Introduction

Treatment strategy of hepatocellular carcinoma (HCC) is a complex decision-making process that considers several variables in clinical practice. A “multiparametric therapeutic approach” has been proposed (Lancet Oncol 2023): the feasibility of all treatments is evaluated systematically in a hierarchical order, to match each patient with the optimal therapy.

Aims & Methods

The aim of the study is to validate the reliability of the new multiparametric decisional framework (MPDF) and the impact of the different variables included in the framework on the final treatment decision in clinical practice. This is a prospective ongoing study with a first monocentric phase (A) and a second multicentric validation phase. Primary endpoints are: 1) adherence to MPDF for treatment allocation, 2) identification of the relative weight of variables involved in this MPDF for treatment decision, while secondary endpoints are overall survival and radiological response after treatment. The study includes all consecutive patients with HCC managed in each center from September 1^st, 2023. The disease is staged according to Barcelona Clinic Liver Cancer (BCLC) classification, while patient’s characterization is conducted by Charlson Comorbidity Index (CCI), Liver Frailty Index (LFI), Child-Pugh (CPT) and ALBI, MELD and skeletal muscle mass index by CT. Each patient is allocated to treatment according to MPDF, BCLC 2022 was the benchmark.

Results

As of November 1st, 2023, 48 patients were recruited into phase A of the study [median age 70 (52-87) years, 81% male, 37% HCV, 27% first treatment for HCC]. Thirteen (27%) patients were classified BCLC 0, 17 (35%) A, 6 (12.5%) B, 6 (12.5%) C and 6 (12.5%) D. The median CCI was 8 (4-13), the median LFI was 4.15 (2.99-4.81). Thirty-nine (81%) patients were in CPT class A, 7 (15%) in B and 2 (4%) in C, while 32 (67%) were classified ALBI grade 1 and 16 (33%) ALBI grade 2, and the median MELD score was 8.5 (6-22). Forty-four (92%) patients received treatment through a multidisciplinary team (MDT). According to the MPDF, 9 (19%) patients were allocated to liver transplant, 5 (10%) to laparoscopic surgical resection, 10 (21%) to microwave thermal ablation (MWTA), 7 (15%) to transarterial treatments (4 TACE and 3 TARE), 3 (6%) to other locoregional therapies (2 MWTA+TACE and 1 electrochemotherapy), 6 (13%) to systemic treatment and 5 (10%) to palliative care only. In 4 cases, the MDT decided for reassessment at 3 months. The treatment allocation was coherent to the BCLC 2022 in 69% BCLC 0, 82% BCLC A, 83% BCLC B, 67% BCLC C and 100% BCLC D.

Conclusion

Preliminary data of our ongoing prospective study suggest that personalized approach for HCC treatment by MPDF is feasible in clinical practice and might help ensuring a customized approach in a standardized framework. Updated data will be presented at the meeting.

Introduction

Tulisokibart is a monoclonal antibody that targets tumor necrosis factor-like cytokine 1A (TL1A), a regulator of inflammation and fibrosis in inflammatory bowel disorders. In the Phase 2 ARTEMIS-UC study, a significantly higher proportion of participants on tulisokibart compared to placebo achieved clinical remission after 12 weeks of induction treatment in patients with moderate to severe active ulcerative colitis (UC) and conventional or advanced treatment failure.¹ In a post hoc analysis from ARTEMIS-UC, re-induction and delayed induction treatment with tulisokibart were effective in participants who did not respond during the initial induction period.² We explored the efficacy and safety of tulisokibart maintenance therapy up to 1 year during the open-label extension of ARTEMIS-UC among patients who responded to induction, re-induction, and delayed induction.

Aims & Methods

Participants (≥18 years) were randomized to intravenous (IV) tulisokibart 1000 mg on day 1 and 500 mg at weeks 2, 6, and 10 or placebo. The study population consisted of 2 cohorts based on a genetic diagnostic test (Dx) assessing likelihood for responding to anti-TL1A treatment. Cohort 1 included participants stratified by Dx results while Cohort 2 included only Dx positive participants. Cohort 1 is the population used in this analysis. Participants were classified as induction responders (defined as reduction of ≥2 points and ≥30% in modified Mayo score from baseline, accompanied by a reduction ≥1 in rectal bleeding subscore or absolute rectal bleeding subscore ≤1 at week 12) or induction nonresponders. Induction non-responders in the tulisokibart and placebo groups were assigned to an open label 12-week re-induction with the aforementioned tulikosibart induction dosing regimen. Cohort 1 induction, re-induction, and delayed induction responders were randomized (stratified by Dx status) to open-label IV tulisokibart 100 mg or 250 mg every 4 weeks.

Results

In Cohort 1, 47 tulisokibart induction responders, 13 re-induction responders, and 29 delayed induction responders were randomized to open-label maintenance treatment with tulisokibart 100 mg or 250 mg every 4 weeks. Maintenance of treatment effect was generally similar with a trend for higher efficacy with tulisokibart 250 mg vs 100 mg maintenance treatment observed across the 3 groups. The safety profile of tulisokibart maintenance treatment was similar across the 3 groups, with no new safety signals being observed.

Conclusion

Tulisokibart was effective as a maintenance treatment in patients with moderate to severe UC. The effect was generally similar across Cohort 1 tulisokibart induction responders, re-induction responders, and delayed induction responders. Tulisokibart maintenance treatment was well tolerated across the 3 groups, with no identified safety signals. Altogether, this suggests that a second induction regimen in tulisokibart non-responders may result in a long-term efficacy benefit without increased safety risk. Larger trials are needed to confirm these findings.

Table: Efficacy and fecal calprotectin

References

1. Sands et al. N Engl J Med 2024; 26;391:1119-1129
2. Hoque et al. J Crohns Colitis 2025; 19(suppl 1):i1158-i1159

Disclosure

Sami Hoque: ​Received Lecture fees and research grants from Tillotts, Ferring and Eli Lilly.
Bruce E. Sands: Travel fees from Eli Lilly and Company, Janssen Research & Development, LLC, Takeda Development Center Americas, Inc.; Received food & beverage from AnX Robotica Corp.; Grant from Janssen Research & Development, LLC; Owns stock options in Ventyx Biosciences; Consulting fees from AbbVie, Abivax, Aclaris Therapeutics, Adiso Therapeutics, Agomab Therapeutics, Alimentiv, Amgen, AnaptysBio, Arena Pharmaceuticals, Artizan Biosciences, Artugen Therapeutics, AstraZeneca, Biora Therapeutics, Boehringer Ingelheim, Bristol Myers Squibb Company, Celltrion Healthcare, Celltrion Healthcare, Connect Biopharm, Cytoki Pharma, Eli Lilly and Company, Entera, Enveda Biosciences, Equilium, Evommune, F. Homann-La Roche AG, Ferring, Fresenius Kabi USA, LLC, Galapagos, Genentech USA, Inc., Gilead Sciences , GlaxoSmithKline, GossamerBio, HMP Acquisition, Imhotex, Immunic, Immunyx Pharma Ltd., Index Pharmaceuticals, Innovation Pharmaceuticals, Janssen Biotech, Inc., Janssen Scientific Affairs, LLC, Johnson & Johnson Health Care Systems Inc., Kaleido, Kallyope, Kyowa Kirin Co., Ltd, Merck & Co., Inc., Microba, Microbiotica Limited, MiroBio, Mobius Care, Morphic Therapeutic, MRM Health, Nexus Therapeutics, Nimbus Therapeutics, Odyssey Therapeutics, Pfizer, Progenity Inc., Prometheus Biosciences, Protagonist Therapeutics, Q32 Bio, Rasayana Therapeutics, Recludix Pharma, Recludix Pharma, SANOFI US SERVICES INC., Spyre Therapeutics, Surrozen, Synlogic Operating Company, Takeda Pharmaceuticals International, Inc., Target RWE, Teva Branded Pharmaceutical Products R&D, Theravance Biopharma Inc., TLL Pharmaceutical, VectivBio AG, Ventyx Biosciences
Brian G. Feagan: Consulting fees from Abbvie, Adiso Therapeutics, Agomab, Alimentiv Inc, Allianthera, Amgen Canada, AnaptysBio, Applied Molecular Transport, Arena Pharma, Biora Therapeutics, Blackbird Laboratories, Boehringer Ingelheim AB, Bristol-Myers Squibb, Celgene, Celsius Therapeutics, Connect BioPharma, Duality, Eli Lilly and Company, Galapagos, Genentech, Gilead Sciences, GlaxoSmithKline, Janssen Global Services, LLC, JapanTobacco, Klick Health, Mage Biologics, Mestag Therapeutics, Mobius Therapeutics, LLC, Monte Rosa Therapeutics, Nexys Therapeutics, Nimbus Discovery Inc, Orphagen Inc., PFIZER CANADA INC, Prometheus, SANOFI US SERVICES INC., Seres Therapeutics, Sobi, Inc, Spyre Therapeutics, Surrozen, Synedgen, Takeda Canada, Takeda Development Center Americas, Inc., Televant, Tillotts, Triastek Inc.
Mark Yen: Employee of Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA.
Bin Dong: Employee of Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA.
Wen Zhou: Employee of Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA.
Silvio Danese: Consulting fees from AbbVie, Allergan, Amgen, AstraZeneca, Biogen, Boehringer Ingelheim, Celgene, Celltrion Healthcare, Ferring, Gilead Sciences, Hospira, Inc., Janssen Biotech, Johnson & Johnson Health Care Systems Inc., Pfizer, Sandoz, UCB Inc., Vifor (International) Ltd.

Introduction

We evaluated celiac disease (CeD)-related symptoms and comorbidities in two US healthcare databases after CeD diagnosis according to the presence/absence of CeD-related symptoms or comorbidities before diagnosis.

Aims & Methods

This matched case–control study used data from two US healthcare databases: 1) MarketScan^® Commercial and Medicare Supplemental database (collection period: 2002−2023); 2) Optum Clinformatics^® Data Mart (Optum CDM; collection period: 2007−2023). Patients with CeD had ≥ 1 inpatient or ≥ 2 outpatient International Classification of Diseases (ICD)-9 or -10 codes (579.0 and K90.0, respectively) at any time during the collection period. Patients were matched in a 1:5 ratio with controls without CeD (no CeD ICD-9/-10 codes at any time during data collection) for age at diagnosis, sex, region, and diagnosis year. All individuals had continuous medical and pharmacy enrollment for 2 years pre/post CeD diagnosis. Odds ratios (ORs) and 95% confidence intervals (CIs) for symptoms/comorbidities after CeD diagnosis were estimated by generalized multivariate logistic regression comparing cases vs matched controls stratified by those with (persistent) or without (new-onset) the relevant symptom/comorbidity pre-diagnosis. Symptoms assessed included abdominal pain, bloating, diarrhea, headache, nausea/vomiting, and tiredness. Comorbidities assessed included depression/anxiety, irritable bowel syndrome (IBS), and iron deficiency anemia (IDA).

Results

The numbers of CeD cases and controls were 39,814 and 198,961 in MarketScan, and 18,357 and 91,754 in Optum CDM, respectively. In both databases, CeD diagnosis was significantly positively associated with having CeD-related symptoms/comorbidities 2 years after CeD diagnosis regardless of pre-diagnosis status; and the strengths of associations were greater among patients with new-onset than with persistent symptoms/comorbidities after CeD diagnosis.
In MarketScan, the ORs (95% CI) in patients with new-onset symptoms were highest for bloating (7.98 [7.32–8.71]), diarrhea (6.83 [6.48–7.19]), and nausea/vomiting (3.75 [3.49–4.04]), and in those with new-onset comorbidities for IBS (8.30 [7.65–9.00]), IDA (6.35 [5.94–6.79]), and depression/anxiety (2.12 [2.04–2.20]). The ORs (95% CI) in patients with persistent symptoms were 2.47 (1.86–3.27) for bloating, 2.47 (2.21–2.76) for diarrhea, and 2.46 (2.05–2.96) for nausea/vomiting, and in those with persistent comorbidities were 1.38 (1.16–1.64) for IBS, 1.93 (1.70–2.18) for IDA, and 1.42 (1.34–1.51) for depression/anxiety.
In Optum CDM, the ORs (95% CI) in patients with new-onset symptoms were highest for bloating (6.15 [5.56–6.80]), diarrhea (5.85 [5.48–6.24]), and nausea/vomiting (3.38 [3.08–3.70]), and in those with new-onset comorbidities for IBS (6.17 [5.56–6.85]), IDA (5.31 [4.89–5.77]), and depression/anxiety (2.10 [1.99–2.21]). The ORs (95% CI) in patients with persistent symptoms were 2.25 (1.70–2.97) for bloating, 2.51 (2.22–2.84) for diarrhea, and 2.77 (2.23–3.43) for nausea/vomiting, and in those with persistent comorbidities were 1.40 (1.15–1.71) for IBS, 2.54 (2.17–2.97) for IDA, and 1.50 (1.38–1.62) for depression/anxiety.

Conclusion

CeD diagnosis was significantly associated with having CeD-related symptoms/comorbidities 2 years after CeD diagnosis, with greater associations among patients with new-onset symptoms/comorbidities than those with persistent symptoms/comorbidities. These results emphasize the clinical burden of CeD and the need for improved treatment options.

Disclosure

FZ has served as a speaker for Werfen and a consultant for Takeda. MA, SM, VG, SG, IP, NC, and DAL are employees and shareholders of Takeda Development Center Americas, Inc. and receive stock or stock options. BL has nothing to disclose.

Introduction

Alcohol use disorder (AUD) remains a significant global health concern. Despite the availability of various pharmacological and behavioural interventions, alcohol-related mortality is rising.

Aims & Methods

This systematic review aimed to critically evaluate the existing literature on the role of glucagon-like peptide receptor 1 analogues (GLP-1 RAs) in people with AUD.
Electronic search was conducted on Ovid Medline, EMBASE, Cochrane Central and PsycINFO from inception to March 2024. Grey literature was searched. The primary outcome of the review was to explore the impact of GLP-1 RAs on self-reported alcohol intake in people with AUD. The extracted data were organized and synthesized to identify key findings and trends evaluating the potential impact of GLP-1 RAs on alcohol consumption in AUD. Narrative synthesis was done to provide a coherent and detailed narrative of the evidence, highlighting significant findings, relationships, and implications. PROSPERO (CRD42024531982).

Results

Seven studies comprising 88,163 participants were included; of these, 38,746 (43.9%) were treated with GLP-1 RAs. Pooled mean age was 49.6 years (SD=10.5), 56.9% (n=50,151/88157) were male, and if the study provided ethnicity 87.5 % (n=426/487) were white. An RCT comparing exenatide with placebo found no significant difference in self-reported alcohol measures at 26 weeks (percentage point reduction: heavy drink days: 6.0, 95% CI -7.4 to 19.4, p=0.37; total alcohol consumption: -42.0 g/30 days, 95% CI -507.7 to 423.7, p=0.86; days without alcohol use: -10.5, 95% CI -2.6 to 23.4, p=0.11). But in participants (n=30) with BMI > 30 kg/m², exenatide compared with placebo was associated with a significant reduction in heavy drinking days (percentage point reduction 23.6, 95% CI, –44.4 to –2.7, P=0.034) and in total alcohol consumed (-1,205 g, 95% CI, –2,206 to –204, P= 0.026). Another RCT showed dulaglutide users compared to placebo were 29% more likely to reduce weekly alcohol consumption (relative effect size: 0.71, p=0.04). In an observational study, authors reported with liraglutide (n=42) 21.4% stopped and 78.6% reduced drinking (p<0.05). A prospective cohort control study showed people on semaglutide or tirzepatide (n=153) compared to no medication significantly reduced number of alcoholic drinks consumed and binge drinking (p<0.05). In a case series (n=6) semaglutide use was associated with a 9.5-point reduction in alcohol use disorder identification test score. A population-based study from Denmark (n=87,676) reported that at 90 days people on GLP-1 RAs compared to one on dipeptidyl peptidase-4 inhibitors (DPP-4) were less likely to experience alcohol-related events ( HR 0. 46, 95% CI 0.24 to 0.86) and effect remained positive long term (356-4619 days: HR 0.72, 95% CI 0.60–0.86). Two qualitative studies examining Reddit social media posts from people on GLP-1 RAs noted over two-thirds addressed a reduction in alcohol consumption. Functional brain imaging suggested significant reductions in cue reactivity were observed in the ventral striatum [F(1,31) = 4.744, P = 0.037, partial η2 = 0.133], dorsal striatum [F(1,31) = 6.124, P = 0.019, partial η2 = 0.165], and putamen [F(1,31) = 4.730, P = 0.037, partial η2 = 0.132] after 26 weeks of exenatide treatment compared with placebo (Table 1). Two studies were of high quality, two medium, and three low.

Conclusion

GLP-1 RAs show promise in reducing alcohol consumption and improving outcomes in individuals with AUD, although heterogeneous study findings warrant further research to establish true effectiveness and safety of GLP-1 RAs in AUD.

References

1. Alcohol use and burden for 195 countries and territories, 1990-2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet (London, England). 2018;392(10152):1015-35.
2. Klausen MK, et al. The role of glucagon-like peptide 1 (GLP-1) in addictive disorders. British Journal of Pharmacology. 2022;179(4):625-41.
3. Klausen MK, et al. Exenatide once weekly for alcohol use disorder investigated in a randomized, placebo-controlled clinical trial. JCI Insight. 2022;7(19).
4. Probst L, et al. Effects of dulaglutide on alcohol consumption during smoking cessation. JCI Insight. 2023;8(22).
5. SANJAY KALRA BK, AMIT SHARMA. Change in Alcohol Consumption Following Liraglutide Initiation: A Real Life Experience. Clinical Diabetes2024.
6. Quddos F, et al. Semaglutide and Tirzepatide reduce alcohol consumption in individuals with obesity. Scientific Reports. 2023;13(1):20998.
7. Bremmer MP, Hendershot CS. Social Media as Pharmacovigilance: The Potential for Patient Reports to Inform Clinical Research on Glucagon-Like Peptide 1 (GLP-1) Receptor Agonists for Substance Use Disorders. Journal of Studies on Alcohol and Drugs. 2023;85(1):5-10.
8. Wium-Andersen IK, et al Osler M. Use of GLP-1 receptor agonists and subsequent risk of alcohol-related events. A nationwide register-based cohort and self-controlled case series study. Basic & Clinical Pharmacology & Toxicology. 2022;131(5):372-9.

Introduction

Upadacitinib, a selective JAK inhibitor, is approved for moderate-to-severe Crohn’s disease (CD). However, real-world data on its effectiveness and safety, particularly in highly refractory populations, remain scarce. The UPITA-Crohn registry was established to evaluate the outcomes of upadacitinib in clinical practice across multiple centers in Andalusia, Spain.

Aims & Methods

This ambispective, multicenter registry included 142 patients with refractory CD treated with upadacitinib in 12 Andalusian hospitals. Clinical and biochemical data were collected at baseline, weeks 12 and 16, and at 6 months. Clinical remission (CR) was defined as a Harvey-Bradshaw Index (HBI) <5; clinical-biochemical remission (CBR) as HBI <5, CRP <5 mg/L, and fecal calprotectin <250 μg/g; and steroid-free remission (SFR) as HBI <5 without corticosteroids from week 12. Outcomes were also analyzed overall and according to prior exposure to advanced therapies.

Results

The UPITA-Crohn registry cohort included 142 patients with refractory Crohn’s disease. The mean age was 40.8 years (range 18–76), with 40% male. The mean disease duration at the time of upadacitinib initiation was 13.3 years (range 1–42), with a mean of 2.7 failed advanced therapies, highlighting a population with longstanding and complex disease. Most patients had ileocolonic involvement (50%), 22% had perianal disease, and 32% had previously undergone resective surgery.
The majority (74.4%) received induction with upadacitinib 45 mg for 12 weeks. At week 12, clinical remission was achieved by 54% of patients, with 15% reaching clinical-biochemical remission. At 6 months, among those still on treatment, 48.3% maintained clinical remission, 16.1% achieved clinical-biochemical remission, and 18.4% achieved steroid-free remission (Table 1).
When stratified by prior therapy, 6-month clinical remission was similar between patients with ≥2 prior advanced therapies (48.5%) and those with only 1 prior therapy (47.4%), with no significant difference. Steroid-free and clinical-biochemical remission rates were also slightly higher among less refractory patients, but differences were modest.
A total of 31 patients (21.8%) discontinued upadacitinib before 6 months, mainly due to primary non-response (70.9%). Adverse events were reported in 21.1% of patients, most commonly infections (9 cases), acne (7), transient transaminase elevation (5), and herpes zoster (2). Seven discontinuations were due to serious adverse events, including thrombosis, stroke, myocardial infarction, fever, and severe infections.

Table 1: Effectiveness Outcomes

Conclusion

Upadacitinib demonstrated meaningful effectiveness and a manageable safety profile in a real-world cohort of refractory Crohn’s disease patients. Notably, clinical remission rates remained stable even in those with multiple prior advanced therapies, suggesting upadacitinib may be a valuable option for patients with limited alternatives.

Introduction

^​​​​​​T cells, are a major regulator and source of pro-inflammatory cytokines, which may promote colitis, but be protective in some cancers^1,2. Nevertheless, limited efforts have been made to explore factors which determine the phenotype of T cells and their impact on the above-mentioned phenomena. The adhesion G protein-coupled receptor F5 (ADGRF5) belongs to seven-transmembrane receptor subfamily and may be related to development of some cancer types³.

Aims & Methods

The aim of the study was to assess the significance of ADGRF5 for T cell function and its impact on colitis and colon cancer progression. WT and ADGRF5 knockout CD4⁺ T cells were examined for cytokine genes and proteins using real-time PCR, Western blot and multiplex analysis. Tumor killing assay and flow cytometry analysis were performed to evaluate tumor killing properties of T cells co-cultured with MC38 colon cancer cells. ADGRF5 WT and KO mice were utilized to examine the impact of ADGRF5 on colitis, colitis-associated cancer development and colon cancer growth. Adoptive transfer of WT and ADGRF5 knockout CD4⁺ and CD8⁺ T cells were examined in tumor models to assess their role in disease progression.

Results

In murine model of colon cancer, animals with ADGRF5 knockout developed significantly smaller tumors which was accompanied by enhanced CD3 expression and hyperproduction of cytokines related to T cell action compared to wild type mice. Ex vivo analysis of two main population of T cells obtained from spleen of wild type and ADGRF5 knockout mice revealed that activated CD4⁺ but not CD8⁺ T cells with ADGRF5 were manifested by increased production of major anti-cancer molecules such as interferon γ (IFN- γ), perforin and granzyme B, and were characterized by secretion profile of cytokines related to mixed phenotype of T helper (T_h) 1 and T_h2 cells. Moreover, we documented higher proportion of apoptotic cells in co-culture of MC38 colon cancer cells with ADGRF5 knockout but not wild type CD4⁺ T cells. Finally, in the murine model of colon cancer adoptive transfer of CD4⁺ T cells with ADGRF5 knockout led to a significant reduction of tumor volume which was reversed by administration of antibody against IFN-γ. Nevertheless, ADGRF5 knockout in mice with colitis-associated colorectal cancer and colitis was related to significantly higher number of colon tumors and worse disease activity index compared to wild type animals, respectively. These results together with the analysis of gene expression signature in inflamed colons suggest opposite role for ADGRF5 in colitis than in the allograft colon cancer.

Conclusion

ADGRF5 is a critical regulator of T cell function and specific knockout of ADGRF5 reprograms CD4⁺ T cells into the pro-inflammatory phenotype. Our in vitro and in vivo studies indicated that ADGRF5 knockout in T cells promote pro-inflammatory cytokines, which prevent gastrointestinal cancer progression while boost both colitis and colitis-associated colorectal cancer development.

References

1. Jacenik D., Karagiannidis I., Beswick E.J. 2022. T_h2 cells inhibit growth of colon and pancreas cancers by promoting anti-tumorigenic responses from macrophages and eosinophils. Br J Cancer. 128: 387-397. doi: 10.1038/s41416-022-02056-2.
2. Shale M., Schiering C., Powrie F. 2013. CD4⁺ T cell subsets in intestinal inflammation. Immunol Rev. 252: 164-82. doi: 10.1111/imr.12039.
3. Jacenik D., Hikisz P., Beswick E.J., Fichna J. 2023. The clinical relevance of the adhesion G protein-coupled receptor F5 for human diseases and cancers. Biochim Biophys Acta Mol Basis Dis. 1869: 166683 doi: 10.1016/j.bbadis.2023.166683.

Disclosure

The work was supported by the National Science Centre, Poland under research project no 2019/32/C/NZ5/00064 (DJ).