Comprehensive analysis of subtype-specific outcomes and management in castleman disease: A 20-year cohort Study

Abstract Introduction: Castleman disease (CD) is a heterogeneous group of lymphoproliferative disorders classified anatomically as unicentric (UCD), oligocentric (OligoCD), or multicentric (MCD). Human herpesvirus 8 (HHV-8)-negative MCD is referred to as idiopathic MCD (iMCD), which includes clinical subtypes with distinct phenotypes and treatment responses: TAFRO (thrombocytopenia, anasarca, fever, renal dysfunction and/or reticulin fibrosis, and organomegaly), idiopathic plasmacytic lymphadenopathy (IPL), and not otherwise specified (NOS). While UCD and iMCD, including TAFRO, have been more clearly defined, OligoCD and iMCD-IPL remain under-characterized, particularly in Western cohorts. This study aimed to validate the updated clinical classification of CD using a large institutional cohort by retrospectively reclassifying patients based on lymph node distribution and clinicopathologic criteria. We further examined whether OligoCD and IPL represent clinically and prognostically distinct entities compared to UCD or iMCD, and other iMCD subtypes. Methods: This study included 217 patients evaluated for CD at our institution between January 2004 and August 2024. All cases were reclassified based on the extent of lymph node involvement and iMCD subtype. UCD was defined as the involvement of a single lymph node region. OligoCD was defined as involvement of ≥2 contiguous lymph node stations without sufficient systemic symptoms to meet iMCD criteria. Patients with POEMS syndrome or HIV/HHV-8–associated CD, or lacking pathologic confirmation, were excluded. We evaluated clinical presentation, laboratory data, histopathology, treatment regimens, and outcomes. The primary endpoint was overall survival (OS), and the secondary endpoint was event-free survival (EFS), defined as time from first-line treatment to progression, change in therapy, or death. Survival outcomes were assessed using the log-rank test. Results: Among 217 CD patients, 123 (57%) had UCD, 43 (20%) had OligoCD, and 51 (23%) had iMCD. iMCD-defining symptoms were more common in iMCD compared to UCD and OligoCD. Among all, only fatigue and organomegaly were more common in OligoCD than in UCD. Laboratory values related to CD disease activities, including hemoglobin, serum creatinine, albumin, CRP, and plasma IL-6, showed significant differences in iMCD compared to UCD and OligoCD. No significant difference in the laboratory findings was observed between UCD and OligoCD, although patients with OligoCD had intermediate values between UCD and iMCD. With median follow-up of 4.0, 7.6, and 4.6 years for UCD, OligoCD, and iMCD, respectively, the 5-year OS was 99.0%, 97.7%, and 89.5% without differences among groups. Corresponding 2-year EFS rates were 95.2% for UCD, 68.0% for OligoCD, and 53.8% for iMCD. With OligoCD being the reference, UCD had a significantly lower hazard ratio (HR) of 0.20 (95% CI 0.10–0.41), and iMCD had a significantly higher HR of 2.0 (95% CI 1.1–3.6). Among OligoCD patients initially treated with surgery (40%), none required systemic therapy thereafter. Those treated with rituximab or glucocorticoid monotherapy required subsequent therapy in 50% and 43%, respectively. Within iMCD subtypes (TAFRO (n=16), IPL (n=9), and NOS (n=26)), the 5-year OS was 80.4%, 100%, and 91.0%, respectively, without significant differences among the subtypes. However, 1-year EFS rates were 40.2%, 100%, and 65.6%, respectively, with an HR of 4.3 (95% CI 1.2–15.2; p = 0.02) in TAFRO and 1.6 (95% CI 0.77–3.5) in NOS compared to IPL. Patients with IPL had a high rate of symptomatic improvement assessed by a physician of 89% with a longer time-to-next treatment. Conclusions: This extensive, retrospective cohort study supports the clinical relevance of updated CD subtype classification. As the present study spans two decades, it reflects evolving diagnostic criteria and therapeutic standards, including the 2014 FDA approval of siltuximab, the iMCD consensus criteria in 2017, with the addition of OligoCD and iIPL. The findings confirm the aggressive clinical nature of iMCD and suggest that OligoCD demonstrated intermediate features between UCD and iMCD, though closer to UCD in outcome. Also, the results reconfirmed that TAFRO is associated with severe symptomatic diseases, and validated that patients with IPL, even in the Western cohort, likely have favorable responses to treatment. Continued prospective efforts are warranted to confirm the clinical distinctions.