Behçet’s disease (BD) is a variable-vessel vasculitis characterized by recurrent oral and genital ulcers and heterogeneous multi-organ involvement. Colchicine is a guideline-recommended first-line treatment for mucocutaneous BD due to anti-inflammatory and anti-neutrophilic effects. In China, clinicians have long used a combined regimen of Total Glucosides of Paeony (TGP) and colchicine—colloquially called the “Two Flowers Therapy”—based on historical TCM practice and over three decades of empirical clinical use. TGP is an approved extract of Paeonia lactiflora composed mainly of four monoterpene glycosides: oxypaeoniflorin, albiflorin, benzoyl paeoniflorin, and paeoniflorin. Preclinical and clinical studies cited in the source report that TGP has anti-inflammatory, antioxidative, and immunoregulatory effects, including modulation of cytokine production, T-cell subsets, neutrophil activation, and NF-κB/MAPK signaling. However, systematic mechanistic data explaining synergy between TGP and colchicine in BD were lacking prior to this study.
This study combined network pharmacology and molecular docking to explore potential mechanisms by which the TGP–colchicine combination may act in BD.
Active compound selection: Four principal TGP constituents (oxypaeoniflorin, albiflorin, benzoyl paeoniflorin, paeoniflorin) were identified via the TCMSP database. Colchicine was included as the fifth bioactive compound. SMILES strings were obtained from PubChem and target prediction performed using SwissTargetPrediction.
Disease target collection and overlap: BD-related targets were sourced from GeneCards (total 837 genes). Mapping compound-predicted targets against BD targets produced 31 overlapping targets visualized in a Venn diagram.
Core target identification: Differential gene expression analysis using GEO dataset GSE209567 (29 BD patients, 15 healthy controls) and GEO2R identified six core genes with P < 0.05: MMP9, ICAM1, FGF2, TLR4, EGFR, NOS3.
PPI network and ranking: The 31 overlapping targets were input to STRING to generate a protein-protein interaction (PPI) network (31 nodes, 197 edges). Cytoscape was used to visualize the network and rank core targets by degree centrality.
Molecular docking: Ligand structures from TCMSP and protein structures from the PDB were prepared and processed (water and native ligands removed). AutoDock Vina was used for docking; grid boxes were set to encompass relevant binding sites for EGFR, FGF2, ICAM1, MMP9, NOS3 and TLR4. Docking results reported hydrogen bonds and hydrophobic interactions with favorable distances (reported ranges in the source text). Visualization was done in PyMol.
Functional enrichment: GO analysis (Microbioinformatics platform) indicated enrichment in inflammation- and immune-related biological processes (top processes visualized). KEGG pathway analysis identified 10 significantly enriched pathways, with notable pathways including Lipid and atherosclerosis and AGE-RAGE, implicating regulation of inflammation, immunity, and vascular function.
A retrospective cohort of 355 newly diagnosed adult BD patients presenting with oral and/or genital ulcers was enrolled. Inclusion required age >18 years and mucocutaneous involvement; patients with major organ involvement (thrombosis, arterial disease, gastrointestinal, nervous system, joint disease), severe infections, malignancy, or pregnancy/lactation were excluded. Baseline characteristics: mean age 37.5 ± 12.8 years; 61.4% female. All patients had oral and genital ulceration at baseline. Baseline biochemical indices and blood counts were reported (e.g., CRP 10.1 ± 9.2 mg/L; ESR 24.9 ± 18.8 mm/h).
Patients were assigned to two groups: combination group (CG, n = 231) receiving TGP plus colchicine, and monotherapy group (MG, n = 124) receiving colchicine alone. The source reports that all patients initially received colchicine; baseline TGP use was 0%. Ethical approval and patient consent were obtained. The source text includes a truncated description of the treatment process and regimens; detailed dosing schedules and duration beyond baseline initiation were not fully reported in the provided material.
Network and docking findings: The integrated network identified 31 overlapping targets and 6 core genes (MMP9, ICAM1, FGF2, TLR4, EGFR, NOS3). Molecular docking showed that colchicine and TGP components could form hydrogen bonds and hydrophobic interactions with these core proteins, suggesting direct binding potential.
Enrichment analyses: GO enrichment implicated immune and inflammation processes; KEGG highlighted pathways involved in inflammation, immune regulation, and vascular biology, including Lipid and atherosclerosis and AGE-RAGE signaling.
Clinical outcomes: Early clinical efficacy favored the combination regimen. At early assessment points (M1–M2), oral ulcer prevalence was lower in CG than MG (reported as 0.0% vs. 21.0% at M1, and 0.0% vs. 100.0% at M2 in the source), and genital ulcer prevalence was also lower in CG at M1 (0.9% vs. 6.5%). Both groups reached complete ulcer resolution by month 3. ESR was significantly lower in CG at month 2 (p < 0.001); CRP showed no significant difference. Safety: CG had transient diarrhea as an adverse event; no drug-associated cytopenia occurred in either group.
The combined TGP–colchicine regimen used empirically in China for more than 30 years was supported by this integrative analysis. Network pharmacology and docking suggest that the regimen can interact with core targets relevant to inflammation and vascular function in BD. Enriched GO and KEGG terms align with mechanisms plausible for modulating mucocutaneous inflammation and vascular responses. Clinically, the combination produced superior short-term control of ulcer recurrence and faster reduction in ESR, with acceptable tolerability; long-term comparative safety details beyond the reported transient diarrhea were not provided in the source.
Limitations noted in the source include the retrospective design of the clinical cohort and that the historical use of the regimen in practice has been largely experiential and underreported in formal literature. Some procedural details (precise dosing schedules and full duration of follow-up beyond month 3 endpoints) were not included in the provided text.
The study integrates computational biology and a retrospective clinical cohort to characterize the Two Flowers Therapy (TGP combined with colchicine) for mucocutaneous BD. Findings indicate probable multi-target modulation of inflammatory and vascular pathways (including core targets MMP9, ICAM1, FGF2, TLR4, EGFR, NOS3) and clinical advantages in early control of mucocutaneous lesions with a favorable safety profile as reported in the source. The source did not report some treatment regimen specifics and longer-term outcome details within the provided sections.