Castleman Disease: Unicentric vs. Multicentric Types

Castleman disease is a rare disorder in which one or more lymph nodes grow abnormally large, driven by an overproduction of inflammatory signals, especially a protein called interleukin-6 (IL-6). It is not a traditional cancer, though it shares some features with lymphoma and can raise the risk of developing one. The condition comes in two strikingly different forms: a localized version that is usually cured with surgery, and a body-wide version that can be life-threatening without ongoing treatment. Because the name covers such different clinical realities, understanding which type a person has matters far more than the label itself.

Two Very Different Diseases Under One Name

The split between unicentric Castleman disease (UCD) and multicentric Castleman disease (MCD) is the single most important distinction. UCD involves a single enlarged lymph node or a cluster in one region. It tends to grow slowly, often causes no symptoms at all, and is discovered incidentally on imaging done for another reason. In a study of 52 patients, all 48 with UCD had benign symptoms and were cured by surgical removal, with excellent long-term outcomes.¹ A separate series of 12 UCD patients reported a mean follow-up of 92 months with no recurrence among those who had complete resection.²

MCD is a fundamentally different situation. It affects lymph nodes in multiple parts of the body at once and triggers widespread inflammation. Patients develop fevers, fatigue, weight loss, anemia, and fluid retention. In the same 52-patient study, three of the four MCD patients relapsed after surgery, and only one of those three survived after additional treatment.³ A larger Chinese cohort of 185 patients found five-year overall survival of about 94% for UCD but roughly 51% for MCD, underscoring the gap.⁴

MCD itself splits further. When the virus HHV-8 (human herpesvirus 8, also known as Kaposi sarcoma-associated herpesvirus) is present, the disease is classified as HHV-8-associated MCD. When the virus is absent and no other cause is found, it is called idiopathic MCD (iMCD). The two share symptoms but differ in biology and treatment, a distinction that matters for everything that follows.

What Drives the Inflammation

The central villain in most forms of Castleman disease is IL-6, an inflammatory signaling molecule normally produced in small amounts during infections or tissue injury. In Castleman disease, the germinal centers of affected lymph nodes churn out large quantities of IL-6, far beyond what the body needs. A landmark study in the late 1980s showed that this overproduction, without significant elevation of other cytokines, correlated directly with the swollen lymph nodes, elevated antibody levels, raised acute-phase proteins, and the constitutional symptoms patients experience.⁵ In that study, surgical removal of a single hyperplastic lymph node led to a drop in blood IL-6 levels and clinical improvement, confirming the node as the source.

In HHV-8-associated MCD, the virus adds a twist. HHV-8 encodes its own version of IL-6, called viral IL-6 (vIL-6). Unlike human IL-6, which needs a specific receptor to activate cells, vIL-6 can stimulate virtually any cell in the body by binding directly to a widely expressed receptor component. Research using transgenic mice showed that vIL-6 alone, at levels comparable to those in infected patients, produced spleen enlargement, widespread lymph node swelling, excess antibody production, and other hallmarks of MCD.⁶ Strikingly, when the mice were bred to lack their own native IL-6, those MCD-like features disappeared, suggesting that the virus needs the body’s own IL-6 as a partner to cause full-blown disease.

IL-6 does not act in isolation. The inflammatory cascade it triggers also pushes plasma cells in affected lymph nodes to produce abnormally high levels of vascular endothelial growth factor (VEGF), a protein that drives the growth of new blood vessels. Researchers have found elevated VEGF in both the blood and the lymph node tissue of Castleman disease patients, and VEGF expression was strong in interfollicular plasma cells while nearly absent in normal lymph nodes.⁷ This vascular overgrowth is what gives some Castleman disease lymph nodes their characteristic appearance under the microscope.

How Castleman Disease Looks Under the Microscope

Pathologists classify Castleman disease lymph nodes into two main patterns. The hyaline-vascular type, which is the more common one, features small follicles laced with glassy (hyaline) material and a prominent tangle of tiny blood vessels between them. The plasma-cell type has large follicles with dense sheets of plasma cells filling the spaces between follicular structures.⁸ A mixed pattern showing features of both also exists.

These histologic types do not map neatly onto UCD and MCD. Most UCD cases are hyaline-vascular, but plasma-cell UCD occurs. Most HHV-8-positive MCD is plasma-cell type. iMCD can be either. The histologic label tells the pathologist what the tissue looks like but does not, by itself, determine treatment. It does, however, affect the clinical picture: the plasma-cell variant is more closely tied to the systemic inflammatory symptoms driven by IL-6.

Diagnosing a Disease That Mimics Others

Getting to a Castleman disease diagnosis can be slow, partly because the symptoms of MCD overlap with many other conditions. Generalized weakness, anemia, swollen lymph nodes, elevated inflammatory markers, and high antibody levels describe a long list of diseases, from lymphoma to autoimmune disorders to chronic infections.⁹ A lymph node biopsy is essential, but even the biopsy findings can be ambiguous.

For iMCD specifically, an international consensus panel established formal diagnostic criteria requiring two major features, characteristic lymph node tissue changes and enlarged lymph nodes in more than one region, plus at least two of eleven minor criteria (including at least one abnormal lab finding). Critically, other conditions that mimic iMCD must be ruled out, including infections, malignancies, and autoimmune diseases.¹⁰

One condition that creates particular diagnostic confusion is IgG4-related disease, an inflammatory disorder that also causes organ enlargement and dense plasma cell infiltration. Case reports have documented patients initially diagnosed with MCD who later developed features of IgG4-related disease, or vice versa, raising the possibility that the two conditions share some underlying biology.¹¹ Clinicians evaluating possible Castleman disease need to keep this overlap in mind.

TAFRO Syndrome

Within the iMCD category, a particularly aggressive subtype has been recognized: TAFRO syndrome, an acronym for its defining features of thrombocytopenia (low platelets), anasarca (severe fluid retention throughout the body), fever, reticulin fibrosis (scarring in the bone marrow), and organomegaly (enlarged organs). First described in a cluster of Japanese patients, TAFRO syndrome has since been identified internationally. In the largest early case series of 25 patients, nearly all had fluid buildup, organ enlargement, and fever, and most had low platelet counts, but none showed the markedly elevated antibody levels that are common in other forms of iMCD.¹² These patients also frequently developed abdominal pain and acute kidney failure.

A later validation study confirmed these patterns, finding that kidney impairment was common enough that roughly a quarter to half of patients in various groups needed dialysis during their disease course.¹³ TAFRO syndrome tends to present more acutely than other forms of iMCD and can deteriorate quickly, making early recognition important.

Treating Unicentric Disease

For UCD, surgery is the treatment and usually the cure. Complete removal of the enlarged lymph node or node cluster resolves the disease in almost all cases. Techniques vary depending on location: video-assisted thoracoscopic surgery for chest masses, open approaches for deeper abdominal or retroperitoneal locations.¹⁴ When the mass wraps around vital structures and cannot be fully removed, partial resection or observation may be considered, but complete removal remains the goal because it offers the best chance of a permanent cure.

After complete resection, recurrence is rare. Long-term follow-up studies consistently show that patients who had their entire mass removed remain disease-free for years or decades.¹⁵ For the small number of UCD patients whose disease cannot be resected, treatments used for MCD, including IL-6-directed therapy, are sometimes tried, though there is less formal evidence guiding these decisions.

Treating Multicentric Disease

MCD cannot be cured with a single surgery because the disease is spread across multiple lymph node sites. Treatment depends heavily on whether HHV-8 is involved.

For HHV-8-positive MCD, which frequently occurs alongside HIV infection, rituximab, an antibody that destroys B cells (the immune cells harboring HHV-8), has transformed outcomes. A prospective cohort of 84 patients treated with rituximab-based approaches reported that 80 achieved remission, with five-year overall survival of about 92%.¹⁶ About a fifth of patients relapsed, but all were successfully retreated with rituximab. Beyond controlling Castleman disease itself, rituximab dramatically lowered the risk of developing lymphoma, a feared complication. In a separate analysis of 113 patients, those treated with rituximab had roughly an eleven-fold lower incidence of non-Hodgkin lymphoma compared with those who had not received it.¹⁷ One notable concern: Kaposi sarcoma can flare after rituximab, so patients need monitoring.

Rituximab has also been used in HHV-8-positive patients who do not have HIV. A case report described an elderly patient achieving sustained remission after a single course of rituximab,¹⁸ and in patients with aggressive presentations, combining rituximab with chemotherapy agents has shown rapid resolution of inflammation and viral clearance.¹⁹

For iMCD, the first-line targeted option is siltuximab, an antibody that directly neutralizes IL-6. In a randomized, placebo-controlled trial, about a third of patients receiving siltuximab achieved durable tumor and symptom responses, compared with none in the placebo group.²⁰ Tocilizumab, which blocks the IL-6 receptor rather than IL-6 itself, has also been used, with clinical reports documenting improvement in anemia and organ function in refractory cases.²¹ Anti-IL-6 therapy improves most symptoms, but it does not work for everyone.

When IL-6 Blockade Is Not Enough

A frustrating reality of iMCD is that some patients do not respond to blocking IL-6. Researchers investigating why found that in at least some refractory cases, a separate signaling pathway called mTOR is overactivated. Analysis of iMCD lymph node tissue from 26 patients showed increased markers of mTOR pathway activity compared to normal lymph nodes.²² Three patients who had failed IL-6 blockade were treated with sirolimus, a drug that inhibits mTOR, and all three entered durable remission lasting over a year.²³

A larger retrospective study of 26 patients with relapsed or refractory iMCD treated with sirolimus-containing regimens confirmed the signal: about 69% achieved symptomatic and biochemical responses, with a median time to response of under two months. No patients died during follow-up, and the drug was well tolerated.²⁴ The mTOR pathway appears to be especially relevant in patients whose disease is not primarily driven by IL-6, pointing toward biological diversity within iMCD that treatment strategies will need to account for.

The Burden of Living with iMCD

Even when treatment controls the disease, the daily experience of living with iMCD can be profoundly difficult. An international survey found that among patients with symptoms, roughly half or more reported moderate to very severe effects on pain, ability to travel, sexual function, emotional wellbeing, finances, daily routine, and social life.²⁵

Fatigue stands out as the most dominant symptom. A Chinese multi-center survey found patients reported a median of nine symptoms, with fatigue scoring highest. Physical and mental health scores were substantially lower than in the general population. About 65% of patients showed signs of depression, with nearly a third meeting criteria for mild to severe major depression. Fewer than half were employed, and about a third reported significant lost work time.²⁶ These numbers reflect a disease that, even when not immediately life-threatening, erodes quality of life in ways that blood tests and imaging do not capture.

On a more encouraging note, clinical trial data suggest that effective treatment does improve how patients feel. In the siltuximab trial, patients on the drug reported better fatigue scores and were more likely to reach quality-of-life levels comparable to the general population than those on placebo.²⁷

Castleman Disease in Children

Although Castleman disease is primarily a condition of adults, it does occur in children. Pediatric cases lean heavily toward UCD: in a two-site retrospective study of 24 pediatric patients, 75% had UCD, and all MCD cases were HHV-8-negative. Systemic symptoms like fevers and weight loss were common even in UCD, appearing in about 44% of those patients. All UCD patients except two underwent complete resection as definitive therapy, and no UCD patients relapsed. Among the six with MCD, three experienced disease progression before achieving lasting remission, but there were no deaths.²⁸

A Chinese pediatric cohort confirmed that hyaline-vascular UCD is overwhelmingly the most common type in children. The neck and abdomen were the most frequent sites. After a median follow-up of four years, overall survival was 100%, and no patients relapsed.²⁹ One notable difference from adults is that systemic inflammation, such as anemia and elevated inflammatory markers, appeared more frequently in pediatric UCD than might be expected, which can complicate the initial diagnostic picture. Still, the overall outlook for children with Castleman disease, especially UCD, is very good.

Associated Conditions and Malignancy Risk

Castleman disease does not exist in a vacuum. MCD, particularly the HHV-8-positive form, carries a meaningful risk of developing non-Hodgkin lymphoma, a complication thought to arise from chronic B-cell stimulation by the virus. Before rituximab became standard, the lymphoma incidence in HHV-8-positive MCD was roughly 70 per 1,000 person-years. With rituximab treatment, that rate dropped to about 4 per 1,000 person-years.³⁰

Another condition linked to Castleman disease is POEMS syndrome, a rare disorder involving polyneuropathy, organomegaly, endocrine abnormalities, a monoclonal protein, and skin changes. Some POEMS patients have Castleman-type lymph node findings, and this overlap is recognized as the Castleman disease variant of POEMS syndrome.³¹ The presence of POEMS changes both the treatment approach and the monitoring strategy, and anyone diagnosed with Castleman disease who develops nerve symptoms or other POEMS features should be evaluated for overlap.

Emerging Biomarkers for Tracking Disease Activity

One of the practical challenges in managing iMCD is knowing whether a treatment is actually working before clinical symptoms clearly improve or worsen. Traditional inflammatory markers like C-reactive protein (CRP) are elevated in active disease but are not specific to Castleman disease. Researchers have been searching for better biomarkers, and one of the most promising candidates is CXCL13, a chemokine involved in organizing immune cell activity in lymph nodes.

In patients treated with siltuximab, a 17% drop in CXCL13 blood levels within the first few weeks of treatment predicted who would respond to therapy with high accuracy. In the initial cohort, this cutoff correctly identified responders with about 82% sensitivity and 77% specificity, and a validation cohort showed perfect separation between responders and non-responders.³² If validated further, this kind of early readout could spare non-responders months of ineffective treatment and push them toward alternatives like sirolimus sooner.

Beyond CXCL13, researchers have identified a broader panel of elevated inflammatory molecules in iMCD and TAFRO syndrome, including VEGF, certain interleukins, and serum amyloid A.³³ The hope is that measuring multiple biomarkers simultaneously will eventually allow clinicians to match each patient’s disease biology to the right treatment pathway from the start, rather than relying on trial and error.