How can Rosai‑Dorfman disease be differentiated from Castleman disease using immunohistochemistry?

Immunohistochemical Differentiation of Rosai-Dorfman Disease from Castleman Disease

The key immunohistochemical distinction is that Rosai-Dorfman disease shows strong S100 positivity (both nuclear and cytoplasmic) in the characteristic large histiocytes, while Castleman disease lacks S100-positive histiocytes and instead shows follicular dendritic cell abnormalities.

Core IHC Panel for Rosai-Dorfman Disease

The diagnostic immunophenotype of RDD histiocytes includes 1:

• S100: Strongly positive (nuclear AND cytoplasmic) - this is the most critical marker
• CD68: Positive
• CD163: Variable positive
• CD14: Variable positive
• CD1a: Negative (distinguishes from Langerhans cell histiocytosis)
• CD207/Langerin: Negative

Recent evidence suggests additional markers may aid diagnosis 2:

• CD30: 100% positive in RDD
• OCT2: 90% positive in RDD
• Cyclin D1: 100% positive in RDD

Castleman Disease IHC Profile

Castleman disease, particularly hyaline-vascular type, shows fundamentally different features 3:

• No S100-positive histiocytes (this is the critical negative finding)
• Abnormal follicular dendritic cell networks (use CD21, CD23, CD35)
• Plasma cell predominance in plasma cell variant (use CD138, kappa/lambda light chains)
• In multicentric CD, look for HHV-8 positivity in plasmablasts

Morphologic Context Matters

While IHC is definitive, recognize the morphologic patterns 1, 3:

RDD shows:

• Sinus expansion by large pale histiocytes with "watery-clear" cytoplasm
• Emperipolesis (intact leukocytes within histiocyte cytoplasm) - helpful but not required
• Lymphoplasmacytic background
• Variable IgG4+ plasma cells (evaluate IgG4/IgG ratio)

Castleman disease shows:

• Hyaline-vascular type: abnormal germinal centers with hyaline-vascular transformation, "onion-skinning" of mantle zones
• Plasma cell type: sheets of plasma cells in interfollicular areas
• No characteristic large histiocytes with emperipolesis

Critical Diagnostic Algorithm

1. First-line stain: S100 on suspected histiocytes

   • If strongly positive (nuclear + cytoplasmic): proceed with RDD workup
   • If negative: consider Castleman disease or other entities

2. Confirm RDD: Add CD68, CD163, CD1a

   • CD68+/CD163+/CD1a- with S100+ = RDD confirmed

3. If S100 negative: Evaluate follicular architecture

   • CD21/CD23 for follicular dendritic cell networks (Castleman disease)
   • CD138 for plasma cells
   • HHV-8 if multicentric presentation

4. Supplementary markers (if diagnosis unclear): CD30, OCT2, cyclin D1 strongly favor RDD over reactive processes 2

Common Pitfalls to Avoid

• Don't rely on emperipolesis alone - it can be focal or absent, especially in extranodal RDD, and occurs in other histiocytoses 1
• Extranodal RDD is harder - shows more fibrosis, fewer histiocytes, less emperipolesis; immunostains are essential 1
• IgG4+ plasma cells occur in both - RDD can have increased IgG4+ cells, but typically shows IgG4/IgG ratio <40% (unlike true IgG4-related disease) 1
• >10% of specimen must show RDD morphology to constitute true RDD rather than reactive histiocytosis 1

When Both Coexist

RDD can occur with lymphomas including those mimicking Castleman disease 1. If you see both patterns, ensure >10% of the specimen demonstrates RDD histology to diagnose neoplasia-associated RDD as a distinct entity 1.

The S100 stain is your definitive discriminator - strong diffuse positivity in large histiocytes confirms RDD, while its absence points toward Castleman disease or other lymphoproliferative disorders.