Australian clinical practice guideline: diagnosis and treatment of idiopathic multicentric Castleman disease

Abstract

Idiopathic multicentric Castleman disease (iMCD) is a rare condition. The pathogenesis is incompletely understood; however, interleukin‐6 (IL‐6) is a major mediator. The clinical presentation is heterogeneous, from mild constitutional symptoms to severe multi‐organ failure. The diagnosis is challenging, as it incorporates clinicopathologic criteria and requires careful evaluation to exclude various systemic disorders. Targeting IL‐6 activity forms the cornerstone of modern therapy for iMCD, with siltuximab recommended as first‐line therapy. Rituximab‐based regimens are recommended for second‐line therapy. However, many patients do not achieve adequate responses with limited evidence to guide further therapy. In the context of these substantial challenges, herein we provide a multidisciplinary Australasian clinical practice guideline to characterise clinical and pathological features, summarise treatment pathways and discuss clinical outcomes of the condition. The objective is to develop a multidisciplinary clinical practice guideline in the diagnosis and management of iMCD in Australia.

Introduction

Castleman disease (CD) was first described in 1956 by Dr Benjamin Castleman, who published reports of a single enlarged lymph node, most commonly in the mediastinum, which demonstrated distinct vascularity and hyalinisation. ¹ To this day, CD remains a poorly understood entity, encompassing a variety of lymphoproliferative disorders presenting with lymphadenopathy and sharing characteristic histologic features. The CD Collaborative Network (CDCN) recently proposed a classification system dividing CD into unicentric CD (UCD), with a single lymph node or one lymph node region being affected, and multicentric CD (MCD), in which at least two lymph node regions are affected. ² MCD is further divided into three subtypes based on aetiological driver: (i) Kaposi's sarcoma‐associated herpesvirus (KSHV)/human herpesvirus 8 (HHV8)‐associated MCD; (ii) polyneuropathy, organomegaly, endocrinopathy, monoclonal plasma cell disorder, skin changes (POEMS)‐associated MCD and (iii) idiopathic MCD (iMCD). Moreover, iMCD is divided by phenotype into thrombocytopenia, anasarca, fever, reticulin fibrosis, renal dysfunction, organomegaly (TAFRO) and not otherwise specified (NOS) subtypes. While HHV8 causes HHV8‐associated MCD in immunosuppressed patients and monoclonal plasma cells are the likely driver of POEMS‐associated MCD, the aetiology and pathogenesis of iMCD is poorly understood. ³

This guideline focusses on iMCD, a rare condition accounting for 33% to 58% of published MCD cases. ⁴ In 2018, the prevalence in the United States was 9.7 (5.6–17.8) cases per million, totalling 3172 (1820–5835) cases. ⁴ However, the incidence is higher among Polynesians. The diagnosis of iMCD is challenging because it is based on lymph node histopathology, which has significant inter‐pathologist discordance, and there is a heterogeneous clinical presentation that overlaps with closely related disorders. ² Aetiology and pathogenesis are as yet unknown; however, interleukin‐6 (IL‐6) has been found to be a major driver in patients. ⁵ A proportion of patients experience a rapid disease onset and aggressive clinical course that requires urgent intervention, meeting criteria for the TAFRO subtype. ² Other patients who do not meet TAFRO criteria tend to experience a milder disease course that sometimes includes thrombocytosis, hypergammaglobulinaemia and plasmacytosis. These patients are considered not otherwise specified (NOS), including a subset which has been referred to as idiopathic plasmacytic lymphadenopathy. ³ , ⁶ , ⁷

Treatment guidelines for iMCD were developed by an international expert panel in 2018 based on the review of a limited number of clinical trials and small case series, and recommendations were stratified by disease severity. ⁸ In both severe and mild to moderate disease, siltuximab, a monoclonal antibody directed against IL‐6 and the only medication approved for treating iMCD in Australia, is recommended for first‐line therapy, demonstrating a 34% response rate in a randomised, double‐blind, placebo‐controlled trial. This was defined as a tumour and symptom response that persisted for at least 18 weeks during masked treatment. Tumour response includes complete or partial response (CR or PR) by modified Cheson criteria, ⁹ and symptomatic response constitutes an improvement in or stabilisation of symptoms related to iMCD. ¹⁰ Adjunctive corticosteroids are recommended as needed. The addition of cytotoxic chemotherapy is recommended for patients with severe disease who progress on anti‐IL‐6 therapy. Rituximab‐based regimens are recommended for second‐line therapy, demonstrating a response rate of 26.9%. ⁷ Data are more limited for alternative treatments such as chemotherapeutic agents. ⁸

In the context of the substantial challenges involved in diagnosing and managing iMCD, herein we provide a multidisciplinary Australasian clinical working guideline with input from expert haematologists, rheumatologists, immunologists and pathologists to characterise clinical and pathological features, summarise treatment pathways and discuss clinical outcomes of iMCD. Our aim is to develop a multidisciplinary clinical working guideline in the diagnosis of iMCD in Australia with the following objectives:

1. To generate awareness of iMCD to improve the diagnostic workflow for and treatment of patients in Australia.

2. To create awareness of the need for a multidisciplinary approach to the diagnosis.

3. To propose a clear referral pathway.

4. To address diagnostic challenges of iMCD in Australia.

5. To discuss the differential diagnoses of iMCD and the role of lymph node histology.

6. To discuss considerations unique to Polynesian patients.

7. To discuss treatment of severe and non‐severe iMCD in the frontline and refractory settings.

The key recommendations discussed in this guideline are summarised in Table 1. The levels of evidence and grades of recommendation follow the National Health and Medical Research Council (NHMRC) guidelines, as shown in Table 2. ¹¹ Statements without grading were considered justified standard clinical practice by the experts.

Table 1.

Summary and key recommendations

Recommendations	Level	Grade
The diagnosis and management of iMCD is challenging. Therefore, a multidisciplinary approach leveraging relevant expertise is recommended.
2The clinical presentation is heterogeneous, ranging from mild constitutional symptoms to severe multi‐organ dysfunction.
3It is important to be aware that the incidence of iMCD is higher among Polynesians. The phenotype may be milder with excellent response to IL‐6‐directed therapy.
4The 2017 international consensus criteria are recommended to establish the diagnosis.	IV	C
5Histopathological assessment of the lymph node is an essential part of the diagnosis.	IV	C
6A core needle biopsy may provide information on the lymph node architecture, but is generally suboptimal compared to an open excisional biopsy.
7Fine needle aspiration can only provide cytological analysis and should not be used in the diagnosis.
8A bone marrow biopsy is warranted to assess for differential diagnoses and may demonstrate histological changes that support the diagnosis.	IV	D
9In addition to histopathology, a panel of other investigations is required, including cross‐sectional imaging, blood and urine tests.	IV	C
10Assessment of disease severity according to the CDCN guidelines is recommended to tailor therapy.	IV	C
11Targeting IL‐6 activity forms the cornerstone of modern therapy for iMCD, with siltuximab recommended as first‐line therapy.	II	B
12The recommended regimen for siltuximab is 11 mg/kg every 3 weeks.	II	B
13Patients who achieve at least a PR should remain on siltuximab maintenance therapy indefinitely until disease progression.	IV	C
14Patients with severe disease should receive first‐line siltuximab in combination with high‐dose steroids.	IV	D
15Rituximab‐based regimens are recommended for second‐line therapy.	IV	C
16Third‐line agents include cyclosporin (especially for TAFRO), sirolimus, anakinra, and thalidomide.	IV	D
17Chemotherapy is recommended as a treatment option in severe, refractory iMCD or TAFRO	IV	D
18We recommend using the CDCN‐defined response criteria to guide, rather than dictate, treatment selection and duration.

The levels of evidence and grades of recommendation follow the NHMRC guidelines. ¹¹ Statements without grading were considered justified standard clinical practice by the experts.

CDCN, Castleman Disease Collaborative Network; IL, interleukin; iMCD, idiopathic multicentric Castleman disease; PR, partial response; TAFRO, thrombocytopenia, anasarca, fever, reticulin fibrosis, renal dysfunction, organomegaly.

Table 2.

NHMRC level of evidence and grades of recommendations

Levels of evidence
Intervention
I	Evidence from a systematic review of level II studies
II	Evidence from a randomised controlled trial
III‐1	Evidence from a pseudorandomised controlled trial
III‐2	Evidence from a comparative study with concurrent controls, including a non‐randomised, experimental trial, cohort study, case–control study, or interrupted time series with a control group
III‐3	Evidence from a comparative study without concurrent controls, including a historical control study, two or more single‐arm studies, and interrupted time series without a parallel control group
IV	Evidence from case series with either post‐test or pre‐test/post‐test outcomes
Diagnosis
I	Evidence from a systematic review of level II studies
II	Evidence from a study of test accuracy with an independent, blinded comparison with a valid reference standard, among consecutive persons with a defined clinical presentation
III‐1	Evidence from a study of test accuracy with an independent, blinded comparison with a valid reference standard, among non‐consecutive persons with a defined clinical presentation
III‐2	Evidence from comparison with a reference standard that does not meet the criteria required for Level II and III‐1 evidence
III‐3	Evidence from a diagnostic case–control study
IV	Evidence from a study of diagnostic yield (no reference standard)
Grades of recommendation
A	Body of evidence can be trusted to guide practice
B	Body of evidence can be trusted to guide practice in most situations
C	Body of evidence provides some support for recommendation(s) but care should be taken in its application
D	Body of evidence is weak and recommendation must be applied with caution

Methods

The guideline incorporates multidisciplinary input from 17 expert haematologists, rheumatologists, immunologists and pathologists in Australasia. Multiple meetings were held, which were followed by email exchanges during the drafting process.

Epidemiology

Given the rarity of iMCD, the lack of consensus guidelines and the likely underdiagnosis of cases, it has been difficult to ascertain the true incidence of this disease. Fajgenbaum and colleagues used a claims‐based algorithm based on the CD‐specific International Classification of Diseases‐10th Revision (ICD‐10) code (D47.Z2; introduced in 2017) and ICD‐10 codes corresponding to two of the minor criteria for iMCD to determine an incidence of 3.4 and prevalence of 6.9 cases per million (in 2017). ⁴ The mean age of diagnosis was 51.4 years, and there was a slight female preponderance (56.7% vs 43.3%). In the ethnically uniform population of Japan, the estimated annual incidence of MCD (i.e. not broken down into subtypes) was comparable at 2.4–5.8 per million, while TAFRO incidence was estimated at 0.9–4.9 per million. ¹²

In an analysis of the ACCELERATE Natural History Registry, 65% were White, 13.7% were Asian, and 12% were Black/African American, with the remainder other ethnicities or not stated. Given that this was a patient self‐reported registry, there is significant potential for these rates to be biased towards one or other ethnicities. ⁷ There has been little further research into the demographics of iMCD. In an analysis of a New Zealand database, there did appear to be a higher incidence of iMCD among Polynesians (New Zealand Maori, Samoan, Niuean) of 80 per million, though these estimates were based on a small sample size. ¹³ Nevertheless, the findings were taken to suggest there may be a genetic predisposition in this group, although many factors could have led to this observation. It was also noted that this group appeared to have a milder disease phenotype than their Caucasian counterparts, with Polynesian patients having an excellent response to IL‐6‐directed therapy.

Diagnosis

Common presentations

The cardinal feature of iMCD is the presence of lymphadenopathy (100% of cases), often associated with classical ‘B’ symptoms of fever, weight loss and night sweats (52%–72%) mediated by increased IL‐6 activity. The severity of constitutional symptoms can range from mild to severe, and in a small subset of patients, they may present with the so‐called TAFRO presentation, a constellation of thrombocytopenia (<100 × 10⁹/L), anasarca, fever, reticulin fibrosis on bone marrow biopsy, renal impairment and organomegaly. Anasarca, the presence of generalised oedema, ascites and pleural effusions, is common in iMCD (78%) and thought to be driven by vascular endothelial growth factor (VEGF)‐mediated capillary leak. Secondary phenomena such as lymphocytic interstitial pneumonitis and skin changes (eruptive cherry haemangiomatosis) may also occur. ¹⁴ Presentations may be mild, with weight loss, fatigue and intermittent episodes of fever and night sweats with lymphadenopathy, hepatosplenomegaly and oedema on physical examination. On the most severe end of the spectrum, patients may present with rapid‐onset multi‐organ dysfunction, often seen in TAFRO subgroup. ⁷ Interestingly, the lymphadenopathy in TAFRO cases is usually modest (<1.5 cm), and laboratory findings typically include elevated alkaline phosphatase and the absence of polyclonal hypergammaglobulinaemia. ¹⁵ , ¹⁶

The extent of other end‐organ involvement determines disease severity, which informs decisions about urgency and intensity of treatment. Pulmonary involvement, such as the presence of interstitial pneumonitis, severe renal dysfunction or significant anaemia (haemoglobin ≤80 g/L), presents markers of severity. The patient's functional status, as measured by the Eastern Cooperative Oncology Group (ECOG) performance status scale, determines the urgency of treatment and whether patients need to be hospitalised, often in a high‐dependency setting.

Diagnostic criteria

Consensus diagnostic criteria for iMCD were proposed in 2017 by the CDCN, an international group comprising pathologists, paediatricians, and physicians with expertise in iMCD. ² These criteria were endorsed by the World Health Organisation (WHO) Classification criteria published in 2024. ¹⁷ In the WHO classification, idiopathic multicentric Castleman disease (iMCD) is categorised as ‘tumour‐like lesions with B‐cell predominance’. This classification distinguishes iMCD from malignant B‐cell lymphomas, while emphasising a B‐cell‐driven pathogenesis and the pathognomonic feature of lymph node proliferation.

The major criteria require the presence of lymphadenopathy affecting at least two lymph node stations and characteristic histopathological features. The histopathological features are best assessed on an excisional biopsy of the abnormal lymph node. Two recognised patterns of lymph node pathology include regressed germinal centres and sheet‐like polytypic plasmacytosis in the interfollicular space. These histopathological variants may co‐exist, or a single dominant pattern may be observed. Additional supportive histological features include prominent endothelium or blood vessels penetrating into the germinal centres, creating a ‘lollipop’ appearance, expanded mantle zones with concentric rings of lymphocytes, known as ‘onion skinning’ pattern, and in some cases, the germinal centre may appear hyperplastic.

In addition to the major criteria, systemic inflammation and end‐organ involvement demonstrated through laboratory or clinical features serve as minor criteria (Table 3). KSHV/HHV8 latency‐associated nuclear antigen (LANA‐1) immunohistochemistry (IHC) must be negative. Before confirming a diagnosis of iMCD, other potential causes, including infectious, immunologic and malignant disorders, must be excluded. ²

Table 3.

Diagnostic criteria for iMCD‐NOS and iMCD‐TAFRO according to the World Health Organisation Classification of Tumours, fifth edition ¹⁷

iMCD‐NOS	iMCD‐TAFRO
Major criteria (all three required)
1 Enlarged lymph nodes in at least two lymph node stations
2 Morphological features consistent with Castleman disease spectrum, including:   ○ Grade 2–3 regressed follicles OR   ○ Grade 2–3 plasmacytosis
3 Negative KSHV/HHV8 LANA‐1 IHC
Major criteria for further subclassification of iMCD‐TAFRO
May demonstrate some, but not all five required features for iMCD‐TAFRO	All five are required:
	Thrombocytopenia (<100 × 109/L)
	Anasarca (includes pleural effusions, ascites, subcutaneous oedema)
	Systemic inflammation: fever >37.5°C and/or CRP ≥20 mg/L
	At least one of the following: bone marrow with reticulin fibrosis and/or megakaryocytic hyperplasia without another cause OR renal insufficiency (pretreatment; eGFR <60 mL/min/1.73 m2; creatinine >115 μmol/L in male patients, >97 μmol/L in female patients) or renal failure (dialysis)
	Organomegaly (CT): Small‐volume lymphadenopathy (often <10 mm in greatest diameter) in more than two lymph node stations and/or hepatomegaly and/or splenomegaly
Minor criteria (need at least two of the criteria below, with at least one being a laboratory criterion)
1 Laboratory criteria   ○ Anaemia (<125 g/L in male patients, <115 g/L in female patients)   ○ Thrombocytopenia (platelet count <150 × 109/L) or thrombocytosis (platelet count >400 × 109/L)   ○ CRP > 10 mg/L   ○ Renal dysfunction or proteinuria   ○ Polyclonal hypergammaglobulinaemia	If the above major criteria are met, then the criteria for iMCD‐TAFRO are fulfilled
2 Clinical criteria   ○ Constitutional symptoms   ○ Large spleen and/or liver   ○ Fluid accumulation (oedema, effusions, anasarca)   ○ Eruptive cherry haemangiomatosis or violaceous papules   ○ Lymphocytic interstitial pneumonitis	If the above major criteria are met, then the criteria for iMCD‐TAFRO are fulfilled
Additional supportive histopathological features (not required)
1 Grade 2–3 vascularity is often present (hypervascular > hyaline‐vascular) 2 Lymph nodes borderline enlarged (often ≤10 mm)	Grade 2–3 hyperplastic germinal centres often present (mixed/plasmacytic histopathological features) At least one lymph node measuring >10 mm on the short axis
Other supportive features (not required)
1 Elevated levels of IL‐6, sIL‐2R, VEGF, and/or B2M 2 Elevated alkaline phosphatase without elevation in bilirubin or transaminases	Elevated levels of IL‐6, sIL‐2R, VEGF, IgA, IgE, LDH, and/or B2M Reticulin marrow fibrosis Typically associated with thrombocytosis and hypergammaglobulinaemia Disorders associated with iMCD†
Exclusion criteria‡
1 Infection:   ○ EBV, COVID‐19, HIV, KSHV/HHV8, tuberculosis, etc.	Infection: ○ EBV, COVID‐19, HIV, KSHV/HHV8, tuberculosis, etc.
2 Autoimmune, inflammatory or immunodeficiency disorders   ○ SLE, RA, Still's disease, SjD ALPS, haemophagocytic lymphohistiocytosis, IgG4‐RD, etc.	2 Autoimmune, inflammatory, or immunodeficiency disorders ○ SLE, RA, Still's disease, SjD, ALPS, IgG4‐RD, etc.
3 Malignancy:   ○ Non‐Hodgkin lymphoma, plasma cell myeloma/plasmacytoma, metastatic cancer, POEMS syndrome, etc.	3 Malignancy: ○ Non‐Hodgkin lymphoma, Hodgkin lymphoma, plasma cell myeloma/plasmacytoma, follicular dendritic cell sarcoma, POEMS syndrome, etc.

Modified from World Health Organisation classification of Tumours, fifth edition. ¹⁷

^†Paraneoplastic pemphigus, bronchiolitis obliterans organising pneumonia, autoimmune cytopenia, polyneuropathy (without diagnosing POEMS), glomerular nephropathy, inflammatory myofibroblastic tumour. ²

^‡Careful clinical evaluation is needed as some patients with Castleman disease may also have some of these processes in addition to Castleman disease.

ALPS, autoimmune lymphoproliferative syndrome; APS, anti‐phospholipid syndrome; B2M, beta‐2‐microglobulin; CMV, cytomegalovirus; CRP, C‐reactive protein; EBV, Epstein–Barr virus; eGFR, estimated glomerular filtration rate; GC, germinal centre; HIV, human immunodeficiency virus; Ig, immunoglobulin; IgG4‐RD, Immunoglobulin G4‐related disease; IHC, immunohistochemistry; IL‐6, interleukin 6; iMCD, idiopathic multicentric Castleman disease; KSHV/HHV8, Kaposi's sarcoma‐associated herpesvirus/Human Herpesvirus 8; LANA‐1, latency‐associated nuclear antigen; LDH, lactate dehydrogenase; NOS, not otherwise specified; POEMS, polyneuropathy, organomegaly, endocrinopathy, monoclonal paraprotein, skin changes; RA, rheumatoid arthritis; sIL‐2R, soluble interleukin‐2 receptor; SjD, Sjögren disease; SLE, systemic lupus erythematosus; TAFRO, thrombocytopenia, anasarca, fever, reticulin fibrosis and organomegaly; VEGF, vascular endothelial growth factor.

Histology

CD is divided into unicentric and multicentric subtypes. UCD is a benign disorder involving a single nodal group. Some consider hyaline vascular (HV) morphology synonymous with this condition. MCD is defined as the involvement of two or more nodal groups with lymph nodes greater than 1 cm in short axial diameter. It is divided into HHV‐8, POEMS‐associated and idiopathic MCD subtypes. The first two entities are relatively well defined in terms of pathogenesis and morphology (Table 4). The diagnosis of iMCD, on the other hand, requires compatible pathological and clinical features. The histology is reproducible but essentially non‐specific without corroborating clinical and laboratory findings.

Table 4.

Comparison of histopathology of different MCD subtypes ¹⁷ , ⁶⁰ , ⁶¹ , ⁶² , ⁶³

	KSHV/HHV8‐associated MCD	POEMS‐associated MCD	iMCD‐NOS	iMCD‐TAFRO
General	Mainly plasmacytic or mixed pattern Regressed follicles Vascularity may be significantly increased Prominent high endothelial venules in the paracortical space	Variable pattern (mixed, plasmacytic, or hypervascular) Regressed follicles Germinal centre hyperplasia may be absent	More commonly mixed pattern Less atrophic follicles Vascularity often increased (grade 2–3)	More commonly mixed pattern Atrophic follicles more common Germinal centre hyperplasia often present (grade 2–3) More high endothelial venules, mainly in the interfollicular areas
Plasma cells	Interfollicular polyclonal plasmacytosis is common Plasmablasts present in variable numbers, mainly located in the mantle zone; polyclonal but express monotypic lambda light chain; may form aggregates	May show prominent plasmacytosis in sheets; atypia (e.g. binucleation) may be seen Monoclonal plasma cells may be present, more commonly expressing lambda light chain	More interfollicular plasma cells, which are polyclonal	Less interfollicular plasma cells, which are polyclonal
KSHV/HHV8 LANA‐1 IHC	Positive	Negative	Negative	Negative
Other	Co‐existing Kaposi's sarcoma may be seen		Bone marrow may show megakaryocytic hyperplasia	Bone marrow may show megakaryocytic hyperplasia or reticulin fibrosis

IHC, immunohistochemistry; iMCD, idiopathic multicentric Castleman disease; KSHV/HHV8, Kaposi's sarcoma‐associated herpesvirus/Human Herpesvirus 8; LANA‐1, latency‐associated nuclear antigen; MCD, multicentric Castleman disease; NOS, not otherwise specified; TAFRO, thrombocytopenia, anasarca, fever, reticulin fibrosis and organomegaly; VEGF, vascular endothelial growth factor.

The spectrum of lymph node morphology is based on the degree of germinal centre regression, hyperplasia, plasmacytosis, follicular dendritic cell prominence and nodal vascularity. These changes can vary in severity. For the purpose of diagnosing iMCD, a moderate degree of follicular atrophy in terms of size and cellularity is required. Similarly, a moderate degree of plasmacytosis may be seen, as shown by increased plasma cell numbers in the interfollicular spaces, leading to the expansion of this zone. Occasional binucleated plasma cells may be present but are not dominant.

Different clinical subtypes of iMCD can have some differences in the dominant histological findings of the lymph node. Most notably, patients who present with TAFRO often have germinal centre hyperplasia, more prominent vascularity and fewer plasma cells. It is important to note that KSHV/HHV8‐associated MCD can have identical histopathological features, but concomitant Kaposi's sarcoma could be a clue, and immunohistochemical stain will be positive for LANA‐1.

Various aspects of lymph node germinal centre morphology (i.e. regression, hyperplasia, follicular dendritic prominence), nodal vascularity and plasmacytosis are graded on a scale of 0–3. Grade 2 or 3 regressed germinal centres or plasmacytosis is a minimum requirement for diagnosis. Pathological features representing the extreme ends of the disease spectrum can be seen. At one extreme, germinal centres are typically large and reactive, with marked background polyclonal plasmacytosis. The degree of germinal centre regression and nodal vascularity may be subdued in these cases. At the other end of the spectrum, germinal centres are regressed, associated with follicular dendritic cells and vascular proliferation, resembling the HV pattern seen in UCD. A proportion of the patients with this pathology present with features of TAFRO syndrome.

The differential diagnosis of iMCD‐type histology includes POEMS‐associated MCD (nodal plasma cells may or may not be clonal in these cases) and HHV‐8‐associated MCD, in addition to a variety of MCD mimics, including autoimmune diseases, IgG4‐related disease (IgG4‐RD), Human Immunodeficiency Virus (HIV) associated lymphadenopathy and lymphoma of varying types.

Other investigations

While histological findings are key to the diagnosis of iMCD, a panel of other investigations is required to work through the major, minor and exclusion criteria of this entity. ²

Cross‐sectional imaging with either CT scan or PET/CT is warranted to identify lymphadenopathy (major criteria) as well as identifying organomegaly (spleen and liver), fluid accumulations (including pleural effusions or ascites) or interstitial pneumonitis, the latter of which are all minor criteria.

Essential bloodwork includes full blood count (to identify anaemia or thrombocytopenia), creatinine (including glomerular filtration rate (GFR) calculations), liver function testing to identify low albumin, immunoglobulin testing (to identify polyclonal hypergammaglobulinaemia) as well as testing inflammatory markers (i.e. erythrocyte sedimentation rate (ESR) and C‐reactive protein (CRP)). All these factors represent minor criteria. Important exclusions will, of course, require testing for HHV‐8, as well as testing for Epstein–Barr virus (EBV) and other uncontrolled infections identified in the diagnosis criteria. Additional bloodwork may include IL‐6, soluble interleukin‐2 receptor (sIL‐2R) and VEGF where appropriate.

While a bone marrow biopsy is not strictly indicated to account for the major and minor diagnostic criteria, it is warranted to assess for other considerations, such as malignancy or POEMS‐associated MCD and to look for the reticulin fibrosis of TAFRO. In any case, this will often form part of the workup for a fever of unknown origin, which may be the genesis of suspicion for iMCD in the first place. ² , ¹⁸

Diagnostic workup table or box

The suggested diagnostic workup for iMCD is summarised in Table 5.

Table 5.

Suggested diagnostic workup for iMCD

Group	Test or clinical factor	Comments
History and examination	Presence of lymphadenopathy Classical ‘B’ symptoms: fevers, night sweats or weight loss Generalised oedema, ascites or pleural effusions Skin changes (e.g. eruptive cherry haemangiomatosis) Polyneuropathy (in the case of POEMS syndrome)	Lymphadenopathy is found in 100% of patients with iMCD ‘B’ symptoms represent a ‘minor’ clinical criterion
Blood tests	Routine: Full blood count and blood film Urea and electrolytes, calcium, phosphate, LDH, uric acid Liver function tests, including albumin Inflammatory markers (i.e. ESR and CRP) Ig levels, serum protein electrophoresis and immunofixation, serum‐free light chains Specialised: IL‐6, sIL‐2R, VEGF and/or B2M† Infectious diseases testing: HIV, hepatitis B and hepatitis C serology, CMV, EBV, toxoplasmosis serology Tuberculosis screen KSHV/HHV8 PCR (peripheral blood) Blood culture Tests to exclude other differentials: ANA, ENA (such as anti‐Sm, RNP, ribosomal, Ro/SSA, La/SSB ab), anti‐dsDNA, C3 and C4. RF, anti‐CCP ANCA Lupus anticoagulant, anti‐cardiolipin, anti‐beta‐2‐glycoprotein I Direct Coomb's test ACE Ferritin IgG and subclasses	Most of these parameters represent the ‘minor’ clinical criteria.
Urine tests	Protein‐creatinine ratio, albumin‐creatinine ratio Protein electrophoresis and immunofixation	Proteinuria is a ‘minor’ laboratory criteria.
Histology	Lymph node biopsy‐strong preference for excisional biopsy Bone marrow aspirate and trephine	Lymph node findings are instrumental in the diagnosis of iMCD. Reticulin fibrosis on bone marrow biopsy is a supportive finding.
Radiology	CT scan of neck, chest, abdomen and pelvis 18F‐FDG PET/CT scan	These identify lymphadenopathy (major criteria) and minor criteria, including organomegaly (spleen and liver), fluid accumulations (including pleural effusions or ascites) and interstitial pneumonitis.

^†IL‐6, sIL‐2R, VEGF are non‐Medicare rebatable.

ACE, angiotensin‐converting enzyme; ANA, antinuclear antibody, ANCA, antineutrophil cytoplasmic antibodies; B2M, beta‐2‐microglobulin; C3, complement component 3; C4, complement component 4; CMV, cytomegalovirus; CRP, C‐reactive protein; ENA, extractable nuclear antigen; EBV, Epstein–Barr virus; ESR, erythrocyte sedimentation rate; HIV, human immunodeficiency virus; Ig, immunoglobulin; IL‐6, interleukin 6; iMCD, idiopathic multicentric Castleman disease; KSHV/HHV8, Kaposi's sarcoma‐associated herpesvirus; LDH, lactate dehydrogenase; sIL‐2R, soluble interleukin‐2 receptor; SLE, systemic lupus erythematosus; VEGF, vascular endothelial growth factor.

Differential diagnosis

iMCD needs to be distinguished from the paraneoplastic syndrome POEMS. ² Polyneuropathy and monoclonal plasma cell disorder (typically lambda light chain expressing) are mandatory to consider POEMS. ¹⁹ The polyneuropathy is initially sensory with distal symmetrical involvement and can progress to motor deficits.

The clinical presentation of iMCD is inherently similar to many autoimmune and autoinflammatory conditions in rheumatology. Autoimmune manifestations are common in iMCD, including autoimmune haemolytic anaemia (AIHA), immune thrombocytopenic purpura (ITP), demyelinating peripheral neuropathy, renal dysfunction with proteinuria or paraneoplastic pemphigus. ²⁰ The presence of autoantibodies such as positive antinuclear antibody (ANA), antibodies targeting extractable nuclear antigens (ENA) or antineutrophil cytoplasmic antibodies (ANCA) can be found in 30% of iMCD cases. ²⁰ iMCD can also coincide with other autoimmune conditions, such as myasthenia gravis and psoriasis, or autoinflammatory diseases, such as familial Mediterranean fever. ²⁰

Common to many rheumatological conditions, the exact aetiology of iMCD is not well defined; however, the consensus definition of iMCD involves findings of a non‐malignant lymphoproliferative process in multiple lymph glands, with or without hepatosplenomegaly and systemic inflammatory illness that is characterised by cytokine (predominantly IL‐6) driven end‐organ effects. As this process is similarly observed in several autoimmune and autoinflammatory conditions in rheumatology, the diagnostic workup requires careful consideration of these other conditions in order to understand the overall prognosis and other treatment options.

Among these autoimmune and autoinflammatory conditions, the most notable to consider are systemic lupus erythematosus (SLE), Sjögren disease (SjD), rheumatoid arthritis (RA), IgG4‐related disease (IgG4‐RD), sarcoidosis, Still disease, VEXAS syndrome and ANCA‐associated vasculitis (Table 6). While lymphadenopathy and systemic inflammatory response are common across all these conditions, each has defining clinical features which usually help clinicians differentiate between them. For the assessment of the individual patients, their specific end‐organ manifestations, together with their serological profile, are key to diagnosis. Of note, although specific, not all patients with SLE have anti‐double‐stranded DNA (dsDNA) or anti‐Sm antibodies or low complement components 3/4 (C3/C4); various other clinical, serological, haematologic and pathologic classification criteria enable confident diagnosis.

Table 6.

Autoimmune and autoinflammatory conditions to consider as differential diagnosis of iMCD

	Predominant clinical features	Typical serological profile	Typical lymph node histology
SLE	Skin, joints, renal and haematological	Anti‐dsDNA, anti‐Sm, anti‐RNP, anti‐ribosomal, anti‐cardiolipin, anti‐Beta‐2‐glycoprotein I; low C3/C4; positive direct Coomb's test	Reactive follicular hyperplasia; possible variants include hyaline vascular, plasmacytic infiltrative and histiocytic necrotising patterns
SjD	Exocrine gland dysfunction. Extraglandular manifestations in 30–50%: joints, skin, muscle, haematological, renal, pulmonary, nervous system	Anti‐Ro60, anti‐Ro52, anti‐La, RF; 25% negative for ANA/ENA	Reactive follicular hyperplasia; similar pattern as SLE can be seen
RA	Small +/− large joint synovitis	Anti‐CCP, RF	Reactive follicular hyperplasia; occasionally neutrophilic infiltrate
IgG4‐RD	Classic Mikulicz syndrome with lacrimal, parotid and submandibular gland enlargement, most typically associated with systemic features. Other variants such as pancreato‐hepato‐biliary disease, retroperitoneal fibrosis with or without aortitis, or head and neck limited disease	Elevated serum IgG4 levels	Lymphoplasmacytic infiltrate with >10 IgG4 positive plasma cells per high‐power field, including a ratio of ≥40% IgG4 positive cells of all IgG positive. Other features, such as storiform fibrosis and obliterative phlebitis, are useful histological findings but are not universally present. Can be indistinguishable from iMCD pattern.
Sarcoidosis	Bilateral hilar lymphadenopathy.	Serum ACE level	Non‐necrotising granulomatous inflammation
ANCA‐associated vasculitis	Predominantly renal, pulmonary, upper respiratory tract (ear, nose, throat), ophthalmic or peripheral nerve involvement.	ANCA, anti‐PR3 or MPO	Lymph node involvement is rare but can reveal granulomatous vasculitis or follicular hyperplasia.
Still's disease	Persistent/recurrent fever, arthritis/arthralgia, skin rash, pharyngitis, lymphadenopathy, hepatosplenomegaly, HLH	RF or ANA positive in <10% of cases; hyperferritinaemia	Reactive follicular hyperplasia or non‐specific chronic inflammation
VEXAS syndrome	Predominantly males >50 years old. Fever, neutrophilic dermatitis, arthritis, chondritis, vasculitis, pulmonary infiltrates and periorbital oedema. Haematological abnormalities such as MDS.	Negative serology. Pathogenic somatic variant in the UBA1 gene	Lymphadenopathy is common. Reactive follicular hyperplasia, lymphoplasmacytic infiltrate, fibrinoid necrosis, and extra medullary haematopoiesis reported.

ACE, angiotensin‐converting enzyme; ANA, antinuclear antibody; ANCA, antineutrophil cytoplasmic antibody; C3, complement component 3; C4, complement component 4; CCP, cyclic citrullinated peptide; dsDNA, double‐stranded DNA; ENA, extractable nuclear antigen; HLH, haemophagocytic lymphohistiocytosis; IgG4, Immunoglobulin G4; IgG4‐RD, IgG4‐related disease; iMCD, idiopathic multicentric Castleman disease; MDS, myelodysplastic syndrome; MPO, myeloperoxidase; PR3, proteinase 3; RA, rheumatoid arthritis; RF, rheumatoid factor; RNP, ribonucleoprotein; SJD, Sjögren disease; SLE, systemic lupus erythematosus; VEXAS, vacuoles, E1 enzyme, X‐linked, autoinflammatory, somatic.

The diagnosis of SjD is frequently missed because of incomplete clinical evaluation: it requires objective confirmation of dry eyes or mouth, in the knowledge that many confirmed patients do not report subjective dryness, ²¹ and either positive serum anti‐Ro or a positive salivary gland biopsy, which is the gold standard for confirming or excluding SjD. Furthermore, serum ANA and ENA are poor screening tools for SjD, with upwards of 25% of confirmed cases testing negative for each of these tests. Anti‐Ro‐negative patients and patients presenting with clinical manifestations suggestive of SjD such as cytopenia, polyclonal hypergammaglobulinaemia, lymphocytic interstitial pneumonitis or tubulointerstitial nephritis, but not yet manifesting objective dryness, require salivary gland biopsy for confirmation/exclusion of SjD. ²²

While autoimmune serologies may help to differentiate iMCD from other conditions, it is important to recognise the limitations. For example, in ANCA‐associated vasculitis, ANCA testing may be negative. ²³ Additionally, patients with CD may have concomitant autoimmune conditions. Some patients with CD without concomitant autoimmune conditions may also demonstrate a positive autoimmune serology. Sun et al. reported that a positive ANA, ENA, anti‐dsDNA, ANCA or anticardiolipin antibody was found in 44.4% and 12.9% of CD patients with and without concomitant autoimmune conditions respectively. ²⁴

Widespread lymphadenopathy mirrors disease activity in SLE and SjD, improving with the treatment of the disease. Lymph node biopsy in these diseases is usually only performed to rule out complicating infection or malignancy. However, specific histopathological findings at the lymph node may assist with the diagnostic process. Follicular hyperplasia is the usual histological finding in SLE, SjD and RA. In contrast, in IgG4‐RD, findings can be indistinguishable from iMCD, but findings of storiform fibrosis or obliterative phlebitis would suggest IgG4‐RD. While sarcoidosis typically involves findings of bilateral hilar lymphadenopathy, in rare cases, granulomatous inflammation can be found in patients with neurological, gastrointestinal and renal involvement without this classic feature.

The recently described VEXAS (vacuoles, E1 enzyme, X‐linked, autoinflammatory, somatic) syndrome is a monogenic adult‐onset autoinflammatory disease that has presented in a small number of cases as a clinical and histopathological mimic of iMCD. ²⁵ Genetic testing for UBA1 somatic mutation will differentiate the two conditions. Still disease (formerly divided by age of onset into Adult‐onset Still disease and systemic juvenile idiopathic arthritis) is a polygenic autoinflammatory disease characterised by features of macrophage activation syndrome and associated upregulation of macrophage‐derived cytokines. It can clinically mimic iMCD with lymphadenopathy and systemic inflammatory symptoms. Laboratory findings such as hyperferritinaemia, transaminitis and hypertriglyceridaemia would support the diagnosis of Still disease. Additionally, histological features of the lymph node and the use of established diagnostic criteria may help differentiate between these conditions and iMCD. ²⁰ , ²⁶ , ²⁷ , ²⁸

Investigation of lymphadenopathy

As mentioned in the previous section, generalised lymphadenopathy is associated with many conditions, including lymphoproliferative disorders, infections, particularly viral and rheumatologic conditions. Hence, a pragmatic approach incorporating clinical examination, imaging and histologic findings is required. ²⁹ In addition to the site, the size, consistency, tenderness and fixation to adjacent tissues are relevant features on clinical examination, with supraclavicular or axillary lymphadenopathy raising the possibility of tumour metastases, hilar and mediastinal lymphadenopathy, a common feature of sarcoidosis, and large, matted, non‐tender nodes suspicious for malignancy. Imaging can be helpful in investigating lymphadenopathy, such as ultrasound findings of nodal matting, surrounding soft tissue oedema and prominent hilar vascularity suggestive of reactive inflammation, while nodal hypoechogenicity and alteration in nodal vascularisation, including avascular areas, displacement of vessels and peripheral vascularity suggestive of malignancy. CT and MRI allow better characterisation of nodal tissue, and 18F‐FDG PET can often discriminate between reactive and malignant lymph nodes based on the levels of standardised uptake values (SUVs) and selecting the most active lymph node for diagnostic biopsy. It may also uncover unsuspected malignant or infective lesions, large‐vessel vasculitis or polymyositis. Ultrasound and CT are also used to guide lymph node biopsy or fine‐needle aspirate.

Histopathological assessment of the lymph gland is an essential part in the consideration of these differential diagnoses. While core needle biopsy can provide information on the lymph node architecture, it is generally suboptimal compared to an open excisional biopsy. Fine needle aspiration can only provide cytological analysis and should not be used in the diagnostic workup for iMCD.

Proposed referral pathway

Given the diagnostic and management challenges associated with iMCD, we propose a streamlined referral pathway (Fig. 1). The manifestations of iMCD are protean; therefore, a high index of suspicion and careful consideration of the differential diagnoses are integral to the diagnosis. For patients where iMCD is suspected, referral to haematology is recommended. After completing the necessary investigations, we recommend a multidisciplinary team meeting to review the diagnosis and management plan.

Figure 1.

Proposed referral pathway. iMCD, idiopathic multicentric Castleman disease; MDT, multidisciplinary team.

Treatment of iMCD

The overall aim of treatment in iMCD is to suppress aberrant cytokine activity, control disease symptoms and prevent organ dysfunction. Siltuximab, an IL‐6 receptor antagonist, is the only approved therapy for iMCD in Australia, based on results from randomised controlled studies, and remains the mainstay of treatment for both severe and non‐severe iMCD. In this section, we discuss criteria for defining severe and non‐severe disease, use of siltuximab in both these scenarios, and provide recommendations on a range of other therapies that can be used for patients who do not respond or are intolerant to siltuximab.

Figure 2 summarises the recommended treatment of severe and non‐severe iMCD, as discussed in the subsequent sections.

Figure 2.

Treatment of non‐severe (A) and severe (B) iMCD. ⁷ , ⁸ , ⁴⁰ , ⁴⁴ d, day; iMCD, idiopathic multicentric Castleman disease; IVMP, intravenous methylprednisolone; mo, month; w, week. ¹Prednisone should be used for short periods only (4–6 weeks) while awaiting symptom response to siltuximab. ²Siltuximab recommended as first‐line therapy (Level II, Grade B). ³Weekly siltuximab for the first 2–4 doses can be considered. ⁴No evidence to favour one regimen over another, but CHOP or etoposide most commonly used. ⁵Rituximab‐based regimens recommended as second‐line therapy (Level IV, Grade C). ⁶Third‐line agents include cyclosporin (esp. TAFRO), sirolimus, anakinra and thalidomide (Level IV, Grade D). *There is insufficient evidence to recommend the extension of siltuximab to 6‐weekly responding patients at time of writing (see section Treatment of severe iMCD).

Defining severe and non‐severe disease

Assessment of disease severity is recommended to tailor therapy, as discussed in the following sections. The severity of iMCD is determined according to the CDCN guidelines (Table 7). ⁸ Patients meeting at least two out of the five criteria are categorised as having severe disease.

Table 7.

Criteria for defining severe^† iMCD according to the CDCN guidelines ⁸

Eastern Cooperative Oncology Group (ECOG) performance status ≥2

Stage IV renal dysfunction (eGFR <30 mL/min/1.73 m2; creatinine >265 μmol/L)

Anasarca AND/OR ascites AND/OR pleural/pericardial effusion

Haemoglobin ≤80 g/L

Pulmonary involvement/interstitial pneumonitis with dyspnoea

Reproduced from van Rhee et al. ⁸

^†Patients meeting at least two of the five criteria are categorised as having severe disease.

CDCN, Castleman Disease Collaborative Network; eGFR, estimated glomerular filtration rate; iMCD, idiopathic multicentric Castleman disease.

Siltuximab

The cornerstone of modern therapy for CD is targeting IL‐6 activity, a major driver of iMCD, with monoclonal antibodies, such as siltuximab and tocilizumab, with and without steroids. The single‐arm phase I study on siltuximab treated 37 patients with CD, including 34 patients with iMCD. Thirty‐six patients were evaluated for disease response using the modified Cheson criteria. ⁹ Twelve out of 36 evaluable patients responded, giving an overall response rate (ORR) of 33%. One (8.3%) and 11 (91.7%) had CR and PR respectively. Clinical benefit response, including haemoglobin, fatigue, anorexia, fever, night sweats, weight and lymph node response, was observed in 32 (87%) patients. Grade 3 adverse events were reported in 11%. ³⁰ An extension phase II study of 19 out of 34 patients with iMCD, including one patient with CR, 11 with PR and seven with stable disease (SD) at the start of the study, showed that treatment response could be maintained for a median of 5.1 (3.4–7.2) years. ³¹

To date, only one large randomised clinical trial has examined the use of siltuximab in iMCD, randomising 79 patients with HIV and HHV‐8 negative iMCD patients to intravenous (IV) siltuximab 11 mg/kg every 3 weeks (n = 53) versus placebo (n = 26), with all patients receiving best supportive care. ¹⁰ The primary outcome was a tumour and symptom response that persisted for at least 18 weeks during masked treatment, as assessed by investigators. Tumour response includes either a CR or PR by modified Cheson criteria, ⁹ and symptomatic response constitutes an improvement in or stabilisation of symptoms related to iMCD. Tumour response was also assessed by an independent radiological review. Symptomatic response was assessed using the sum of the severity of 34 iMCD‐related signs and symptoms. ¹⁰

The primary endpoint occurred in 18 (34%) patients in the siltuximab group, including one and 17 (2 and 32%) patients with CR and PR, respectively, and 0% in the placebo group (difference 34.0%, 95% CI 11.1–54.8, P = 0.0012). Patients randomised to receive siltuximab had a longer duration of therapy (375 (1–1031) days) compared to placebo (152 (23–666) days). Four (31%) of 13 and one (11%) of nine patients in the siltuximab and placebo groups were able to discontinue corticosteroids (difference 19.7%, 95% CI −23.6–56.7, P = 0.3602). The rate of serious adverse events was similar between siltuximab and placebo groups (23% vs 19%). However, some adverse events were more commonly seen with siltuximab, including pruritus, weight gain, maculopapular rash and upper respiratory tract infections. ¹⁰ Additional real‐world evidence has been published since (see Table 8 for a summary). ³² , ³³ , ³⁴ , ³⁵ , ³⁶ As a result of the randomised clinical trial, siltuximab was FDA‐approved and is the only Pharmaceutical Benefits Advisory Committee (PBAC)‐reimbursed medication for iMCD in Australia. International CDCN and NCCN guidelines suggest that all patients with iMCD should commence therapy with siltuximab 11 mg/kg every 3 weeks. ⁸ , ³⁷

Table 8.

Comparison of efficacy and safety of different regimens in iMCD

Regimen	Study design	Line of therapy	Therapy	Response criteria	Efficacy	Adverse events	References
Siltuximab‐based	Randomised, double‐blind, international, multicentre, placebo‐controlled trial	First‐line (n = 33) or previously treated (n = 46)	Siltuximab (n = 53) versus placebo (n = 26)	Response in tumour (CR or PR by modified Cheson criteria) and iMCD‐associated symptoms that persisted for at least 18 weeks during masked treatment 9	ORR: 34 versus 0% in placebo; CR: 2%, PR: 32%	Siltuximab versus placebo: SAE (23 vs 19%); all‐grades pruritus (42 vs 12%), URTI (36 vs 15%), maculopapular rash (34 vs 12%), peripheral oedema (32 vs 23%), weight gain (21 vs 0%) Siltuximab group: infusion reactions (8%), anaphylaxis (1.9%)	Van Rhee et al. 10
	Phase II, open‐label, multicentre, extension study	First‐line (n = 7) or previous systemic therapy (n = 12)	Siltuximab (n = 19)	Modified Cheson criteria 9	At the start of the study, CR was 5.3%, PR 57.9%, and SD 36.8%. All patients had sustained disease control; 42.1% had prolonged CR and changed from 3‐ to 6‐weekly dosing	Over phase I and II: all‐grades URTI (90%), nausea (63%), vomiting (58%), diarrhoea (53%), raised total cholesterol (47%), hypertriglyceridaemia (42%), headache (42%), rash (42%), abnormal LFT (42%)	Van Rhee et al. 31
	Retrospective, single‐centre	First‐line (n = 10) or previously treated (n = 11)	Siltuximab (n = 21)	≥50% improvement in iMCD‐associated symptoms and laboratory abnormalities	CR: 42.86%, PR: 33.33%, NR: 23.81%	Not reported	Yu et al. 5
	Retrospective, international, multicentre	First (n = 5) or second line (n = 4)	Siltuximab (n = 9)	CDCN‐defined response criteria 8	ORR: 55.56%, NR: 22.22%, treatment failure†: 22.22%	Not reported	Yu et al. 43
	ACCELERATE international natural history registry	Not reported	Siltuximab ± corticosteroids (n = 42)	Proportional change in iMCD‐associated clinical and laboratory derangements 2	ORR: 52.4%	Percentage of total events (n = 90): rash (10%), nausea (6.7%), headache (6.7%), diarrhoea (5.6%), fatigue (5.6%), weight gain (4.4%), pain in extremities (3.3%), infusion reactions (1.1%)	Pierson et al. 7
	Retrospective, multicentre	Relapsed/refractory	Siltuximab (n = 9)	Modified Cheson criteria and clinical response (haemoglobin, fatigue, fever) 9	ORR/CR: 33.3%, SD: 44.5%, PD: 22.2%	No grade ≥3 AEs; URTI (33%), fatigue grade 1–2 (22.2%), pruritus grade 1–2 (22.2%), localised oedema grade 2 (22.2%), weight gain grade 1 (11.1%)	Tonialini et al. 32
	Retrospective, multicentre	First‐line (n = 4) or previously treated (n = 7)	Siltuximab (n = 11)	Radiological (modified Cheson criteria) and symptomatic (CTCAE version 4) response 9 , 64	Radiological‐ ORR: 73%, CR: 18%, PR: 55%, PD: 27% Symptom‐ most cases had best response after the 4th (3rd‐9th) cycle 3‐year OS: 81 (95% CI 81 ± 11) % 3‐year PFS: 45 (95% CI 45 ± 15) %	URTI (64%), recurrent diverticulitis (9%), diarrhoea (9%), haemoglobin >175 g/L (9%), raised transaminases (9%), acute renal failure (9%)	Ostrowska et al. 33
	Retrospective	First‐line (n = 6) or previously treated (n = 9)	Siltuximab (n = 15)	Response in tumour (modified Cheson criteria) and symptoms that persisted for at least 3 months during treatment 9	Radiological – CR: 20%, PR: 46.7%, SD: 20% Symptom – 64.7%	URTI (26.7%), maculopapular rash (20%), nephropathy ≥grade 3 (13.4%), hepatopathy ≥grade 3 (13.4%)	Min et al. 34
	Retrospective, single centre	First‐line (n = 15), previously treated (n = 12)	Siltuximab (n = 27)	CDCN‐defined response criteria 8	CR: 66.7%, PR: 14.8%, SD: 7.4%, PD: 11.1% Symptom – 81.5%	URTI grade 1–2 (22.3%), maculopapular rash grade 2 (14.8%), peripheral neuropathy grade 2 (11.2%), hepatopathy grade 3 (11.1%), nephropathy grade 3 (7.4%), neutropenia grade 2 (7.4%), diarrhoea grade 2 (3.7%), weight gain grade 2 (3.7%), reactivation of pulmonary tuberculosis (3.7%)	Min et al. 35
	Retrospective, international, multicentre cohort study	First‐line (n = 3) or previously treated (n = 44)	Siltuximab (n = 48)	CDCN‐defined response criteria 8	CR: 55.3%, PR: 15.8%, NR/SD: 15.8%, PD: 13.2%	SAE (16.7%); abnormal LFTs (10.4%), anxiety (10.4%), allergic reactions (6.3%), nausea/diarrhoea (6.3%), anaemia (4.2%), thrombocytopenia (4.2%), hypertension (4.2%), bleeding (2.1%), atrial fibrillation (2.1%)	Jitaru et al. 36
Rituximab‐based	Retrospective, single‐centre	First‐line (n = 16) or previously treated (n = 9)	Rituximab or rituximab‐based therapy (n = 25)	≥50% improvement in iMCD‐associated symptoms and laboratory abnormalities	CR: 20%, PR: 48%, NR: 32%	Not reported	Yu et al. 5
	Retrospective, international, multicentre	First (n = 4) or second line (n = 1)	Rituximab (n = 5)	CDCN‐defined response criteria 8	ORR: 60%, NR: 0%, Treatment failure†: 40%		Yu et al. 43
		First (n = 18) or second line (n = 4)	R‐CHOP (n = 22)		ORR: 63.64%, NR: 13.64%, Treatment failure†: 22.73%
	ACCELERATE international natural history registry	Not reported	Rituximab ± corticosteroids (n = 26)	Proportional change in iMCD‐associated clinical and laboratory derangements 2	ORR: 26.9%	Percentage of total events (n = 83): rigours (9.6%), chills (7.2%), fatigue (4.8%), headache (4.8%), chest pain (3.6%), infusion reactions (3.6%), rash (3.6%), urticaria (1.2%)	Pierson et al. 7
			Rituximab + immunomodulator(s) ± corticosteroids (n = 8)		ORR: 12.5%
Tocilizumab‐based	Retrospective, international, multicentre	First (n = 1) or second line (n = 2)	Tocilizumab (n = 3)	CDCN‐defined response criteria 8	ORR: 33.33%, NR: 33.33%, Treatment failure†: 33.33%	Not reported	Yu et al. 43
	ACCELERATE international natural history registry	Not reported	Tocilizumab ± corticosteroids (n = 18)	Proportional change in iMCD‐associated clinical and laboratory derangements 2	ORR: 44.4%	Percentage of total events (n = 26): anaphylactic reaction (15.4%), diarrhoea (7.7%), headache (7.7%), rash (7.7%), hypotension (7.7%)	Pierson et al. 7
Chemotherapy‐based	Retrospective, international, multicentre	First (n = 45) or second line (n = 10)	CHOP or similar (n = 55)	CDCN‐defined response criteria 8	ORR: 41.82%, NR: 27.27%, treatment failure†: 30.91%	Not reported	Yu et al. 43
	ACCELERATE international natural history registry	Not reported	Chemotherapy (n = 25)	Proportional change in iMCD‐associated clinical and laboratory derangements 2	ORR: 52%		Pierson et al. 7
Corticosteroids	Retrospective, international, multicentre	First‐line	Prednisone (n = 17)	CDCN‐defined response criteria 8	ORR: 47.06%, NR: 41.18%, treatment failure†: 11.76%	Not reported	Yu et al. 43
	ACCELERATE international natural history registry	Not reported	Corticosteroids (n = 36)	Proportional change in iMCD‐associated clinical and laboratory derangements 2	ORR: 2.8%		Pierson et al. 7
Immunomodulators	Single‐centre, single‐arm phase II trial	First‐line	Thalidomide, cyclophosphamide, and prednisolone (n = 25)	Response in tumour (modified Cheson criteria) and iMCD‐associated symptoms that persisted for at least 24 weeks 9	ORR: 48%, CR: 4%, PR: 44% Symptom‐ ORR: 48%, CR: 28% 1‐year OS: 88% 1‐year PFS: 60%	1 (4%) mortality due to pulmonary infection. Only grade 3 AE is rash (4%); grade 1–2 constipation (40%), pruritus (20%), peripheral sensory neuropathy (16%), rash (16%), nausea (16%), irregular menses (12.5%)	Zhang et al. 48
	Retrospective, international, multicentre	First (n = 5) or second line (n = 2)	Cyclophosphamide (n = 7)	CDCN‐defined response criteria 8	ORR: 28.57%, NR: 42.86%, treatment failure†: 28.57%	Not reported	Yu et al. 43
		First (n = 2) or second line (n = 3)	Thalidomide (n = 5)		ORR: 40%, NR: 40%, Treatment failure†: 20%
	Prospective, multicenter, single‐arm, phase II trial	Relapsed/refractory	Bortezomib, cyclophosphamide, and dexamethasone (n = 24)	Response at 6‐months as defined according to CDCN criteria 8	ORR/PR: 66.7%, SD: 12.5%, PD: 20.8% Symptom‐ CR: 41.7%, PR: 41.7%, PD: 16.7% 1‐year OS: 92% 1‐year PFS: 79%	No grade ≥ 3 AEs or treatment‐related mortality; irregular menstruation (18.2%), insomnia (16.7%), constipation (16.7%), nausea (16.7%), URTI (12.5%), peripheral sensory neuropathy (12.5%)	Zhang et al. 47
	ACCELERATE international natural history registry	Not reported	Immunomodulators± corticosteroids (n = 21)	Proportional change in iMCD‐associated clinical and laboratory derangements 2	ORR: 19%	Not reported	Pierson et al. 7
	Retrospective, single‐centre	Relapsed/refractory	Sirolimus, monotherapy (n = 15) or combination (n = 11) therapy	CDCN‐defined response criteria 8	CR: 15.4%, PR: 53.8%, SD: 23.1%, PD: 7.7% Symptom‐ CR: 57.7%, PR: 30.8%, SD: 3.8%, PD: 7.7%	No grade ≥ 3 AEs or treatment‐related mortality; mucositis grade 1 (23.1%), rash grade 1–2 (15.4%), diarrhoea grade 1 (3.8%), hyperlipidaemia grade 1 (3.8%)	Liu et al. 51

Where reported, the grade of adverse events is according to the National Cancer Institute Common Terminology Criteria for Adverse Events version 4. ⁶⁴

AE, adverse events; CR, complete response; CDCN, Castleman Disease Collaborative Network; CTCAE, Common Terminology Criteria for Adverse Events; iMCD, idiopathic multicentric Castleman disease; LFT: liver function tests; NR, no response; ORR, overall response rate; OS, overall survival; PD, progressive disease; PFS, progression‐free survival; PR, partial response; R‐CHOP, rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone; SAE, serious adverse events; SD, stable disease; URTI, upper respiratory tract infection.

^†Treatment failure in Yu et al., 2020 refers to relapse after CR, progression from PR or during therapy, or death resulting from any cause.

The side effect profile of siltuximab makes it a well‐tolerated treatment, and one of its main benefits is that it reduces or eliminates steroids from the treatment regimen and reduces steroid‐related side effects. The main concern when treating patients with siltuximab is the prevention of a febrile response to infection because of IL‐6 blockade. Thus, patients should be made aware of the risk of and symptoms of infection in the absence of fever. Other less common side effects include pruritus, hypertension, infusion reactions, hyperuricaemia, peripheral neuropathy and fluid retention, which are usually mild. ¹⁰ , ³⁸ Regular blood pressure monitoring is recommended. Management options for pruritus include montelukast and antihistamines.

Treatment of severe iMCD

Patients with severe disease should receive first‐line siltuximab in combination with high‐dose steroids (e.g. IV methylprednisolone 500 mg daily for 3 days). Patients with TAFRO syndrome may be critically unwell, and the use of cyclosporin A has been suggested in combination with steroids and anti‐IL‐6 antibodies. ¹⁶

The recommended regimen for siltuximab is 11 mg/kg every 3 weeks. However, the use of more frequent dosing of weekly siltuximab for 4 weeks has also been suggested for severe disease. ⁸ , ³⁹ This variation in dosing should ideally require either retrospective registry‐based studies and/or prospective randomised observational studies comparing dosing regimens. ³⁹

Patients who achieve at least a PR should remain on siltuximab maintenance therapy indefinitely until disease progression.

A clinical need which remains to be addressed is whether the dosing window could be extended beyond 3 weeks, for example to 6 weeks, in responding patients. ³⁹ Any change in dosing would ideally require the use of retrospective registry‐based studies and/or prospective randomised observational studies comparing dosing regimens. The only data available on 6‐weekly dosing are from a long‐term pre‐specified, open‐label extension analysis of 60 patients enrolled in the dose finding phase I and the phase II placebo‐controlled study and had stable or improved disease. ⁴⁰ Twenty‐five (42%) of these 60 patients who showed CR or PR for longer than 6 months received siltuximab at a dose of 11 mg/kg every 6 weeks, of which 8 (32%) were commenced on 6‐weekly dosing at the time of entry into the study, and 17 (68%) transitioned to a 6‐weekly regimen during the study. Only one (4%) patient who transitioned at cycle 5 returned to the original dosing after cycle 29 due to suspected progressive disease. ⁴⁰

However, it should be noted that no consensus has been reached with regard to dose extension in non‐severe iMCD in the international consensus guidelines, ³⁹ , ⁴¹ and the effect of 6‐weekly dosing on organ function remains unknown. As far as possible, 3‐weekly dosing should be continued.

Second‐line therapy consists of rituximab and chemotherapeutic regimens (e.g. cyclophosphamide, doxorubicin, vincristine and prednisone (CHOP) or etoposide). However, if disease control is achieved, maintenance therapy with siltuximab is suggested after chemoimmunotherapy.

Treatment of non‐severe iMCD

Although non‐malignant, iMCD is a progressive disease. As such, asymptomatic patients with normal organ function do not require treatment, but close monitoring to screen for signs or symptoms of disease progression is needed, initially 6‐weekly, then extending to 3‐monthly.

A thorough clinical history, physical examination and basic blood work (including a full blood count, electrolyte, renal, and liver function tests and C‐reactive protein) should be performed at each review. Prompt treatment initiation is recommended at the onset of symptoms or organ dysfunction attributable to iMCD (e.g. cytopenias, renal impairment).

Symptomatic patients with non‐severe disease should receive first‐line siltuximab (11 mg/kg every 3 weeks) in combination with lower doses of steroids (e.g. prednisolone 1 mg/kg). Steroids should be used for short periods only (4–6 weeks) while awaiting symptom response to siltuximab. Alternatively, patients with minimal systemic symptoms may start on single‐agent siltuximab therapy. Responding patients should remain on siltuximab maintenance therapy indefinitely as discussed in ‘Histology’ section, and a steroid wean should be undertaken to minimise long‐term steroid side effects. Alternative therapies supported by case series and anecdotal evidence include tocilizumab (IL‐6 receptor antagonist, reimbursed in Japan) at 8 mg/kg every 2 weeks. ⁴²

If anti‐IL‐6 agents are not available or patients are not responding to therapy, the use of second‐line agents, including rituximab and steroids, is suggested.

Third‐line therapy is discussed in the section ‘Refractory iMCD’.

Drug access and reimbursement in Australia

Agents used to treat iMCD are categorised into anti‐IL‐6 antibodies, immunomodulatory agents and off‐label treatment targeting plasma cells and B‐cell lymphoma.

Siltuximab is funded through the Pharmaceutical Benefits Scheme (PBS) for initial and continuing treatment without evidence of disease progression. Tocilizumab is an alternative if siltuximab is not available. However, tocilizumab is not PBS‐reimbursed for this indication.

Most immunomodulatory agents, including corticosteroids, cyclophosphamide and cyclosporin, are available through the general schedule. Off‐label agents include treatments targeting plasma cells (e.g. thalidomide, bortezomib) and B‐cell lymphomas (e.g. rituximab with or without chemotherapy).

Comparison of different agents

The main treatment regimens for iMCD, including anti‐IL‐6 therapies, corticosteroids, rituximab‐based therapy and chemotherapy, were compared in recent studies (Table 8). ⁵ , ⁷ , ⁴³ Anti‐IL‐6 therapies were the most effective. Yu et al. reported a twofold higher rate of CR with siltuximab than rituximab or rituximab‐based therapy (42.86 vs 20%, P = 0.034). ⁵ Similarly, Pierson et al. reported a twofold higher rate of response with anti‐IL‐6 (siltuximab and tocilizumab) with or without corticosteroids in comparison to rituximab with or without corticosteroids (50% vs 26.9%). In non‐TAFRO patients, siltuximab treatment was associated with better PFS than rituximab or rituximab‐based therapies (P = 0.048), chemotherapy or corticosteroids (P = 0.053). ⁷ Yu et al. reported a similar response rate with siltuximab therapy (55.56%). ⁴³ The rate of response with tocilizumab‐based therapy was 33.3%–44.4%. ⁷ , ⁴³ Overall, these data support the use of anti‐IL‐6 antibodies as first‐line therapy.

Beyond first‐line therapy, Yu et al. reported a CR rate of 20% with rituximab or rituximab‐based therapy, similar to the response rate of 26.9% in patients who received rituximab with or without corticosteroids reported by Pierson et al. In contrast, the response rate in patients who received immunomodulators (e.g. sirolimus, immunoglobulin replacement therapy, cyclosporin) with or without corticosteroids was lower, at 19%. These data form the basis for recommending rituximab‐based regimens as second‐line therapy. Yu et al. reported that rituximab or rituximab‐based therapy did not confer better rates of CR and PFS than chemotherapy or corticosteroids. In contrast, Pierson et al. found that chemotherapy‐based regimens achieved a twofold higher response rate than rituximab with or without corticosteroids (52% vs 26.9%). However, patients with TAFRO were the predominant group who received chemotherapy‐based regimens, resulting in a 47.8% response rate. Yu et al. observed that most patients who received CHOP (or a similar regimen) and R‐CHOP had intermediate‐ to high‐risk disease, with response rates of 41.82% and 63.64% respectively. Overall, these data support the use of chemotherapy as an alternative in severe, refractory iMCD or TAFRO. ⁷ Regarding corticosteroid monotherapy, Yu et al. noted a response rate of 47.06% in patients. However, primarily low‐risk patients were treated, and recurrence was noted in many patients within 1–2 years. Pierson et al. noted that patients treated with corticosteroid monotherapy had the lowest response rate (2.8%); hence, this regimen is not recommended.

Regarding adverse events, skin and subcutaneous tissue events were most common in siltuximab‐treated patients (20%), most commonly rashes (50%). Gastrointestinal events were the second most common (17.8%), in particular nausea (37.5%), diarrhoea (31.3%) and abdominal pain (12.5%). This is followed by nervous system events (11.1%), most commonly headaches (60%) and neuropathy (20%). Similarly, gastrointestinal as well as skin and subcutaneous events are common in tocilizumab‐treated patients (23.1 and 19.2% respectively), specifically rashes (40%), pruritus (20%) and diarrhoea (33.3%). In contrast to siltuximab, tocilizumab‐treated patients had more immune system events (15.4% vs 2.2%), specifically anaphylaxis (100% vs 0%). Rituximab‐treated patients mainly had musculoskeletal and connective tissue events (24.1%), less common in the siltuximab group (10%), and not reported in tocilizumab‐treated patients. Specifically, rigour was the most common (40%), followed by chills (30%). General and administration site events were the next most common (14.5%), including infusion reactions and fatigue (33.3% each). ⁷

Response criteria

In an effort to standardise response assessment in iMCD, the CDCN has published consensus criteria that incorporate three domains, ⁷ , ³² including biochemistry, lymph node and symptoms as shown in Tables 9 and 10. Of note, the assessment of lymph node is based on the Cheson criteria, commonly used in lymphoma, modified to include the assessment of cutaneous manifestations. ⁸ , ⁹

Table 9.

CDCN‐defined response criteria ⁸

Overall response	Biochemical	Lymph node†	Symptoms
CR	Normal CRP, haemoglobin, albumin, GFR	CR	Normalisation to baseline
PR	>50% improvement in all of CRP, haemoglobin, albumin, GFR	PR	Improvement in all four symptom categories, but not to baseline
SD	<50% improvement (or <25% worsening) in all of CRP, haemoglobin, albumin, GFR	No PR or CR	Improvement in at least one (but not all) symptoms
PD	>25% worsening in any of CRP, haemoglobin, albumin, GFR	>25% increase	Any symptoms worse on ≥2 assessments

Reproduced from van Rhee et al. ⁸

^†Assessment of lymph node based on the Cheson criteria, modified to include cutaneous manifestations. ⁸ , ⁹

CDCN, Castleman Disease Collaborative Network; CR, complete remission; CRP, C‐reactive protein; GFR, glomerular filtration rate; PD, progressive disease; PR, partial remission; SD, stable disease.

Table 10.

CDCN‐defined symptoms response criteria ⁸

Symptom	Criteria
Fatigue	Decrease of ≥1 CTC grade point relative to baseline
Anorexia	Decrease of ≥1 CTC grade point relative to baseline
Fever	Decrease of ≥1°C relative to baseline
Weight improvement	Increase of ≥5% relative to baseline

Reproduced from van Rhee et al. ⁸

CDCN, Castleman Disease Collaborative Network; CTC, common toxicity criteria.

Strict application of these criteria may be useful for uniform response reporting in the clinical trial setting; however, in clinical practice, we recommend using them to guide, rather than dictate, treatment selection and duration. Symptom response and organ improvement should be prioritised over nodal responses when making decisions about changing therapy.

In patients treated with siltuximab, resolution of fever and thrombocytopenia is the earliest marker of treatment response, often normalising within the first month. Improvements in haemoglobin, CRP and albumin generally occur within 3 months of therapy commencement, while lymphadenopathy can take 6–12 months to resolve. ¹⁰ , ⁴⁴

Clinical and biochemical response assessment is recommended weekly until evidence of improvement, and then with each 3‐weekly treatment cycle. Nodal response assessment using CT is recommended at 6 and 12 months following the commencement of treatment.

Refractory iMCD

Non‐responders to siltuximab are patients experiencing a sustained increase in grade 2 or more symptoms persisting for at least 3 weeks or onset of any new disease‐related symptoms grade 3 or higher or sustained deterioration in performance status, radiological progression, as measured by modified Cheson criteria or the need to start any other therapy. ⁴⁵

Up to half of patients receiving first‐line therapy will not achieve an adequate response. ⁸ This represents a difficult‐to‐treat population in whom there is limited evidence to guide the selection of second‐line therapy. Inadequate response to glucocorticoids with IL‐6 inhibition should prompt a review of the diagnosis. There is evidence in some cases of response to chemotherapy, such as rituximab‐CHOP (R‐CHOP) and other low‐toxicity lymphoma protocols. ⁸ Others have reported success with haemophagocytic lymphohistiocytosis‐directed regimens. ⁴⁶ More recently, a multicentre prospective trial of bortezomib, cyclophosphamide and dexamethasone in 24 relapsed/refractory (R/R) patients reported a promising 1‐year progression‐free survival and overall survival of 79% and 92% respectively. Of note, none of these patients had been exposed to IL‐6‐directed therapy, with the majority (n = 18) receiving thalidomide‐based treatment as the first line and five having been treated with R‐CHOP. ⁴⁷ The same group showed efficacy in a subset of patients who were treated up front with thalidomide, cyclophosphamide and prednisone (TCP). Durability at 24 weeks was observed in 48% of patients, making this another approach worthy of consideration in the R/R population. ⁴⁸

Sirolimus, an mTOR inhibitor, is becoming recognised as a viable option for treatment. Pathogenic PI3K/Akt/mTOR pathway signalling has been identified in patients resistant to IL‐6 inhibition and, as such, could provide an alternative target for treatment. ⁴⁹ Sirolimus is immunosuppressive and antiproliferative, reduces CD8+ T cells and leads to a decrease in vascular endothelial growth factor levels. ⁵⁰

Liu et al. published a retrospective study involving 26 R/R iMCD patients treated with sirolimus which showed 18/26 (69.2%) achieving symptomatic and biochemical responses. ⁵¹ Other case reports likewise have shown improved symptoms and biochemical responses in refractory patients treated with sirolimus. ⁵² , ⁵³ This has resulted in two current clinical trials of sirolimus in anti‐IL‐6‐refractory iMCD patients. In clinical trial NCT03933904, the dosing regimen for sirolimus in iMCD patients is initial dose, 7.5 mg/m², and maintenance dose, 2.5 mg/m², whereas the Koga et al. trial uses 2 mg/day. ⁵⁴ , ⁵⁵

Other combinations that have been shown to be effective are RCD (lenalidomide, cyclophosphamide and dexamethasone) and RD (lenalidomide and dexamethasone). ⁵⁶

Despite these studies, the optimal treatment is unknown and must be guided by local availability, the patient's clinical condition and the clinician's judgement. Though chemotherapy can be used, the use of immunomodulatory and immunosuppressive therapies is preferable because of the lower toxicity and similar efficacy. An exception is for patients with severe, rapidly progressive disease, where chemotherapy is recommended, as discussed in the section ‘Siltuximab’.

Recent findings on JAK–STAT involvement in iMCD‐TAFRO pathogenesis promoted JAK inhibitors as a possible new chemo‐free salvage option, with few cases described in the literature; in addition, inhibition of the tumour necrosis factor (TNF) pathway holds promise for some refractory patients. ⁵⁷ , ⁵⁸ , ⁵⁹

Conclusion

iMCD is a rare condition with a heterogeneous clinical presentation. Given the substantial challenges involved in diagnosing and managing the condition, we provide a multidisciplinary Australasian clinical practice guideline. Given the overlap with various conditions, a comprehensive evaluation, including cross‐sectional imaging and lymph node biopsy, cannot be overemphasised. Anti‐IL‐6 antibodies, in particular siltuximab, form the cornerstone of modern therapy of iMCD. They have a good safety profile; however, suboptimal responses to therapy remain common and pose a management challenge due to limited evidence to guide further therapy. Hence, further research will be helpful to address this unmet need and improve care for patients.

Data availability statement

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.