Page 44 - 2019_03-Haematologica-web

L.Y.C. Chen et al.
Erdheim-Chester disease have retroperitoneal fibrosis, which may raise the suspicion of IgG4-RD; however, more than 95% of patients with Erdheim-Chester disease have skeletal involvement, which, aside from rare cases of IgG4-related angiocentric eosinophilic fibrosis (midline destructive lesions of the head and neck), is generally not seen in IgG4-RD.50 The histology of Erdheim-Chester dis- ease shows a CD68+S100–CD1a– histiocyte infiltration often with “foamy histiocytes”.51 Extra-pulmonary sar- coidosis may share clinical features similar to those of IgG4-RD, including polyclonal hypergammaglobulinemia, lymphadenopathy, pulmonary nodules, sclerosing mesen- teritis and pachymeningitis. The association between IgG4-RD and malignant lymphoma has been studied extensively. In Asian patients, mucosal-associated lym- phoid tissue (MALT) lymphoma, particularly of ophthal- mological tissues has been described, whereas in western populations a variety of histologies (diffuse large B-cell, follicular, lymphoplasmacytic, and MALT) have been reported.52,53 There are also case reports of IgG4-RD con- comitant with autoimmune lymphoproliferative syn- drome.54,55
Case continued
The patient was suspected to have IgG4-RD, so serum IgG subclasses were analyzed. His serum IgG4 level was markedly elevated at 11.6 g/L (normal values <1.35). The tissue blocks from his previous nephrectomy were retrieved and pathology review showed a lymphoplasmacytic infiltrate, moderate tissue eosinophilia and interstitial fibrosis and atrophy. Staining for IgG4 and IgG revealed abundant IgG4+ plasma cells with >40 IgG4+ plasma cells per high power field and an IgG4/IgG ratio >40%. Computed tomography scans of the neck, chest, abdomen and pelvis revealed multiple pulmonary nodules, carinal lym- phadenopathy and a soft tissue density encasing the main coro- nary arteries, previous right nephrectomy and pancreatic atrophy. His IgG4-RD Responder Index activity score was 12.
Diagnosis and staging
A careful history and thorough physical examination, attending to clues such as a history of waxing and waning glandular swelling, sicca symptoms and unexplained pan- creatitis or jaundice must be accompanied by serum pro- tein studies. Histological confirmation of the disease is essential, and once a diagnosis has been established, inves- tigations to assess for symptomatic and subclinical organ involvement, such as early retroperitoneal fibrosis and albuminuria, are important for management planning. The IgG4-RD Responder Index is a standardized, validat- ed tool for the evaluation of disease activity at initial eval- uation and subsequent follow up.56,57 Suggested laboratory and imaging tests are summarized in Table 3.
Serum protein studies
Approximately 70% of patients with IgG4-RD have an elevated serum IgG4 level. Serum IgG subclasses should be investigated in conjunction with serum protein elec- trophoresis to exclude a monoclonal paraprotein (Figure 3). Serum IgG4 level has a diagnostic sensitivity ranging from 83-97% and specificity from 60-85% with a general cut-off of “above the upper limit of normal”.58-60 Typically, 1.35 g/L is used as the biomarker cut-off for IgG4-RD (which corresponds to the upper limit of normal for one
common commercial method but not another) without specifying the methodology. While mildly elevated serum IgG4 can be seen in many conditions, markedly elevated serum IgG4 (>5 g/L) is approximately 90% specific for IgG4-RD. Aside from methodological differences, serum IgG4 levels in IgG4-RD can vary greatly depending on eth- nicity and degree of organ involvement. In the Boston cohort of patients (76% White), only 53 of 103 patients had elevated serum IgG4 levels.16 In contrast, in a cohort of 334 Japanese patients, more than 95% had elevated serum IgG4.17 In our multi-ethnic cohort, we found that Asians have a higher serum IgG4 than non-Asians (medi- an 11.2 g/L versus 2.9 g/L, P=0.0094), and elevated serum IgG4 had a sensitivity of 96% in Asians compared to 67% in non-Asians.61 Patients with multi-organ involvement or of Asian ethnicity typically have elevated serum IgG4, sometimes markedly so, such as the patient in this illustra- tive case. The serum IgG4/IgG ratio is typically >0.2 in patients with IgG4-RD, although the ratio does not increase the diagnostic specificity of serum IgG4 alone. Flow cytometric detection of plasmablasts may offer a more sensitive modality for diagnosing IgG4-RD, with a reported sensitivity of 95% and specificity of 82% using a cut-off of 900/mL.62 However, the flow cytometry method used to detect plasmablasts is not widely available.
Most centers use immunonephelometry to measure IgG subclasses, which can cause some challenges with inter- pretation. The two most common immunonephelometric methods (Siemens and Binding Site) correlate well with regard to IgG4, but the absolute IgG4 values differ by approximately 50% at the upper limit of normal.63 IgG4 levels may also be markedly under-reported in cases of extreme IgG4 elevations due to the hook effect. The hook effect, or prozone phenomenon, occurs when an excessive amount of analyte prevents binding of the capture anti- body in a sandwich assay, yielding a falsely low or normal result. Erroneously low measurements of serum IgG4 reported in the literature reflect this error.64 Furthermore, IgG4 itself interferes with the nephelometric measure- ment of IgG1 and IgG2, in particular, which can obscure the immunoglobulin profile that would otherwise high- light the disproportionate elevation of serum IgG4.65 Because of the traditional errors in immunonephelometry, some have mistakenly reported increased serum IgG2 lev- els as a marker of IgG4-RD.66-68 Our group has recently demonstrated that mass spectrometry is an alternative that eliminates these analytical errors and is more cost- effective than immunonephelometry.65
Histopathology
A firm diagnosis of IgG4-related disease requires histopathological confirmation, except in the case of autoimmune pancreatitis, in which radiological features (diffuse “sausage-like” enlargement of the pancreas with featureless borders and delayed enhancement with or without a capsule-like rim or “halo”) may be sufficiently specific to exclude requirement for tissue biopsy.3,69 As in sarcoidosis, in which non-caseating granulomas may be seen in any of the organs affected by the disease, IgG4-RD demonstrates common histology in most of the multitude of organs that may be affected.
The three major histological features of IgG4-RD in tis- sue are: (i) a dense, polyclonal lymphoplasmacytic infil- trate enriched with IgG4+ plasma cells; (ii) fibrosis; and (iii) obliterative phlebitis. With regards to the lymphoplasma-
450
haematologica | 2019; 104(3)