In this study, we retrospectively collected the information of 203 hospitalized patients with CD-like pathologically features in the past 15 years and re-analyzed these patients according to the 2017 CDCN criteria. 93.1% (n = 189) of these patients were confirmed to be CDs, of which 37.6% (n = 71) were HHV8-negative MCD and finally only 47 cases (29%) met the CDCN diagnostic criteria of iMCD.The ratio of iMCD was comparable to those reported previously. This suggested that the differential diagnosis of iMCD was still a challenge for pathologists and clinicians. In this cohort, eight patients with CD-like pathological features were subsequently diagnosed as FDCSs. Due to the variety of the pathological manifestations of CD, sometimes it is difficult to distinguish CD from thymoma, FDCS, and Hodgkin's lymphoma9–11. The diagnosis of iMCD requires the exclusion of infections, autoimmune diseases, primary or acquired immunodeficiency syndromes, and malignancies, of which the differentiation of PC-MCD and IgG4RD had always been a challenge in clinical practice12–15. Both of them have an appearance of generalized lymph node enlargement and extra-nodal involvement, and iMCD patients may have high serum IgG4 levels, whilst some IgG4RD patients may have CD-like pathological changes 16. According to the most recent diagnosis and exclusion criteria for IgG4RD 7, 8, six patients with IgG4RD were excluded, and two patients previously diagnosed with IgG4RD were re-diagnosed as iMCD. The 2021 IgG4RD guidelines7 state that even if the diagnostic criteria for IgG4RD are met, the presence of elevated CRP, IgA, or IgM in clinical indicators or pathology with any of the manifestations of lamellar-like mature plasma cell proliferation, high iron-containing heme deposition, and neutrophil infiltration cannot directly diagnose IgG4RD, the diagnosis of MCD takes precedence over IgG4RD. To summarize, diagnosing IgG4RD solely based on lymph node pathology without evidence of organ involvement should be interpreted cautiously. Clinical response to glucocorticoid therapy will favor the diagnosis of IgG4RD, and the IgG4/IgG ratio of serum and tissues will be also informative17, while the most recent IgG4RD diagnostic and exclusion criteria can help distinguish IgG4RD from MCD effectively.
Controversy remains regarding the transformation of CD into other tumors. Generally, iMCD patients are at increased risk of the second tumorigenesis. One PC variant UCD patient in this study transformed to T-lymphoblastic leukemia/lymphoma, with the onset of transformation occurring 2 years after the CD diagnosis. It is reported in the literature that iMCD patients have an increased risk of lymphoma, which is usually regarded as typical Hodgkin's lymphoma (usually mixed cell type), diffuse large B-cell lymphoma, mantle cell lymphoma, and peripheral T-cell lymphoma18. Three other patients with iMCD in this cohort transformed to marginal zone B-cell with large B-cell lymphoma transformation, and myelodysplastic syndrome, with transformation times of 2, 10, and 1.2 years, respectively. In the four cases of CD in which tumor transformation occurred in this study, the diagnosis of the second tumor was more than 1 year after the initial diagnosis of CD, and retrospective analysis of pathological biopsy specimens of the initial CD diagnosis revealed no evidence of the presence of subsequent tumors. The specific mechanism by which tumor transformation occurs in CD patients is unclear and may be related to the production of IL-6 and other important cytokines. Further clonality analysis of both tumor cells was needed to clarify the possible mechanisms of CD transformation.
The treatment of UCD in our patients was divided into two categories: those UCD patients without complications are usually followed up after surgery or biopsy. Patients complicated with PNP/BO still require vigorous treatment before and after surgical resection. But these patients still have a poorer prognosis than those without complications, mostly dying due to pulmonary infection and respiratory failure, consistent with previous studies19. In a former study conducted in our center, PNP was shown to be an independent risk factor for poor prognosis in UCD19,20. The proportion of PNP and BO in our group of CD patients was high than in other centers, which may be due to a fact that the dermatology department of our hospital is a PNP referral center in China. This bias in single-center studies should be rectified by multicenter data. The treatment of MCD varies significantly in different regions, siltuximab has become the first-line treatment in Europe and the United States3. A study that included7,8 patients with iMCD showed that siltuximab significantly improved patients' anemia, the inflammatory marker levels, and systemic symptoms compared to placebo21. In China, owing to a lack of anti-IL-6-targeted drug accessibility, front-line treatment for MCD patients is rarely based on drugs targeting IL-6 or its receptors. In our cohort, the lymphoma-like regimen was the main treatment option, with rituximab- or bortezomib-based regimen for refractory and relapsed patients. A nationally reported phase II clinical trial on 25 patients with primary iMCD treated with oral thalidomide, cyclophosphamide, and prednisone showed that 48% of patients achieved oncologic and clinical symptomatology remission for more than 24 weeks22. Another prospective study of 24 patients with relapsed refractory iMCD demonstrated the safety and efficacy of the bortezomib-cyclophosphamide-dexamethasone (BCD) regimen, with an estimated 1-year PFS and OS of 79% and 92%, respectively23. Previous four large studies of iMCD cases reported 5-year OS of 55%, 55%, 65%, and 77%, respectively24–27, whereas the 5-year OS of the iMCD population in this study was over 80%, possibly related to the higher rate of loss of follow-up rates, geographical differences, changes in diagnostic criteria, and small sample size. In short, Patients with MCD have a relatively poor prognosis. We did not find differences on survival between patients with MCD and UCD in this study, which may be attributed to a higher proportion of UCD patients combined with PNP.
There are some limitations in our study. First, as a retrospective study, the data were quite heterogeneous. Second, the available data were somewhat biased as a single-center study, for example, the percentage of UCD patients complicated with PNP was higher than reported in the literature. Future multicenter and larger sample data will be more informative and accurate in illustrating the real-world status of CD diagnosis and treatment.