Predictive Model for Idiopathic Multicentric Castleman Disease Supporting Treatment Decisions

Short abstract

The International Castleman Disease Consortium recently presented the first idiopathic multicentric Castleman disease prognostic model correlating with clinical outcome. This commentary summarizes the value of this international prognostic index tool and the group's next research goals.

Castleman disease (CD) is a rare and heterogeneous lymphoproliferative disorder of unknown etiopathogenesis. Although it was first described by Benjamin Castleman as early as 1954 [1], most of the understanding of its etiology and progress in treatment options were obtained during the last 10 years. According to the different clinical manifestations, CD is classified as unicentric and multicentric CD [2]. Unicentric CD occurs at a single anatomic site, is usually asymptomatic and not life‐threatening, and can be successfully removed by surgical resection. In contrast, multicentric CD (MCD) involves multiple regions of enlarged lymph nodes, commonly has mild to severe systemic inflammatory symptoms, cytopenias, and multiorgan dysfunction, and usually needs systemic therapy with drugs. MCD can occur in immunocompromised patients with human herpes virus‐8 (HHV‐8) infection [3] or in individuals without HHV‐8 infection [4], which was named idiopathic multicentric Castleman disease (iMCD) by Fajgenbaum et al. in 2014, as its etiology and pathogenesis are not clear [5].

There are an estimated 1,650 cases of MCD cases diagnosed per year in the U.S. and 33%–58% of them are iMCD [6, 7]. The overall prognosis of iMCD is poor: approximately 23%–45% patients die within 5 years of diagnosis [6, 8], and approximately 60% die within 10 years [9]. Because of the rarity of iMCD, the etiopathogenesis and optimal treatment strategy for iMCD are still not fully illustrated. The tradition treatment regimens for iMCD usually involve glucocorticoids and chemotherapy, which can achieve complete response in only 27%–45% of patients [6]. Moreover, iMCD responding to initial treatment often experiences recurrence within 1 to 2 years. The newer therapy siltuximab, an interleukin (IL)‐6 antibody, also only produced a 34% overall response rate in a phase II clinical trial [10], although half of the non‐responders also had elevated IL‐6 levels which showed significant correlation with the severity of the iMCD symptoms [11]. Elevated mTOR signaling has been identified to be a treatment resistance mechanism, and the mTOR inhibitor sirolimus induced clinical benefit responses in three patients with anti–IL‐6–refractory iMCD in a pilot study. The efficacy of sirolimus in previously treated iMCD is being evaluated in a phase II clinical trial [12, 13, 14].

These data demonstrate the diverse clinical presentation and prognosis of iMCD and the urgent need of more effective therapies for refractory/relapsed patients. However, reliable risk stratification of iMCD with therapeutic guidance implication is lacking because of the rarity of this disease. The recently published paper in The Oncologist by Yu et al. from The International Castleman Disease Consortium presented the first iMCD prognostic model correlating with clinical outcome [15]. Based on prognostic analysis in a large international cohort of iMCD (176 patients) with long‐term follow‐up (median, 12 years), the first iMCD international prognostic index (iMCD‐IPI) was established based on presence or absence of five risk factors: age over 40 years, plasma cell variant, hepatomegaly and/or splenomegaly, hemoglobin <80 g/L, and pleural effusion. The iMCD‐IPI is able to classify patients into three risk groups with different prognoses (Fig. 1). All five clinical parameters are routinely evaluated in clinical practice, thereby making the risk stratification easily accessible. The predictive value of this iMCD‐IPI was confirmed in a large validation cohort of 197 patients with iMCD organized from The International Castleman Disease Consortium. Yu and colleagues further analyzed relationships between the risk stratification groups and the therapy choices. The authors found that even with watchful waiting, 98% of the low‐risk patients achieved long‐term survival. The intermediate‐risk patients were more likely to be treated (mostly with cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) or rituximab‐CHOP (R‐CHOP) chemotherapy) than undergo watchful waiting, but their survival rate was significantly lower than the low‐risk patients. Most (82%) of the high‐risk patients were treated with combination chemotherapy, but their survival rate remained to be the lowest, suggesting an unmet clinical need. However, comparing between watchful waiting and systemic chemotherapy R‐CHOP/CHOP for the same risk group, the clinical benefit is obvious for the intermediate‐risk patients (5‐year survival rate: 22% vs. 68%), indicating the importance of timely risk prediction for proper treatment decision.

Figure 1.

Schematic illustration of risk stratification of iMCD (idiopathic multicentric Castleman disease, HIV/HHV‐8‐negative) and proposed optimal treatment

Therefore, the novel iMCD‐IPI provides us a valuable tool for iMCD prognostic prediction and treatment decision making (Fig. 1), especially considering the rarity of the disease and that most physicians have never encountered iMCD in their career. However, the iMCD‐IPI has limited ability to identify high‐risk patients (only 7.4%–10.7%). Also, because of the small number of cases treated with anti–IL‐6, single‐agent rituximab, or immunomodulatory agent, the study was not able to identify the most optimal treatment for intermediate‐risk iMCD or perform a multivariate analysis for all the prognostic clinical and treatment factors. In addition, treatment resistance mechanisms and the therapeutic targets in intermediate‐ and high‐risk patients with iMCD still need to be discovered. Future studies on molecular biomarkers may further refine the iMCD risk stratification and help develop targeted therapy strategies. Several previous studies in 3 to 34 patients with iMCD have identified a few mutations possibly relevant for the iMCD pathogenesis; however, none of these mutations were reproducible across studies [16, 17, 18, 19]. Therefore, our next research goal is to perform genomic and gene‐expression profiling studies in a large cohort of patients with iMCD, including more cases of patients treated with anti–IL‐6–based or immunomodulatory agents to gain insight into iMCD pathogenesis and biomarker‐guided therapy supporting clinical trials. Once the association of specific oncogenic pathways and therapeutic targets in different iMCD‐IPI risk groups is precisely identified, physicians can easily make clinical decisions on the optimal therapy for patients with iMCD and ultimately improve the clinical outcome of iMCD.

Disclosures

The authors indicated no financial relationships.