A case of multicentric Castleman’s disease in HIV infection with the rare complication of acquired angioedema

Introduction

Multicentric Castleman’s disease (MCD), a polyclonal lymphoproliferative disorder of unknown aetiology, is well-recognised as presenting in association with HIV infection. We report the first case of HIV-related MCD causing the rare complication of acquired angioedema (AAE), and also, unusually, first manifesting in the context of an immune restoration inflammatory syndrome (IRIS).

Case report

Our patient, a 37-year-old African woman, was first diagnosed with HIV in 2002. She initiated highly active antiretroviral therapy (HAART) in March 2004, with efavirenz, lamivudine and tenofovir, for symptoms of recurrent genital herpes. She was otherwise well at the time her CD4 count was 438 cells/mm³ and viral load (VL) 102,000 copies/ml; other routine blood results were unremarkable.

Three weeks after HAART start, the patient was admitted to hospital with her haemoglobin (Hb) having fallen from 127 to 43 gm/l. Bloods tests (including a direct Coombs test) suggested immune-haemolysis; a glucose 6-phosphate dehydrogenase (G6PD) test was negative. Multiple blood, stool, urine and sputum culture investigations, and serological tests (including for Epstein-Barr virus, cytomegalovirus and parvovirus) were negative for infection. Radiological investigations identified mild splenomegaly, but no indication of lymphoma. Bone marrow aspirate and trephine revealed only a hyperplastic marrow.

Our patient was initially treated with blood transfusions and a reducing course of oral steroids, for a diagnosis of presumed idiopathic haemolysis. She continued on HAART, and by May 2004 her VL was undetectable (<40 copies/ml). However, with gradual steroid withdrawal, the anaemia recurred. The patient therefore underwent splenectomy in July 2004, as definitive management of the anaemia, with resultant stabilisation of Hb. The spleen pathology report noted ‘mild hyperplasia of the white pulp and a congested red pulp’, but no evidence of infection or lymphoma.

In August 2004 (48 h after stopping steroids), the patient developed her first episode of transient facial swelling and itch, which settled spontaneously. Over the next few years she remained stable on HAART, with consistently undetectable VLs and CD4 counts of >500. However, by early 2008, she experienced recurrence of episodic facial swelling and itch, worst around her eyes and lower jaw region. There was no associated stridor. There were no identified precipitating factors, including medications. The patient was reviewed by both dermatology and haematology and, on the basis of classical clinical symptoms and signs (in photos provided by the patient), given a diagnosis of idiopathic acquired angioedema (AAE). However, a complement screen undertaken between episodes was normal: C3 1.6 (0.73–1.4), C4 0.25 (0.12–0.3) and C1-esterase inhibitor 0.29 (0.22–0.36); this was not repeated during a symptomatic episode. Other investigations undertaken included unremarkable CT skull, abdomen and pelvis. Amyloidosis and myeloma screens were negative, and anti-nuclear factor (ANF) levels were normal. The angioedema episodes gradually increased in frequency and severity over the next few years. Antihistamine treatment was of minimal benefit, and episodes usually settled spontaneously within 48 h.

In August 2012, the patient developed leg swelling, and radiological investigations identified iliac vein compression, secondary to widespread pelvic (and also co-existent abdominal and chest) lymphadenopathy. An axillary lymph node biopsy now revealed MCD. The peripheral blood human herpes virus-8 (HHV8) load was 87,000 copies/ml.

In light of these new developments, pathologists re-examined the 2004 splenectomy specimens. The original report was revised; appearances were now identified as indicative of HIV-related MCD, and also stained positive for HHV-8.

The patient’s MCD was treated with a 4-week course of rituximab with good response: there was significant radiological improvement of the widespread lymphadenopathy and peripheral HHV-8 levels fell to undetectable. Interestingly, the recurrent AAE episodes also resolved completely following MCD treatment.

Discussion

Our case raises a number of interesting points. Firstly, considering timing of symptom onset and the revised pathology report, it appears unquestionable that our patient’s MCD initially arose in 2004 – in the context of IRIS, following HAART commencement. MCD is a polyclonal, usually B-cell type, lymphoproliferative disorder of unknown aetiology, almost invariably associated with the presence of HHV8 infection in HIV patients. ¹ While HHV8 is associated with other lymphoproliferative disease processes in HIV-positive patients, specifically Kaposi’s sarcoma (KS) and primary effusion lymphoma (PEL), ² blood HHV8 levels are comparatively higher in patients with MCD, ³ and in keeping with the level found in our patient.

There are few other published reports describing the occurrence of MCD on a background of IRIS.^4,5 Our patient had an unusually high CD4 count ⁶ but also, as more conventionally observed in IRIS, a rapid decrease in VL post-HAART initiation. ⁷

The second interesting point in our case is that, despite the MCD diagnosis initially being missed, the patient serendipitously received an accepted (albeit limited) therapeutic intervention – namely, splenectomy, ⁸ which served to control her disease for many years. MCD recurrence is common following management with splenectomy alone, ⁹ but our patient took significantly longer to relapse than in previous case reports. This is likely because her initial MCD, somewhat unusually, was restricted to a great extent to the spleen (radiological investigations in 2004 failed to reveal any lymphadenopathy). Unfortunately, there were inadequate stored serum samples remaining from the time of the initial presentation for retrospective testing of HHV8 levels, for further confirmation of disease bulk.

Finally and most intriguingly, our patients’ first episode of AAE onset during initial MCD occurrence in 2004, and (despite a pattern of frequent and increasing recurrence in 2011–2012) resolved completely post-MCD therapy. AAE (in comparison to the more familiar hereditary angioedema) onsets in adulthood, in individuals without a family history of the same. Patients commonly have underlying B-cell lymphoproliferative disorders (or occasionally autoimmune conditions), causing disruption of their C1-esterase inhibitor (C1-INH) levels ¹⁰ and/or function. ¹¹ Dysregulation of C1-INH activity results in activation of the classic complement pathway, with consequent angioedema. As was the case with our patient, control of the causal condition usually results in partial or complete resolution of the symptoms of AAE. ¹⁰

It is unusual that our patient’s complement screen was normal, despite classical profile, symptoms and signs of AAE. In AAE associated with underlying lymphoproliferative malignancy (traditionally known as AAE type-1), it was initially considered that the principle defect was an absolute reduction of C1-INH levels, secondary to destruction by malignant cells or consumption in immune-complex formation. More recently however, it has been recognised that even in some cases of lymphoproliferative malignancy-associated AAE, the impairment may be in C1-INH function, secondary to production of autoantibodies by malignant B-cell clones. ¹² It is noteworthy that our patient also developed an immune-haemolysis early on in the course of her MCD, likely by a similar mechanism of autoantibody production. Unfortunately, C1-INH functional activity level was never checked in our patient.

To our knowledge, none of the previously reported cases of AAE have been linked with MCD as the contributory underlying disease. In our patient’s case, considering the factors: that MCD is a recognised lymphoproliferative disorder, the association between her AAE and MCD symptom onset, and complete remission of AAE symptoms post MCD treatment, we suggest this is the first report of MCD-associated AAE. We recommend therefore that MCD is considered in the list of underlying causes, particularly in HIV-positive individuals, when investigating new onset AAE.