research paper

Clinical characteristics and healthcare utilization of patients with multicentric Castleman disease

Corey Casper,1 Dana Y. Teltsch,2 Don Robinson Jr,3 Marie-Pierre Desrosiers,4 Philip Rotella,2 Angela Dispenzieri,5 Ming Qi,6 Thomas Habermann5 and Matthew W. Reynolds2 1

Fred Hutchinson Cancer Research Center, Seat-

tle, WA, 2Evidera, Lexington, MA, 3Janssen Global Services, Raritan, NJ, USA 4United BioSource Corporation, Dorval, QC, Canada 5

Mayo Clinic, Rochester, MN, and 6Janssen

Research and Development, Spring House, PA, USA Received 28 May 2014; accepted for publication 29 July 2014 Correspondence: Matthew W. Reynolds, Evidera, 430 Bedford Street, Suite 300, Lexington, MA 02420, USA. E-mail: [email protected]

Summary Multicentric Castleman disease (MCD) is a rare lymphoproliferative disease. Little is known about how patient clinical features and healthcare utilization varies by human immunodeficiency virus (HIV) status and disease subtype. Data of MCD patients identified between 2000 and 2009 were collected from medical records at two United States treatment centres. Clinical, demographic, and biochemical characteristics, drug therapies and medical utilization were descriptively reported by HIV status and cell histology, and statistically compared with the Fisher’s Exact and Kruskal–Wallis tests. Patients (n = 59) had a pathologically and clinically confirmed MCD diagnosis: plasmacytic (42%), hyaline vascular (29%) and mixed (15%); 10% had HIV infection. In the first year after diagnosis, MCD patients on average saw a healthcare provider more than six times, were hospitalized at least once and underwent frequent radiological and laboratory testing. Rituximab was the most commonly used drug therapy, followed by corticosteroids and conventional chemotherapy. One- and 2-year survival was excellent in HIV-negative patients (100% and 97%, respectively) but inferior for HIV-positive cases (67% and 67%, respectively). Heterogeneous treatment decisions were observed in this MCD study; HIV status was the only distinguishing clinical criteria associated with pharmacotherapies. Additional research is necessary to guide treatment of this rare lymphoproliferative disorder. Keywords: lymphoproliferative disorder, Castleman disease, treatment, healthcare utilization.

Castleman disease (CD), also known as angiofollicular lymph node hyperplasia, is a rare and poorly understood disease characterized by over-expression of interleukin-6 and subsequent proliferation of the lymphoid tissue (Brandt et al, 1990). CD can present as two distinct clinical entities: the localized form (unicentric Castleman disease [UCD]) and the multicentric form [multicentric Castleman disease (MCD)] (Waterston & Bower, 2004). Patients with UCD typically have localized, enlarged lymph nodes and are generally asymptomatic or mildly symptomatic. In contrast, MCD is a more severe form where patients have multiple and dispersed enlarged lymph nodes and are often heterogeneous in signs and symptoms: fatigue, night sweats, fevers and anaemia are some of the more frequent. In addition to the two clinical subtypes of CD, three cell histological variants have been recognized: hyaline vascular, plasmacytic and mixed type. The hyaline vascular type is First published online 11 September 2014 doi: 10.1111/bjh.13111

found in 90% of UCD patients but rarely in those with MCD, while the plasma type is found in only 10% of UCD patients but in 80–90% of those with MCD (Waterston & Bower, 2004; Casper, 2005; Dispenzieri, 2008; van Rhee et al, 2010; El-Osta & Kurzrock, 2011). MCD has a high prevalence in human immunodeficiency virus (HIV)-positive patients, affecting mainly men of a median age of 40 years; this is probably due to the association of MCD with human herpesvirus 8 (HHV-8) among HIV-infected individuals (Waterston & Bower, 2004; Mylona et al, 2008). Few studies have reported the epidemiology and natural history of MCD patients, with a focus on how subsets of patients may differ when defined by histological subgroups, HIV status or ethnicity (Mylona et al, 2008; Stebbing et al, 2008). A number of therapeutic interventions have been described in retrospective case series, with the selection of regimens often based on the predominant symptoms exhibited by the patient or, ª 2014 John Wiley & Sons Ltd British Journal of Haematology, 2015, 168, 82–93

Clinical Characteristics of Multicentric Castleman Disease perhaps, the histological subtype. To date, no definitive standard treatment for patients with MCD has been defined. Because of the complex, chronic and often serious nature of MCD, patients afflicted with the disease frequently have multiple encounters with the health care system over the course of their illness. Limited information is available on the health services that MCD patients use when under medical care. We sought to describe the care provided to patients with MCD at two major referral centres for this disease to characterize potential predictors of clinical and healthcare resource utilization.

Kruskal–Wallis test was used to compare continuous and discrete variables. All comparisons with a P-value 0
05). Biological agents were utilized for a majority of both hyaline vascular and plasmacytic patients, though no hyaline vascular patients were prescribed immune modulators. A majority (60%) of the five deaths in the cohort occurred among plasmacytic patients, including both of the deaths during the first year of follow-up, though survival did not differ significantly between hyaline vascular and plasmacytic subtypes (1-year: 100% vs. 90
0%; P = 0
516).

Discussion A decades worth of data from two large referral centres for patients with the rare lymphoproliferative disorder MCD were examined to characterize the natural history of the disease and examine healthcare utilization patterns. Based on a small number of patients, we observed that MCD appeared to be a disease with greater morbidity and mortality among HIV-positive persons, a result that is consistent with previous studies (Bower et al, 2011). HIV-positive patients tended to have anaemia, higher use of certain healthcare services and poor survival over 2 years. In contrast, HIV-negative MCD 89

90

Bilirubin, lmol/l, n Mean (SD) Median (IQR)

AST, u/l, n Mean (SD) Median (IQR)

LFTs, n ALT, u/l, n Mean (SD) Median (IQR)

LDH, u/l, n Mean (SD) Median (IQR)

CRP, mg/l, n Mean (SD) Median (IQR)

Hb concentration, g/l, n Mean (SD) Median (IQR) Platelet count, 109/l, n Mean (SD) Median (IQR)

Biochemical measures*

24

125 (27) 121 (108–146) 24

294
0 (191
8) 242
5 (155
5–415
3) 13 25
8 (26
1) 19
1 (12
0–38
8) 14 213
6 (238
3) 153
0 (134
8–179
8) 22 20 34
3 (32
5) 18
0 (11
5–41
8) 22 26
8 (20
8) 17
0 (15
0–31
0) 22

15
6 (15
2) 12
0 (8
6–15
4)

84 (15) 84 (77–92) 3

90
0 (49
5) 91
0 (65
5–115
0) 3 96
3 (154
0) 13
1 (7
5–143
6) 2 167
0 (19
8) 167
0 (160
0–174
0) 5 5 28
2 (8
1) 24
0 (22
0–37
0) 5 36
2 (13
9) 42
0 (30
0–44
0) 4

11
5 (4
3) 11
1 (9
4–13
3)

HIV-negative (N = 26)

3

HIV-positive (N = 5)

HIV status

10
3 (3
8) 10
3 (8
6–12
0)

10 17
2 (8
5) 15
0 (12
3–22
5) 10

8 155
9 (78
5) 140
0 (104
0–179
8) 10 1 4
0 (NA) 4
0 (4
0–4
0)

378
6 (246
8) 318
0 (224
5–407
0) 2 0
4 (0
1) 0
4 (0
4–0
5)

130 (28) 125 (124–144) 11

11

HIV unknown (N = 13)

16
7 (19
0) 12
0 (9
4–13
7)

11

11 24
1 (19
0) 16
0 (14
5–23
0)

8 252
4 (317
0) 148
5 (133
8–163
3) 11 7 36
0 (36
5) 18
0 (13
5–48
0)

306
5 (165
5) 251
0 (221
3–414
3) 6 61
7 (104
8) 21
2 (14
8–35
7)

129 (23) 124 (118–152) 12

12

Hyaline vascular (N = 13)

5 170
2 (91
2) 158
0 (114
0–167
0) 6 5 22
2 (24
0) 15
0 (10
0–18
0) 6 29
8 (29
1) 16
0 (14
3–29
8) 6 17
1 (12
9) 12
0 (10
7–15
8)

11
2 (5
5) 10
3 (6
8–15
4)

345
5 (248
0) 328
0 (213
0–399
5) 3 6
8 (11
4) 0
3 (0
2–10
1)

132 (27) 135 (126–140) 6

6

Mixed (N = 7)

252
7 (193
0) 218
0 (95
0–350
0) 7 30
8 (34
0) 13
1 (7
2–45
7) 8 153
1 (46
0) 165
5 (116
0–187
8) 16 12 35
8 (29
7) 31
0 (20
5–37
5) 16 25
2 (14
3) 21
0 (14
5–36
5) 15

117 (30) 112 (94–132) 15

15

Plasmacytic (N = 18)

Cell histology subtypes

Table IV. Biochemical profile (first observation during first year post-study entry).

9
4 (3
6) 9
4 (7
7–11
1)

4

4 23
3 (13
2) 19
0 (14
5–27
8)

3 158
7 (23
9) 166
0 (149
0–172
0) 4 2 18
5 (12
0) 18
5 (14
3–22
8)

389
8 (333
0) 318
0 (161
0–424
0) 2 9
8 (13
2) 9
8 (5
2–14
5)

120 (41) 108 (84–145) 5

5

Unknown (N = 6)

13
7 (12
2) 11
1 (8
6–14
1)

36

37 25
4 (18
1) 17
5 (14
0–31
8)

24 190
5 (186
5) 154
5 (116
0–181
8) 37 26 31
9 (29
2) 21
5 (12
8–37
0)

302
4 (211
9) 249
5 (144
5–410
5) 18 34
8 (64
3) 15
3 (2
0–35
7)

123 (29) 124 (101–144) 38

38

Study total (N = 44)†

0
858

0
126

0
454

0
634

0
946

0
034

0
019

HIV-positive vs. HIV-negative

P-values

0
514

0
603

0
447

0
834

0
775

0
262

0
204

Hyaline vascular vs. plasma-cytic

C. Casper et al

ª 2014 John Wiley & Sons Ltd British Journal of Haematology, 2015, 168, 82–93

147
5 (145
4) 107
0 (71
8–151
8) 129
3 (69
6) 140
0 (97
5–166
5) 174
0 (112
5) 126
0 (88
0–279
0) 126
1 (64
2) 107
0 (85
0–143
3) 174
0 (248
6) 80
5 (66
3–151
8) 139
8 (76
2) 142
0 (86
0–153
0) 158
2 (178
1) 105
0 (70
0–147
0) 108
8 (51
2) 93
0 (84
0–117
8)

HIV, human immunodeficiency virus; ALT, alanine transaminase; AST aspartate transaminasee; IQR, intraquartile range; LDH, lactate dehydrogenase; LFT, liver function test; MCD, multicentric Castleman disease; SD, standard deviation. *Biochemical profile based on first biochemical measures documented in the patient’s medical chart during the first year post-study entry date. Frequency reported is likely underestimated as would not capture information that is missing because of lack of recording. †Three patients had biochemical measurements recorded from the MCD diagnosis date which was prior study period; one patient did not have data collected during the study period and therefore has been excluded from biochemical profile analysis. For the other two patients, the first year of data available in the study period was used for the biochemical profile analysis. Among all study population (N = 58), a total of 44 patients (76%) had data on biochemical measurements documented in their medical chart during the first year post-study entry date. For each sub-population, the corresponding total number of patients with biochemical measurements documented in the medical chart during the first year post-study entry date was used as the denominator.

0
493 0
824 34 3 5 16 10 9 21 4

Alkaline phosphatase, u/l, n Mean (SD) Median (IQR)

HIV-positive vs. HIV-negative Plasmacytic (N = 18) Hyaline vascular (N = 13) HIV-positive (N = 5) Biochemical measures*

HIV status

Table IV. (Continued)

HIV-negative (N = 26)

HIV unknown (N = 13)

Cell histology subtypes

Mixed (N = 7)

Unknown (N = 6)

Study total (N = 44)†

P-values

Hyaline vascular vs. plasma-cytic

Clinical Characteristics of Multicentric Castleman Disease

ª 2014 John Wiley & Sons Ltd British Journal of Haematology, 2015, 168, 82–93

patients attended medical visits frequently and experienced a markedly better survival over the 2 years of observation. To our knowledge, our study is the first to describe the treatment patterns among patients with MCD by HIV status and disease type. Other rare diseases similar to MCD have been studied more thoroughly. One such rare disease, systemic lupus erythematosus (SLE), shares a number of clinical features with MCD. It has an unknown aetiology that involves a dysfunctional autoimmune system with diverse symptoms that vary in frequency and severity. MCD is sometimes mistakenly diagnosed as SLE, and as with MCD, the treatment goal in SLE is disease control through ongoing, often life-long treatment (Wallace & Hahn, 2013; Azevedo et al, 2014). Another relevant rare disease is indolent lymphoma, which is a particularly useful comparison given that MCD can transform into lymphoma. Healthcare service utilization in the US was reported for indolent non-Hodgkin lymphoma (NHL) patients in a study of newly diagnosed patients in the first 2 years after diagnosis (Kutikova et al, 2006), and several studies have reported on the same for SLE patients (Panopalis et al, 2008; Pelletier et al, 2009; Furst et al, 2013; Garris et al, 2013; Kan et al, 2013). Indolent NHL patients had a mean of 16 outpatient radiology procedures and 61 office visits per year (Kutikova et al, 2006), while for SLE, office visits ranged from 10 to 24 per year (Panopalis et al, 2008; Pelletier et al, 2009; Furst et al, 2013; Garris et al, 2013; Kan et al, 2013). These estimates are higher than we observed for MCD patients, who had a mean of five radiological examinations and six physician visits per patient during the first year after diagnosis. The mean number of hospitalizations, however, was higher in our study of MCD patients (4
7 per year) than has been reported for indolent NHL (1
2) (Kutikova et al, 2006) or SLE (0
3–1
6) (Panopalis et al, 2008; Pelletier et al, 2009; Furst et al, 2013; Garris et al, 2013; Kan et al, 2013). MCD patients also tended to have more laboratory procedures (mean: 48 per year) than indolent NHL (28 per year) (Kutikova et al, 2006) or SLE (12–33 per year) (Pelletier et al, 2009; Garris et al, 2013). A study that compared healthcare utilization by SLE patients in three countries found that 22% of all patients had visits to the emergency department over 1 year, and the percentage was significantly higher for the US patients (Sutcliffe et al, 2001); however, only two MCD patients had a record a visit to the emergency department. Variations in reporting may have influenced these differences in healthcare utilization between MCD, indolent lymphoma and SLE. For example, our definition of ‘physician visits’ included outpatient clinic visits, office visits, and emergency department visits, while other studies separated outpatient hospital visits from office visits (Garris et al, 2013; Kan et al, 2013) and most reported emergency department visits separately (Kutikova et al, 2006; Pelletier et al, 2009; Furst et al, 2013; Garris et al, 2013; Kan et al, 2013). In addition, the mean hospitalizations per patient in our study was heavily influenced by a few patients who had a high number 91

C. Casper et al (around 30) of individual hospitalizations on consecutive days, with codes for stem cell transplantation; these may have been reported as outpatient encounters in other studies. We reported the number of individual laboratory measurements per patient, whereas other studies reported the number of ‘laboratory services’ (Pelletier et al, 2009; Garris et al, 2013) and may have grouped certain tests as panels (e.g., complete blood count), counting multiple measurements as one test or service. There are several limitations associated with rare disease data taken from medical records from referral centres, and these should be considered when interpreting our results. As reported previously (Robinson et al, 2014), data were unavailable if patients presented for disease-related care at a treatment facility other than the referral centre (Mayo or FHCRC) that enrolled them into the study. It was also impossible to distinguish between no use of resources and the lack of recording the use of resources in patients’ records. Therefore, our results are likely to be an underestimate of the medical services used by US patients with MCD. In addition, the completeness of records was highly variable and might have been lower for those patients with unknown HIV status. Given that MCD is a rare disease with no established standard of care or treatment guidelines, it was not surprising to see a wide variety of medical therapies provided to MCD patients. While rituximab was the most commonly used treatment, provided to more than half of patients in whom drug therapy was recorded, it has been described only in retrospective cohort studies and small single-arm clinical trials (Marcelin et al, 2003; Bower et al, 2007, 2011) and there are no randomized controlled trials to support its use in this patient population. In addition, apart from case reports (Ide et al, 2006), there is no data for the use of rituximab in HIV-negative MCD patients. Similarly, the next most common medical therapies were corticosteroids or conventional chemotherapy, which also have been evaluated only in retrospective case series in this patient population. Tocilizumab, which was recently reported as effective in achieving sustained remission when combined with other therapies in a paediatric case study (Turcotte et al, 2014), did not appear in the records of the study patients during this period. Given this low level of available evidence, prospective randomized trials of MCD therapies are necessary to help establish the most effective treatment. Based on these data and consistent with a previous report (Dispenzieri et al, 2012), the only distinguishing clinical characteristic in MCD that clearly and adversely affects short-term outcomes is HIV status. Also, there were no clear differences in the treatment patterns of MCD by histology. Differences by histology in gender, age and, possibly, symptoms were observed, however, particularly when the hyaline vascular and plasmacytic subtypes are contrasted. Little is known about the biological basis of the histopathological variants of MCD and, in some cases, mixed hyaline 92

vascular and plasma cell variants may be seen, suggesting that perhaps these variants are not distinct. However, the number of patients with each histological variant was small and further studies are warranted to follow up on our observations. Our study may have been confounded by the types of patients who were referred to the specialty centres that participated in our study or by the small number of patients within each subset of MCD and the limited amount of information that could be retrospectively abstracted from medical records. In addition, data analysis for all patients was restricted to 2 years of follow-up. Additional retrospective and prospective research is needed to more fully determine the external validity of these results. Until that time, these findings may help care providers evaluate their patients by providing a bigger picture or broader understanding of the disease population. Future studies are needed to better understand the natural history and appropriate medical management of this rare lymphoproliferative disorder (Fajgenbaum et al, 2014).

Acknowledgements Financial support: Funding for this research was provided by Janssen Global Services. Financial disclosures: Dana Y. Teltsch, Philip Rotella, and Matthew W. Reynolds are employees of Evidera, a consulting company which received funding from Janssen Global Services for their participation in the research and manuscript. Marie-Pierre Desrosiers was an employee of United BioSource Corporation, which received funding for her participation. Angela Dispenzieri and Thomas Habermann are employees of the Mayo Clinic in Rochester MN and Corey Casper is an employee of the Fred Hutchinson Cancer Research Center in Seattle, WA. Ming Qi is an employee of Janssen Research and Development in Spring House PA and Don Robinson, Jr. is an employee of Janssen Global Services in Raritan, NJ. Author contributions: DR and CC developed the concept for the study. DR, MR and MPD designed the study. MPD conducted the study, and CC, AD and TH were involved with the acquisition of data. MPD, DT, PR analysed the data; DR, CC, MR, AD, MQ, DT, PR and MPD contributed to the interpretation of the results. The manuscript was drafted by DR, CC, DT and PR. MR, AD, MQ, TH and MPD reviewed and critically revised the manuscript. Appreciation: The authors appreciated the initial exploratory work provided by employees of UBC, an Express Scripts Company: Noreen Lordan, Data Management; Bernardo Duran, Custom Software; and Jack Ishak, Biostatistics, Evidera. We would like to thank the study nurses Judy Schramm and Kay Ristow, whose help was invaluable throughout the project. Doctor Joseph Connors of the British Columbia Cancer Agency provided early inspiration, information and support for this study. ª 2014 John Wiley & Sons Ltd British Journal of Haematology, 2015, 168, 82–93

Clinical Characteristics of Multicentric Castleman Disease

References Azevedo, P.C., Murphy, G. & Isenberg, D.A. (2014) Pathology of systemic lupus erythematosus: the challenges ahead. Methods in Molecular Biology, 1134, 1–16. Bower, M., Powles, T., Williams, S., Davis, T.N., Atkins, M., Montoto, S., Orkin, C., Webb, A., Fisher, M., Nelson, M., Gazzard, B., Stebbing, J. & Kelleher, P. (2007) Brief communication: rituximab in HIV-associated multicentric Castleman disease. Annals of Internal Medicine, 147, 836–839. Bower, M., Newsom-Davis, T., Naresh, K., Merchant, S., Lee, B., Gazzard, B., Stebbing, J. & Nelson, M. (2011) Clinical features and outcome in HIV-associated multicentric Castleman’s disease. Journal of Clinical Oncology, 29, 2481–2486. Brandt, S.J., Bodine, D.M., Dunbar, C.E. & Nienhuis, A.W. (1990) Dysregulated interleukin 6 expression produces a syndrome resembling Castleman’s disease in mice. The Journal of Clinical Investigation, 86, 592–599. Casper, C. (2005) The aetiology and management of Castleman disease at 50 years: translating pathophysiology to patient care. British Journal of Haematology, 129, 3–17. Dispenzieri, A. (2008) Castleman disease. In: Rare Hematologic Malignancies (ed. by S. M. Ansell), pp. 293–330. Springer Science & Business Media, LLC, New York. Dispenzieri, A., Armitage, J.O., Loe, M.J., Geyer, S.M., Allred, J., Camoriano, J.K., Menke, D.M., Weisenburger, D.D., Ristow, K., Dogan, A. & Habermann, T.M. (2012) The clinical spectrum of Castleman’s disease. American Journal of Hematology, 87, 997–1002. El-Osta, H.E. & Kurzrock, R. (2011) Castleman’s disease: from basic mechanisms to molecular therapeutics. The Oncologist, 16, 497–511. Fajgenbaum, D.C., van Rhee, F. & Nabel, C.S. (2014) HHV-8-negative, idiopathic multicentric

Castleman disease: novel insights into biology, pathogenesis, and therapy. Blood, 123, 2924– 2933. Furst, D.E., Clarke, A., Fernandes, A.W., Bancroft, T., Gajria, K., Greth, W. & Iorga, S.R. (2013) Resource utilization and direct medical costs in adult systemic lupus erythematosus patients from a commercially insured population. Lupus, 22, 268–278. Garris, C., Jhingran, P., Bass, D., Engel-Nitz, N.M., Riedel, A. & Dennis, G. (2013) Healthcare utilization and cost of systemic lupus erythematosus in a US managed care health plan. Journal of Medical Economics, 16, 667–677. Ide, M., Kawachi, Y., Izumi, Y., Kasagi, K. & Ogino, T. (2006) Long-term remission in HIV-negative patients with multicentric Castleman’s disease using rituximab. European Journal of Haematology, 76, 119–123. Kan, H.J., Song, X., Johnson, B.H., Bechtel, B., O’Sullivan, D. & Molta, C.T. (2013) Healthcare utilization and costs of systemic lupus erythematosus in Medicaid. Biomed Research International, doi: 10.1155/2013/808391. Kutikova, L., Bowman, L., Chang, S., Long, S.R., Arning, M. & Crown, W.H. (2006) Medical costs associated with non-Hodgkin’s lymphoma in the United States during the first two years of treatment. Leukemia & Lymphoma, 47, 1535– 1544. Marcelin, A.G., Aaron, L., Mateus, C., Gyan, E., Gorin, I., Viard, J.P., Calvez, V. & Dupin, N. (2003) Rituximab therapy for HIV-associated Castleman disease. Blood, 102, 2786–2788. Mylona, E.E., Baraboutis, I.G., Lekakis, L.J., Georgiou, O., Papastamopoulos, V. & Skoutelis, A. (2008) Multicentric Castleman’s disease in HIV infection: a systematic review of the literature. AIDS Reviews, 10, 25–35. Panopalis, P., Yazdany, J., Gillis, J.Z., Julian, L., Trupin, L., Hersh, A.O., Criswell, L.A., Katz, P. & Yelin, E. (2008) Health care costs and costs

ª 2014 John Wiley & Sons Ltd British Journal of Haematology, 2015, 168, 82–93

associated with changes in work productivity among persons with systemic lupus erythematosus. Arthritis and Rheumatism, 59, 1788–1795. Pelletier, E.M., Ogale, S., Yu, E., Brunetta, P. & Garg, J. (2009) Economic outcomes in patients diagnosed with systemic lupus erythematosus with versus without nephritis: results from an analysis of data from a US claims database. Clinical Therapeutics, 31, 2653–2664. van Rhee, F., Stone, K., Szmania, S., Barlogie, B. & Singh, Z. (2010) Castleman disease in the 21st century: an update on diagnosis, assessment, and therapy. Clinical Advances in Hematology & Oncology, 8, 486–498. Robinson, D. Jr, Reynolds, M., Casper, C., Dispenzieri, A., Vermeulen, J., Payne, K., Schramm, J., Ristow, K., Desrosiers, M.P., Yeomans, K., Teltsch, D., Swain, R., Habermann, T.M., Rotella, P. & Van de Velde, H. (2014) Clinical epidemiology and treatment patterns of patients with multicentric Castleman disease: results from two US treatment centres. British Journal of Haematology, 165, 39–48. Stebbing, J., Pantanowitz, L., Dayyani, F., Sullivan, R.J., Bower, M. & Dezube, B.J. (2008) HIV-associated multicentric Castleman’s disease. American Journal of Hematology, 83, 498–503. Sutcliffe, N., Clarke, A.E., Taylor, R., Frost, C. & Isenberg, D.A. (2001) Total costs and predictors of costs in patients with systemic lupus erythematosus. Rheumatology, 40, 37–47. Turcotte, L.M., Correll, C.K., Reed, R.C. & Moertel, C.L. (2014) Sustained remission of severe Multicentric Castleman disease following multiagent chemotherapy and tocilizumab maintenance. Pediatric Blood & Cancer, 61, 737–739. Wallace, E.J. & Hahn, B.H. (2013) Dubois’ Lupus Erythematosus and Related Syndromes. Lippincott Williams & Wilkins, Philadelphia, PA. Waterston, A. & Bower, M. (2004) Fifty years of multicentric Castleman’s disease. Acta Oncologica, 43, 698–704.

93