Five biopsies, one diagnosis: challenges in idiopathic multicentric Castleman disease
- 1 Department of Medicine, University of Toronto, Toronto, Ontario, Canada
- 2 Department of Medicine, Division of Nephrology, University of Toronto, Toronto, Ontario, Canada
- 3 Department of Laboratory Medicine & Pathology, University of Toronto, Toronto, Ontario, Canada
- 4 Department of Medicine, Medical Oncology & Hematology, University of Toronto, Toronto, Ontario, Canada
- Correspondence to Dr Julie Semenchuk; julie.semenchuk@mail.utoronto.ca
Abstract
A previously healthy 29-year-old man initially presented to the hospital with pleuritic chest pain and shortness of breath. Over the next 2 months he developed ongoing fevers and night sweats with recurrent exudative pleural effusions and ascites. He had an extensive infectious and autoimmune workup that was unremarkable. He had an initial lymph node biopsy that showed reactive changes only. He had an acute kidney injury and his renal biopsy revealed thrombotic microangiopathy. His liver biopsy showed non-specific inflammatory changes. His bone marrow biopsy showed megakaryocyte hyperplasia and fibrosis, which raised suspicion for the thrombocytopenia, ascites, reticulin fibrosis, renal dysfunction and organomegaly syndrome subtype of multicentric Castleman disease. This prompted a repeat lymph node biopsy, showing changes consistent with mixed type Castleman disease that fit with his clinical picture. He was initiated on steroids and siltuximab with significant clinical improvement.
Background
Castleman disease describes a group of heterogeneous lymphoproliferative disorders that share common lymph node histopathology.1 Advances in diagnosis and classification have been substantial since the disease was first described by Benjamin Castleman in 1954.2 Castleman disease is separated clinically into unicentric and multicentric subgroups. Unicentric Castleman disease is characterised by an asymptomatic mass in patients with one region of lymphadenopathy.2 Multicentric Castleman disease (MCD) is characterised by constitutional symptoms, multicentric lymphadenopathy, hepatosplenomegaly and lab abnormalities such as anaemia, hypergammaglobulinaemia and bone marrow plasmacytosis.3
MCD is separated based on the aetiological driver thought to cause polyclonal proliferation in lymph nodes. This includes human herpes virus 8 (HHV8)-associated MCD, seen frequently in patients with HIV where HHV8 escapes host immune cells causing cytokine release.4 Another subgroup is polyneuropathy, organomegaly, endocrinopathy, M proteins and skin changes (POEMS)-associated MCD where the aetiological driver is a monoclonal gammopathy, often in genetically predisposed individuals.4 Finally, there is idiopathic MCD (iMCD), accounting for approximately 50% of MCD, where the aetiological driver is unknown.4
The limited understanding of aetiology, cell types and signalling pathways involved in iMCD makes this disease difficult to recognise. The median age at diagnosis of iMCD is 50 years, although the age range is large and affects individuals anywhere from 2 to 80 years old.1 Historically, these patients have a threefold increased prevalence of malignancy and 35% die within 5 years of diagnosis.1 iMCD can be further subclassified into iMCD thrombocytopenia, ascites, reticulin fibrosis, renal dysfunction and organomegaly (TAFRO) or iMCD not otherwise specified, which often causes thrombocytosis and hypergammaglobulinaemia.4
Here we describe a patient who was ultimately diagnosed with iMCD after presenting to the hospital with pleuritic chest pain and eventually returning with recurrent serositis, fevers and night sweats. This case highlights the clinical heterogeneity that can be seen in iMCD; it demonstrates the challenges in obtaining an accurate diagnosis and emphasises lymphoproliferative diseases as an important cause of renal thrombotic microangiopathy (TMA) and exudative effusions.
Case presentation
A previously healthy 29-year-old man of Canadian descent initially presented to the hospital with a 1-week history of severe pleuritic chest pain, dyspnoea and fever. Prior to the onset of his symptoms, he experienced an influenza-like illness with rhinorrhoea and cough. Initial blood work demonstrated leucocytosis (15.4×109/L) and elevated C reactive protein (196.19 mg/L). He was admitted to the hospital with suspected pericarditis based on diffuse ST changes and PR depression on ECG. He was treated for community-acquired pneumonia with amoxicillin/clavulinic acid, and pericarditis with aspirin and colchicine.
Over the next 2 months, he developed progressive dyspnoea on exertion, orthopnoea, persistent fevers, night sweats, diarrhoea and lethargy. He also reported increasing abdominal distension and a weight gain of 10 kg. He denied any chest pain, rashes, oral ulcers, arthralgias and visual changes, and had no recent travel and no sick contacts. He was seen in the hospital 2 weeks prior where he was found to have bilateral pleural effusions and ascites. He had an ultrasound-guided biopsy of a right axillary lymph node. The flow cytometry and pathology evaluation of this right axillary lymph node showed reactive changes only and no evidence of lymphoma.
He presented to our centre because of recurrent serositis, ongoing fevers and night sweats despite multiple thoracenteses and paracenteses. At the time of admission, he was febrile (38.5°C) and hypertensive (blood pressure 157/83 mm Hg) with abdominal distension and bilateral pitting oedema.
Investigations
His initial blood work was significant for normocytic anaemia (haemoglobin=113 g/L, mean corpuscular volume=83.8 fL) and thrombocytopenia (platelets=132×109/L). He had an acute kidney injury with a creatinine of 134 μmol/L with hyperkalaemia (K=5.8 mmol/L), hypocalcaemia (2.01 mmol/L) and hyperphosphataemia (1.84 mmol/L). CT demonstrated hepatosplenomegaly, large volume ascites, pleural collections, trace pericardial effusion and non-specific lymphadenopathy in the axillary, paratracheal, mammary and hilar regions. An echocardiogram showed a small 1.8 cm pericardial effusion and normal systolic function.
His pleural effusions were exudative with a pleural lactate dehydrogenase (LDH) of 143 U/L, serum LDH 141 U/L, pleural protein 34 g/L and serum protein 54 g/L. He had blood cultures, urine culture, stool cultures, Clostridium difficile, pleural cultures including acid fast and fungal cultures, and ascitic fluid cultures, all with no growth. He had no monoclonal spike on a serum protein electrophoresis. Of note, on his first admission, C4 was low at 0.08 g/L (0.10–0.40) and C3 was within normal range. C3 and C4 were both within normal limits following this. A list of other investigations to rule out infectious and autoimmune diseases can be found in table 1.
A list of relevant investigations sent off to rule out infectious, autoimmune and malignant processes
| Anti-nuclear antibody | Negative | |
| Anti-double-stranded DNA antibody | Negative | <1 IU/mL |
| Anti-Ro antibody | Negative | <0.2 AI |
| Anti-La antibody | Negative | <0.2 AI |
| Anti-Smith antibody | Negative | <0.2 AI |
| Anti-centromere antibody | Negative | <0.2 AI |
| Anti-myeloperoxidase | Negative | <0.2 AI |
| Anti-proteinase 3 | Negative | <0.2 AI |
| Anti-phospholipid antibody | Negative | <2 GPL U/mL |
| Rheumatoid factor | Negative | <7 IU/mL |
| Anti-cyclic citrullinated peptide | Negative | <0.5 U/mL |
| Anti-cardiolipin antibody | Negative | |
| Beta-2 glycoprotein | Negative | |
| IgG | 6.7 g/L | 7.0–16.0 g/L |
| IgM | 1.53 g/L | 0.7–4.0 g/L |
| IgA | 0.90 g/L | 0.4–2.3 g/L |
| Serum protein electrophoresis | Diffuse gamma region | |
| Ferritin | 559 μg/L | 30–250 μg/L |
| Strongyloides serology | Negative | <0.2 |
| HIV serology | Negative | |
| HIV p24 antigen | Negative | |
| Lyme IgM/IgG | Negative | |
| Parvovirus B19 IgG | Negative | |
| Parvovirus B19 IgM | Negative | |
| Q fever antibody | Negative | <1:16 |
| Blastomycosis antibody | Negative | |
| Histoplasmosis antibody | Negative | |
| Bartonella henselae IgG | Negative | <1:64 |
| Antistreptolysin O titre | Negative | <200 IU/L |
| Hepatitis B surface antigen | Negative | |
| Hepatitis B surface antibody | Immunity | >1000.00 IU/L |
| Hepatitis B core antibody | Negative | |
| Hepatitis C antibody | Negative | |
| Cytomegalovirus IgG | Positive | |
| Cytomegalovirus IgM | Low reactivity | |
| Cytomegalovirus blood PCR | Negative | |
| Epstein-Barr virus viral load | 2.84×103 IU/mL | |
| Rickettsia serology | Negative | <1:64 |
Differential diagnosis
The differential diagnosis on admission to our centre was based around the patient’s recurrent exudative effusions. It involved ruling out malignancies such as lung and breast cancer, infections including fungi, parasites, viruses, tuberculosis, autoimmune inflammatory diseases, including lupus and rheumatoid arthritis, asbestosis, pulmonary embolism and intra-abdominal processes such as pancreatitis.
When his renal biopsy showed evidence of renal TMA in the absence of a systemic microangiopathy, suspicion for a secondary TMA due to a lymphoproliferative disorder was high.
Castleman disease is a diagnosis of exclusion that involves ruling out multiple diseases outlined in table 2.
Diagnostic criteria for idiopathic multicentric Castleman disease, adapted from Fajgenbaum et al 5
Outcome and follow-up
Once iMCD was diagnosed, the patient was started on prednisone and siltuximab 11 mg/kg every 3 weeks. His acute kidney injury had resolved after 4 weeks of treatment. His prednisone was slowly tapered over a 5-month period.
At 6 months follow-up, he felt well and had returned to work. He had more energy and his shortness of breath, abdominal distension and lower extremity oedema had completely resolved. He remains on siltuximab infusions.
Discussion
Castleman disease can present with a wide spectrum of pathological findings and manifestations which can lead to delays or inaccurate diagnoses. The challenges in diagnosis arise from the large differential, clinical heterogeneity and our limited understanding around the pathophysiology.
In MCD, the systemic manifestations and multiple areas of lymphadenopathy can look like multiple autoimmune diseases, acute infections and malignancies. HHV8 and M-protein can help in diagnosis if present but it is ultimately a diagnosis of exclusion.5 6 In iMCD, there are no known risk factors or biological markers.5 After exclusion of other etiologies, obtaining lymph node histology is key to diagnosis. A bone marrow biopsy can help support the diagnosis of iMCD-TAFRO and should be done if this is a concern. In our case, the focal mild fibrosis and megakaryocyte hyperplasia were key to raising the suspicion for iMCD-TAFRO. There are no specific findings on liver biopsy to aid with diagnosis. Our patient had a liver biopsy where the sinusoidal dilatation was likely due to congestion and was not helpful for ruling in the disease.
Renal complications may occur in up to 50% of patients with Castleman disease.7 The most common finding on renal biopsy is TMA followed by amyloid A amyloidosis.7 8 There have also been reports of membranoproliferative glomerulonephritis, IgA nephropathy, focal segmental glomerulosclerosis and membranous glomerulopathy, although these are less common.9–12 Unfortunately, many case reports describing renal complications were done before the classification system of Castleman disease was fully developed. Understanding the complications of renal disease within each subgroup is important and could one day help aid in diagnosis. In one study, a renal disorder was diagnosed before Castleman disease and led to the diagnosis in nearly 50% of the studied cohort.7 One study has shown that the link between renal TMA and Castleman disease is related to increased serum VEGF levels and decreased VEGF expression at the level of the glomeruli.12 Since the definition of TAFRO syndrome in 2015, a few case reports have found renal TMA to be a common aetiology of renal insult in this subgroup.12–14 Our case helps support the idea that when a diagnosis is unclear, evidence of renal TMA should heighten the suspicion for MCD.
Little is still known about the subgroup iMCD-TAFRO, which was only recently described in 2015 based on 28 patients, all originating from Japan.15 In 2016, Iwaki et al performed a comprehensive clinicopathological analysis of 24 Asian and 1 Caucasian patient with iMCD-TAFRO and proposed modified diagnostic criteria, accepted by the American Society of Hematology.4 6 Our patient meets criteria for iMCD-TAFRO shown in box 1.
Diagnostic criteria for idiopathic multicentric Castleman disease with thrombocytosis, ascites, fever, reticulin fibrosis, renal dysfunction and organomegaly, adapted from Dispenzieri and Fajgenbaum.4
Histopathological criteria (need all)
-
Typical lymph node pathology (atrophic germinal centres with enlarged nuclei of endothelial cells, proliferation of endothelial venules and small numbers of mature plasma cells).
-
Negative latency associated nuclear antigen-1 for human herpes virus 8.
Major criteria (need three out of five)
-
Thrombocytopenia (<100 000/μL).
-
Anasarca (pleural effusions and ascites on CT).
-
Fever (>38°C).
-
Reticulin fibrosis.
-
Organomegaly.
Minor criteria (need at least one)
-
Hyperplasia/normoplasia of megakaryocytes.
-
High alkaline phosphatase without markedly elevated transaminases.
-
Republished with permission of the American Society of Hematology from Overview of Castleman disease, Dispenzieri and Fajgenbaum, 2020,4 permission conveyed through Copyright Clearance Center.
To our knowledge, there have been few reports of iMCD-TAFRO that are not of Asian descent.16 17 In Western countries, most cases of MCD are associated with HHV8.17 Our review of the literature found one case report of a 22-year-old Caucasian man in the USA with a very similar presentation, mainly severe pleuritic chest pain and recurrent serositis.16 However, this patient eventually required intubation and dialysis in contrast to our patient who remained stable off treatment for an extended period.16 This contrast emphasises the clinical heterogeneity that needs to be further explored.
The pathophysiology of iMCD is poorly understood, contributing to delays in diagnosis. The systemic involvement in all MCD is thought to arise from a cytokine storm primarily driven by interleukin-6. In iMCD, the cytokine storm is hypothesised to be caused by either an uncontrolled infection (pathogen hypothesis), autoantibodies or autoreactive T cells with germline mutations (autoimmune hypothesis) and/or somatic mutations in monoclonal lymph nodes, leading to ectopic cytokine secretion.1 iMCD in our patient may have been precipitated by an acute infection. His symptoms were preceded by an influenza-like illness, and earlier in his work-up, he had a transiently low C4 . However, our only definitive infectious finding was low level Epstein-Barr virus (EBV) viremia. There have been several case reports that have found an association between EBV and iMCD.16 18 One study, attempting to use virome capture sequencing on patients with Castleman disease found there was no clear association with between any viral agents in iMCD; however, there were a very limited number of patients (n=11) and the relationship between viral aetiologies and iMCD should be further elucidated.19
Learning points
-
Multicentric Castleman disease (MCD) is a rare condition that can present with recurrent serositis, fevers and night sweats in a previously healthy patient.
-
Renal thrombotic microangiopathy is the most common cause of kidney injury in this disease and is a useful biopsy finding to help support the diagnosis.
-
Lymph node histology is required for diagnosis and physicians should not hesitate to repeat this procedure if suspicion is high.
-
Very few cases of the subgroup thrombocytopenia, ascites, reticulin fibrosis, renal dysfunction and organomegaly syndrome have been described in the North American population and the clinical heterogeneity needs to be further explored.
-
While it is understood that a cytokine storm leads to Castleman disease, the aetiology of idiopathic MCD remains elusive and it is unclear whether infectious agents have a role.
Footnotes
-
Twitter @doc_merc
-
Contributors JS was responsible for writing the complete manuscript draft and revising it. AM was responsible for critically revising the work, providing guidance regarding which content to include as well as approving the final draft. AS was responsible for critically revising the work, providing pathology slides and approving the final draft. VK was responsible for critically revising the work, providing guidance regarding which content to include as well as approving the final draft.
-
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
-
Competing interests None declared.
-
Patient consent for publication Obtained.
-
Provenance and peer review Not commissioned; externally peer-reviewed.
- © BMJ Publishing Group Limited 2020. No commercial re-use. See rights and permissions. Published by BMJ.
References
Use of this content is subject to our disclaimer