Acquired angioedema: an unusual presentation of haematological malignancy

Abstract

A previously healthy man in his 60s being worked up for splenomegaly presented to the emergency department with recurrent episodes of angioedema. Each episode was attributed to a precipitating cause, and consequently, the predisposing C1 esterase inhibitor (C1-INH) deficiency remained undiagnosed until the third presentation. The aetiology of acquired C1-INH deficiency would be primarily obscure and require further investigations to identify. A clonal B cell population was finally isolated by flow cytometry after multiple repeat marrow samples, and a diagnosis of splenic marginal zone lymphoma was subsequently reached. Response to single-agent rituximab was observed with resolution of splenomegaly, disappearance of the antibody and restoration of C1-INH levels.

Background

Angioedema refers to the histamine or bradykinin-induced rapid onset oedema, seen most obviously in the face. In this case, we highlight the relationship between splenic marginal zone lymphoma (SMZL) and recurrent attacks of angioedema occurring in the absence of an obvious trigger. Angioedema is subdivided into hereditary and acquired causes, with C1 esterase inhibitor (C1-INH) deficiency falling into both categories. Thus, this case emphasises both the importance of considering an acquired cause of C1-INH deficiency for unexplained angioedema, as well as the need for a high index of suspicion for a B cell lymphoproliferative disorder, even when bone marrow aspirate and trephine are seemingly normal.

Case presentation

A previously well man in his 60s was referred to haematology outpatients because of clinical splenomegaly identified during a routine check-up by the family doctor. While under investigation, the patient presented to the emergency department with periorbital oedema and mild upper lip swelling shortly after starting esomeprazole for gastro-oesophageal reflux. There was no history of ACE inhibitor use, no history of allergy and no relevant family history. Consequently, he was given intravenous hydrocortisone, started on oral antihistamines and discharged home on a 2-week tailing course of oral prednisolone. Around a month after presenting to emergency, a bone marrow aspirate together with a trephine was obtained as part of the work up for his splenomegaly. The sample was of adequate quality but failed to identify a clonal B cell population on flow cytometry of the aspirate and immunohistochemistry (IHC) staining of the trephine.

Three months later, he re-presented with isolated tongue swelling, which woke him up a few hours after he had consumed fish for dinner. An indirect laryngoscopy at the time did not reveal any laryngeal oedema, and the patient had no breathing difficulties. His symptoms resolved after intravenous hydrocortisone and chlorphenamine, and he was again discharged home. Two weeks later, the patient presented with isolated lip swelling and was again given intravenous hydrocortisone and discharged home, only for him to come back a few hours later with difficulty in breathing for which he was admitted for further investigation and monitoring. An otolaryngological examination revealed uvular, palatine, tongue and epiglottic oedema. A C1-INH level was checked together with tryptase and IgE levels among other tests shown in table 1. The patient was discharged again on an oral antihistamine and an epinephrine rescue pen.

Table 1.

Investigations performed prior to and following treatment

Investigations	At time of diagnosis	Reference range	Four months post-rituximab induction
White cell count	7.14×109/L	4.30×109–11.40×109/L	5.97×109/L
Lymphocytes	1.94×109/L	1.30×109–3.60×109/L	1.50×109/L
Haemoglobin	136 g/L	12.0–15.5 g/dL	149 g/L
Platelets	156×109/L	132×109–349×109/L	245×109/L
IgA	1.52 g/L	0.70–4.00 g/L	1.09 g/L
IgG	8.08 g/L	7.01–16.0 g/L	8.38 g/L
IgM	1.10 g/L	0.40–2.30 g/L	0.97 g/L
Serum protein electrophoresis/free light chain analysis	No detectable monoclonal band
C1 esterase inhibitor antigen	<0.05 g/L	0.21–0.39 g/L	0.27 g/L
C1 esterase inhibitor activity	<25%	70%–130%	88%
C1 esterase inhibitor antibody	IgM positive		IgM negative
C3 complement	733 mg/L	900–1800 mg/L
C4 complement	<20 mg/L	100–400 mg/L	234 mg/L
C1q complement	23 mg/L	127–201 mg/L	131 mg/L
Tryptase	3.6 µg/L	<11.4 µg/L
IgE	47 IU/mL	≤100 IU/mL
Hepatitis C antibody	Negative
Hepatitis B core antibody	Negative
Hepatitis B surface antigen	Negative
HIV p24Ag/T antibody	Negative
Antinuclear antibody	Negative
Antidouble stranded DNA	Negative
Flow cytometry	Mature B cell population: Positive: CD19+++, CD20+++, Kappa+, CD79b+, CD11c+, CD81+, CD31dim+, CD95++, CD49d++. CD22+ and CD185+++ Negative: CD5, CD38, Lambda, CD3, CD4, CD10, CD200, CD43, CD23, CD305, CD103 and IgM		No detectable clonal B cell population.

The diagnosis of an acquired C1-INH deficiency secondary to a B cell lymphoproliferative disorder, in this case SMZL, was finally reached once low C1-INH levels were found, and a CD5 negative B cell population was identified on flow cytometry as detailed below.

Investigations

C1 esterase inhibitor levels checked during the third admission were shown to be low, with less than 25% and 0.05 g/L of C1-INH activity and antigen, respectively. Given the presence of splenomegaly, an impression of acquired C1-INH autoantibody from a possible lymphoproliferative disorder was strongly suspected. However, given that marrow examination was unrevealing, core biopsies of the spleen were taken. This failed to identify any clonal disorder and showed mild congestion and small foci of granulomatous inflammation.

A second bone marrow aspirate and trephine were carried out 4 months after the first. A clonality of CD5 negative B cells was revealed by flow cytometry in 7% of the total lymphoid population, corresponding to a suspicious intrasinusoidal pattern of infiltration on trephine as shown in figure 1. A qualitative anti-C1-INH antibody assay was positive for an IgM inhibitor. Thus, we did not perform Serpin Family G Member 1 (SERPING1) gene analysis given the strong evidence in keeping with an acquired rather than hereditary angioedema. Subsequent positron emission tomography/CT (PET/CT) scan revealed increased splenic tracer uptake but was otherwise unremarkable as shown in figure 2. The presence of splenic tracer uptake is significant in that it allows for the differentiation of SMZL from other forms of marginal zone lymphoma (MZL). In this case, we had excluded other lymphomas by virtue of the unique IHC phenotypes of B cell clone found in bone marrow. This alone however only proves the presence of MZL. Studies have shown that SMZL shows the highest rates of PET/CT avidity and that a diffuse splenic uptake is the most common pattern.^{1 2} Thus, in the absence of histological evidence of MZL from splenic biopsy, the PET/CT result showing diffuse splenic uptake in the absence of significant lymphadenopathy and/or extranodal uptake would strongly indicate the diagnosis of SMZL in this case.

Figure 1.

Composite image where images (A) and (B), respectively, represent ×100 and ×400 magnification images of the splenic biopsy showing congestion. Images (C) and (D), respectively, represent ×100 and ×400 magnification images of bone marrow showing a subtle lymphocytic intrasinusoidal infiltrate in a single filing pattern on CD20 staining.

Figure 2.

Increased diffuse splenic tracer uptake and splenomegaly on positron emission tomography/CT scan.

Differential diagnosis

During the initial presentation, we considered whether or not this could be an acquired or inherited case of angioedema. Given the patients lack of a family history and any history of previous attacks, it was thought that an inherited cause was unlikely. One of the common differential diagnoses was allergic angioedema secondary to a culprit drug. Although the absence of hives or urticarial rashes is not usually in keeping with an allergic aetiology, the angioedema seemed to respond to steroids and the administration of antihistamines pointing towards a histaminergic mechanism. However, the concomitant administration of steroids likely produced a response in our patient due to the lymphoproliferative mechanism underlying the pathophysiology of these episodes. This had the dual effect of masking the bradykinin-induced angioedema typical of C1-INH deficiency, as well as possibly hindering our diagnostic efforts to identify a clonal B cell disorder.

A diagnosis of scombroid was shortly entertained in the second presentation given the consumption of Coryphaena hippurus commonly known as Mahi mahi, a locally popular dish which is seasonally consumed and commonly implicated in such cases. Scombroid toxicity typically presents within an hour of consumption with several different symptoms, with tongue swelling rarely being among them and only occurring in cases of severe toxicity.

The absence of an elevated IgE and tryptase pointed towards the possibility of a C1-INH deficiency which was confirmed on testing. C1-INH deficiency can be either acquired or hereditary. Acquired cases occur as a consequence of antibody formation secondary to a lymphoproliferative or autoimmune condition, resulting in direct inhibition or overactivation of complement with C1-INH consumption. Alternatively, congenital cases result due to reduced levels or function of C1-INH secondary to SERPING1 gene mutations. In this case, given the lack of family history, response to steroids and coexistence of splenomegaly, a lymphoproliferative disorder-induced acquired C1-INH deficiency was favoured.

Treatment

Treatment of the condition was composed of two parts: one aimed at preventing and preparing for potential further attacks of angioedema and the other targeting the underlying B cell clonal disorder, treatment of which should cure the patients with C1-INH deficiency.

To best prepare for potential further attacks until initiation of rituximab and response is achieved, a medic-alert card was issued, and a plan was given for the patient to be administered plasma-derived C1-INH concentrate in case of a recurrence of symptoms as an ‘open door’, on demand policy, that is, the patient can turn up to a dedicated ward for administration of concentrate at any point of symptom recurrence. Since C1-INH concentrate is derived from human plasma, vaccination against hepatitis A and B was ensured, and baseline virology assessment was performed. Response to C1-INH concentrate in the context of a C1-INH autoantibody may be unreliable due to risk of neutralisation of transfused concentrate. Should C1-INH replacement have proven to be insufficient, emergency plasmapheresis could be considered.³ Other treatment modalities such as bradykinin and kallikrein inhibitors are not accessible at our centre, and C1-INH concentrate remained the best available option for the treatment of an acute attack as an interim measure until the underlying disorder was diagnosed and treated, and response was ascertained. We do not have access to attenuated androgens such as danazol; therefore, tranexamic acid prophylaxis at 1 g two times per day was used as per international guidance regarding acquired angioedema (AAE).^{3 4}

The treatment options considered for SMZL were splenectomy, rituximab combined with cytotoxic chemotherapy or isolated use of rituximab single agent. After discussion with the patient, the latter option was initiated with the use of single-agent rituximab at 375 mg/m² weekly for 4 weeks. C1-INH concentrate was kept readily available during his treatment in case of further attacks of angioedema, and the first dose of rituximab was administered as an inpatient for fear of a severe infusion reaction.

Outcome and follow-up

A partial response was obtained following rituximab therapy, with partial resolution of splenomegaly on repeat ultrasound at 4 months after rituximab induction, disappearance of the inhibitor antibody and normalisation of C4 and C1q levels. Flow cytometry of a post-treatment bone marrow aspirate did not reveal any clonal B cell population. Consequently, it was decided to proceed with maintenance rituximab every 2 months for a total of 2 years. On a repeat ultrasound done after 6 months of maintenance, the spleen size was normal in keeping with a complete response (<13 cm). The patient is currently continuing his second year of maintenance rituximab and being followed up at outpatients. He remains in possession of an alert card, with a plan in place for C1-INH concentrate administration in case of symptom recurrence. No further episodes of AAE have occurred since rituximab induction.

Discussion

SMZL is a relatively rare lymphoma accounting for 0.6% of all non-Hodgkin’s lymphomas (NHLs) and has consequently often been included in combination with other low-grade NHL in clinical trials.⁵ The occurrence of this lymphoma in association with AAE grossly exceeds that expected for overall incidence. NHL is found in up to 33% of cases of AAE, with up to 66% of these cases being SMZL.⁶

Treatment of AAE in the setting of a lymphoproliferative disorder has been explored in a number of case reports and series.^7–10 Current recommendations based on the existing literature state that acute treatment of attacks should be in the same manner as hereditary C1-INH deficiency, while an overall cure of the condition requires definitive treatment of the underlying malignancy, in this case SMZL.¹¹

Rituximab has been used successfully in acquired C1-INH deficiency, and not only in the context of a lymphoproliferative disorder.¹² There seem to be no reports detailing increased incidence or severity of infusion reactions with the use of rituximab in AAE; however, we felt that the administration of the drug should be done in a setting where resuscitation facilities and C1-INH concentrate are immediately at hand. In the absence of any evidence of high-grade disease, we considered rituximab monotherapy because of its low toxicity and good overall response rates exceeding 80%.¹³ An important trial by Kalpadakis et al supported a progression-free survival but not an overall survival benefit in patients with SMZL who responded to rituximab monotherapy.¹³ A major caveat of this trial is that only one patient with a secondary autoimmune manifestation was included, and we believe that prolonging progression-free survival provides us with a major justification for maintenance, as it probably translates to an important benefit in terms of reducing the likelihood of C1-INH autoantibody recurrence.

An interesting issue relates to the frequency and method of follow-up of such patients at clinic in the absence of any recommendations. No reliable data exists as of yet regarding the relationship between C1-INH autoantibody levels and disease activity. This raises the question as to whether C1-INH autoantibody recurrence may be used as a marker of disease recurrence and thus early re-treatment. Furthermore, criteria for complete response in SMZL include a negative direct antiglobulin test, and thus it would be useful to note whether a negative C1-INH autoantibody test can be used as a surrogate for disease status.

SMZL cells underlying autoimmune phenomena such as autoimmune haemolysis and immune thrombocytopenia were found to have a stereotyped immunoglobulin heavy-chain rearrangement.¹⁴ This was also found to be true in a cohort of SMZL causing AAE with the same IGHV1-2*04 rearrangement identified.¹⁵ This suggests that an autoantigen, perhaps a particular epitope of the C1-INH or a structurally related molecule, is eliciting an antigenic response with the consequent formation of a progressively monoclonal B cell population secondary to chronic B cell receptor stimulation. This antigenic mechanism is already appreciated in MZL caused by hepatitis C or Helicobacter pylori infection.

In conclusion, while recommendations exist as to the treatment of AAE due to C1-INH deficiency (C1-INH-AAE) in the context of malignancy, there exists a lack of data regarding the exact relationship between C1-INH autoantibodies and disease activity. Furthermore, there exists no official guidance regarding the efficacy of different standard treatments of SMZL in cases displaying C1-INH deficiency. While studies in SMZL are ongoing, further clinical trials in SMZL-associated AAE are therefore warranted.

Learning points.

• C1 esterase inhibitor (C1-INH) deficiency should be suspected in the absence of wheals or lack of biochemical evidence of a histaminergic mechanism, that is, a normal IgE and tryptase.

• Due to the fact that non-Hodgkin’s lymphoma (NHL), particularly SMZL, is associated with acquired angioedema due to C1-INH deficiency (C1-INH-AAE), a high index of suspicion for occult or coexisting SMZL should be given for patients who present with AAE.

• Treatment of AAE with rituximab monotherapy aimed at inducing remission of the underlying SMZL successfully caused a complete resolution of C1-INH-AAE

• Steroid therapy can mask occult underlying NHL as the cause of AAE.

Ethics statements

Patient consent for publication

Consent obtained directly from patient(s).