Idiopathic multicentric Castleman disease (iMCD) is a rare and heterogeneous lymphoproliferative disorder characterized by systemic inflammation, multicentric lymphadenopathy, and multi-organ dysfunction.1 iMCD can be classified into three clinical subtypes: iMCD-idiopathic plasmacytic lymphadenopathy (IPL); iMCD-thrombocytopenia, anasarca, fever, renal dysfunction/reticulin fibrosis, and organomegaly (TAFRO); and iMCD-not otherwise specified (NOS).2-8 iMCD-TAFRO patients suffer from significant cytokine storm and require intensive therapy.9, 10 On the other hand, iMCD-IPL subtype (a subtype formerly incorporated in the NOS subtype) has emerged as a more benign clinical and biological entity.1, 6 It is characterized by elevated serum immunoglobulin G (IgG) levels, thrombocytosis, and plasmacytic or mixed histopathology.6, 11 Notably, despite higher systemic inflammation levels than iMCD-NOS, iMCD-IPL patients often have significantly longer survival, with an estimated 5-year overall survival rate of 97% compared to 85.5% for iMCD-NOS patients.6 Current treatment recommendations for iMCD are largely guided by a severity-based model proposed by the Castleman Disease Collaborative Network (CDCN) in 2018, regardless of clinical subtype-specific disease behavior. According to the CDCN guidelines, severe iMCD patients were recommended to receive intensive treatment, involving anti-interleukin-6 (IL-6) agents combined with high-dose steroids and/or cytotoxic chemotherapy.12 In contrast, non-severe iMCD cases were managed with less aggressive therapies, such as anti-IL-6 monotherapy. However, the recommendations were developed before iMCD-IPL was formally recognized as a distinct clinical subtype and therefore may not fully account for its relatively indolent disease behavior. Emerging evidence suggested that, apart from disease severity, clinical subtype could also substantially influence prognosis and therapeutic response in iMCD.6, 13, 14 Accordingly, in this study, we aimed to evaluate whether the current CDCN's disease-severity-based treatment framework was suitable for iMCD-IPL patients. This retrospective, single-center study was approved by the Medical Ethics Committee of Peking Union Medical College Hospital (PUMCH). Patients diagnosed with iMCD-NOS and iMCD-IPL from January 2000 to March 2025 at PUMCH were consecutively included. The diagnosis of iMCD-IPL was defined as described by Gao et al.6 including (1) fulfillment of eligibility for the diagnostic criteria of iMCD-NOS; (2) elevated serum IgG level (>17.4 g/L); (3) plasmacytic or mixed pathological subtypes; and (4) elevated platelet count (>350 × 109/L). According to the CDCN treatment consensus, patients were classified as having severe or non-severe disease.12 Severe iMCD should have at least two of the five following criteria: (1) Eastern Cooperative Oncology Group performance status (ECOG-PS) score ≥ 2 points, (2) Stage IV renal dysfunction, (3) anasarca, (4) hemoglobin (Hb) level of ≤80 g/L, and (5) pulmonary involvement/interstitial pneumonitis with dyspnea. Disease severity was assessed at the time of initial diagnosis, before the initiation of systemic therapy. Missing clinical and laboratory data were handled using a complete-case approach. A total of 338 patients were included in the study, comprising 152 with iMCD-IPL and 186 with iMCD-NOS. According to the CDCN-defined severity criteria, patients were stratified into four groups: severe iMCD-IPL (n = 37), non-severe iMCD-IPL (n = 115), severe iMCD-NOS (n = 52), and non-severe iMCD-NOS (n = 134). Figure 1 provides an overview of the major clinical, pathological, and laboratory characteristics of these four subgroups of patients. Focusing on severe subgroups (Table S1), patients with severe iMCD-IPL were younger at diagnosis than those with severe iMCD-NOS (median 40 vs. 45 years, P = 0.012) and predominantly exhibited plasmacytic pathology (n = 32, 86.5%), whereas hyaline vascular pathology subtype was exclusively present in severe iMCD-NOS (n = 13, 25.0%). Although the prevalence of systemic symptoms did not differ significantly between groups, patients with severe iMCD-IPL showed significantly higher platelet counts (median, 473 vs. 221; P < 0.001) and inflammatory markers, including C-reactive protein (CRP) (149.4 vs. 69.1; P < 0.001), erythrocyte sedimentation rate (ESR) (114.0 vs. 91.5; P = 0.001), IgG (42.8 vs. 28.9; P = 0.001), and IL-6 (63.5 vs. 22.7; P = 0.007). In contrast, renal function was better preserved in severe iMCD-IPL compared with severe iMCD-NOS (median eGFR 109.0 vs. 87.0; P = 0.002). The coexistence of Hb ≤ 80 g/L and pulmonary involvement was more frequently observed in severe iMCD-IPL than in severe iMCD-NOS (45.9% vs. 25.0%; P = 0.045), whereas the other two-criterion combinations showed no significant differences (Table S2). Overall survival (OS) was analyzed using the Kaplan–Meier method across the four subgroups (Figure 2A). During a median follow-up of 45.1 months (range, 0.4–295.2), the median OS was not reached in any group. Two of 152 patients (1.3%) with iMCD-IPL and 9 of 186 patients (4.8%) with iMCD-NOS deceased from disease progression during follow-up. Severe iMCD-NOS was associated with significantly worse OS, whereas severe iMCD-IPL patients showed no statistical difference to those of non-severe iMCD-IPL (log-rank P = 0.506) and non-severe iMCD-NOS (log-rank P = 0.402). By univariate Cox regression analysis, severe iMCD-NOS was associated with a significantly higher risk of death compared with severe iMCD-IPL (hazard ratio HR = 9.61, 95% CI: 1.25–73.94), non-severe iMCD-IPL (HR = 24.34, 95% CI: 3.16–187.66), and non-severe iMCD-NOS (HR = 3.99, 95% CI: 1.56–10.18). In contrast, the comparison between severe and non-severe iMCD-IPL did not reach statistical significance (HR = 2.49, 95% CI: 0.16–39.91). Given the overall favorable survival of iMCD-IPL patients, time to next treatment (TTNT) was utilized as a substitute endpoint. Three non-severe iMCD-IPL patients who remained untreated by the end of follow-up were excluded from the TTNT analysis. Among the remaining 149 patients with iMCD-IPL, during a median follow-up of 23.3 months (range, 0.7–289.5), 14 of 37 patients (37.8%) with severe iMCD-IPL and 53 of 112 patients (47.3%) with non-severe iMCD-IPL initiated next-line treatment. As shown in Figure 2B, no significant difference in TTNT was observed between severe and non-severe iMCD-IPL (log-rank P = 0.10). The prognostic value of the CDCN severity criteria for TTNT was subsequently evaluated. For patients with iMCD-IPL, neither individual criteria nor their cumulative number (≥1, ≥2, or ≥3) were significantly associated with treatment outcomes in iMCD-IPL patients (Table S3), consistent with the OS analysis. Among 37 patients with severe iMCD-IPL, treatment regimens were heterogeneous. Five patients (13.5%) received immunomodulatory therapy (siltuximab combined with high-dose steroids or with bortezomib–cyclophosphamide–dexamethasone BCD), eight patients (21.6%) received lymphoma-based therapy (rituximab ± cyclophosphamide–doxorubicin–vincristine–prednisone CHOP or rituximab–cyclophosphamide–vincristine–prednisone), and the remaining 24 patients (64.9%) were treated with myeloma-like therapy (thalidomide–cyclophosphamide–prednisone/dexamethasone, BCD, cyclophosphamide plus prednisone). In line with Chinese guideline recommendations8 and real-world clinical practice, myeloma-based regimens are more commonly used in non-severe iMCD, whereas lymphoma-like or immunomodulatory therapies (combined with BCD/high-dose steroids) are preferentially applied in patients with severe iMCD. Therefore, patients were grouped according to treatment intensity as myeloma-like therapy versus immunomodulatory/lymphoma-like therapy. Baseline characteristics were generally comparable between the two groups (Table S4). Patients were young, most showing plasmacytic histopathology and frequent systemic symptoms such as fatigue, fever, and weight loss. Laboratory tests revealed hypoalbuminemia, elevated inflammatory markers, and preserved renal function. No significant differences were detected. With a median follow-up of 30.1 months (range: 2.6–289.5 months), one patient deceased due to disease progression, which occurred in the myeloma-like therapy group. As shown in Figure 2C, no significant difference in TTNT was observed between the two treatment groups (log-rank P = 0.52). The median TTNT was not reached in the immunomodulatory/lymphoma-like therapy group, whereas patients receiving myeloma-like therapy had a median TTNT of 87.4 months. There were no significant differences in short-term efficacy between groups (3-month overall response rate ORR: 77.8% vs. 60.0%, P = 0.40; 6-month ORR: 64.7% vs. 55.6%, P = 0.69), as were the best overall response rates (78.3% vs. 70.0%, P = 0.67). Regarding safety, three patients in the immunomodulatory/lymphoma therapy group experienced adverse events. One patient receiving the CHOP regimen developed leukopenia (Grade 3), while another experienced vomiting (Grade 1). Additionally, one patient treated with siltuximab plus high-dose steroids developed pneumonia (Grade 2). In the myeloma-based therapy group, three patients experienced mild (Grade 1 or 2) skin-related adverse events, including rashes and pruritus; one patient exhibited swelling in the interphalangeal joints of the hands (Grade 1). The CDCN has played a pivotal role in establishing internationally recognized standards for the diagnosis and treatment of iMCD.1, 12 However, the guidelines were developed before iMCD-IPL was formally recognized as a distinct clinicopathologic subtype.12 Thus, the applicability of the current CDCN severity-based treatment framework to iMCD-IPL patients warrants investigation. In our cohort, patients with severe iMCD-IPL exhibited clinical and biological features distinct from those with severe iMCD-NOS. Despite higher inflammatory burden and more pronounced anemia, severe iMCD-IPL patients were younger and had better preserved renal function. Importantly, their clinical outcomes showed no significant differences to those of non-severe iMCD-IPL/iMCD-NOS, and were significantly more favorable than those observed in severe iMCD-NOS. Notably, nearly half of the severe iMCD-IPL patients were primarily classified as “severe” due to Hb ≤ 80 g/L combined with pulmonary involvement (Table S1), features that have not consistently been associated with poor prognosis in prior studies. For example, Gao et al. demonstrated that although iMCD-IPL patients exhibited more severe anemia than iMCD-NOS patients without IPL, their overall survival was longer.6 Similarly, a recent study indicated that pulmonary involvement did not predict poor outcomes in iMCD patients.15 Thus, iMCD-IPL patients who were classified as “severe” based on anemia and pulmonary involvement might not be really severe. Consistent with these observations, CDCN-defined severity criteria were not associated with OS or TTNT among iMCD-IPL patients. Moreover, treatment intensification in severe iMCD-IPL did not appear to confer additional clinical benefit. Patients receiving lymphoma-like or immunomodulatory regimens achieved response rates and TTNT comparable to those treated with myeloma-like therapy but experienced more frequent and severe adverse events. Taken together, our findings suggest that disease severity, as currently defined by CDCN criteria, may not adequately reflect prognosis or guide treatment in patients with iMCD-IPL. Our previous single-center retrospective study demonstrated that the oral Bruton's tyrosine kinase (BTK) inhibitor orelabrutinib achieved an overall response rate of 70% in relapsed/refractory iMCD, with the majority of responders (7 of 10) belonging to the iMCD-IPL subtype.16, 17 Similarly, Nishikori et al. reported that iMCD-IPL patients were more likely to respond favorably to anti-IL-6 receptor antibody monotherapy.13 These findings underscored the need to incorporate disease subtype, alongside severity, into future treatment recommendations. This study has several limitations. First, a few patients received IL-6–targeted therapy because siltuximab was not available in China until 2022, while tocilizumab was not approved for iMCD in China. Although treatment in our cohort followed the CCDN framework and was fundamentally consistent with CDCN guideline recommendations,8, 12 this limited and delayed access to IL-6–targeted therapy may have influenced treatment assignment and outcomes, representing a potential time-period and treatment-availability bias. Second, the wide confidence intervals for iMCD-IPL stemmed from the scarcity of fatal events due to its benign and indolent nature rather than sample insufficiency, especially considering our substantial sample size for a rare disease. Given the generally favorable prognosis of iMCD-IPL, overall survival could not be used as a good endpoint for iMCD-IPL patients. Thus, TTNT was used as a surrogate endpoint, but TTNT had inherent limitations. For instance, some patients in partial remission switched therapy when siltuximab became available or for personal reasons, artificially shortening TTNT. Third, standardized quality-of-life assessments were lacking. Given the chronic nature of iMCD-IPL and the substantial symptom burden reported in iMCD,18, 19 future prospective studies should prioritize the well-being of both patients and caregivers. In conclusion, the current CDCN severity-based stratification does not adequately reflect the prognostic characteristics of the iMCD-IPL subtype. “Severe” iMCD-IPL defined by CDCN guidelines might not be “severe.” Future treatment studies and clinical guidelines should consider disease subtype alongside severity when guiding management of iMCD. Siyuan Li: Investigation; data curation; formal analysis; visualization; writing—review and editing; writing—original draft. Yuhan Gao: Investigation; data curation; formal analysis; visualization; writing—original draft; writing—review and editing. Yue Dang: Data curation; visualization; writing—original draft. Haoyi Xu: Data curation; writing—original draft. Hongxiao Han: Data curation; writing—original draft. Lu Zhang: Conceptualization; methodology; formal analysis; supervision; writing—review and editing. Jian Li: Conceptualization; methodology; writing—review and editing; formal analysis; supervision. The authors declare no conflicts of interest. This work was supported by the Peking Union Medical College Hospital Talent Cultivation Program (Category C) (Grant Number UBJ10800 to L.Z.), CAMS Innovation Fund for Medical Sciences (ClFMS) (Grant Number 2023-I2M-C&T-B-045 to L.Z.), and National High Level Hospital Clinical Research Funding (Grant Number 2025-PUMCH-A-184 to L.Z.). The data that support the findings of this study are available from the corresponding author upon reasonable request. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.
Li et al. (Sun,) studied this question.