Renal Manifestations of Common Variable Immunodeficiency

Abstract

Background 

Common variable immunodeficiency (CVID) is one of the most common primary immunodeficiency syndromes, affecting one in 25,000–50,000 people. Renal insufficiency occurs in approximately 2% of patients with CVID. To date, there are no case series of renal biopsies from patients with CVID, making it difficult to determine whether individual cases of renal disease in CVID represent sporadic events or are related to the underlying pathophysiology. We performed a retrospective analysis of renal biopsy specimens in our database from patients with a clinical history of CVID (n=22 patients, 27 biopsies).

Methods 

Light, immunofluorescence, and electron microscopy were reviewed. IgG subclasses, PLA2R immunohistochemistry, and THSD7A, EXT1, and NELL1 immunofluorescence were performed on all membranous glomerulopathy cases. CD3, CD4, CD8, and CD20 immunohistochemistry was performed on cases of tubulointerstitial nephritis.

Results 

AKI and proteinuria were the leading indications for renal biopsy in patients with CVID. Immune-complex glomerulopathy was present in 12 of 22 (54.5%) cases, including nine cases with membranous glomerulopathy, one case with a C3 glomerulopathy, and one case with membranoproliferative GN with IgG3κ deposits. All membranous glomerulopathy cases were PLA2R, THSD7A, EXT1, and NELL1 negative. The second most common renal biopsy diagnosis was chronic tubulointerstitial nephritis, affecting 33% of patients. All tubulointerstitial nephritis cases showed tubulitis and a lymphocytic infiltrate with >90% CD3⁺ T cells. Other renal biopsy diagnoses within our cohort included acute tubular injury (n=1), amyloid light-chain amyloidosis (n=1), diabetic glomerulosclerosis (n=1), thin basement membranes (n=1), pauci-immune GN (n=1), and arterionephrosclerosis (n=1).

Conclusions 

Membranous glomerulopathy and tubulointerstitial nephritis were the predominant pathologic findings in patients with CVID. Membranous glomerulopathy cases in patients with CVID were IgG1 subclass dominant and showed mesangial immune deposits. Four of the membranous glomerulopathy cases had associated proliferation, with mesangial and/or endocapillary hypercellularity, with or without crescent formation. CVID should be considered as a potential cause when membranous glomerulopathy or chronic tubulointerstitial nephritis is seen in a young patient with a history of recurrent infections.

Visual Abstract

Introduction

Common variable immunodeficiency (CVID) is one of the most common primary immunodeficiencies, yet it is overall rare, affecting one in 25,000–50,000 whites and one in 100,000 individuals worldwide (¹) and has near equivalent prevalence in men and women. CVID consists of a heterogenous group of primary immunodeficiency disorders characterized by a marked reduction in antibody production (IgG, IgM, and IgA), poor humoral responses to protein or polysaccharide vaccines, and an exclusion of other causes of hypogammaglobulinemia. A genetic cause for CVID can be identified in a minority of patients, with only 2%–10% of patients showing monogenic forms (²). CVID is most often diagnosed between ages 20 and 40 years, usually with a long delay from onset of symptoms to clinical recognition (average 6–8 years), as the symptoms are not specific, and infections are common in children.

Patients with CVID have a variety of disease manifestations, including recurrent bacterial infections, autoimmunity, chronic lung disease, gastrointestinal disease, lymphoid hyperplasia, granulomatous disease, and an increased risk of malignancy (³). Treatment of CVID is aimed at life-long Ig replacement, which reduces risk of bacterial infections, but does not affect autoimmune and inflammatory manifestations of the disease (³).

Renal failure is a rare complication of CVID, affecting approximately 2% of patients. In a cohort of 240 patients with CVID, five were found to have CKD (⁴). Ig replacement therapy in hypertonic solutions has been shown to be nephrotoxic and can result in reversible acute tubular injury. The presence of proteinuria or an active urinary sediment is not typical for intravenous Ig (IVIg)–induced renal injury, and is an indication for renal biopsy in patients with CVID. To date, there are no case series of renal biopsies from patients with CVID, making it difficult to determine whether individual case reports of renal disease in CVID represent sporadic events or are related to the underlying pathophysiology. Here, we present the first case series and review of the literature of the nephropathology in patients with CVID.

Materials and Methods

Case Selection

Approval for this study was obtained by the Solutions Institutional Review Board, and the ethical principles highlighted by the Declaration of Helsinki were followed. A retrospective analysis of 10 years of renal biopsy specimens using our database of patients with a clinical history of CVID was performed (n=22 patients, comprising 27 renal biopsy specimens) from January 1, 2010 to December 31, 2019. During this time period, there were 27 renal biopsies from patients with CVID out of a total of 110,918 cases (0.02%). Patients with primary immunodeficiency of unknown cause, as well as patients with a clinical or family history with x-linked (Bruton) agammaglobulinemia, severe combined immunodeficiency, Wiskott–Aldrich syndrome, DiGeorge syndrome, IgA deficiency, chronic granulomatous disease, and hyper-IgE syndrome were also studied, but only a small overall number of patients were available for analysis (n=18 cases in total).

Clinical notes at the time of biopsy and at follow-up were also reviewed. Data abstracted from the medical records included history of infections, comorbid health conditions, medications (including IVIg therapy), serum creatinine, and proteinuria. Patients without proteinuria on urinalysis did not have quantitative proteinuria or serum albumin values available. Follow-up serum creatinine and quantitative proteinuria values were also obtained. A history of infections, serum creatinine, and proteinuria values were available for the majority of patients. Serum creatinine values at the time of biopsy were available from 18 patients, 14 of which had follow-up creatinine values. Proteinuria data was available for 18 patients, ten of which had quantitative proteinuria. Seven of the ten patients with quantitative proteinuria values had follow-up quantitative proteinuria data. Quantitative proteinuria included both spot urine protein-to-creatinine ratio values and 24-hour urine collections.

A literature review was performed of all reported cases of native renal biopsies in patients with CVID that were indexed on PubMed, Scopus, or Google Scholar (n=24 manuscripts and two conference abstracts).

Renal Biopsy

All cases were processed for light, immunofluorescence, and electron microscopy. Formalin-fixed tissue was prepared for light and electron microscopy. For light microscopy, biopsies were fixed in formalin, dehydrated using alcohols, and embedded in paraffin wax after microwave processing. Paraffin sections were cut at 3 µm thickness, and stained with hematoxylin and eosin, periodic acid–Schiff, Masson trichrome, Jones methenamine silver, and Congo Red. For preparation of electron microscopy, 1 mm cubes of formalin-fixed tissue were removed, dehydrated with alcohols, and embedded in an epon–araldite resin. An ultramicrotome was used to cut 1-µm thick sections, stained with toluidine blue, and examined by light microscopy. Thin sections were examined by electron microscopy.

Samples for immunofluorescence microscopy were received in Michel transport medium, washed with Michel wash buffer, and frozen with optimal cutting medium at −20°C. Sections were cut at 4 µm thickness, rinsed, and stained with fluorescein-tagged rabbit anti-human polyclonal antibodies. Evaluation for IgG, IgA, IgM, C3, C1q, fibrinogen, and κ and λ light chain were performed for all cases. Dako antibodies were added for 30 minutes, rinsed, and cover-slipped in mounting medium. The staining intensity was semiquantitative on a 0–3+ scale. For membranous glomerulopathy (MG) cases, if sufficient frozen tissue was available, sections were stained with mouse anti-human antibodies to IgG1, IgG2, IgG3, and IgG4.

Immunohistochemical and Immunofluorescence Staining

For MG cases, immunohistochemistry for phospholipase A2 receptor (catalog number HPA012657, PLA2R, PLA2R rabbit polyclonal antibody; Sigma), and immunofluorescence for thrombospondin type 1 containing 7A (catalog number AMAB91234, THSD7A, THSD7A mouse monoclonal antibody; Atlas antibodies), exostosin-1 (catalog number PA5–60699, EXT1, EXT1 rabbit polyclonal antibody; Invitrogen), and neural EGF-like 1 (catalog number PA5–27958, NELL1, NELL1 rabbit polyclonal antibody; Invitrogen) were performed. PLA2R staining was performed on frozen tissue by immunohistochemistry. THSD7A, EXT1, and NELL1 staining was performed on formalin-fixed, paraffin-embedded tissue sections after pro-K digestion. THSD7A antibodies were crosslinked with Alexa Fluor 488 goat anti-mouse IgG for detection by immunofluorescence. EXT1 or NELL1 antibodies were crosslinked with Rhodamine Red-X-AffiniPure goat anti-rabbit IgG antibodies at 1:100 for 30 minutes (catalog number 111–295–144; Jackson Immunoresearch). Granular capillary loop staining was considered a positive result, with an absence of staining indicating a negative result. IgG subclasses were performed on cases of MG. IgG antibodies were all direct FITC conjugates (IgG1, Jackson Immunoresearch catalog number 115–095–205; IgG2, Jackson Immunoresearch catalog number 115–095–207; IgG3, Sigma catalog number F4641; and IgG4, Sigma catalog number F9889).

Immunophenotyping of interstitial lymphoid infiltrates in cases of tubulointerstitial nephritis (TIN) was done by CD3 (catalog number 103R-98; Cell Marque), CD4 (catalog number 104R-18; Cell Marque), CD8 (catalog number 108R-18; Cell Marque), CD20 (catalog number 120M-88; Cell Marque), and CD68 (catalog number 168M-98; Cell Marque) immunohistochemistry on formalin-fixed, paraffin-embedded tissue of patients with TIN. The relative percentage of T cells (CD3⁺) to B cells (CD20⁺), and the CD4/CD8 ratio for T cell–rich infiltrates was assessed, as were estimates by evaluating multiple ×400 microscopic fields.

For statistical analyses, t tests were used to evaluate differences between groups.

Results

Clinical Manifestations

All 22 patients in our cohort had a clinical diagnosis of CVID. Two patients had a genetic diagnosis, with one patient having LPS-responsive beige-like anchor gene deficiency and the other patient having an anomaly in chromosome pairs 3 and 18. In the majority of patients in our cohort, CVID was diagnosed by hypogammaglobulinemia for IgG, IgM, and IgA, identified in a work-up initiated because of recurrent or unusual infections. These included sinusitis, otitis media, pharyngitis, pneumonia, colitis (Clostridium difficile, cytomegalovirus), and skin infections. Eighteen of 22 patients received long-term IVIg therapy every 3–4 weeks. One patient underwent splenectomy, one patient underwent rituximab therapy, and one patient was deceased at follow-up.

The mean age of patients with CVID at the time of renal biopsy was 37.7 years (range, 2–69 years; SD±20.6 years). The mean serum creatinine was 1.97±1.76 mg/dl, which was higher on average for patients with TIN (mean, 2.75 mg/dl) than MG (mean, 1.62 mg/dl), although it was not statistically significant (P=0.54, unpaired t test). Mean proteinuria was 3.25 g±4.82 g/24 h.

Patients with CVID presented with a variety of immunodeficiency-related manifestations and comorbid conditions, including developmental, hematologic, cardiovascular, pulmonary, gastrointestinal, endocrine, oncologic, and rheumatologic disorders (Supplemental Table 1).

Patients with CVID with MG were more likely to have IgE-mediated comorbidities, including allergies, asthma, and eosinophilic esophagitis, than patients with acute or chronic TIN. As IgE is formed from cross-switching from IgG1 in germinal centers (⁵,⁶), there may be a link between IgG1-dominant MG and IgE-mediated disease manifestations.

Recurrent infections were frequent within the CVID cohort, but there was no stratification between type of infection and renal manifestations. Patients with MG and CVID had a history of chronic otitis media, upper respiratory infections (including pharyngitis [in two patients] and sinusitis), lower respiratory infections (including pneumonia in five patients and bronchiectasis in two patients), colitis (cytomegalovirus and C. difficile), and skin infections (in two patients). Patients with CVID with TIN similarly had a history of pharyngitis, pneumonia, and enteritis.

AKI and proteinuria were the leading indications for renal biopsy in CVID. Proteinuria was present at the time of biopsy for 18 of 27 biopsies (67%). AKI was present at the time of biopsy for ten patients (36%), and two patients had anuria (9%). Four patients underwent repeat biopsy, and the findings for three patients were similar, with only slight differences in activity and chronicity compared with the initial biopsy. In one case, a patient with a proliferative immune-complex GN developed MG in addition to proliferative immune-complex GN at follow-up.

Immune-Complex GN in Patients with CVID

Thirteen of 22 (59%) patients had immune-complex GN with capillary loop immune deposits. Nine patients had subepithelial deposits and met the criteria for MG, one patient had subendothelial deposits and was diagnosed with membranoproliferative GN with IgG3κ deposits, one patient had C3 glomerulopathy, and the remaining two patients had a proliferative immune-complex GN. In the patient with membranoproliferative GN with IgG3κ deposits, serum immunofixation and free light-chain assays were performed to exclude a paraprotein-related process, although presence of a clone below the limits of detection cannot be excluded. Immunofixation was negative for a paraprotein and a serum free light-chain assay showed a normal κ/λ ratio. One of the biopsy specimens containing mesangioproliferative GN had coexistence of a severe thrombotic microangiopathy with impending cortical necrosis in addition to an immune-complex GN. This biopsy specimen had ischemic, hypoperfused glomeruli, severe tubular injury with focal acute tubular necrosis, and severe microangiopathic changes within glomeruli and vessels.

Of the MG cases, eight of 13 cases contained mesangial deposits in addition to subepithelial and intramembranous deposits, whereas the remaining five had no mesangial or paramesangial deposits (Table 1). All cases were PLA2R, THSD7A, EXT1, and NELL1 negative (Supplemental Figure 1). IgG subclass staining shows strong IgG1 positivity along glomerular capillary loops in all four tested cases. IgG2, IgG3, and IgG4 were negative (Figure 1). IgG deposits were present within all cases and there was variable expression of IgA (seven of 13, 54%), IgM (six of 13, 46%), C3 (nine of 13, 69%), and C1q (four of 13, 23%). Global glomerulosclerosis ranged from 2% to 37%. Interstitial fibrosis was mild in 11 of 13 patients, moderate in one of 13 patients, and severe in one of 13 patients. There was moderate arteriosclerosis in three of 13 cases (23%), with no cases showing severe vascular disease. Four cases showed an associated proliferative component, with mesangial hypercellularity, endocapillary hypercellularity, and/or fibrinoid necrosis or crescent formation (Figure 2).

Table 1. - Immune-complex GN in patients with CVID

Parameter	Frequency of Finding, N=13
Male	6/13
Female	7/13
Proteinuria	13/13
Capillary loop immune deposits	13/13
Subepithelial	13/13
Subendothelial	2/13
Membranous glomerulopathy	9/13
With no mesangial deposits	1/9
With mesangial deposits	8/9
Proliferative GN with monoclonal IgG3κ deposits	1/13
C3-dominant membranous-like glomerulopathy	1/13
IgG	12/13
IgA	7/13
IgM	7/13
C3	12/13
C1q	4/13

Membranous glomerulopathy in patients with CVID. Clinical and renal biopsy characteristics from eight patients with CVID with capillary loop immune deposits are shown. The frequency of immunofluorescence positivity is identified. CVID, common variable immunodeficiency.

Figure 1.: MG in patients with CVID is IgG1 subclass restricted. A PAS stain showing a glomerulus with mild mesangial expansion and prominent capillary loops (A). A JMS stain showing thickened glomerular capillary loops with formation of holes (B). IgG subclasses showing IgG1 granular capillary loop deposits (C), which were negative for IgG2 (D), IgG3 (E), and IgG4 (F) immune deposits. CVID, common variable immunodeficiency; JMS, Jones methenamine silver; MG, membranous glomerulopathy; PAS, periodic acid–Schiff.

Figure 2.: Proliferative changes are identified within glomeruli in patients with CVID with MG. (A) PAS stain showing a membranoproliferative pattern of glomerular injury, magnification ×400. (B) PAS stain showing mesangial and endocapillary hypercellularity, magnification ×400. (C) Cellular crescent formation on JMS stain, magnification ×400. (D) Focus of fibrinoid necrosis in a glomerulus with endocapillary hypercellularity, hematoxylin and eosin stain, magnification ×400; arrow indicates area of fibrinoid necrosis. (E) Segmental sclerosis, PAS stain, magnification ×400; arrow indicates region of segmental sclerosis. (F) Subepithelial and mesangial electron-dense deposits seen by electron microscopy.

Electron microscopy was evaluated from the patients with CVID with MG to evaluate the stage of the electron-dense deposits. All cases showed at least stage 2 electron-dense deposits (range, stages 2–4), with reaction of the underlying glomerular basement membrane (Figure 2F).

At follow-up, patients with CVID with MG had a mean serum creatinine of 1.06±0.59 mg/dl. Proteinuria persisted in six patients. One patient died of complications arising from a concurrent acute thrombotic microangiopathy.

TIN in Patients with CVID

The second most common diagnosis in our CVID cohort was acute or chronic TIN, affecting six of 22 (27%) patients with CVID (Table 2). The lymphocytic infiltrates all contained >90% CD3⁺ T cells, and showed a normal CD4/CD8 ratio in four of six cases, whereas two cases had a reduced ratio (Figure 3). The normal ratio of CD4/CD8 T cells circulating within is approximately 1.5–2.5, and a ratio <1 is considered abnormal. Infiltrates within the interstitium should reflect that seen within the peripheral blood; however, this could not be confirmed, as only one patient had peripheral blood flow cytometry performed (one of the CD8⁺ T cell–dominant cases with a low CD4/CD8 ratio).

Table 2. - Acute tubulointerstitial nephritis in patients with CVID

Parameter	Frequency of Finding, N=6
Male	2/6
Female	4/6
Clinical presentation
AKI	4/6
Proteinuria	2/6
Interstitial inflammatory infiltrates
Patchy	1/6
Severe, diffuse	5/6
CD3+ >> CD20+	6/6
CD4+ > CD8+	4/6
CD8+ > CD4+	2/6
Tubulitis	6/6
Interstitial fibrosis
Mild	2/6
Severe	4/6
Arteriolosclerosis	3/6
None/mild	2/6
Moderate/severe	3/6
Unable to assess	1/6
Glomerular immune complexes	3/6 have IgG, κ, λ in mesangium

Acute tubulointerstitial nephritis in patients with CVID. Clinical and renal biopsy characteristics from six patients with CVID with acute tubulointerstitial nephritis are shown. Immunohistochemical staining to characterize lymphoid infiltrates included CD3, CD20, CD4, and CD8, and all cases displayed T cell dominance with >90% CD3⁺ T cells. CVID, common variable immunodeficiency; CD, cluster of differentiation.

Figure 3.: Some cases of tubulointerstitial nephritis in CVID are CD8 ⁺ T cell–predominant. A haematoxylin and eosin (H&E) stained section shows replacement of renal parenchyma with a diffuse lymphocytic infiltrate (A). Scattered CD68⁺ histiocytes are seen within the infiltrate (B). Immunohistochemistry shows that the infiltrate is comprised of predominantly CD3⁺ T cells (C), with rare CD20⁺ B cells (D). The infiltrate shows few CD4⁺ T cells (E) and is CD8⁺ T cell dominant (F), with a reduced CD4/CD8 ratio.

Mesangial immune complexes for IgG, κ, and λ were concurrently present in three of six patients with CVID with TIN (50%). Global glomerulosclerosis ranged from 0% to 87%. Interstitial fibrosis was moderate to severe in three of six cases (50%). In previously reported cases of TIN owing to CVID, all showed T cell–rich infiltrates, and four of seven (57%) were CD8⁺ T cell dominant (^{17–1819202112}). There was mild arteriosclerosis in one case and there were no cases that showed moderate or severe vascular changes. At follow-up, four of six patients had persistent CKD stage 3 or above (83%). One patient who had a CD8⁺ T cell–rich TIN underwent bone marrow biopsy and was found to have CD8⁺ natural killer–like T cell lymphoma. The other case of CD8⁺ T cell–rich TIN was found to be a benign, reactive infiltrate.

Other Renal Biopsy Findings in Patients with CVID

Other renal biopsy diagnoses included acute tubular injury (n=1), amyloid light-chain amyloidosis (n=1), diabetic glomerulosclerosis (n=1), thin glomerular basement membranes (n=1), pauci-immune sclerosing GN (n=1), and arterionephrosclerosis (n=1). Three of these cases showed mesangial immune complexes of IgM and/or C3. Congo Red staining was positive for the case of amyloid light-chain amyloidosis, with apple-green birefringence on polarization, but was negative for all of the remaining cases. A summary of the clinical and histopathologic features of all CVID cases in our cohort is provided in Supplemental Table 2.

Renal Biopsies of Primary Immunodeficiency in Patients without CVID

To compare CVID renal biopsy findings with those of other primary immunodeficiencies, we performed a database search of cases of Bruton agammaglobulinemia (n=2), Wiskott–Aldrich syndrome (n=1), severe combined immunodeficiency (n=2), DiGeorge syndrome (n=2), IgA deficiency (n=4), chronic granulomatous disease (n=1), hyper-IgE syndrome (n=1), and primary immunodeficiency not otherwise specified (n=5). The mean age of the patients was 29.7 years (range, <1 month to 86 years). Despite finding a low number of overall cases (n=18), nine patients with non-CVID primary immunodeficiency had a proliferative GN (mesangial and/or endocapillary proliferative, with or without the presence of crescents) and two had an MG. A single patient had acute TIN. Therefore, similar kidney biopsy manifestations may be seen in patients with other primary immunodeficiencies other than CVID. This data is provided in Table 3.

Table 3. - Histopathologic features of non-CVID primary immunodeficiencies

Patient	Immunodeficiency	Diagnosis	% GGS	IF/TA	AS	AH	Immunofluorescence
1	X-linked agammaglobulinemia	Proliferative immune-complex GN	36	20–30	Mild	None	Trace IgA, 1+ IgG, 2+ C3, mesangial + capillary loop
2	X-linked agammaglobulinemia	Proliferative immune-complex GN	12	None	None	None	Trace IgA, 3+ IgG, trace C3, 1–2+ C1q, mesangial + capillary loop
3	Wiskott–Aldrich	Proliferative GN plus membranous glomerulopathy	50	20–30	None	Mild	3+ IgA, 3+ IgG, trace IgM, 2–3+ C3, trace C1q, mesangial + capillary loop
4	Severe combined immunodeficiency	ATI	71	>50	None	Moderate	None
5	Severe combined immunodeficiency	Congenital nephrotic syndrome, Finnish type	4	10–20	None	None	None
6	DiGeorge syndrome	Diffuse proliferative immune-complex GN	11	30	None	None	3+ IgM, 1+ C3, 1–2+ C1q, mesangial + capillary loop
7	DiGeorge syndrome	Crescentic GN	17	<10	None	None	Pauci-immune w/positive ANCA serology
8	Chronic granulomatous disease	Crescentic GN	43	20	Severe	Moderate	Pauci-immune w/positive ANCA serology
9	Job syndrome/hyper-IgE	Advanced sclerosing glomerulopathy	74	80	Severe	Severe	IgA 2+, IgG 3+, IgM 3+, κ 2+, λ 2+, mesangial and capillary loop
10	Primary immunodeficiency, not otherwise specified	Membranous glomerulopathy	0	None	None	None	IgG 2+, κ 2+, λ 2+ capillary loop deposits
11	Primary immunodeficiency, not otherwise specified	FSGS	0	None	None	None	C3 2+ mesangial
12	Primary immunodeficiency, not otherwise specified	Acute tubular injury	0	10	No arteries for evaluation	None	Negative within glomeruli
13	Primary immunodeficiency, not otherwise specified	Moderate global glomerulosclerosis, IF/TA, and arteriosclerosis	29	40	Moderate	None	None
14	Primary immunodeficiency, not otherwise specified	Mesangial immune-complex deposition, increased GGS	39	10	Moderate	Mild	IgM 3+, C3 1+, κ 2+, λ 3+ mesangial
15	IgA deficiency	Acute interstitial nephritis, mesangial immune-complex deposition	0	5–10	None	None	IgG 2+, IgM 2+, C3 1–2+, κ 2+, λ 2+ mesangial
16	IgA deficiency	Mild global glomerulosclerosis, IF/TA, and arteriosclerosis	5	20–30	Mild to moderate	Mild	None
17	IgA deficiency	Minimal change disease	0	None	None	None	None
18	IgA deficiency	Mesangiopathic glomerulopathy with IgG deposits	44	20	Severe	Moderate	IgG 2+, κ 1+, λ 1+ mesangial

The diagnoses; glomerular, tubulointerstitial, and vascular scarring; and immunofluorescence staining are shown. + refers to positive staining and is graded on a 1+ to 3+ scale. 1+ indicates mild staining, 2+ indicates moderate, and 3+ indicates severe. CVID, common variable immunodeficiency; GGS, global glomerulosclerosis; IF/TA, interstitial fibrosis/tubular atrophy; AS, arteriosclerosis; AH, arteriolar hyalinosis; ATI, acute tubular injury.

Discussion

The most common findings on renal biopsy in patients with CVID were MG and acute or chronic TIN. MG in patients with CVID often had mesangial immune deposits and were PLA2R, THSD7A, EXT1, and NELL1 negative with IgG1 restriction (¹³). Identification of MG in patients with CVID is important as nephrotic-range proteinuria results in increased renal losses of Ig, worsening their immunocompromised state and increasing susceptibility to infections. In these patients, it is difficult to achieve IgG levels >500 mg/dl despite high doses of IgG replacement (¹⁴). Persistently low Ig levels increases risk of recurrent infections and their complications. A total of 26 CVID biopsies were reported, including 24 manuscripts and two abstracts. Of these cases, eight patients had an immune-complex GN (^{9–10111213141522}), including six patients with MG, and two with a membranoproliferative GN.

Eight patients had an acute interstitial nephritis (^{7–1819202112}), one of which was granulomatous and one of which occurred concurrently with MG. Eight cases showed amyloid A amyloidosis (^{23–24252627282930}). Other reported cases include minimal change disease (n=1) (³¹), antiglomerular basement membrane nephritis (n=1) (³²), and acute tubular necrosis (³³) (n=1, see Table 4). Although our findings of an increased incidence of MG and acute interstitial nephritis parallel that reported in the literature, we did not have any cases in our series of amyloid A amyloidosis. This may represent improved control of chronic infections in this patient cohort.

Table 4. - Previously reported cases of CVID-associated kidney disease in the literature

Case No.	Age	Sex	Diagnosis	Creatinine	Proteinuria	Pathology	Treatment	Outcomes	Reference
1	13	M	Membranous	0.39	7.7 g	IgG, IgM, C3, C4, C1q deposits	Methylprednisolone, cyclosporine, IVIg	Reduced proteinuria	(15)
2	15	M	Membranous and AIN	NA	NA	IgG capillary loop deposits	Antimicrobial agents	Died at age 15 yr; pneumonia	(16)
3	17	F	MPGN	1.1	3 g	IgM, C3, C1q deposits	Methylprednisolone, IVIg, rituximab	No proteinuria at 4 mo	(17)
4	33	F	Membranous	222 μmol/L	2.7 g	IgG, IgM, C3 deposits	Methylprednisolone	Died at age 38 yr; pneumonia	(22)
5	36	F	Membranous	0.6	8 g	IgG, C3 deposits	Corticosteroids, rituximab	Reduced proteinuria	(18)
6	55	M	Membranous	0.92	0.9	IgG capillary loop deposits	IVIg	Reduced proteinuria	(19)
7	59	M	Membranous	NA	25 g	IgG capillary loop deposits	Methylprednisolone	Died at age 61 yr	(20)
8	22	M	MPGN	NA	0.9 g	IgA, IgM, IgG, C3, C1q deposits	Corticosteroids	Improved renal function	(21)
9	8	F	AIN	10.1	NA	Mononuclear infiltrate, 50% IF	Peritoneal dialysis	Listed for renal transplantation	(7)
10	21	F	AIN	164 μmol/L	NA	CD8+ T cell infiltrate	IVIg	Stable renal function	(8)
11	24	F	AIN	NA	NA	CD8+ T cell infiltrate; severe IF	Prednisone, MMF, cyclosporine	Dialysis, listed for transplant	(9)
12	28	F	AIN	2.7	0.4 g	CD3+ T cell infiltrate, CD4>8	Corticosteroids	Recovery, then recurrence	(10)
13	31	F	AIN	NA	NA	CD8+ T cell infiltrate	Corticosteroids	Improved renal function	(11)
14	44	M	AIN	WNL	NA	CD8+ T cell infiltrate	Corticosteroids, azathioprine	Improved renal function	(12)
15	47	M	GIN	2.1	NA	Non-necrotizing granulomas	Corticosteroids	Improved renal function	(34)
16	24	F	Amyloid A amyloidosis	1.1	4.5	Amyloid A deposition	IVIg	NA	(23)
17	24	F	Amyloid A amyloidosis	NA	NRP	Amyloid A deposition (thyroid)	IVIg	NA	(24)
18	28	M	Amyloid A amyloidosis	NA	9	Amyloid A deposition	Ciprofloxacin, metronidazole	No improvement	(25)
19	29	F	Amyloid A amyloidosis	NA	9	Amyloid A deposition	Antibiotics, IVIg, losartan, cilazapril	Persistent NRP	(26)
20	29	M	Amyloid A amyloidosis	1.8	11.8	Amyloid A deposition	Losartan, ramipril, IVIg	Reduced proteinuria	(27)
21	40	M	Amyloid A amyloidosis	NA	3.4	Amyloid A deposition	Angiotensin receptor blocker	NA	(28)
22	48	M	Amyloid A amyloidosis	NA	“Massive”	Amyloid A deposition	Antibiotics, IVIg	No infections in 2 yr	(29)
23	49	F	Amyloid A amyloidosis	233 μmol/L	5.5	Amyloid A deposition	IVIg	Stable renal function	(30)
24	32	M	ATN (suspected)	1.33	NA	Biopsy not performed	Cessation of IVIg	Improved renal function	(33)
25	15	F	Anti-GBM nephritis	486.2 μmol/L	29.4	Necrotizing crescents, linear IgG staining	Corticosteroids, rituximab, cyclophosphamide	Died, age 15 CMV encephalomyelitis	(32)
26	12	M	Minimal change	0.5	8.2	Mesangial hypercellularity, diffuse FPE	Corticosteroids	No recurrence at 2 yr	(31)

CVID, common variable immunodeficiency; M, male; IVIg, intravenous Ig; AIN, acute interstitial nephritis; NA, not applicable; F, female; MPGN, membranoproliferative GN; IF, immunofluorescence; MMF, mycophenolate mofetil; WNL, within normal limits; GIN, granulomatous interstitial nephritis; NRP, nephrotic-range proteinuria; ATN, acute tubular injury; GBM, glomerular basement membrane; CMV, cytomegalovirus; FPE, foot process effacement.

Patients with CVID with acute or chronic TIN had CD3⁺ T cell–predominant inflammatory infiltrates and increased tubulointerstitial fibrosis. Patients with chronic TIN had a higher creatinine level and worsened kidney function than patients with an immune-complex GN. In a study of bone marrow biopsy specimens from patients with CVID, diffuse and nodular CD3⁺ T cell infiltrates were common within the marrow, which correlates with an increase in CD4, CD45RO, and soluble IL-2R expression on T cells. This correlates with increased memory T cells, as well as a block in B cell development from the pre–B-1 to the pre–B-II phase (³⁵).

Autoimmunity is common in patients with CVID because of several possible mechanisms. There are B cell receptor-editing defects, which lead to a loss of checkpoint control. These defects can lead to persistence of autoreactive B cell clones that escape apoptosis. In addition to B cell defects, there is defective T cell function, with impaired CD4⁺ T cell activation and a depletion of regulatory T cells. Chronic and recurrent infections could also induce autoimmune responses by bystander lymphocyte activation, superantigen activation, and via epitope spreading. As CVID is a heterogeneous disease resulting from a loss of checkpoint inhibition, its histopathology mimics nephrotoxicities arising from checkpoint inhibitors. Interestingly, in a 10-year, single-center case series of renal pathology of checkpoint inhibitor therapies, the findings included acute TIN, MG, C3 glomerulopathy, and amyloid A amyloidosis (³⁶)—the same findings seen in patients with CVID.

In summary, CVID is a primary immunodeficiency disorder characterized by hypogammaglobulinemia, recurrent infections, and systemic autoimmune/inflammatory manifestations. A majority of patients with CVID do not display renal manifestations, but those who do show a disproportionate increase in MG and T cell–rich acute TIN. These were the leading causes of nephropathology in patients with CVID who underwent biopsy in our cohort and within the literature. CVID should be considered as a potential cause of PLA2R-, THSD7A-, EXT1-, and NELL1-negative MG or acute interstitial nephritis, especially when seen in a young patient without a known cause, who has a history of repeated infections.

Disclosures

All authors have nothing to disclose.

Funding

None.

Supplemental Material

This article contains supplemental material online at http://kidney360.asnjournals.org/lookup/suppl/doi:10.34067/KID.0000432020/-/DCSupplemental.

Supplemental Figure 1. Membranous glomerulopathy in CVID patients is PLA2R (A), THSD7A (B), EXT (C), and NELL1 (D) negative for granular capillary loop staining.

Supplemental Table 1. Co-morbid conditions in CVID patients with kidney disease.

Supplemental Table 2. Histopathology and follow-up data from patients with common variable immunodeficiency