the sample diluent was added.
(4) The sample and standard wells, but not the blank wells, were incubated in a 37 °C incubator for 60 minutes with 100 μl of an antibody labeled with horseradish peroxidase.
(5) The liquid in the wells was discarded, the plate was dried via patting with absorbent paper, each well was lled with diluted wash buffer, the wells were oscillated for 30 seconds, the wash buffer was removed by forceful shaking, and the plates were dried via patting with absorbent paper; the process was repeated 5 times.
(6) The reaction substrates A (50 μl) and B (50 μl) were added to all the wells, aluminum foil was added to protect the reaction from light, and the reactions were incubated in a 37 °C incubator for 15 minutes.
The optical density value of each well was determined at a wavelength of 450 nm within 15 minutes after adding 50 μl of the reaction termination solution to all the wells.
(8) The standard curve was drawn, and the concentration of each sample was calculated according to the regression curve equation (Fig. 1A).

Standard solution preparation
(1) Kyn standard solution preparation The prepared 10 mmol/L Kyn storage solution was absorbed and diluted to 10 mL with phosphate buffer saline (PBS). The standard solution of 50 μmol/L Kyn was obtained by vigorous stirring and shaking. The Kyn standard solutions of 10, 5, 2, 1, and 0.5 μmol/L were diluted from the standard sequentially.
(2) Trp standard solution preparation The 10 mmol/L Trp storage solution was added into an Eppendorf tube with a pipette, and 495 μl of PBS buffer was then added. The Trp standard solution of 100 μmol/L was obtained by shaking and mixing. The standard solution of 100 μmol/L Trp was diluted to 100, 80, 50, 30, and 10 μmol/L sequentially.

Pretreatment of serum samples
The patient serum was removed from the refrigerator at -80 °C, thawed at room temperature, and placed in a 1.5 mL Eppendorf tube. Each tube was mixed with 150 μL of the trichloroacetic acid solution. The supernatant was centrifuged at 4 °C for 20 minutes at maximum speed. The supernatant was ltered by a microporous membrane. Finally, 400 mL of the ltrate was absorbed and added into an HPLC automatic sampling tube.

Pretreatment with standard solutions
The standard solutions of Kyn and Trp at different concentrations were removed from the refrigerator at -20 °C and thawed at room temperature, and 500 μL was added into a 1.5 mL Eppendorf tube, mixed with prepared 150 μL trichloroacetic acid, and centrifuged at 4 °C for 20 minutes at maximum speed. The supernatant was ltered with a microporous membrane, and then 400 μL of the ltrate was transferred into an HPLC automatic sampling tube.

Establishment of the ultraviolet detection wavelength
By measuring the maximum absorption wavelength of Kyn in the range of 300 to 400 nm, it was found that there was an expected absorption peak of Kyn at 360 nm. Therefore, 360 nm was selected as the detection wavelength for quantitative analysis of Kyn in this experiment.

Determination of the wavelengths for uorescence detection
According to previous experiments, the excitation wavelength and emission wavelength of Trp were 286 nm and 360 nm, respectively. (2) Creation of the Trp standard curve.
The standard solution of Trp was also analyzed by HPLC. Using the uorescence absorption peak areas as longitudinal coordinates, the actual concentrations of Trp solutions (76.92, 61.54, 38.46, 23.08, and 7.692 μmol/L) were used as abscissa. Linear regression was used to obtain regression equations and correlation coe cients, and the resulting linear regression equation was y = 938,719x-997,092, with R 2 = 0.99929 (as shown in Fig. 1E).

Calculation of IDO Activity
The concentration of Kyn and Trp in the samples was determined according to the standard curve drawn by the external standard method. The calculated concentration was multiplied by the dilution (1.3) during pretreatment, and that represented the Trp concentration in the sample. The activity of IDO was expressed as the ratio of Trp concentration to Kyn concentration (Trp/Kyn). When increases in IDO activity induced enhanced Trp metabolism, the concentration of Kyn was increased, and the ratio was reduced.

Statistical analysis
We used a chi-square test or Fisher's exact test to compare the clinical characteristics of all patients. Using the Kolmogorov-Smirnov test and nonparametric testing of two independent samples (Mann-Whitney test), we compared the tumor growth rate. We used a paired Student's t test to compare IDO concentration and activity before and after treatment, and the correlation analyses comparing the differences in IDO concentration and activity before and after treatment with the differences in AFP before and after treatment used the Spearman test. Each statistical test was two-sided; P < 0.05 was deemed to have statistical signi cance. SPSS version 19.0 (IBM Corp, Armonk, NY, USA) was employed in performing all statistical analyses.

Baseline characteristics of the patients in this retrospective study
Of the 66 patients in the retrospective analysis, 46 (69.7%) patients received one RFA treatment (treatment group), and 20 (30.3%) patients did not receive RFA treatment (control group) between the two imaging evaluations. 39 patients in the treatment group and 17 patients in the control group had hepatitis B virus (HBV) and/or hepatitis C virus (HCV) infection. In the treatment group, 9 (19.6%) patients had surgery, and 8 (40%) patients had surgery in the control group. In the treatment group, 9 (19.6%) patients had venous tumor embolism, and 2 (4.3%) patients had ascites. The details are shown in Table 1. The mean non-ablated lesion tumor growth rate was 0.0291 ± 0.0965 mm/d in the treatment group and 0.0947 ± 0.0754 mm/d in the control group. Patients who had RFA treatment showed slower tumor growth than those who were not treated, and the difference was statistically signi cant (P = 0.001) (Fig. 2 and Fig. 3).   Table 3.

The correlation of AFP and IDO
The median AFP levels before and after treatment were 174.5 ng/ml (16.6950-4780.5 ng/ml) and 75.18 ng/ml (4.615-713.65 ng/ml), respectively, and the difference was statistically signi cant (P = 0.026).
There was no correlation between the change in AFP before and after treatment and the change in IDO activity and concentration (activity P = 0.169, R = 0.209; concentration P = 0.504, R = 0.102).

Discussion
In this retrospective case-controlled study, the baseline characteristics of the treatment group and control group were balanced. The growth rate of non-ablated lesions in the treatment group was signi cantly slower than that in the control group (P = 0.001), and we veri ed that RFA can stimulate the abscopal effect; these results are in agreement with the results of Dromi et al. [5]. In the IDO analyses, the patients had indications for RFA, and most early-stage patients had undergone surgery, so there were no patients with stage IV disease and few patients with stage I-II disease. Most patients had infection with HBV, and the infection rate agreed with that in China [15,16]. According to the ELISA results, HPLC test results and clinical AFP analysis, before and three weeks after RFA, the AFP value and the IDO concentration and activity were signi cantly different. These results suggest that RFA can affect tumor burden, destroy the IDO-related immunosuppressive microenvironment and reduce the systemic immune tolerance induced by IDO. In addition, the changes in IDO concentration and activity did not correlate with tumor burden, which demonstrates that the changes in IDO were induced by RFA.
According to previous studies, local tumor radiotherapy can stimulate the abscopal effect through several mechanisms. During the process of tumor cell death, calreticulin is transferred to the cell surface and captured by DCs, thereby presenting tumor antigens to activate cytotoxic T lymphocytes. Released high-mobility group box 1 protein (HMGB1) acts as a pro-in ammatory mediator that stimulates monocytes to secrete the cytokines TNF, IL-1, IL-6 and IL-8 [17]. HMGB1 also binds to Toll-like receptor 4 (TLR4) on DCs.
Degradation of tumor antigens captured by DCs is prevented to enhance the tumor antigen presentation process [18][19][20]. Released adenosine triphosphate binds to the purinergic receptors on DCs, leading to activation of the in ammatory process and release of interleukin-1β (IL-1β) [20]. After tumor cell disintegration, DNA can also promote the production of type I interferon (IFN) by activating the stimulator of interferon gene (STING) pathway in DCs [21]. RFA can also kill tumor cells and lead to the release of tumor-related cytokines, including tumor necrosis factor-α (TNF-α), IFN-γ, and IL-1β, which are signi cantly reduced and induce the production of tumor in ltration-related cytokines [4,22]. Shi Liangrong et al. showed that RFA treatment can induce modest and short-lived growth inhibition and Treg and myeloid-derived suppressor cell expansion in the tumor microenvironment and upregulate the expression of PD-1 and programmed cell death 1 ligand 1(PDL-1) in distant tumors [23]. In addition, cytotoxic T lymphocyte-associated antigen-4 (CTLA-4) has also been observed to activate antitumor immunity in ablation treatment [2]. At present, there is no relevant report on the mechanism of IDO participating in the abscopal effect caused by RFA.
Because RFA can activate immunity, people began to seek immunosuppressive agents combined with RFA to treat tumors, thereby obtaining a sustained immune response, achieving the purpose of treating tumors and prolonging survival. Some clinical trials use RFA combined with PD-1/PDL-1 or CTLA-4 inhibitors. In NCT01853618, tremelimumab (a CTLA-4 inhibitor) was used in combination with ablation in patients with advanced HCC, and the results showed that the combination group had better survival than the monotherapy group [24]. In NCT03101475, RFA was combined with PDL-1 or CTLA-4 in patients with incurable liver metastases from colorectal cancer [25]. IDO1, as a key enzyme of the Kyn pathway mediating immunosuppression, is considered a potential small molecule target for cancer immunotherapy. Several IDO1-related immunosuppressant drugs have been developed, such as epacadostat, indoximod, navoximod (GDC-0919), and BMS 986205 [26]. In NCT01685266 and NCT00567931, IDO1 inhibitor monotherapy did not show any bene t in solid tumors [27,28]. A previous in vivo study by Stefani Spranger et al. showed that the expression of IDO1 was upregulated with anti-CTLA-4 or anti-PD-1/PD-L1 treatment; as such, researchers are gradually exploring combination therapy [29]. In ECHO-202/KEYNOTE-037, patients received epacadostat combined with pembrolizumab (a PD-1 inhibitor), which showed good tolerability and safety [30]. In addition, combination therapies in the rstline setting in melanoma and in the second-line setting in non-small-cell lung cancer and head & neck cancer showed superior overall response rates compared with pembrolizumab [31]. Clinical trials of other IDO inhibitors combined with anti-angiogenesis agents, peptide vaccines, etc., are also in full swing given the interest in the potential synergistic interactions of these therapies [31]. We expect future clinical trials of IDO1 inhibitors combined with RFA treatment in liver cancer.
We used a retrospective analysis to con rm that RFA can reduce non-ablated lesions and simulate abscopal effects. ELISA and HPLC were used to dynamically track the immunosuppressive factors related to IDO in patients with hepatic carcinoma, and values before and after RFA treatment were compared. It was con rmed that RFA could effectively reduce the tumor burden, the concentration of IDO and the activity of IDO in serum to a certain extent. Some of these ndings are of great signi cance for reducing the immunosuppression of tumors. This study also had several limitations, including errors in the ELISA data, inadequate collection of blood samples, and the retrospective nature of this study, which looked at the tumor growth rate in untreated lesions, which will be improved in future studies.

Conclusions:
RFA can slow the growth of non-ablated lesions and activate abscopal effects. Additionally, it can reduce tumor burden and decrease the concentration and activity of IDO in peripheral blood, reducing immunosuppression. Declarations Ethics approval and consent to participate

Abbreviations
The current study was approved by the Institutional Review Board of Sun Yat-sen University Cancer Center, and all procedures met the basic standards of the Declaration of Helsinki.
Ethics, consent, and permissions: All patients were clearly informed of the potential risks and bene ts before undergoing the procedure. All patients stated their understanding and signed the informed consent form voluntarily.
Consent to participate: All patients were clearly informed of the potential risks and bene ts before undergoing the procedure. All patients stated their understanding and signed the informed consent form voluntarily.

Consent for publication
All patients consented for publication voluntarily.
Availability of data and material The dataset(s) supporting the conclusions of this article is (are) available in the public Research Data Deposit platform (www.researchdata.org.cn) with the approval RDD number RDDA2019001297.   The tumor growth rate in the treatment group was slower than in the control group, and the difference was statistically signi cant (P=0.001).