Low immunoglobulin A levels detected via the tissue transglutaminase assay can reveal previously undetected monoclonal proteins

Abstract

Increased awareness of coeliac disease and the 2009 NICE guidance has led to an increase in patients being screened for Immunoglobulin A deficiency. We have shown previously that this provides an opportunity for the early identification of other underlying primary immunodeficiency, e.g. common variable immunodeficiency. In this context, the underlying gastrointestinal problem appears to be related to bacterial overgrowth. Here, we demonstrate that in addition this also provides an opportunity to reveal underlying secondary immunodeficiency due to other causes in patients with gastrointestinal presentation, notably lymphoproliferative disorders. In one 3-month period, of 60 cases reviewed for low Immunoglobulin A, we found four new paraproteins through this testing route; one symptomatic multiple myeloma, one asymptomatic multiple myeloma, one monoclonal gammopathy of uncertain significance and one in a known chronic lymphocytic leukaemia patient.

Introduction

In common with many laboratories in the UK, we have seen a huge rise in requests for antitissue transglutaminase testing (TTG). ¹ In 2008, we had 6696 requests; this had risen to 18,184 by 2014. As immunoglobulin A (IgA) deficiency is found in approximately 1 in 50 cases of coeliac disease, NICE (National Institute for Health and Care Excellence) have recommended that all laboratories have in place a mechanism to detect IgA deficiency. ² This has led to uncovering IgA deficiency in patients without other obvious immunological disease. In order to confirm that the patient has selective IgA deficiency, which is seen in approximately 1 in 500 persons ³ (most of whom are asymptomatic), it is necessary to confirm the other immunoglobulins are normal (IgG and IgM). This in turn has uncovered other diagnoses, including primary antibody deficiency. ⁴

We present a short series of patients found over a three-month period during 2014 in whom previously undetected paraproteins were revealed.

Methods and testing protocol

In accordance with NICE clinical guidelines 86, ² when we observe a low background absorbance on the anti-TTG ELISA method (QUANTlite Rh tTG IgA ELISA, Inova Diagnostics Inc., San Diego, CA, USA), we initiate a formal measurement of serum IgA using a sensitive immunochemical IgA assay on the BN II nephelometer (Siemens Healthcare Diagnostics Ltd., Camberley, Surrey, UK) with a limit of detection of <0.07 g/L. The OD cut-off of <0.060 was initially determined by probability plotting (n = 390). ⁵ It was confirmed as sensitive for IgA deficiency in a confirmatory study quantifying 114 samples for IgA by nephelometry. All four IgA-deficient patients (IgA < 0.07 g/L) had OD ≤ 0.050. Seventeen of 114 had OD ≤ 0.060; 10/17 had IgA below 0.5 g/L. Conversely, 14 of 114 had IgA < 0.8 g/L and only 1 of these (0.7 g/L IgA) had OD > 0.060. No sample with OD ≥ 0.060 had an IgA < 0.2 g/L. Therefore, the cut-off is able to identify all samples requiring further analysis as determined by the NICE and ESPGHAN guidelines.

In the ESPGHAN guidelines, ⁶ it is recommended that serum samples with <0.2 g/L of IgA be further tested for IgG class coeliac disease-related autoantibodies. We have adopted a higher cut off for further investigation (<0.50 g/L) than that suggested as a means of uncovering underlying immunodeficiency. ⁴ We acknowledge that the choice of 0.5 g/L is somewhat arbitrary. Given the prevalence of partial IgA deficiency in the UK population, we feel that analysing those in the 0.5–0.8 g/L range is likely to have a low yield. Where a low IgA is confirmed, we proceed to measure IgG, IgM (BN II Nephelometer) and perform capillary zone electrophoresis (CZE) (Capillarys 2, Sebia, Lisses, France).

During the three-month investigation period, we received 4589 samples for anti-TTG. Of these, 191 had an OD < 0.060. Nineteen of 191 were IgA deficient (9.9%). A further 41 had detectable IgA but <0.5 g/L (21.5%), and hence 60 samples were investigated further as per the described protocol.

The final tier of investigation is initiated where there is suspicion of a possible underlying monoclonal immunoglobulin, i.e. obvious additional peak on CZE or hypogammaglobulinaemia (indicative of possible monoclonal free light chain (FLC) disease). This includes serum (and urine where available) gel immunofixation (Hydrasys, Sebia) and serum FLC analysis (Binding Site FreeLite Assay; Birmingham, UK, run on BN2 nephelometer using Katzmann’s reference ranges ⁷ ), as appropriate.

Case histories and results

Case 1

A 49-year-old male presented to his general practitioner (GP) with a two-month history of stomach cramps, diarrhoea, a 10 kg weight loss and occasional rectal bleeding. Anti-TTG was requested and testing showed a low background absorbance; IgA was added on and found to be 0.34 g/L (age-related reference range 0.80–4.00 g/L). Electrophoresis, IgG and IgM were added on as per lab protocol to confirm selective IgA deficiency. Electrophoresis showed no clear band, but there was immune paresis present so we proceeded to gel immunofixation which showed a clear band in the Lambda lane. Further gel immunofixation analysis using anti IgD and IgE antisera (Dako UK Ltd., Ely, Cambridgeshire, UK) was negative. We then went on to serum FLC analysis: Kappa 6.5 mg/L (reference range 3.3–19.4 mg/L), Lambda 2120 mg/L (reference range 5.7–26.3 mg/L), Kappa/Lambda ratio 0.003 (reference range 0.26–1.65). ⁷ This confirmed the finding of Lambda monoclonal FLC and subsequent urine testing showed monoclonal free Lambda light chains (Bence Jones Protein) to be present at 4779 mg/L.

Bone marrow biopsy showed 40% plasma cells consistent with plasma cell myeloma. Fluorescent in situ hybridisation on material enriched for plasma cells demonstrated an IGH-MAFB (t[14;20]) rearrangement; this conveys a worse prognosis. Given a diagnosis of asymptomatic myeloma, not meeting the CRAB (Calcium, renal failure, anaemia, bone lesions) criteria, this patient was put on an active monitoring programme. In view of the new International Myeloma Working Group guidelines, ⁸ the serum FLC ratio of >100, his diagnosis was changed to symptomatic multiple myeloma with an International scoring system of 1. He is due for treatment with combination chemotherapy including a proteosome inhibitor, thalidomide and steroids once the colitis is controlled.

Case 2

A 72-year-old lady visited her GP complaining of tiredness and back pain. Coeliac disease was considered among the differentials for tiredness. Low absorbance on the anti-TTG assay led to serum IgA quantification which was low at 0.26 g/L. Electrophoresis showed a clear monoclonal protein which was typed as an IgG Kappa of 9 g/L. This lady is asymptomatic with normal calcium, renal function and full blood count. She is being followed up in primary care with four monthly paraprotein checks and clinical reviews with a presumptive diagnosis of monoclonal gammopathy of uncertain significance (MGUS). She has a low chance of progression to multiple myeloma.

Case 3

A 71-year-old man attended his GP surgery complaining of persistent diarrhoea. An IgA of 0.34 g/L was found following a low absorbance on the anti-TTG assay. Electrophoresis showed a significant monoclonal band, which was subsequently typed as an IgG Kappa of 17 g/L. He has normal adjusted calcium, renal function and full blood count, but his bone marrow demonstrated 20% plasma cells with a neoplastic phenotype. His current working diagnosis is asymptomatic multiple myeloma and using the scoring system for this condition, his predicted median time to progression/treatment is eight years.

Case 4

An 84-year-old man with a preceding diagnosis of chronic lymphocytic leukaemia presented to his GP complaining of diarrhoea. He had received fludarabine-based chemotherapy in 2012 to good clinical effect. Following a low absorbance on the TTG assay, his IgA was found to be 0.14 g/L. Electrophoresis showed a small band which gel immunofixation revealed to be an IgM Lambda of 1 g/L. He is currently being monitored without treatment.

Discussion

The issue of adding on tests which are not directly related to the original request has provoked much debate. On the one hand, we are potentially causing confusion and performing unnecessary tests; on the other hand, as these cases illustrate, we can reveal significant and unsuspected pathological conditions.

A diagnosis of myeloma may be significantly delayed, particularly in patients with non-specific symptoms which may include back pain, tiredness and infections. With the improvement in recent treatments, delays in diagnosis can lead to significantly poorer survival and quality of life outcomes. It can also be important to pick up the asymptomatic condition MGUS; high-risk cases require monitoring and 1% of these patients evolve to myeloma annually.

In the context of patients presenting with gastrointestinal problems, primary or secondary immunodeficiency may result in bacterial overgrowth and the development of symptoms. We have previously demonstrated the value of cascade testing of immunoglobulins in this clinical scenario ⁴ and now extend that to uncovering immune paresis in association with previously undiscovered paraproteinaemia. This has allowed appropriate therapy or monitoring to be initiated at a much earlier phase in the disease than would have otherwise been possible.