Abstract
Castleman disease (CD) has been reported as a group of poorly understood lymphoproliferative disorders, including unicentric CD (UCD) and idiopathic multicentric CD (iMCD) which are human immunodeficiency virus (HIV) negative and human herpes virus 8 (HHV-8) negative. The clinical and independent prognostic factors of CD remain poorly elucidated. We retrospectively collected the clinical information of 428 patients with HIV and HHV-8 negative CD from 12 large medical centers with 15-year follow-up. We analyzed the clinicopathologic features of 428 patients (248 with UCD and 180 with iMCD) with a median age of 41 years. The histology subtypes were hyaline-vascular (HV) histopathology for 215 patients (56.58%) and plasmacytic (PC) histopathology for 165 patients (43.42%). Most patients with UCD underwent surgical excision, whereas the treatment strategies of patients with iMCD were heterogeneous. The outcome for patients with UCD was better than that for patients with iMCD, 5-year overall survival (OS) rates were 95% and 74%, respectively. In further analysis, a multivariate analysis using a Cox regression model revealed that PC subtype, hepatomegaly and/or splenomegaly, hemoglobin ≤ 80 g/L, and albumin ≤ 30 g/L were independent prognostic factors of CD for OS. The model of iMCD revealed that age > 60 years, hepatomegaly and/or splenomegaly, and hemoglobin ≤ 80 g/L were independent risk factors. In UCD, single-factor analysis identified two significant risk factors: hemoglobin ≤ 100 g/L and albumin ≤ 30 g/L. Our study emphasizes the distinction of clinical characteristics between UCD and iMCD. The importance of poor risk factors of different clinical classifications may direct more precise and appropriate treatment strategies.
Keywords: Castleman disease, clinical characteristics, treatment, risk prognostic factors
Introduction
Castleman disease (CD) is a heterogeneous group of rare lymphoproliferative disorders that was first described by Dr. Benjamin Castleman in 1956 [1]. According to clinical features and the distribution of enlarged lymph nodes, CD can be subclassified into unicentric CD (UCD) and multicentric CD (MCD). MCD is more aggressive and is frequently found in patients with human herpes virus 8 (HHV-8) and human immunodeficiency virus (HIV) infection, while MCD without HHV-8 and HIV infection is classified as idiopathic MCD (iMCD) [2]. Pathologically, CD can be classified into the hyaline-vascular (HV) variant and the plasmacytic (PC) variant [3]. The HV type is more common in patients with UCD, and the PC type is more common in patients with MCD.
As the etiology and pathogenesis of CD are unclear, the clinical characteristics and treatment options remain vague. The symptoms of UCD are often lighter, and most are related to enlarged lymph nodes and subsequent compression of adjacent tissues. Complete resection of the involved lesion is considered as the golden standard treatment [4]. The clinical presentation of HHV-8-associated MCD and iMCD varies from mild symptoms to life-threatening cytokine storms, which can both present with recurrent episodes of diffuse lymphadenopathy with systemic inflammatory symptoms, polyserositis, anemia, hypoalbuminemia, and multiple organ system dysfunctions. It can be fatal if improperly treated, and it easily develops into hematologic malignancies [5]. Because of the increasing attention on CD, especially iMCD, which has poor prognosis, an international expert group proposed an international consensus on the diagnosis and treatment of iMCD [6,7].
Because of its rarity, current studies are mostly retrospective or case reports from single institutions. Here, we analyzed the clinical, laboratory, pathologic, and treatment data and the prognosis of 428 patients with CD from 12 major academic medical centers with 15-year follow-up in China and the US to better understand the disease. Through our retrospective study, we hope to improve our understanding of CD and optimize the diagnosis and prognosis. To the best of our knowledge, this study comprises the largest sample size currently analyzed.
Materials and methods
Patients
To establish the study cohort, we collected data of 428 patients with CD from 10 large medical centers in China (Sun Yat-Sen University Cancer Center, Sun Yat-sen Memorial Hospital, The First Affiliated Hospital of Sun Yat-Sen University, Guangdong Provincial People’s Hospital, The First Affiliated Hospital of Zhejiang University, The Second Affiliated Hospital of Zhejiang University, Tianjin Medical University Cancer Hospital, Shandong Provincial Cancer Hospital, Jiangsu Nanjing University Hospital, and Shanxi Provincial Cancer Hospital) and two in the US (University of Texas M. D. Anderson Cancer Center and Indiana School of Medicine) between January 1994 and December 2017, which included 355 Asian and 73 American patients. All patients presented with an enlarged lymph node area, and CD diagnosis was confirmed based on international, evidence-based consensus diagnostic criteria for UCD and iMCD [7-9]. We excluded patients with lymph node hyperplastic disease, rheumatic disease, malignant tumor, and polyneuropathy, organomegaly, endocrinopathy, monoclonal protein, and skin changes (POEMS) syndrome and other diseases.
We collected the clinical information, including gender, age, race, clinical characteristics, laboratory results, radiological findings, pathologic types, and treatment strategies from the medical records. B symptoms were defined as fever above 38°C, night sweats, or weight loss ≥ 10% in the past 6 months. Treatment options included surgery, corticosteroids, cyclophosphamide, chemotherapy, rituximab, anti-interleukin-6 (IL-6) monoclonal antibody, thalidomide, radiation therapy, and watching and waiting. The dose, order, and regimen of drugs administered varied across patients.
Follow-up
All patients were followed up by outpatient services or over telephonic conversations (follow-up every 1 to 3 months). The last follow-up date was January 2018, and detailed follow-up records are available for 365 of total 428 patients. The median follow-up time was 41 months (range, 1 to 279 months). Overall survival (OS) was defined as the time from pathological diagnosis until death, loss to follow-up, or last follow-up. Treatment failure was defined as recurrence, disease progression, or death of any cause. Relapse or progression was determined by the Lugano classification. The response was defined as complete response (CR) or partial response (PR), which was evaluated by the Castleman Disease Collaborative Network (CDCN). Some cases in this group were not included due to the lack of objective examination data (such as computerized tomography detection of changes in mass size), resulting in the number of efficacy evaluations being lower than the number of follow-up visits. Patients with survival times longer than 15 years were not included in the follow-up evaluation of risk factors affecting prognosis.
Statistical analysis
Statistical analyses were carried out using SPSS 22.0 (IBM, Armonk, NY, USA). Descriptive statistics were used to analyze the characteristics and treatment outcomes. Different groups were compared using chi-square and Fisher’s exact. We used the Kaplan-Meier method to perform univariate analyses of possible prognostic factors for OS, and the log-rank test was used to analyze the survival rate between the two groups. GraphPad Prism software (GraphPad Software, San Diego, California, USA) was used to draw survival curves and to analyze the influence of single factors on OS. Finally, the Cox regression model was used to analyze independent prognostic factors in CD or iMCD. A logarithmic rank test was used to calculate the P value, and P < 0.05 was considered to indicate statistical significance.
Results
UCD vs. iMCD
We compared the clinical characteristics between patients with UCD and iMCD. The results are shown in Table 1. Histopathological and radiological findings were used to classify the patients as UCD (n = 248) and iMCD (n = 180). The group consisted of Asians (n = 355), Americans (n = 73), and patients of unknown nationality (Figure 1). Of the available data (n = 394), the median age was 41 years. Patients with UCD (median age, 38 years; range, 7 to 84 years) were significantly younger than patients with iMCD (median age, 47 years; range, 11 to 75 years). There was also a significant difference in age (age ≤ 40 vs. 40 < age ≤ 60 vs. age > 60) between patients with UCD and iMCD.
Table 1.
The baseline demographic clinical features of patients with CD in our cohort (N = 428)
| Total | UCD | iMCD | P value | |||
|---|---|---|---|---|---|---|
|
|
|
|
||||
| N | N | No. of patients N (%) | N | No. of patients N (%) | ||
| Gender | 428 | 248 | 180 | 0.2237 | ||
| Male | 112 (45.16) | 92 (51.11) | ||||
| Female | 136 (54.84) | 88 (48.89) | ||||
| Age (years) | 394 | 235 | 159 | < 0.0001 | ||
| ≤ 40 | 136 (57.87) | 54 (33.96) | ||||
| 40 < age ≤ 60 | 75 (31.91) | 77 (48.43) | ||||
| > 60 | 24 (10.21) | 28 (17.61) | ||||
| Ethnicity | 428 | 248 | 180 | 0.9379 | ||
| American | 42 (16.94) | 31 (17.22) | ||||
| Asian | 206 (83.06) | 149 (82.78) | ||||
| Histological subtype | 380 | 221 | 159 | < 0.0001 | ||
| HV | 170 (76.92) | 45 (28.30) | ||||
| PC | 51 (23.08) | 114 (71.70) | ||||
| Clinical manifestation | ||||||
| Fever | 428 | 248 | 5 (2.02) | 180 | 33 (18.33) | < 0.0001 |
| Fatigue | 428 | 248 | 11 (4.44) | 180 | 34 (18.89) | < 0.0001 |
| Pain | 428 | 248 | 39 (15.73) | 180 | 30 (16.67) | 0.7938 |
| B symptom | 428 | 248 | 16 (6.45) | 180 | 58 (32.22) | < 0.0001 |
| ECOG | 273 | 122 | 102 | < 0.0001 | ||
| 0-2 | 122 (100) | 78 (76.47) | ||||
| 3-5 | 0 | 24 (23.53) | ||||
CD, Castleman disease; UCD, unicentric Castleman disease; iMCD, idiopathic multicentric Castleman disease; HV, hyaline-vascular variant; PC, plasmacell variant; MIX, mixed cellular variant. B symptom include fever (above 38°C), night sweats, weight loss (more than 10% within 6 months). American include White, Black and Hispanic. The N(x) presents the number of patients and the percentage of each group. UCD and iMCD groups were compared by chi-square test. There is statistic difference when P < 0.05.
Figure 1.
Correlation between the location and nationality of the affected lymph node and UCD/iMCD. Two patients who survived for more than 15 years were excluded. UCD, unicentric castleman disease. iMCD, idiopatic multicentric castleman disease.
The lymph nodes of most patients with UCD were histologically classified as HV subtype (76.92%). Most lymph nodes of patients with iMCD were of the PC subtype (71.7%), with only 28.3% being of the HV subtype (Table 1). Patients with iMCD commonly had symptoms of systemic inflammation. Fever (33/180, 18.33%), fatigue (34/180, 18.89%), and B symptoms (58/180, 32.22%) occurred more frequently in patients with iMCD. In patients with UCD, the most common invasions were head/neck (36.3%) and abdomen (34.7%), and in some patients the mediastinum was also involved (15.4%). iMCD was mainly found in the head/neck (26%) and armpit (20.7%) (Figure 1). However, UCD had larger masses than iMCD, with a mean size of 46 mm (maximum diameter range, 20-110 mm) compared with the 23 mm masses of iMCD (maximum diameter range, 10-100 mm) (P < 0.01). There was also a rare case of family history of iMCD in the US, one brother with UCD and one daughter with iMCD.
Laboratory findings
At presentation, patients with iMCD commonly had symptoms of systemic inflammation (Table 2). iMCD were much more likely to experience anemia than UCD (P < 0.01). Of the 409 patients whose platelet counts were measured, platelet count less than 100 × 109/L occurred in 19 of the iMCD, which compared with 5 of UCD (P < 0.01). Elevation of β2-microglobulin, alkaline phosphatase, C-reactive protein, and erythrocyte sedimentation rate were significantly more common in iMCD than in UCD (P < 0.01). Among patients with iMCD, 58.24% had decreased albumin, whereas in patients with UCD, the percentage was 22.73% (P < 0.01). Of the 428 patients who had the record data of computed tomography or color ultrasonic, patients with iMCD frequently presented with serous effusion (46/180, 25.55%) and hepatomegaly and/or splenomegaly (71/180, 39.44%; P < 0.01).
Table 2.
The baseline laboratory results of patients with CD in our cohort (N = 428)
| Total | UCD | iMCD | P value | |||
|---|---|---|---|---|---|---|
|
|
|
|
||||
| N | N | No. of patients N (%) | N | No.of patients N (%) | ||
| WBC | 410 | 233 | 177 | 0.0141 | ||
| ≤ 4 × 109/L | 18 (7.73) | 23 (12.99) | ||||
| 4-10 × 109/L | 195 (83.69) | 127 (71.75) | ||||
| > 10 × 109/L | 20 (8.58) | 27 (15.25) | ||||
| Hemoglobin | 410 | 233 | 177 | |||
| ≤ 100 g/L | 12 (5.15) | 56 (31.64) | < 0.0001 | |||
| ≤ 80 g/L | 3 (1.29) | 27 (15.25) | < 0.0001 | |||
| Platelets | 409 | 232 | 177 | < 0.0001 | ||
| ≤ 100 × 109/L | 5 (2.16) | 19 (10.73) | ||||
| 100-300 × 109/L | 186 (80.17) | 96 (54.24) | ||||
| > 300 × 109/L | 41 (17.67) | 62 (35.03) | ||||
| Serous effusion | 428 | 248 | 5 (2.02) | 180 | 46 (25.55) | < 0.0001 |
| Hepatomegaly and/or splenomegaly | 428 | 248 | 6 (2.42) | 180 | 71 (39.44) | < 0.0001 |
| Mass > 5 cm | 331 | 194 | 88 (45.36) | 137 | 24 (17.52) | < 0.0001 |
| Elevated LDH | 355 | 201 | 13 (6.47) | 154 | 23 (14.94) | 0.0088 |
| Elevated β2-MG | 226 | 113 | 13 (11.50) | 113 | 81 (71.68) | < 0.0001 |
| Elevated ESR | 169 | 62 | 14 (22.58) | 107 | 56 (52.34) | 0.0001 |
| Decreased Alb | 245 | 154 | 35 (22.73) | 91 | 53 (58.24) | < 0.0001 |
| Elevated CRP | 176 | 78 | 17 (21.79) | 98 | 56 (57.14) | < 0.0001 |
| Elevated AKP | 218 | 119 | 10 (8.40) | 99 | 26 (26.26) | 0.0004 |
WBC, white blood cells. Serous effusion includes pleural effusion, abdominal effusion, pericardial effusion, and pelvic effusion. Elevated LDH, lactate dehydrogenase over test baseline. Elevated β2-MG, β2 microglobulin > 2.4 mg/L. Elevated ESR, erythrocyte sedimentation rate > 20 mm/h. Decreased ALB, albumin < 40 g/L. Elevated CRP, C-reactive protein > 10 mg/L. Elevated AKP, alkaline phosphatase > 126 U/L. The N(x) presents the number of patients and the percentage of each group. UCD and iMCD groups were compared by chi-square test. P < 0.05 was considered to indicate significant differences.
Histopathological findings
Considering that this study was conducted over a considerable period of time, some pathological diagnoses were not covered in a standardized or comprehensive manner in the early years of the study. Therefore, in the histopathological analysis, we only included 380 cases with detailed pathological diagnosis records (Table 3). Of 380 CD cases, 215 (56.58%) could be classified as HV subtype, and 165 (43.42%) could be classified as PC subtype. Patients with HV subtype (median age, 37 years; range, 7 to 81 years) were younger than patients with PC subtype (median age, 48 years; range, 10 to 84 years). Most patients with PC subtype had obvious systemic symptoms and showed significant changes in laboratory indicators. Elevation of lactate dehydrogenase, β2-microglobulin, alkaline phosphatase, and C-reactive protein were significantly more common in patients with PC subtype (P < 0.01). Of 68 (85%) cases with PC subtype also accompanied with decreased albumin and this patient group often presented with anemia and thrombocytopenia (P < 0.01). In contrast, patients with HV subtype had fewer symptoms and inflammatory changes.
Table 3.
The baseline demographic clinical features of patients with CD in our cohort (n = 380)
| Total | HV | PC | P value | |||
|---|---|---|---|---|---|---|
|
|
|
|
||||
| N | N | No. of patients N (%) | N | No. of patients N (%) | ||
| Clinical subtype | 380 | 215 | 165 | < 0.0001 | ||
| UCD | 170 (79.07) | 51 (30.91) | ||||
| iMCD | 45 (20.93) | 114 (69.09) | ||||
| Gender | 380 | 215 | 165 | 0.3155 | ||
| Male | 97 (45.12) | 83 (50.30) | ||||
| Female | 118 (54.88) | 82 (49.70) | ||||
| Age (years) | 360 | 212 | 148 | < 0.0001 | ||
| ≤ 40 | 127 (59.91) | 47 (31.76) | ||||
| 40 < age ≤ 60 | 65 (30.66) | 72 (48.65) | ||||
| > 60 | 20 (9.43) | 29 (19.59) | ||||
| Ethnicity | 380 | 215 | 165 | 0.1344 | ||
| American | 47 (21.86) | 26 (15.76) | ||||
| Asian | 168 (78.14) | 139 (84.24) | ||||
| Clinical manifestation | ||||||
| Fatigue | 380 | 215 | 12 (5.58) | 165 | 29 (17.58) | 0.0001 |
| Pain | 380 | 215 | 31 (14.42) | 165 | 27 (16.36) | 0.6012 |
| B symptom | 380 | 215 | 20 (9.30) | 165 | 54 (32.73) | < 0.0001 |
| ECOG | 212 | 132 | 80 | < 0.0001 | ||
| 0-2 | 131 (99.24) | 58 (72.50) | ||||
| 3-5 | 1 (0.76) | 22 (27.50) | ||||
| WBC | 364 | 207 | 157 | 0.0373 | ||
| ≤ 4 × 109/L | 18 (8.70) | 16 (10.19) | ||||
| 4-10 × 109/L | 172 (83.09) | 115 (73.25) | ||||
| > 10 × 109/L | 17 (8.21) | 26 (16.56) | ||||
| Hemoglobin | ||||||
| ≤ 100 g/L | 364 | 207 | 11 (5.31) | 157 | 51 (32.48) | < 0.0001 |
| ≤ 80 g/L | 364 | 207 | 5 (2.42) | 157 | 28 (17.83) | < 0.0001 |
| Platelets | 363 | 207 | 156 | < 0.0001 | ||
| ≤ 100 × 109/L | 1 (0.48) | 17 (10.90) | ||||
| 100-300 × 109/L | 165 (79.71) | 86 (55.13) | ||||
| > 300 × 109/L | 41 (19.81) | 53 (33.97) | ||||
| Serous effusion | 380 | 215 | 9 (4.19) | 165 | 38 (23.03) | < 0.0001 |
| Hepatomegaly and/or Splenomegaly | 380 | 215 | 12 (5.58) | 165 | 54 (32.73) | < 0.0001 |
| Mass > 5 cm | 311 | 194 | 82 (42.27) | 117 | 27 (23.08) | 0.0006 |
| Elevated LDH | 308 | 173 | 8 (4.62) | 135 | 23 (17.04) | 0.0003 |
| Elevated β2-MG | 185 | 78 | 15 (19.23) | 107 | 66 (61.68) | < 0.0001 |
| Elevated ESR | 149 | 58 | 14 (24.14) | 91 | 47 (51.65) | 0.0008 |
| Elevated CRP | 158 | 89 | 15 (16.85) | 69 | 46 (66.67) | < 0.0001 |
| Decreased Alb | 245 | 165 | 71 (43.03) | 80 | 68 (85.00) | < 0.0001 |
| Elevated AKP | 176 | 78 | 11 (14.10) | 98 | 19 (19.39) | 0.4220 |
The N(x) presents the number of patients and the percentage of each group. HV and PC groups were compared by chi-square test. P < 0.05 was considered to indicate significant differences.
Treatment outcome
Among the 417 patients with relatively complete treatment data, there were 243 UCD cases and 174 iMCD cases. The curative effect was calculated according to the treatment options (Table 4). Information with respect to treatment outcome was analyzed to effectively assess treatment effects. Only some patients had a detailed treatment evaluation.
Table 4.
Treatment outcomes of patients with CD
| First-line | Patients (N) | Patients with follow-up data (N) | Patients with treatment evaluation data (N) | Response* (N) | No response* (N) | Treatment failure* (N) |
|---|---|---|---|---|---|---|
| UCD | 243 | 201 | 145 | 129 | 3 | 13 |
| Surgery (total excision) | 230 | 189 | 136 | 126 | 1 | 9 |
| Drugs or radiation with surgical biopsy | 10 | 9 | 9 | 4 | 1 | 4 |
| Watch and wait | 3 | 3 | 3 | 3 | / | / |
| iMCD* | 170 | 153 | 120 | 55 | 27 | 38 |
| Corticosteroid monotherapy | 13 | 13 | 7 | 2 | 4 | 1 |
| CTX/Thalidomide/Rituximab + Corticosteroid | 13 | 9 | 7 | 4 | 3 | / |
| Cytotoxic chemotherapy | 75 | 71 | 59 | 32 | 13 | 14 |
| CHOP or CHOP-like | 56 | 54 | 46 | 22 | 14 | 10 |
| Rituximab based (R-CHOP or R-COP) | 17 | 15 | 14 | 9 | 1 | 4 |
| Others* | 2 | 2 | 1 | 1 | / | / |
| Anti-interleukin 6 monoclonal antibody* | 8 | 8 | 8 | 4 | 1 | 3 |
| Tocilizumab with other agents | 1 | 1 | 1 | / | / | / |
| Siltuximab with or without any agent | 7 | 7 | 6 | 3 | 1 | 2 |
| Watch and wait (include biopsy) | 61 | 52 | 26 | 13 | 7 | 6 |
The treatment of patients with iMCD excluded one patient who received biopsy and radiotherapy, one patient who received cyclophosphamide monotherapy, and two patients who received rituximab monotherapy, because of the low numbers of cases. CHOP includes cyclophosphamide, doxorubicin, vincristine, and prednisone. CHOP-like is based on COP.
Others include fludarabine/cyclophosphamide and cyclophosphamide/Vindesine cytotoxic chemotherapy, excluding CHOP-like or R-CHOP-like. Anti-interleukin-6 monoclonal antibody was always used in combination with drugs such as thalidomide and corticosteroid. The proportion alive at follow-up only includes patients with available treatment information.
Response = complete remission and partial remission;
No response = stable disease;
Failure = disease progression, invalidity, or death.
Complete surgical resection was performed in 230 (94.65%) patients diagnosed with UCD as first-line treatment. Ten patients received drugs or radiation with surgical biopsy, and two patients chose to watch and wait. Of these patients, 126 (92.65%) who underwent complete surgical resection achieved a response, and 10 (7.35%) patients suffered from disease recurrence or progression. Of these 10 patients, two patients who underwent re-operation achieved CR, and two patients achieved CR and PR after cyclophosphamide/doxorubicin/vincristine/prednisone (CHOP) treatment or radiotherapy. Among the patients who received drug or radiation treatment with surgical biopsy (R-CHOP/CHOP, rituximab monotherapy, and radiotherapy), four patients achieved CR/PR, two patients achieved a stable condition, and two patients progressed (one patient who received cyclophosphamide combined with hormone and one patient who received rituximab). Two other patients chose to watch and wait; the mass shrank spontaneously, and the patients achieved CR.
Treatments differed considerably among patients with iMCD. Among the 13 patients treated with glucocorticoid monotherapy, two patients achieved CR/PR. The curative effects of glucocorticoid monotherapy and of glucocorticoid combined with other drugs were not significantly different (P > 0.05). Most patients (75/174, 43.1%) chose chemotherapy. Among the 17 patients treated with rituximab-based chemotherapy (R-CHOP or R-COP), 9 patients (64.28%) showed a response, and treatment failed in four (28.57%) patients (P > 0.05). Among the 56 patients who received CHOP or a CHOP-like regimen, 22 patients (47.8%) showed a response, and treatment failed in 10 (21.73%) patients (P > 0.05). In the present study, seven patients were treated with siltuximab monotherapy or siltuximab combined with other drugs; of these patients, three achieved CR/PR, and treatment failed in one patient, despite autologous stem cell transplantation after combination therapy. Only one patient chose tocilizumab with other drugs who achieved CR. Among the 61 patients who chose to watch and wait, 13 cases improved spontaneously and 6 did not. Due to the lack of detailed clinical and laboratory data, it was difficult to further analyze these patients.
Univariate survival analysis
In the present study, follow-up data were available for 365 patients with CD (203 with UCD and 160 with iMCD), excluding two patients with CD who survived for more than 15 years. The median follow-up time for CD was 41 months (range, 0.53 to 173.6 months). By the last follow-up, 50 patients were dead (7 UCD cases and 43 iMCD cases). In the group of UCD, one patient died of unknown cause who was a 33-year-old female with severe anemia, high ESR, and low protein at onset; a 27-year-old female with paraneoplastic pemphigus who died of disease recurrence and progression; a 50-year-old male with renal insufficiency died of disease recurrence and progression; a 28-year-old male who died of recurrent disease progression; a 54-year-old male, died of cerebrovascular accident and the remaining two patients died of unknown cause who were of no obvious special. In the group of iMCD, 22 patients died of disease progression and recurrence, 2 patients died of infection. The cause of death in 5 cases was unknown, but they all complicated other diseases or diagnosed severe iMCD. There were 2 patients died of cardiovascular accidents and the remaining 12 cases had unknown causes of death. The 5-year OS of patients with CD was 87%. For patients with UCD, the 5-year OS was 95%. For patients with iMCD, the 5-year OS was 74%. The log-rank test for OS showed a significant difference between UCD and iMCD (P < 0.001).
Using the Kaplan-Meier method and the log-rank test, we conducted a univariate analysis and identified 10 significant risk factors of CD: multicentricity (Figure 2), age > 60 years, PC subtype, fever, B symptoms, serous effusion, hepatomegaly and/or splenomegaly, hemoglobin ≤ 80 g/L, albumin ≤ 30 g/L, and elevation of erythrocyte sedimentation rate and β2-microglobulin (P < 0.05; some results showed in Figure S1). The survival rate was lower among the Asian population than among the American population (P < 0.05; Figure 2).
Figure 2.
Overall survival of multicentricity and nationality in patients with CD. Patients with iMCD (A) was correlated with significantly poorer overall survival. Asian patients with CD (B) or iMCD (C) were correlated with inferior overall survival. (D) There was no difference between patients with UCD for Asian and Amierican. Two patients who survived for more than 15 years were excluded. CD, castleman disease. iMCD, idiopatic multicentric castleman disease.
UCD had a good prognosis after primary lesion resection, while iMCD had a poor prognosis. In UCD, we identified two significant risk factors: hemoglobin ≤ 100 g/L and albumin ≤ 30 g/L (Figure 3). In this study, serous effusion, hepatomegaly and/or splenomegaly, leukopenia, and albumin ≤ 30 g/L were significantly associated with OS of patients with iMCD (P < 0.001; Figure 4). Age > 60 years, B symptoms, and hemoglobin ≤ 80 g/L were also associated with OS of patients with iMCD (P < 0.05), but PC subtype did not correlate with better OS, although a trend toward better survival is suggested among HV subtype (Figure 4). The patients with iMCD who chose to watch and wait had a poorer prognosis than patients with iMCD who were treated (P < 0.05).
Figure 3.
Overall survival of hemoglobin and albumin in patients with UCD. (A) Hemoglobin ≤ 100 g/L and (B) Albumin ≤ 30 g/L were associated with poor prognosis in patients with UCD. One patient who survived for more than 15 years were excluded. CD, castleman disease. UCD, unicentric castleman disease.
Figure 4.
Overall survival rate of patients with iMCD. (A) PC subtype did not correlate with better OS. (B) age, (C) Elevated or decreased white blood cell counts, (D) hemoglobin ≤ 80 g/L, (E) Albumin ≤ 30 g/L, and (F) elevated erythrocyte sedimentation rate > 20 mm/h, (G) B symptoms, (H) serous effusion, and (I) hepatomegaly and/or splenomegaly were correlated with poor overall survival in patients with iMCD. One patient who survived for more than 15 years were excluded. iMCD, idiopatic multicentric castleman disease. PC subtype, plasmacytic subtype.
Multivariate analysis of risk factors
The Cox proportional hazards model was used to perform a multivariate analysis of factors affecting CD, including multicentricity, pathological subtype, age, B symptoms, hepatomegaly and/or splenomegaly, hemoglobin ≤ 80 g/L, and albumin ≤ 30 g/L. The results showed that PC subtype, hepatomegaly and/or splenomegaly, hemoglobin ≤ 80 g/L, and albumin ≤ 30 g/L were independently associated with OS (Table 5). Further, we generated a model of iMCD, including pathological subtype, age, B symptoms, serous effusion, hepatomegaly and/or splenomegaly, leukopenia, hemoglobin ≤ 80 g/L, and albumin ≤ 30 g/L. The results showed that age > 60 years, hepatomegaly and/or splenomegaly, and hemoglobin ≤ 80 g/L were independent risk factors for the prognosis of iMCD (Table 5).
Table 5.
Multivariate analysis of CD and iMCD
| Factor | OS in patients with CD | ||
|
| |||
| HR | 95% CI | P value* | |
|
| |||
| PC subtype | 3.853 | 1.587-9.357 | 0.0029 |
| Hepatomegaly and/or splenomegaly | 3.495 | 1.626-7.511 | 0.0013 |
| Hemoglobin (≤ 80 g/L) | 2.290 | 1.028-5.101 | 0.0042 |
| Albumin (≤ 30 g/L) | 1.671 | 1.068-2.614 | 0.0247 |
| iMCD | 1.655 | 0.490-5.589 | 0.3617 |
|
| |||
| Factor | OS in patients with iMCD | ||
|
| |||
| HR | 95% CI | P value* | |
|
| |||
| Age (>60 years) | 2.288 | 1.307-4.003 | 0.0037 |
| Hepatomegaly and/or splenomegaly | 3.248 | 1.532-7.314 | 0.0024 |
| Hemoglobin (≤ 80 g/L) | 2.699 | 1.133-6.431 | 0.0250 |
| PC subtype | 1.360 | 0.388-4.763 | 0.1948 |
P values ≤ 0.05 shown as bold indicated a statistically significant difference.
Discussion
Most studies on clinical, laboratory, treatment, and prognosis data of patients with CD are single randomized and controlled trials, small series research, and case reports [10,11]. In order to obtain more information, we performed a large study of patients with CD from Chinese and American medical centers, which comprises the largest sample size reported thus far and different ethnic groups with 15-year follow-up. In particular, our study provides valuable and comprehensive information on clinical characteristics and prognostic factors that advance our understanding of CD, especially iMCD.
We show the heterogeneity between patients with UCD and iMCD as well as within each pathological subtype. CD affects patients of all ages, with a peak frequency during adulthood. The median age of patients with UCD (38 years) was significantly lower than that of patients with iMCD (47 years), which is consistent with the median age of Japanese (43 years, n = 342) and American (55 years, n = 59) patients with MCD [12,13]. Consistent with previous studies, the changes in symptoms and laboratory indicators were more pronounced in iMCD, and the masses of UCD were larger than those of iMCD [13-15]. A few cases of UCD also had systemic inflammation, aggressive symptoms, and abnormal laboratory results, although these indicators lacked diagnostic specificity. There is a certain proportion of crossover between UCD and iMCD in all aspects. Pathological changes are the basis of clinical manifestation. The HV subtype was more prevalent in patients with UCD (76.92%), and the PC subtype was more prevalent in patients with iMCD (71.7%), as we reported before [15,16]. The differences in clinical manifestations between patients with PC and patients with HV were similar to those between patients with iMCD and patients with UCD. The pathological changes are the basis of the clinical manifestations. It may related to the differences between the pathogeneses of CD and elderly people often accompanied by immunodeficiency. It has been hypothesized that it involves autoimmunity/auto-inflammation, paraneoplastic syndrome, and viral infection [16,17]. In our analysis, there was a patient with one brother who had UCD and one daughter who had iMCD. A family history of iMCD has also been found in a previous study, with two patients who had a FAS mutation [18]. Another study found that mutation of the MEFV gene may contribute to systemic reactions of MCD [19]. Due to the lack of second-generation sequencing, we were unable to identify the pathogenesis in this study.
Great differences with respect to treatment were observed between patients with UCD and iMCD. For patients with UCD, complete resection of the involved lesion is considered as the golden standard treatment, and the 5-year OS rate approaches 100% [8]. Unlike UCD, no standard treatment has been established for iMCD, and the outcome is less favorable. A variety of agents have been used to treat MCD, including corticosteroids, cytotoxic chemotherapy, thalidomide, immunoglobulin, rituximab, anti-IL-6 antibody (siltuximab and tocilizumab), and mTOR pathway inhibitors [20-22]. Corticosteroids can quickly and effectively relieve clinical symptoms, which is a major advantage over other drugs. However, the CR rate has been reported to be low. In the present analysis, the effective rate of 13 patients who underwent glucocorticoid monotherapy was about 30% and 50% of patients with iMCD are resistant to glucocorticoids. The CR and PR rates with rituximab or rituximab-based chemotherapy regimens as first-line therapy were 20% and 48%, respectively. Cytotoxic chemotherapy may induce responses, and the overall remission rate of chemotherapy can be as high as 78%, but many patients will progress or experience infectious toxicities [6,15,20]. In our study, only two patients chose rituximab monotherapy, compared with 75 patients who chose chemotherapy. Of these 75 patients, 17 (22.67%) patients chose R-CHOP/R-CHOP-like treatment, and others chose CHOP/CHOP-like treatment. The remission rate of cytotoxic chemotherapy was over 50%, which was not significantly different from that of other treatment regimens, but considerable toxicity and frequent relapses deter its use [23,24]. Considering the increased expression of IL-6, anti-IL-6 therapy may be more beneficial [9]. Due to drug inaccessibility, few patients received the new drugs. Only 11 patients were treated with anti-IL-6 therapy; of these, eight cases were treated with siltuximab, with an effective rate of 42.85% (3/7), similar to the 34% remission rate of siltuximab observed in the only randomized controlled trial of iMCD [22], but there was no obvious advantage compared with other treatment groups. For iMCD with no obvious symptoms, watching and waiting can be chosen. However, the failure rate of watching and waiting was 50% (13/26), suggesting that long-term observation and timely treatment are necessary. In conclusion, the treatment regimen must be selected based on drug accessibility and experiences. The severity, prognostic factors, and biomarkers that are available to select patients who will respond to these treatments are limited, which need more studies.
In the present study, the different prognostic factors between CD and iMCD may be related to the overall effect of UCD inclusion in the analysis or statistical bias. By single-factor analysis, anemia and hypoproteinemia also affect UCD, as reported by Lan X et al [16,25]. Patients with iMCD had significantly worse survival rates, with a 5-year OS of 74% (P < 0.001), which is consistent with previous reports [26,27]. Due to the lack of an optimal treatment for iMCD, it is important to identify prognostic factors to help determine treatment strategies. Meta-analysis of 416 cases has suggested five prognostic factors by univariate analysis, including multicentricity, pathology type, the presence of symptoms, gender, and age [27]. A study reported by the MD Anderson Cancer Center suggested that multicentricity, PC subtype, and anemia were associated with poor prognosis in patients from North America [15]. In our study, PC subtype and anemia were associated with poor prognosis. A retrospective study executed by Peking University obtained similar results [28]. Hepatomegaly and/or splenomegaly is a common symptom in iMCD [13,15,29,30]. Furthermore, previous studies found that hypoproteinemia is an independent prognostic risk factor in HIV-negative patients with CD (n = 71) and patients with MCD (n = 185) [25,31]. We also found that hypoproteinemia (≤ 30 g/L) is an independent risk factor for patients with CD and for patients with iMCD. Other unfavorable prognostic factors have been reported, including increased IL-6 levels, the presence of extravascular fluid accumulation, clinical complications, and splenomegaly [28,31-33]. The prognosis of CD in Asian patients was worse than that in American patients, of which may be due to statistical bias or factors associated with geography and race, but future studies are needed to confirm this. Combined with literature reports and the above analysis, the pathological types and clinical classification have not changed, and to obtain the best treatment effect under existing conditions, the control of systemic inflammation should be enhanced, anemia and hypoalbuminemia should be prevented, and liver function should be protected.
This study has several limitations. First, as a retrospective study, there may be a bias for patient selection and data collection. We excluded some laboratory tests because data were not available, including C-reactive protein levels and the erythrocyte sedimentation rate. Second, because of the heterogeneous treatments of patients with iMCD, we could not further compare the effects of different treatment strategies. In addition, only a small number of patients with iMCD in our study received anti-IL-6 or rituximab as first-line therapy, mostly for drug inaccessibility. The role of these new agents in iMCD treatment requires further investigation. The present study has several strengths. First of all, with the sample size being the largest reported to date, this multicenter study identified the clinical characteristics and prognostic factors of CD, especially iMCD. Second, the study indicated that the unfavorable prognostic factors of UCD and MCD were not identical, which suggests that different therapeutic strategies need to be followed for different clinical classifications. Finally, the results indicated that PC subtype, hepatomegaly and/or splenomegaly, hemoglobin ≤ 80 g/L, and albumin ≤ 30 g/L are independent prognosis factors of CD. Age > 60 years, hepatomegaly and/or splenomegaly, and anemia were independent risk factors for the prognosis of iMCD. Therefore, except chemotherapy and resection, appropriate medical supportive treatment should be considered, such as splenectomy, correction of anemia, and supplementation of albumin. Further studies are needed to confirm these prognostic factors and investigate the optimal treatment for iMCD.
In general, CD is a highly heterogeneous disorder, and little is known about its pathogenesis, clinical manifestation, treatment, and prognosis. No standard treatment for CD has been established. In the present study, we identified differences in clinical features and prognostic factors between UCD and iMCD. We found significant risk factors of CD and iMCD, which were not identical. Comparison of the results of our study with other case reports or small-scale studies can provide a better understanding of the pathophysiology of CD, enabling faster and more efficient diagnosis and effective treatment. We hope our study provides a clear understanding of the differences between UCD and iMCD, and the risk factors identified in the present study can provide a reference for therapeutic decision-making and prognosis assessment. It is essential to continue to explore the etiology, classification, treatment, and prognosis of CD in future investigations.
Acknowledgements
We thank the patients who participated in this study and LetPub for its linguistic assistance during the preparation of this manuscript. This work was supported by Hagemeister Lymphoma Foundation, Gundersen Lutheran Medical Foundation, Roche Molecular System, Gilead Sciences, Seattle Genetics, Dai Sanyo, Adaptive Biotechnology, Incyte Pharmaceutical, and HTG Molecular Diagnostics. LY is supported by National Natural Science Foundation of China 81460030 and 81770221.
Disclosure of conflict of interest
None.
Supporting Information
References
- 1.Castleman B, Iverson L, Menendez VP. Localized mediastinal lymphnode hyperplasia resembling thymoma. Cancer. 1956;9:822–830. doi: 10.1002/1097-0142(195607/08)9:4<822::aid-cncr2820090430>3.0.co;2-4. [DOI] [PubMed] [Google Scholar]
- 2.Fajgenbaum DC, van Rhee F, Nabel CS. HHV-8-negative, idiopathic multicentric Castleman disease: novel insights into biology, pathogenesis, and therapy. Blood. 2014;123:2924–2933. doi: 10.1182/blood-2013-12-545087. [DOI] [PubMed] [Google Scholar]
- 3.Dispenzieri A, Fajgenbaum DC. Overview of Castleman disease. Blood. 2020;135:1353–1364. doi: 10.1182/blood.2019000931. [DOI] [PubMed] [Google Scholar]
- 4.Talat N, Belgaumkar AP, Schulte KM. Surgery in Castleman’s disease: a systematic review of 404 published cases. Ann Surg. 2012;255:677–684. doi: 10.1097/SLA.0b013e318249dcdc. [DOI] [PubMed] [Google Scholar]
- 5.Liu AY, Nabel CS, Finkelman BS, Ruth JR, Kurzrock R, van Rhee F, Krymskaya VP, Kelleher D, Rubenstein AH, Fajgenbaum DC. Idiopathic multicentric Castleman’s disease: a systematic literature review. Lancet Haematol. 2016;3:e163–175. doi: 10.1016/S2352-3026(16)00006-5. [DOI] [PubMed] [Google Scholar]
- 6.van Rhee F, Voorhees P, Dispenzieri A, Foss A, Srkalovic G, Ide M, Munshi N, Schey S, Streetly M, Pierson SK, Partridge HL, Mukherjee S, Shilling D, Stone K, Greenway A, Ruth J, Lechowicz MJ, Chandrakasan S, Jayanthan R, Jaffe ES, Leitch H, Pemmaraju N, Chadburn A, Lim MS, Elenitoba-Johnson KS, Krymskaya V, Goodman A, Hoffmann C, Zinzani PL, Ferrero S, Terriou L, Sato Y, Simpson D, Wong R, Rossi JF, Nasta S, Yoshizaki K, Kurzrock R, Uldrick TS, Casper C, Oksenhendler E, Fajgenbaum DC. International, evidence-based consensus treatment guidelines for idiopathic multicentric Castleman di sease. Blood. 2018;132:2115–2124. doi: 10.1182/blood-2018-07-862334. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Fajgenbaum DC, Uldrick TS, Bagg A, Frank D, Wu D, Srkalovic G, Simpson D, Liu AY, Menke D, Chandrakasan S, Lechowicz MJ, Wong RS, Pierson S, Paessler M, Rossi JF, Ide M, Ruth J, Croglio M, Suarez A, Krymskaya V, Chadburn A, Colleoni G, Nasta S, Jayanthan R, Nabel CS, Casper C, Dispenzieri A, Fossa A, Kelleher D, Kurzrock R, Voorhees P, Dogan A, Yoshizaki K, van Rhee F, Oksenhendler E, Jaffe ES, Elenitoba-Johnson KS, Lim MS. International, evidence-based consensus diagnostic criteria for HHV-8-negative/idiopathic multicentric Castleman disease. Blood. 2017;129:1646–1657. doi: 10.1182/blood-2016-10-746933. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.van Rhee F, Oksenhendler E, Srkalovic G, Voorhees P, Lim M, Dispenzieri A, Ide M, Parente S, Schey S, Streetly M, Wong R, Wu D, Maillard I, Brandstadter J, Munshi N, Bowne W, Elenitoba-Johnson KS, Fössa A, Lechowicz MJ, Chandrakasan S, Pierson SK, Greenway A, Nasta S, Yoshizaki K, Kurzrock R, Uldrick TS, Casper C, Chadburn A, Fajgenbaum DC. International evidence-based consensus diagnostic and treatment guidelines for unicentric Castleman disease. Blood Advances. 2020;4:6039–6050. doi: 10.1182/bloodadvances.2020003334. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Carbone A, Borok M, Damania B, Gloghini A, Polizzotto MN, Jayanthan RK, Fajgenbaum DC, Bower M. Castleman disease. Nat Rev Dis Primers. 2021;7:84. doi: 10.1038/s41572-021-00317-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Soumerai JD, Sohani AR, Abramson JS. Diagnosis and management of Castleman disease. Cancer Control. 2014;21:266–278. doi: 10.1177/107327481402100403. [DOI] [PubMed] [Google Scholar]
- 11.Munshi N, Mehra M, van de Velde H, Desai A, Potluri R, Vermeulen J. Use of a claims database to characterize and estimate the incidence rate for Castleman disease. Leuk Lymphoma. 2015;56:1252–1260. doi: 10.3109/10428194.2014.953145. [DOI] [PubMed] [Google Scholar]
- 12.Robinson D, Reynolds M, Casper C, Dispenzieri A, Vermeulen J, Payne K, Schramm J, Ristow K, Desrosiers MP, Yeomans K, Teltsch D, Swain R, Habermann TM, Rotella P, Van de Velde H. Clinical epidemiology and treatment patterns of patients with multicentric Castleman disease: results from two US treatment centres. Br J Haematol. 2014;165:39–48. doi: 10.1111/bjh.12717. [DOI] [PubMed] [Google Scholar]
- 13.Murakami M, Johkoh T, Hayashi S, Ohshima S, Mizuki M, Nakatsuka SI, Tomobe M, Kuroyanagi K, Nakasone A, Nishimoto N. Clinicopathologic characteristics of 342 patients with multicentric Castleman disease in Japan. Mod Rheumatol. 2020;30:843–851. doi: 10.1080/14397595.2019.1704983. [DOI] [PubMed] [Google Scholar]
- 14.Casper C, Teltsch DY, Robinson D Jr, Desrosiers MP, Rotella P, Dispenzieri A, Qi M, Habermann T, Reynolds MW. Clinical characteristics and healthcare utilization of patients with multicentric Castleman disease. Br J Haematol. 2015;168:82–93. doi: 10.1111/bjh.13111. [DOI] [PubMed] [Google Scholar]
- 15.Yu L, Tu M, Cortes J, Xu-Monette ZY, Miranda RN, Zhang J, Orlowski RZ, Neelapu S, Boddu PC, Akosile MA, Uldrick TS, Yarchoan R, Medeiros LJ, Li Y, Fajgenbaum DC, Young KH. Clinical and pathological characteristics of HIV- and HHV-8-negative Castleman disease. Blood. 2017;129:1658–1668. doi: 10.1182/blood-2016-11-748855. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Yu L, Shi M, Cai Q, Strati P, Hagemeister F, Zhai Q, Li L, Fang X, Li J, Sun R, Zhang S, Yang H, Wang Z, Qian W, Iwaki N, Sato Y, Zhang L, Li J, Oksenhendler E, Xu-Monette ZY, Young KH. A novel predictive model for idiopathic multicentric Castleman disease: the international Castleman D isease consortium study. Oncologist. 2020;25:963–973. doi: 10.1634/theoncologist.2019-0986. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Fajgenbaum DC, Shilling D. Castleman disease pathogenesis. Hematol Oncol Clin North Am. 2018;32:11–21. doi: 10.1016/j.hoc.2017.09.002. [DOI] [PubMed] [Google Scholar]
- 18.Baker TS, Gambino KJ, Schriefer L, Lim JY, Steinberg KM, Fajgenbaum DC, Martín García-Sancho A, Byun M. A novel FAS mutation with variable expressivity in a family with unicentric and idiopathic multicentric Castleman disease. Blood Adv. 2018;2:2959–2963. doi: 10.1182/bloodadvances.2018023911. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Endo Y, Koga T, Otaki H, Sasaki D, Sumiyoshi R, Furukawa K, Tanaka Y, Katsunori Y, Kawakami A. Idiopathic multicentric Castleman disease with novel heterozygous Ile729Met mutation in exon 10 of familial Mediterranean fever gene. Rheumatology (Oxford) 2021;60:445–450. doi: 10.1093/rheumatology/keaa269. [DOI] [PubMed] [Google Scholar]
- 20.Kapriniotis K, Lampridis S, Mitsos S, Patrini D, Lawrence DR, Panagiotopoulos N. Biologic agents in the treatment of multicentric Castleman disease. Turk Thorac J. 2018;19:220–225. doi: 10.5152/TurkThoracJ.2018.18066. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Stern RM, Berliner N. Targeting the mTOR pathway in idiopathic multicentric Castleman disease. J Clin Invest. 2019;129:4086–4088. doi: 10.1172/JCI131332. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.van Rhee F, Wong RS, Munshi N, Rossi JF, Ke XY, Fosså A, Simpson D, Capra M, Liu T, Hsieh RK, Goh YT, Zhu J, Cho SG, Ren H, Cavet J, Bandekar R, Rothman M, Puchalski TA, Reddy M, van de Velde H, Vermeulen J, Casper C. Siltuximab for multicentric Castleman’s disease: a randomised, double-blind, placebo-controlled trial. Lancet Oncol. 2014;15:966–974. doi: 10.1016/S1470-2045(14)70319-5. [DOI] [PubMed] [Google Scholar]
- 23.Seo S, Yoo C, Yoon DH, Kim S, Park JS, Park CS, Huh J, Suh C. Clinical features and outcomes in patients with human immunodeficiency virus-negative, multicentric Castleman’s disease: a single medical center experience. Blood Res. 2014;49:253–258. doi: 10.5045/br.2014.49.4.253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Zhu SH, Yu YH, Zhang Y, Sun JJ, Han DL, Li J. Clinical features and outcome of patients with HIV-negative multicentric Castleman’s disease treated with combination chemotherapy: a report on 10 patients. Med Oncol. 2013;30:492. doi: 10.1007/s12032-013-0492-0. [DOI] [PubMed] [Google Scholar]
- 25.Lan X, Li Z, Zhang M. Analysis of clinical characteristics and prognosis factors of 71 cases with HIV-negative Castleman’s disease: hypoproteinemia is an unfavorable prognostic factor which should be treated appropriately. J Cancer Res Clin Oncol. 2018;144:1265–1277. doi: 10.1007/s00432-018-2647-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Zhang L, Li Z, Cao X, Feng J, Zhong D, Wang S, Zhou D, Li J. Clinical spectrum and survival analysis of 145 cases of HIV-negative Castleman’s disease: renal function is an important prognostic factor. Sci Rep. 2016;6:23831. doi: 10.1038/srep23831. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Talat N, Schulte KM. Castleman’s disease: systematic analysis of 416 patients from the literature. Oncologist. 2011;16:1316–1324. doi: 10.1634/theoncologist.2011-0075. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Dong Y, Wang M, Nong L, Wang L, Cen X, Liu W, Zhu S, Sun Y, Liang Z, Li Y, Ou J, Qiu Z, Ren H. Clinical and laboratory characterization of 114 cases of Castleman disease patients from a single centre: paraneoplastic pemphigus is an unfavourable prognostic factor. Br J Haematol. 2015;169:834–842. doi: 10.1111/bjh.13378. [DOI] [PubMed] [Google Scholar]
- 29.Liu AY, Nabel CS, Finkelman BS, Ruth JR, Kurzrock R, van Rhee F, Krymskaya VP, Kelleher D, Rubenstein AH, Fajgenbaum DC. Idiopathic multicentric Castleman’s disease: a systematic literature review. Lancet Haematol. 2016;3:e163–e175. doi: 10.1016/S2352-3026(16)00006-5. [DOI] [PubMed] [Google Scholar]
- 30.Luo JM, Li S, Huang H, Cao J, Xu K, Bi YL, Feng RE, Huang C, Qin YZ, Xu ZJ, Xiao Y. Clinical spectrum of intrathoracic Castleman disease: a retrospective analysis of 48 cases in a single Chinese hospital. BMC Pulm Med. 2015;15:34. doi: 10.1186/s12890-015-0019-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Zhang X, Rao H, Xu X, Li Z, Liao B, Wu H, Li M, Tong X, Li J, Cai Q. Clinical characteristics and outcomes of Castleman disease: a multicenter study of 185 Chinese patients. Cancer Sci. 2018;109:199–206. doi: 10.1111/cas.13439. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Shin DY, Jeon YK, Hong YS, Kim TM, Lee SH, Kim DW, Kim I, Yoon SS, Heo DS, Park S, Kim BK. Clinical dissection of multicentric Castleman disease. Leuk Lymphoma. 2011;52:1517–1522. doi: 10.3109/10428194.2011.574759. [DOI] [PubMed] [Google Scholar]
- 33.Choy EH, De Benedetti F, Takeuchi T, Hashizume M, John MR, Kishimoto T. Translating IL-6 biology into effective treatments. Nat Rev Rheumatol. 2020;16:335–345. doi: 10.1038/s41584-020-0419-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.