Epidemiology of Castleman disease associated with AA amyloidosis: description of 2 new cases and literature review

Introduction: HHV8-negative Castleman disease (CD) is classified as hyaline vascular (HV) type, or mixed or plasma cell (PC) types. It may present as multicentric CD (MCD) or unicentric CD (UCD). CD is a rare cause of AA amyloidosis (AAA). We aimed to report the main features of CD with secondary AAA through a description of new cases and a systematic literature review. Patients and methods: New cases were identified from the French National Reference Center for AAA. A systematic literature review was performed to identify HHV8-negative CD cases associated with AAA. Results: Thirty-seven patients were analysed, consisting of two new cases and 35 from literature. Twenty-three had UCD and 14 had MCD. PC was the main histologic subtype (n = 25; 68%) in both UCD and MCD patients. Surgical excision of UCD was performed in 21 patients (91%) with a favourable outcome, except for four patients (19%). Clinical and biologic remission was achieved in six patients with MCD (43%), all of whom were treated with anti-interleukin-6 (IL-6) therapy. Conclusions: AAA is a rare complication of CD, namely idiopathic MCD and UCD presenting with the PC histologic subtype. Surgical excision of UCD should be the first-line treatment whenever possible, while anti-IL-6 therapies seem effective for MCD.

Keywords: AA-amyloidosis; Castleman disease; unicentric; multicentric; anti-interleukin-6 therapy

Introduction

Castleman disease (CD) is a rare benign lymphoproliferative disorder, histologically characterized by a distinctive angiofollicular hyperplasia without (hyaline vascular (HV) subtype) or with plasma cell interfollicular infiltration (mixed or plasma cell (PC) type). It is divided into two distinct pathological entities with their own specific disease course and management: unicentric CD (UCD) and multicentric CD (MCD). UCD presents with the HV subtype in 70–75% of cases. It mainly affects adults in their 4th decade, who exhibit no symptom or symptoms related to compression of adjacent structures by enlarged lymph nodes with usually low inflammation. Increased incidence of paraneoplastic pemphigus has also been reported [[1]]. Surgery is considered as the standard of care for UCD. On the other hand, MCD is divided into two categories based on HHV8 status: HHV8-associated MCD, which especially – but not only – affects HIV-infected patients, and idiopathic MCD (iMCD) [[3]]. Both of these entities present in a more symptomatic fashion, with major constitutional symptoms and increased inflammatory markers [[1]], and in association with the PC or mixed subtype. Pathophysiology of CD remains uncertain. There is a well-established role of interleukin-6 (IL-6) in MCD pathogenesis [[4]] but the origins of IL-6 dysregulation are still unknown in iMCD. Conversely, in HHV8-associated MCD, HHV8 triggers the production of both viral and host IL-6 [[7]]. As a consequence, anti-IL-6 therapies have been evaluated in iMCD since 2000, with both clinical and biological efficacy [[9]].

Both MCD and UCD are classical but rare causes of AA amyloidosis (AAA) [[15]]. This link is easy to understand for MCD, of which the inflammatory profile is partially mediated by increased serum IL-6 that induces predominantly hepatic production of acute phase reactants, including serum amyloid A (SAA) protein. Regarding UCD, the cause of AAA is less clear since this entity leads to only very little inflammation.

Our objectives were to analyse the main features of patients displaying AAA complicating HHV8-negative CD (UCD and iMCD), both diseases, and to describe treatment management and response in this rare situation by performing a systematic review of the literature and description of two new cases.

Patients and methods

New cases identified through French Reference Centers for AAA and CD

All patients with AAA secondary to CD were retrieved from the database of the French Reference Center for Systemic AAA (CEREMAIA) as well as the French National Center for Castleman Disease. An author (AF) manually reviewed the medical records. Patients with UCD or MCD were included if CD and AAA were both histologically confirmed:

• For CD: typical findings consistent with one of the CD histologic subtypes: HV, PC or mixed.
• For AAA: Congo red staining of amyloid with green birefringence under polarized light, and immunohistochemical staining for SAA.

Patients were not included if they had HHV8-associated MCD (positive HHV8 polymerase chain reaction (PCR) on peripheral blood or positive HHV8 immunohistochemistry for latency-associated nuclear antigen (LANA) on tissue sample), positive HIV status or another potential AAA-causing disease (inflammatory rheumatism, chronic inflammatory disease, inflammatory bowel disease, chronic or recurrent infection, monogenic autoinflammatory disease, cancer and immune deficiency).

We collected clinical, biological, histological and demographic data. Information extracted from medical records included ages at onset and diagnosis of both CD and AAA, clinical form and histologic subtype of CD, organs affected by AAA, selected treatment and therapeutic response.

Systematic literature review

The PRISMA guidelines were applied in order to perform the literature review. We searched for original articles, case reports and letters to editor reporting cases of AAA associated with CD in two electronic databases: PubMed and Web of Science. The search was restricted to articles published between 01 January 1996 and 01 January 2018. We chose 01 January 1996 as the earliest date in order to limit the impact of missing data regarding HIV or HHV8 status, since the implication of HHV8 in MCD has first been reported in 1995 [[16]]. The following keywords were used (as free text and MeSH terms): Castleman disease, giant lymph node hyperplasia, amyloidosis, kidney diseases, nephrotic syndrome and heart diseases. Relevant articles were first selected based on their titles. Abstracts and full-texts of selected abstracts were subsequently reviewed. Only articles with full text available in English or French languages were considered. All cases of UCD and MCD with secondary amyloidosis were included if the diagnosis of AA (or secondary) amyloidosis and the clinical form of CD were both clearly mentioned in the article. Exclusion criteria included positive HHV8 testing, positive HIV status or presence of another potential AAA-causing disease. Relevant clinical, biological and histopathological data were extracted from selected articles using a pre-established electronic survey (see Supplementary Table S1).

Results

Description of the new cases identified through French Reference Centers for AAA and CD

The first patient was 70-year-old at the time of diagnosis (Table 1). She was diagnosed with abdominal UCD of PC subtype, revealed by acute kidney failure due to renal AAA. She presented a persistent unexplained inflammatory syndrome for the past 8 years. Despite complete excision of UCD and normalization of inflammatory markers, she rapidly progressed to end-stage renal disease and died in 2005, one year after initial diagnosis.

Table 1. French cases of Castleman disease associated with AA amyloidosis.

Patient	Castleman disease	AA amyloidosis	Treatment	Outcome
Gender	Comorbidities	Clinical form/localization	Histologic subtype	Age at onset (years)	Age at diagnosis (years)	First symptom	Age at diagnosis (years)	CRP at diagnosis (mg/L)	Biopsy site	Organ involvement	Surgery	Other treatments	Therapeutic response	Vital status
F	Hysterectomy, smoking	UCD/abdominal	Plasma cell	62	70	Fever and biological inflammatory syndrome	70	100	Renal	Renal: progression to ESRD	Complete excision (70 yo)	None	Remission of fever and inflammatory syndrome. No improvement of renal function	Death one year after diagnosis
M	None	UCD/mediastinal	Mixed	Missing data	29	Isolated biological inflammatory syndrome	32	<3	Renal	Renal: NS and progression to ESRD	Complete excision (29 yo)	Kidney transplantation	Remission of inflammatory syndrome	Alive, no recurrence of AAA in renal graft

1 UCD: unicentric Castleman disease; AAA: AA amyloidosis; CRP: C-reactive protein; NS: nephrotic syndrome; ESRD: end-stage renal disease; yo: years old.

The second patient was a 29-year-old man diagnosed with mediastinal UCD of the mixed subtype (Table 1). Complete excision was performed, allowing resolution of compressive symptoms and partial correction of the inflammatory syndrome. Three years later, he developed a nephrotic syndrome. Inflammatory markers were normal. Kidney biopsy showed amyloidosis, with immunohistochemistry negative for Ig light chains and positive for SAA (Supplementary Figure S1). He progressed to end-stage renal disease and had a renal transplantation. So far, AAA has not recurred in the graft.

Literature review

One hundred and six original articles, case reports and letters to the editor corresponding to our research criteria were identified. Sixty-six were selected based on their titles, 49 based on their abstracts and 29 based on their full-texts. The selected full-text papers reported 35 patients with CD and AAA (see Supplementary Figure S2). Main characteristics are available in Supplementary Table S2. With the addition of the two patients followed up in French Reference Centers, a total of 37 patients were included in the study: 19 females and 18 males.

Table 2. Characteristic of the 37 CD cases associated with AA amyloidosis.

	UCD	MCD	Total
Total	23	14	37
Histologic subtype
PC	13	12	25
HV	4	1	5
Mixed	6	1	7
Median age at CD onset – years (Q1–Q3)	30 (21.0–44.5)	45.0 (32.5–49.0)	35.0 (23.8–48.0)
Median age at CD diagnosis – years (Q1–Q3)	39.0 (26.0–48.0)	48.0 (34.5–55.0)	41.0 (28.5–51.0)
AAA leading to CD diagnosis – n (%)	17 (73.9)	5 (35.7)	24 (64.9)
Median CRP at AAA diagnosis – mg/L (Q1–Q3)	106 (50–134)	49 (22–101)	89 (39–126)
AAA organ involvement
Kidney (%)	17 (73.9)	10 (71.4)	27 (73.0)
NS (%)	14 (82.4)	8 (80.0)	22 (81.5)
ESRD (%)	6 (35.3)	3 (30.0)	9 (33.3)
Digestive tract (%)	5 (21.7)	5 (35.7)	10 (27.0)
Liver (%)	6 (26.1)	0	6 (16.2)
PNS (%)	1 (4.3)	0	1 (2.7)
Heart (%)	0	2 (14.3)	2 (5.4)

2 CD: Castleman disease; UCD: unicentric Castleman disease; MCD: multicentric Castleman disease; PC: plasma cell; HV: hyaline vascular; NR: not reported; AAA: AA amyloidosis; CRP: C-reactive protein; ESRD: end-stage renal disease; PNS: peripheral nervous system.

Overall

CD diagnosis was made at a median age of 41 years (interquartile range Q1–Q3: 28.5–51). At CD diagnosis, systemic symptoms or persistently elevated inflammatory markers had been present for a median of 2 years. Sixty-five percent of CD cases were diagnosed following the exploration of unexplained AAA. There were 23 UCD and 14 MCD cases. Histological subtypes included 25 PC, five HV and seven mixed. No paraneoplastic pemphigus was diagnosed among UCD patients. AAA involved the kidney in 73% of patients, with a nephrotic syndrome in 82% of them, and the digestive tract in 27% of patients. Table 2 summarizes the characteristics of UCD and MCD cases associated with AAA.

Unicentric Castleman disease

Among the 23 UCD cases, 83% presented histologically with either a PC (57%) or mixed (26%) subtype, while the HV subtype accounted for only 17% of cases. Although systemic symptoms, namely fatigue (n = 7), weight loss (n = 4) and fever (n = 5), or persistently elevated inflammatory markers had been present in 16 patients for a median of 3 years before diagnosis, most UCD cases (74%) were diagnosed concurrently or following the diagnosis of AAA. The most common clinical presentation of UCD involved an isolated mass located in the abdomen or mediastinum in respectively 74% and 17% of patients. Surgical excision of UCD was performed in 21 patients. A favourable outcome was observed in most patients, with the exception of four (19%) whose inflammatory syndrome persisted and lead to progression of AAA. However, complete excision of UCD was confirmed in only two of those four patients, with insufficient data available for the remaining two patients.

Table 3 compares the clinical characteristics of the 23 UCD cases with AAA to the 71 UCD cases without AAA followed by the French National Reference Center for CD. Briefly, UCD complicated by AAA tends to be located in the abdomen, of PC subtype, and with more frequent constitutional symptoms and elevated inflammatory markers when compared to non-AAA associated UCD.

Table 3. Comparison between UCDs without AAA and UCDs with AAA.

	UCD without AAA	UCD with AAA
N = 71	N = 23
Sex, male/female	32/39	10/13
Age (median), years	35	39
Localization
Mediastinum (%)	21 (30)	4 (17)
Abdominal (%)	19 (27)	17 (74)
Cervical (%)	21 (30)	0
Axillary (%)	3 (4)	0
Inguinal (%)	3 (4)	0
Extranodal (%)	4 (5)	0
Not reported (%)	0	2 (9)
Histologic subtype
PC	16 (22)	13 (57)
Mixed	6 (26)
HV	55 (78)	4 (17)
Diagnostic delay (median), months	NR	36.0
Fever (%)	3 (4)	5 (22)
Fatigue (%)	NR	7 (30)
Weight loss (%)	NR	4 (17)
CRP (median), mg/L	2	106
Haemoglobin (median), g/dL	13.6	9.1

3 PC: plasma cell; HV: hyaline vascular; NR: not reported; CRP: C-reactive protein.

Multicentric Castleman disease

Of the 14 iMCD cases, 93% presented histologically with either a PC (86%) or mixed (7%) subtype, while the HV subtype accounted for only 7% of cases. HIV status was negative and reported in 71% of cases. At iMCD diagnosis, systemic symptoms or persistently elevated inflammatory markers had been evolving for a median of 2 years. Five were diagnosed with or after AAA, four before AAA and date of diagnosis was not reported in five cases. Various treatments were administered, including colchicine, corticosteroids and the following immunosuppressive therapies: azathioprine, cyclophosphamide, anti-CD20 agents, anti-IL-6 agents and haematology-derived chemotherapy regimens. The only cases reporting remission of both symptoms and biological inflammation involved the six patients treated with anti-IL-6 therapy (tocilizumab). In the remaining eight patients, only four had improvement of symptoms and decrease in inflammatory markers resulting in stabilization of AAA.

Comparison between iMCD patients with AAA and iMCD patients without AAA from a previously reported series [[1]] is presented in Table 4. Briefly, PC subtype seems more common in iMCD complicated by AAA seems with a longer median diagnosis delay.

Table 4. Comparison between iMCDs without AAA and iMCDs with AAA.

	iMCD without AAA	iMCD with AAA
N = 27a	N = 14
Sex, male/female	20/7	8/6
Age (median), years	47	48
Histologic subtype
PC (%)	21 (78)	12 (86)
Mixed (%)	1 (7)
HV (%)	6 (22)	1 (7)
Diagnostic delay (median), months	3	24
Splenomegaly (%)	10 (37)	4 (29)
Haemoglobin (median), g/dL	10.4	8.3
CRP (median), mg/L	112	71
Gammaglobulin, g/L	20	8.6

• 4 PC: plasma cell; HV: hyaline vascular; CRP: C-reactive protein.
• 5 Data from the iMCD patients series previously reported by Oksenhendler et al. [[1]].

Discussion

This review analysed 37 CD cases associated with AAA, including 35 cases previously reported in the literature, and two new cases followed up in French Reference Centres. Despite limitations intrinsic to the study's design, including selective reporting of cases showing treatment efficacy and substantial missing data, several hypotheses and conclusions can be formulated.

We found a preponderance of the PC histological subtype in both MCD and UCD associated with AAA, which could be linked to the strong IL-6 pro-inflammatory pathway already described in association with this subtype [[18]].

Although UCD usually causes little or no increase in inflammatory markers, we showed that it was the most common clinical form complicated by AAA. However, when compared to UCD in general, we found that those associated with AAA often presented with prominent constitutional symptoms and an inflammatory syndrome. It is consistent with the observed preponderance of PC or mixed histologic subtype, features usually characterizing iMCD. Therefore, we hypothesize that peculiar UCDs characterized by PC histology and IL6 overproduction, maybe closer to iMCD than to classical UCD, can cause AAA. Furthermore, these PC-UCD cases display only mild symptoms allowing the inflammatory disease to go unnoticed for a prolonged period and, subsequently, leading to the development of AAA. Their preferential localization to the abdomen had previously been reported [[1]] and may contribute to a delayed diagnosis.

Regarding MCD, we designed our inclusion and exclusion criteria to select mostly iMCD. The relatively low number of iMCDs among CD associated with AAA may be explained by a frequent aggressive course of iMCD requiring treatment before SAA deposition. AAA occurrence could therefore be related to iMCD resistance to initial management or delayed diagnosis, thus leading to prolonged inflammation and amyloidosis. We may also hypothesize in some patients a specific autoinflammatory condition leading to AAA and iMCD. A last hypothesis is that they could occur in individuals with a genetic predisposition to AAA such as the SAA1 polymorphisms previously described in rheumatoid arthritis patients [[20]].

Regarding treatments, our data suggest that surgical excision is an effective treatment for UCD with AAA. However, the substantial failure rate (19%) of this approach, which may be due in part to incomplete resection, raises the question of anti-IL-6 therapy in the management of UCD with AAA. Although it has not been tested yet, a combined surgical and medical approach with anti-IL-6 agents may be beneficial in selected patient; however, more data are needed to support this hypothesis. Furthermore, no patient underwent radiotherapy, which is an alternative to surgery for UCD in general [[1]]. As a result, the role of radiotherapy in UCD complicated by AAA remains uncertain, especially as radiotherapy induces complete UCD remission in no more than 50% of cases [[22]]. Anyway, the eventuality of an MCD mistaken for a UCD needs to be carefully considered before any therapeutic escalation. The efficacy of tocilizumab already documented in regular iMCD is reinforced by this review, with perhaps an even better response in the context of AAA. It might however be influenced by the limited number of patients and reporting bias. Immuno-chemotherapy could therefore be relegated to second-line therapy, in case of failure of anti-IL-6 agents. Considering these results and known effective treatments for CD in general [[1], [23]], we propose an algorithm for the management of CD with AAA (Figure 1).

Graph: Figure 1. Proposed algorithm for the management of CD with AAA. HIV: human immunodeficiency virus; HHV8: human herpes virus 8; UCD: unicentric Castleman disease; iMCD: idiopathic multicentric Castleman disease; AAA: AA amyloidosis; anti-IL6: anti-interleukin 6 biotherapy; CS: corticosteroids; RTX: rituximab.

As a conclusion, one can state that CD causing AAA tends to be localized (UCD) and of PC or mixed histological subtypes. Although rare, CD should be considered when exploring AAA, as our review showed that CD is most frequently diagnosed in this setting. UCD is a cause of AAA that can be cured with surgical excision alone. Conversely, iMCD associated with AAA is best managed with anti-IL-6 therapy.

Disclosure statement

The authors report no conflict of interest.

Abbreviations

• AAA

• AA amyloidosis

• CD

• Castleman disease

• HHV8

• human herpes virus 8

• HIV

• human immunodeficiency virus

• HV

• hyaline vascular subtype

• IL-6

• interleukin-6

• iMCD

• idiopathic MCD
• MCD: LANA
• latency-associated nuclear antigen; multicentric Castleman disease

• PC

• plasma cell subtype

• PCR

• polymerase chain reaction

• SAA

• serum amyloid A

• UCD

• unicentric Castleman disease

1 Supplemental data for this article can be accessed here. Oksenhendler E, Boutboul D, Fajgenbaum D, et al. The full spectrum of Castleman disease: 273 patients studied over 20 years. Br J Haematol. 2018; 180 : 206 – 216. 2 Yu L, Tu M, Cortes J, et al. Clinical and pathological characteristics of HIV- and HHV-8-negative Castleman disease. Blood. 2017; 129 : 1658 – 1668. 3 Fajgenbaum DC, Shilling D. Castleman disease pathogenesis. Hematol Oncol Clin North Am. 2018; 32 : 11 – 21. 4 Yoshizaki K, Matsuda T, Nishimoto N, et al. Pathogenic significance of interleukin-6 (IL-6/BSF-2) in Castleman's disease. Blood. 1989; 74 : 1360 – 1367. 5 van Gameren MM, Willemse PH, Mulder NH, et al. Effects of recombinant human interleukin-6 in cancer patients: a phase I–II study. Blood. 1994; 84 : 1434 – 1441. 6 van Rhee F, Wong RS, Munshi N, et al. Siltuximab for multicentric Castleman's disease: a randomised, double-blind, placebo-controlled trial. Lancet Oncol. 2014; 15 : 966 – 974. 7 Neipel F, Albrecht JC, Ensser A, et al. Human herpesvirus 8 encodes a homolog of interleukin-6. J Virol. 1997; 71 : 839 – 842. 8 Parravicini C, Corbellino M, Paulli M, et al. Expression of a virus-derived cytokine, KSHV vIL-6, in HIV-seronegative Castleman's disease. Am J Pathol. 1997; 151 : 1517 – 1522. 9 Nishimoto N, Sasai M, Shima Y, et al. Improvement in Castleman's disease by humanized anti-interleukin-6 receptor antibody therapy. Blood. 2000; 95 : 56 – 61. Nishimoto N, Kanakura Y, Aozasa K, et al. Humanized anti-interleukin-6 receptor antibody treatment of multicentric Castleman disease. Blood. 2005; 106 : 2627 – 2632. Song S-NJ, Tomosugi N, Kawabata H, et al. Down-regulation of hepcidin resulting from long-term treatment with an anti-IL-6 receptor antibody (tocilizumab) improves anemia of inflammation in multicentric Castleman disease. Blood. 2010; 116 : 3627 – 3634. van Rhee F, Fayad L, Voorhees P, et al. Siltuximab, a novel anti-interleukin-6 monoclonal antibody, for Castleman's disease. J Clin Oncol. 2010; 28 : 3701 – 3708. Matsuyama M, Suzuki T, Tsuboi H, et al. Anti-interleukin-6 receptor antibody (tocilizumab) treatment of multicentric Castleman's disease. Intern Med. 2007; 46 : 771 – 774. Kurzrock R, Voorhees PM, Casper C, et al. A phase I, open-label study of siltuximab, an anti-IL-6 monoclonal antibody, in patients with B-cell non-Hodgkin lymphoma, multiple myeloma, or Castleman disease. Clin Cancer Res. 2013; 19 : 3659 – 3670. Shimojima Y, Takei Y, Tazawa K, et al. Histopathological regression of systemic AA amyloidosis after surgical treatment of a localized Castleman's disease. Amyloid. 2006; 13 : 184 – 186. Soulier J, Grollet L, Oksenhendler E, et al. Kaposi's sarcoma-associated herpesvirus-like DNA sequences in multicentric Castleman's disease. Blood. 1995; 86 : 1276 – 1280. Dupin N, Gorin I, Deleuze J, et al. Herpes-like DNA sequences, AIDS-related tumors, and Castleman's disease. N Engl J Med. 1995; 333 : 798; author reply 798–799. Veldhuis GJ, van der Leest AH, de Wolf JT, et al. A case of localized Castleman's disease with systemic involvement: treatment and pathogenetic aspects. Ann Hematol. 1996; 73 : 47 – 50. Vinzio S, Ciarloni L, Schlienger J-L, et al. Isolated microcytic anemia disclosing a unicentric Castleman disease: the interleukin-6/hepcidin pathway? Eur J Intern Med. 2008; 19 : 367 – 369. van der Hilst JCH, Yamada T, Op den Camp HJM, et al. Increased susceptibility of serum amyloid A 1.1 to degradation by MMP-1: potential explanation for higher risk of type AA amyloidosis. Rheumatology. 2008; 47 : 1651 – 1654. Obici L, Raimondi S, Lavatelli F, et al. Susceptibility to AA amyloidosis in rheumatic diseases: a critical overview. Arthritis Rheum. 2009; 61 : 1435 – 1440. Noh OK, Lee S-W, Lee JW, et al. Cases report of unicentric Castleman's disease: revisit of radiotherapy role. Radiat Oncol J. 2013; 31 : 48 – 54. Rhee F, van Voorhees P, Dispenzieri A, et al. International, evidence-based consensus treatment guidelines for idiopathic multicentric Castleman disease. Blood. 2018; 132 : 2115 – 2124.

By Antoine Fayand; David Boutboul; Lionel Galicier; Jean-Emmanuel Kahn; David Buob; Jean-Jacques Boffa; Alexandre Cez; Eric Oksenhendler; Gilles Grateau; Stéphanie Ducharme-Bénard and Sophie Georgin-Lavialle

Reported by Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author