Antiglomerular basement membrane (anti-GBM) disease with clinical and histological features that bridge the typical to atypical spectrum

Abstract

We describe antiglomerular basement membrane (anti-GBM) disease with rapidly progressive glomerulonephritis and concurrent parainfluenza pneumonia. Circulating anti-GBM antibodies were barely detectable and disappeared rapidly following corticosteroids, cyclophosphamide and plasma exchange. Kidney biopsy demonstrated strong linear GBM staining for IgG and IgG4 and unusually prominent endocapillary hypercellularity, suggesting ‘atypical anti-GBM disease’, although glomerular necrosis and crescents were also seen. When kidney function deteriorated further, despite persistently absent circulating anti-GBM antibodies, a repeat kidney biopsy was performed, showing crescents in 100% of glomeruli with ongoing endocapillary hypercellularity and strong IgG and IgG4 GBM staining. This case highlights complexities in the diagnosis of anti-GBM disease, with clinical and histological features bridging the atypical to typical anti-GBM disease spectrum. We hypothesise that these findings might be explained by the presence of IgG4 (rather than traditional IgG1 or IgG3) autoantibodies. To our knowledge, this is also the first report of parainfluenza associated with anti-GBM disease.

Background

Antiglomerular basement membrane (anti-GBM) disease typically presents with rapidly progressive glomerulonephritis (RPGN). The disease is caused by autoantibodies directed against the non-collagenous domain of the α3 chain of type IV collagen (α3(IV)NC1).¹ Anti-GBM antibodies are detected in kidney biopsy specimens using immunofluorescence (IF) microscopy, while circulating anti-GBM antibodies are identified by an ELISA. Respiratory infections are known to trigger anti-GBM disease,² although a specific association with parainfluenza infection has not, to our knowledge, been reported. Cases of anti-GBM disease with atypical features have recently been described, in which there is strong linear staining of the GBM for immunoglobulin, supporting a diagnosis of anti-GBM disease, but negative testing for circulating anti-GBM antibodies along with atypical histological (ie, prominent endocapillary hypercellularity and a relative paucity of necrosis and crescents) and clinical (ie, a comparatively indolent clinical course) features.³ Here, we describe a case of anti-GBM disease in which pulmonary-renal syndrome might have been triggered by parainfluenza infection. Clinical and histological features bridged a spectrum between typical and atypical anti-GBM disease, resulting in diagnostic and therapeutic challenges that can inform the management of future patients with a similar presentation.

Case presentation

A young male adult presented to the hospital with a 3-week history of cough and dyspnoea, a 3-day history of blood-tinged purulent sputum and new renal impairment detected in the community. He also reported frothy urine, lower extremity oedema and fatigue. No arthralgias, myalgias, rash, fever or other symptoms. Social history was relevant for recent return from South East Asia, as well as current cigarette smoking. On examination, the patient was afebrile, blood pressure was 150/90 mm Hg, oxygen saturations were 96% on room air and he had 1+bipedal oedema. His chest was clear to auscultation and a complete physical examination was otherwise unremarkable.

Investigations

The patient’s serum creatinine was 228 μmol/L, from 169 μmol/L 1 week and 127 μmol/L 2 weeks, prior (figure 1). Urinalysis showed 4+blood and 4+protein. The patient provided a medical report showing that he had 3+blood and 3+protein by urinalysis performed as part of a work-related medical exam 2 months prior. Unfortunately, the patient was not informed of these abnormalities nor referred to a nephrologist at the time. Further investigations at presentation included normal or negative testing for complement C3 and C4, antinuclear antibodies, antidouble-stranded DNA antibodies, extractable nuclear antigen antibodies, antineutrophil cytoplasmic antibodies (ANCAs) and HIV, hepatitis B and hepatitis C viruses. Circulating anti-GBM antibodies was weakly positive at 15 (normal 0–7) U/mL using a commercial ELISA that quantifies IgG antibodies to a truncated human recombinant α3(IV)NC1 (Phadia 250 EliA GBM). A nasopharyngeal swab was negative for COVID-19 and positive for parainfluenza B by PCR. Chest CT showed small bilateral pleural effusions and subtle areas of ground-glass nodularity in the lingula, left upper, right middle and right lower lobes.

Figure 1.

Timeline of patient’s clinical course, including treatments received. GBM, glomerular basement membrane; PLEX, plasma exchange.

Differential diagnosis

The patient’s clinical presentation was consistent with RPGN, with possible pulmonary haemorrhage suggesting pulmonary-renal syndrome. With negative ANCA testing, and positive (although weakly) anti-GBM testing, a diagnosis of anti-GBM disease was made. However, a kidney biopsy was pursued to confirm the diagnosis (see below), given barely detectable anti-GBM antibodies and an atypical history of indolent haematuria and proteinuria months prior.

Treatment

The patient was treated with high dose corticosteroids (intravenous methylprednisolone 500 mg daily for 3 days converting to oral prednisolone starting at 60 mg daily), oral cyclophosphamide (initially 2 mg/kg) and eight sessions of plasma exchange over the course of 11 days (figure 1).

Outcome and follow-up

The patient’s circulating anti-GBM antibodies disappeared within 2 days of initiating treatment. Thirteen days after initiating treatment, he underwent a kidney biopsy to confirm the clinical diagnosis. Light microscopy identified cortex and medulla incorporating up to 29 glomeruli in one plane of section. Two glomeruli were globally sclerosed and two appeared relatively normal. Cellular crescents were present in 21 glomeruli and there were possible small fibrocellular crescents in two additional glomeruli. Segmental tuft necrosis was seen in five glomeruli. Focal endocapillary proliferation with increased mesangial matrix and focal segmental sclerosis was present in at least nine glomeruli. No extraglomerular vasculitis. Proximal tubules showed focal epithelial degenerative change with cytoplasmic vacuolation and red cell casts. Minimal interstitial fibrosis and tubular atrophy was appreciated (<5% of the total cortical area) (figure 2). IF microscopy confirmed strong (3+) linear staining for IgG along the GBM, along with 3+ staining for IgG4 (immune reactants for IgG1, IgG2 and IgG3 are not available in our clinical laboratory).

Figure 2.

Histological features identified by kidney biopsy. (A) First biopsy: light microscopy (methenamine silver stain) showing global endocapillary hypercellularity and focal disruption to Bowman’s capsule (black arrow). (B) First biopsy: light microscopy (methenamine silver stain) the lower glomerulus contains a cellular crescent (black arrow) and segmental endocapillary hypercellularity (open arrow), in contrast to the uninvolved glomerulus above. (C) Second biopsy: light microscopy (methenamine silver stain) showing a cellular crescent (black arrow) with global endocapillary hypercellularity.(D), (E) Second biopsy: immunofluorescence with strong global linear positivity for IgG4 and IgG along the glomerular basement membrane. Findings were similar for the first biopsy (not shown). (F) Second biopsy: electron microscopy showing endocapillary hypercellularity with extensive foot process effacement (black arrow) and focal microvillous transformation (open arrow) without electron dense immune-type deposits. Findings were similar for the first biopsy (not shown).

The patient’s kidney function continued to decline despite therapy. A repeat kidney biopsy, performed 3 weeks after the first biopsy and 5 weeks after presenting to hospital, showed cortex and medulla incorporating up to 27 glomeruli in one plane of section. Two glomeruli were globally sclerosed and all 25 of the remaining viable glomeruli were involved by crescents, predominantly cellular with focal fibrocellular forms. At least five glomeruli showed segmental tuft necrosis. Focal endocapillary proliferation remained present, with increased mesangial matrix and focal segmental sclerosis. No extraglomerular vasculitis. Proximal tubules showed focal epithelial flattening and cell shedding with occasional tubular mitoses and red cell casts. Mild interstitial fibrosis and tubular atrophy (<10% of the total cortical area). Strong linear staining of the GBM for both IgG and IgG4 persisted by IF microscopy (figure 2).

Testing for circulating anti-GBM antibodies using a commercial ELISA was persistently negative. However, in view of the patient’s deteriorating kidney function and histological evidence of ongoing disease activity, a serum sample collected at the time of his second biopsy was referred to an external reference laboratory for indirect IF testing using monkey kidney tissue. Pending this result (which was later reported as negative for anti-GBM antibodies), he received five further plasma exchange treatments. Despite intensive therapy, he progressed to dialysis-requiring end-stage kidney failure and remains on dialysis since.

Discussion

Anti-GBM disease typically presents with RPGN, with or without pulmonary haemorrhage. Characteristic histological findings include linear staining of the GBM for immunoglobulin (most typically IgG1 or IgG3) and diffuse crescentic and necrotising glomerulonephritis (GN).^{1 2} This case demonstrates features of anti-GBM disease never or rarely reported. First, we identified parainfluenza pneumonia as a potential pathogenic trigger for RPGN in a primed patient, an association that has not, to our knowledge, previously been described. Second, the clinical and histological features of this case bridged a spectrum between ‘typical’ and recently described ‘atypical’ anti-GBM disease.³

The level of circulating anti-GBM antibody in this case was unexpectedly low considering our working diagnosis of anti-GBM disease. When the initial kidney biopsy showed prominent endocapillary hypercellularity, although along with glomerular necrosis and crescents, we considered a diagnosis of atypical anti-GBM disease and hoped that the patient might have a favourable outcome.³ However, when the patient’s kidney function continued to decline, despite intensive immunosuppressive therapy and persistently negative testing for circulating anti-GBM antibodies, we proceeded to perform a second kidney biopsy. When this showed severe crescentic GN and ongoing strong IgG and IgG4 deposition in the GBM, we considered the patient to have more typical anti-GBM disease, but with anti-GBM antibodies that were somehow missed by the commercial ELISA test.

One older report examined three cases of anti-GBM disease in which serological testing for anti-GBM antibodies was negative using standard ELISA; however, two of three had detectable anti-GBM antibodies using more sensitive biosensor techniques. All three had linear IgG staining of the GBM histologically. Interestingly, one of these cases was a young male smoker and another had a preceding viral illness, similar to our described case.⁴ A more recent case series described four Chinese patients with histologically confirmed anti-GBM disease in whom circulating anti-GBM antibodies were not detected by ELISA but were detected using indirect IF and Western blot analyses (under non-reducing conditions to preserve disulfide bonds).⁵ The authors concluded that ELISA might miss anti-GBM antibodies targeting cryptic, conformation-dependent epitopes. Other explanations for absent circulating anti-GBM antibodies despite histological evidence of anti-GBM disease include the presence of antibodies targeting GBM antigens other than α3(IV)NC1, the presence of antibodies with particularly high affinity for the GBM (‘mopped up’ by the GBM), or delays in clinical presentation such that circulating antibodies are no longer being produced. We confirmed that our patient was negative for circulating anti-GBM antibodies using an indirect IF assay with monkey tissue, although not until 5 weeks after initiating immunosuppressive therapy (unfortunately earlier samples were not stored). Nevertheless, we demonstrated intense IgG staining of the GBM by repeat kidney biopsy contemporaneously with absent circulating anti-GBM antibodies by both ELISA and indirect IF testing, raising questions as to whether or not his disease was still immunologically active.

Four subclasses of IgG exist. In typical cases of anti-GBM disease, the IgG1 subclass predominates^6–10 and IgG4 anti-GBM antibodies are infrequently identified in the GBM.¹¹ Further, when IgG4 anti-GBM antibodies predominate in the circulation, a milder clinical course is typically described when compared with IgG3 or IgG1 subclasses of anti-GBM antibodies.^{6 10 12} By contrast, in recently described cases of atypical anti-GBM disease, a more mixed distribution of IgG subclasses has been identified (30% IgG1, 40% IgG2 and 30% IgG4 predominant).³ Although we were unable to examine the IgG subclass of our patient’s transiently circulating anti-GBM antibody, we determined that the IgG deposited in his GBM stained strongly for IgG4. We wonder whether this explained the clinical and histological overlap between typical and atypical anti-GBM disease observed in this case, including an initially indolent clinical course. Nevertheless, we did also identify one recent case report of IgG4 related atypical anti-GBM disease associated with rapid progression to end-stage kidney failure.¹³ This and our present report suggest that IgG4 anti-GBM disease might not always portend a benign prognosis, contrary to prior reports.¹² Further, we identified a case series of four women with IgG4 anti-GBM disease who had absent or low-level circulating antibodies using commercial ELISA but successful detection using a dedicated IgG4 ELISA.¹⁴ Despite a comparatively mild initial clinical course, two of these four women relapsed, which is an uncommon phenomenon in anti-GBM disease. Accordingly, in our patient with strong IgG4 staining of the a GBM, we will monitor carefully for relapse both before and after future kidney transplantation. Further, if relapse is suspected, but commercial anti-GBM ELISA testing is negative, we will hunt for circulating anti-GBM antibodies using indirect IF and dedicated IgG4 ELISA testing at research laboratories.

Our report has some limitations. We were unable to determine the IgG subclass of the transiently identified circulating IgG antibody, and we only had the immune reactant to evaluate for IgG4 but not other IgG subclasses deposited in the GBM. Thus, any hypotheses involving this patient having IgG4-mediated anti-GBM disease are speculative. Second, although a temporal relationship existed, we cannot assign causality to the parainfluenza infection that coincided with the onset of RPGN. Third, the initial and subsequent kidney biopsies were both performed after starting treatment, which might have modified the morphology. Accepting these limitations, this is the first reported case, to our knowledge, of concomitantly diagnosed parainfluenza infection and anti-GBM disease. Additionally, by performing a second kidney biopsy in this patient, we demonstrated both clinical and histological progression in a patient with overlapping features of atypical and typical anti-GBM disease, suggesting that these might not be distinct disease entities but rather different ends of a disease spectrum.

Learning points.

• This case highlights major teaching points as well as more subtle complexities in the pathogenesis, diagnosis and management of antiglomerular basement membrane (anti-GBM) disease.

• Pathogens (eg, parainfluenza) and environmental toxins (eg, cigarette smoking) can trigger the onset of rapidly progressive glomerulonephritis or pulmonary renal syndrome in a predisposed patient.

• The class and subclass of pathogenic anti-GBM autoantibodies might partly determine the histological and clinical features of disease.

• Use of circulating anti-GBM antibody levels as a biomarker of disease activity and response to treatment is complicated by the poor sensitivity of commercial ELISA testing in some settings.

Ethics statements

Patient consent for publication

Obtained.