Home / Deep Research / Mycosis Fungoides

Mycosis Fungoides

Mycosis Fungoides Deep Research

OpenAI MONDO:0009691 Model: gpt-5.4

Mycosis Fungoides Deep Research

Modeling choices

This curation follows the cancer-modeling guidance from dismech issue #1198. The dismech page is the disease-level mechanism graph for Mycosis_Fungoides, not a page-per-subclass decomposition. I kept the disease anchor MONDO-first at MONDO:0009691 and treated WHO-EORTC-recognized variants as flat subtype facets rather than separate disease pages. That means folliculotropic mycosis fungoides, localized pagetoid reticulosis, and granulomatous slack skin are represented as subtype values under one MF entry. Large-cell transformation was modeled as progression biology within MF rather than as a separate disease file.

NCIT was used where it adds oncology-specific value under the current schema, especially for histopathology and treatment terms. I did not add disease-level or subtype-level NCIT mappings because the present schema exposes MONDO disease mapping slots but not a parallel ncit_mappings structure for disease/subtype anchors.

Disease identity and subtype structure

Recent reviews consistently describe MF as the most common cutaneous T-cell lymphoma and as a disease centered on malignant T-cell infiltration of skin (PMID:40721285, PMID:37124514). Clinical staging remains the core disease axis: patch, plaque, and tumor stages are explicit in recent reviews (PMID:40721285, PMID:34541796). WHO-EORTC-recognized clinicopathologic variants include folliculotropic MF, pagetoid reticulosis, and granulomatous slack skin (PMID:37124514, PMID:29726638, PMID:35485962, PMID:11411914).

The resulting subtype model in YAML is intentionally flat:

  • classification: clinicopathologic_variant
  • classification: skin_stage

This preserves clinically meaningful subtype structure without implying that every ontology subclass needs its own dismech page.

Histopathology and diagnosis

The strongest disease-level histopathology signal available in recent MF review abstracts is the characteristic epidermotropic infiltrate of small-to-medium lymphocytes, especially in early patch-stage disease, alongside the need for repeated biopsies and careful clinicopathologic correlation (PMID:40721285, PMID:34541796). Folliculotropism is clinically important because it is tied to recognized variant structure and poorer prognosis in infiltrated head/neck plaques and tumors (PMID:37124514). Diagnostic workflow is anchored to skin biopsy with histopathology and immunohistochemistry (PMID:37124514).

Pathophysiology

I kept the mechanistic graph atomic and avoided bundled nodes. The main disease mechanisms supported by the literature are:

  1. Cutaneous homing and accumulation of malignant T-cell clones in skin (PMID:40721285).
  2. Deregulated TCR/PLCG1 signaling as a core driver axis (PMID:33923722).
  3. Recurrent JAK/STAT activation supported both by integrative review and contemporary MF genomics (PMID:33923722, PMID:41008827).
  4. TNFRSF1B/TNFR2-driven non-canonical NF-kappaB signaling as a recurrent human genomic mechanism (PMID:26258847).
  5. Cancer-associated fibroblast and collagen-remodeling programs during early progression (PMID:40163757).
  6. Interleukin-4-associated transcriptional activation during patch-to-plaque evolution (PMID:40163757).
  7. M2 macrophage enrichment and exhaustion-linked immune suppression in progressive lesions (PMID:40163757).
  8. RAS-signaling acquisition in large-cell transformation, modeled as MF progression rather than a new disease anchor (PMID:34771672).

Genetics

The genetics section emphasizes lesions that help explain MF mechanism or progression:

  • TNFRSF1B recurrent mutation/gain and TNFR2 signaling (PMID:26258847)
  • CTLA4-CD28 fusion (PMID:26258847)
  • recurrent FAT1, KMT2D, TP53, and JAK3 alterations in a 2025 national genomics analysis (PMID:41008827)
  • transformation-associated RAS mutations in aggressive large-cell transformation (PMID:34771672)

This keeps the disease page focused on driver axes rather than turning the entry into an exhaustive mutation inventory.

Phenotypes

The phenotype set emphasizes disease-defining cutaneous morphology and clinically important symptom burden:

Treatments

The treatment section was kept disease-relevant and ontology-grounded, with NCIT preferred when it gives materially better oncology specificity:

This avoids mixing in very broad generic MAXO treatment labels when a more oncology-specific NCIT procedure is available, while still using MAXO:0000058 for systemic pharmacotherapy actions paired with specific antineoplastic agents.