Proliferative Glomerulonephritis With Monoclonal IgG Deposits Recurs or May Develop De Novo in Kidney Allografts

Case Report Proliferative Glomerulonephritis With Monoclonal IgG Deposits Recurs or May Develop De Novo in Kidney Allografts Alia Albawardi, MD, Anjali Satoskar, MD, Jon Von Visger, MD, Sergey Brodsky, MD, Gyongyi Nadasdy, MD, and Tibor Nadasdy, MD Proliferative glomerulonephritis with monoclonal immunoglobulin G (IgG) deposits (PGNMIGD) is a recently recognized glomerular disease. Light microscopy usually resembles membranoproliferative glomerulonephritis. Glomerular deposits are mostly IgG3 ␬; however, unlike in the usual forms of monoclonal immunoglobulin deposition disease, extraglomerular deposits are absent. If PGNMIGD is secondary to the glomerular deposition of circulating monoclonal IgG, it is expected to recur in kidney allografts with the same pattern of monoclonal IgG deposition. We reviewed our kidney biopsy files between January 1, 2003, and January 4, 2010, and identified 21 biopsy specimens with PGNMIGD, mostly with glomerular IgG3 ␬ deposits. Of the 21 biopsy specimens, 4 were from kidney allografts; 2 were recurrent and the other 2 were de novo diseases. Recurrent PGNMIGD develops rapidly, causing proteinuria. This rapid recurrence of PGNMIGD in kidney allografts provides further proof that PGNMIGD is secondary to the glomerular deposition of circulating monoclonal IgG. Am J Kidney Dis. 58(2):276-281. © 2011 by the National Kidney Foundation, Inc. INDEX WORDS: Proliferative glomerulonephritis with monoclonal IgG deposit disease; renal kidney allograft; immunoglobulin G3 (IgG3) ␬.

P

roliferative glomerulonephritis with monoclonal immunoglobulin G (IgG) deposits (PGNMIGD) is a recently described disease entity.1,2 PGNMIGD represents a form of proliferative glomerulonephritis with electron-dense deposits localized to glomeruli and immunofluorescence findings indicating monoclonal IgG deposits. Most patients present with heavy proteinuria and variable degrees of decreased kidney function. The most commonly seen light microscopic pattern is hypercellular lobular-nodular glomerulonephritis (membranoproliferative glomerulonephritis [MPGN] pattern), followed by the pattern of endocapillary proliferative glomerulonephritis. Using immunofluorescence, most patients have IgG3 ␬ deposits.2 Complement deposits, including C3 and C1q, are frequent and are noted in most biopsy specimens; however, serum complement levels are within reference values in most patients.2 Electron microscopy (EM) shows mainly subendothelial and mesangial electron-dense deposits. PGNMIGD is clearly different from monoclonal immunoglobulin deposition disease, including light chain deposition disease, heavy chain deposition disFrom the Ohio State University Medical Center, Columbus, OH. Received November 8, 2010. Accepted in revised form May 6, 2011. Originally published online June 27, 2011. Address correspondence to Tibor Nadasdy, MD, Renal Pathology Laboratory, M015/018 Starling Loving Hall, 320 West 10th Ave, Columbus, OH 43210-1240. E-mail: [email protected] © 2011 by the National Kidney Foundation, Inc. 0272-6386/$36.00 doi:10.1053/j.ajkd.2011.05.003 276

ease, or heavy and light chain deposition disease. In PGNMIGD, deposits are localized to glomeruli only, unlike in other forms of monoclonal immunoglobulin deposition disease, in which the monoclonal immunoglobulin is deposited along extraglomerular structures, including the tubular basement membrane, vasculature, peritubular capillary basement membrane, and interstitium.3-5 Also, Nasr et al1 did not find a detectable deletion in any constant domains of the IgG heavy chain in PGNMIGD. Unlike monoclonal immunoglobulin deposition disease, most patients have no detectable monoclonal spike in serum or urine, although the frequency of monoclonal spike is more common than in the general population.2 It is possible that reactive proliferating B cells may produce undetectable monoclonal IgG, which (particularly IgG3) can be absorbed quickly by the glomerular matrix and self-aggregate there as deposits. Other differential diagnostic possibilities include MPGN, cryoglobulinemic glomerulonephritis, and infection-related glomerulonephritis.2,5 The differential diagnosis usually is easy if immunofluorescence studies with antibodies to the IgG subclasses are performed. If PGNMIGD is secondary to glomerular deposition of circulating monoclonal IgG, one would expect recurrence of the disease with the same glomerular pattern of monoclonal IgG deposition in kidney allografts. Therefore, we reviewed our kidney biopsy files from January 1, 2003, to March 15, 2010, and identified all glomerulonephritides with monoclonal-appearing IgG deposits. Using antibodies to IgG subclasses (reagents obtained from The Binding Site, www.theAm J Kidney Dis. 2011;58(2):276-281

Monoclonal IgG Deposits in Kidney Allografts

bindingsite.com), as described previously,2 we identified 4 kidney allograft recipients with PGNMIGD. Next, we detail our findings in these 4 cases (Table 1).

CASE REPORTS Case 1 A 59-year-old obese white man developed end-stage renal disease (ESRD) secondary to biopsy-proven PGNMIGD with glomerular IgG3 ␬ deposits. He received a deceased donor kidney transplant on April 12, 2006. He was administered antithymocyte globulin (Thymoglobulin) induction with rapid prednisone taper. In May 2007, serum creatinine level increased to 2.7 mg/dL (238.7 ␮mol/L; corresponding to estimated glomerular filtration rate [eGFR] of 26 mL/min/1.73 m2 [0.43 mL/s/1.73 m2] using the MDRD [Modification of Diet in Renal Disease] Study equation), and the patient was found to have nephrotic syndrome. A kidney transplant biopsy specimen showed glomeruli with an MPGN pattern similar to that in the patient’s native kidney biopsy specimen (Fig 1A and B). Signs of acute rejection were not seen. Immunofluorescence studies showed strong glomerular staining for IgG3 and ␬ light chain (Fig 1C). Using EM, numerous subendothelial and mesangial electron-dense deposits were seen, with no patterned substructure, just like in the patient’s native kidney biopsy (Fig 1D and E). The diagnosis of recurrent PGNMIGD was made. After the biopsy, the patient’s serum creatinine level varied between 1.7 mg/dL (150.3 ␮mol/L; eGFR, 41 mL/min/1.73 m2 [0.68 mL/s/1.73 m2]) and 2.5 mg/dL (221.0 ␮mol/L; eGFR, 27 mL/min/1.73 m2 [0.45 mL/s/1.73 m2]). The patient sustained a fatal heart attack 20 months after the kidney biopsy and died with a functioning kidney allograft.

Case 2 A white man presented to an outside institution in November 1997 at the age of 57 years with decreased kidney function and nephrotic-range proteinuria. A kidney biopsy was performed. There were 4 glomeruli for light microscopy. They showed diffuse intracapillary hypercellularity with lobular/nodular mesangial expansion. The biopsy specimen was small and no tissue was available for immunofluorescence. Only 1 globally sclerotic glomerulus was available for EM. According to the hospital policy at that institution, the biopsy material was discarded 10 years after the biopsy. The patient received a living related kidney transplant on December 14, 2004, with antithymocyte globulin induction and rapid steroid taper. His serum creatinine level increased to 3.9 mg/dL (397.8 ␮mol/L; eGFR, 40 mL/min/1.73 m2 [0.27 mL/s/1.73 m2]) by early 2006. Urine cytology tests were positive for polyomavirus inclusions. Two kidney transplant biopsies were performed on October 3, 2006, and December 19, 2006, with similar findings in both. Polyomavirus inclusions were noted in tubular epithelial cells, a finding confirmed by immunostaining for polyomavirus (using a mouse monoclonal antibody directed against the simian virus 40 large T antigen; Calbiochem, www.emdchemicals.com). Immunofluorescence studies showed prominent coarsely granular mesangial staining for IgG (IgG3 only), C3, and ␬ light chain. Focal mild granular tubular basement membrane IgG containing deposits also was seen, but these tubular basement membrane deposits stained for IgG1, IgG2, and IgG3, as well as ␬ and ␭ light chains. Such tubular basement membrane deposits are common in polyomavirus nephropathy. Diagnoses of progressive polyomavirus nephropathy and probable recurrent PGNMIGD with IgG3 ␬ deposits were made. The patient’s immunosuppression therapy was changed to mycophenolic acid (Myfortic), 360 mg, twice daily and prednisone, 10 Am J Kidney Dis. 2011;58(2):276-281

mg/d. Serum creatinine level decreased slightly and was between 3.3 mg/dL (295.8 ␮mol/L; eGFR, 19 mL/min/1.73 m2 [0.31 mL/s/1.73 m2]) and 3.7 mg/dL (327.1 ␮mol/L; eGFR, 17 mL/min/ 1.73 m2 [0.28 mL/s/1.73 m2]) for the next 11 months. Serum BK viral titers decreased from 17 billion copies to 12,000. In March 2008, the patient developed cryptococcal meningitis and died with a functioning kidney allograft.

Case 3 This 68-year-old white woman developed ESRD secondary to polycystic kidney disease. She received a deceased donor kidney on September 8, 1992, with antilymphocyte and steroid induction. In November 2005, she noticed increasing peripheral edema and had a serum creatinine level of 2.6 mg/dL (229.8 ␮mol/L; eGFR, 18 mL/min/1.73 m2 [0.30 mL/s/1.73 m2]) and 24-hour proteinuria with protein excretion of 2.2 g. Kidney biopsy showed hypercellular glomeruli with an MPGN pattern. Immunofluorescence showed strong, diffuse, linear, ribbon-like glomerular capillary staining for IgG (IgG1 only) and ␬ light chain. Ultrastructurally, mild, segmentally moderate, subendothelial, electron-lucent widening was seen with segmental mesangial cell interposition in some glomerular capillaries. Finely granular electron-dense deposits were present along the endothelial aspect of the lamina densa of the glomerular basement membrane. The diagnosis of PGNMIGD with features of transplant glomerulopathy was made. Graft function gradually deteriorated and hemodialysis therapy was started 20 months after the kidney transplant biopsy.

Case 4 This 24-year-old white man with diabetes mellitus type 1 since the age of 6 years underwent simultaneous kidney-pancreas transplant on August 2, 2006, for end-stage diabetic nephropathy. In February 2009, serum creatinine level increased to 2.8 mg/dL (247.5 ␮mol/L; eGFR, 23 mL/min/1.73 m2 [0.38 mL/s/1.73 m2]). An allograft biopsy specimen with more than 30 glomeruli showed acute tubular necrosis, but no rejection. Immunofluorescence was nonspecific, with only mild segmental glomerular IgG, IgM, and ␬ and ␭ light chains. Immunofluorescence using antibodies to IgG subclasses showed mild segmental mesangial staining for IgG3 only. C4d stain was negative in peritubular capillaries. EM did not show electron-dense immune-type deposits. The patient was noted to have proteinuria with protein excretion of 1.7 g/24 h and serum creatinine level of 2.8 mg/dL (247.5 ␮mol/L; eGFR, 23 mL/min/ 1.73 m2 [0.38 mL/s/1.73 m2]) 6 months later. Therefore, another biopsy was performed in September 2009. Biopsy findings were similar to those in February 2009, except acute tubular necrosis was not obvious now. Ultrastructurally, no discrete electron-dense immune-type deposits were noted. At the end of February 2010, serum creatinine level increased to 3.3 mg/dL (291.7 ␮mol/L; eGFR, 19 mL/min/1.73 m2 [0.32 mL/s/1.73 m2]). A third allograft biopsy was performed on March 8, 2010. Glomeruli now showed an MPGN pattern with IgG3 ␬ deposits (Table 1).

DISCUSSION Recurrent glomerular diseases usually develop early (within the first 2 years) posttransplant, whereas de novo glomerular diseases usually develop several years after kidney transplant.6 PGNMIGD in kidney allografts appears to follow the same pattern. The 4 recurrent PGNMIGD cases recently published by Nasr et al7 were diagnosed on protocol kidney allograft biopsy specimens obtained 3-5 months posttransplant. Our cases also indicate that PGNMIGD recurs early 277

278 Table 1. Clinical and Kidney Biopsy Findings in Kidney Transplant Recipients With PGNMID Patient 1

Age at biopsy (y) Race/sex Cause of ESRD

61 White/man PGNMID (IgG3 ␬)

Tx type Tx date Diagnostic biopsy date Immunosuppression PRA levels Baseline SCr (mg/dL) SCr at diagnostic biopsy (mg/dL) eGFR (mL/min/1.73 m2) Proteinuria (g/d) Hematuria Monoclonal spikea Serum complement (mg/dL) HBV/HCV infection Outcome after allograft biopsy

␬

3

4

68 White/woman PKD

24 White/man T1DM

Deceased donor 4/12/2006 5/7/2007 Mycophenolic acid, sirolimus

68 White/man GN w/ MPGN pattern (probably PGNMID) Living related donor 12/14/2004 10/3/2006 Mycophenolic acid, sirolimus

Deceased donor 9/8/1992 11/7/2005 Prednisone, cyclosporine

0 1.3 2.7

0 1.5 3.2

0 1.5 2.6

Simultaneous kidney-pancreas 8/2/2006 3/8/2010 Mycophenolate mofetil, tacrolimus, prednisone 0 1.8 3.3

24 8.7 2⫹ None ND Negative Fatal MI w/ functioning graft at 20 mo

19 0.5 2⫹ None ND Negative Fatal cryptococcal meningitis w/ functioning graft at 18 mo 1 y 10 mo 27/1 Mesangioproliferative 3⫹ 2⫹ (polyomavirus) 2⫹/2⫹ 6/0

18 1.5 1⫹ None C3, 103; C4, 22 Negative Dialysis at 20 mo 13 y 8/5 MPGN-like 3⫹ 1⫹ 3⫹/2⫹ 6/1

23 1.7 3⫹ None ND Negative Alive w/ functioning graft (short follow-up) 3 y 7 mo 13/8 MPGN-like 2⫹ 1⫹ Present/NAb 7/2

3⫹ diffuse granular mesangial (IgG3) 3⫹ diffuse granular mesangial

3⫹ diffuse linear glomerular capillaries (IgG1) 3⫹ linear glomerular capillaries

3⫹ diffuse granular to smudgy mesangial (IgG3) 1⫹ diffuse granular to smudgy paramesangial

1y 9/1 MPGN-like 1⫹ 1⫹ None/1⫹ 1/0 2⫹ diffuse granular glomerular capillaries & mesangial (IgG3) 3⫹ diffuse granular glomerular capillaries & mesangial

(Continued)

Albawardi et al

Am J Kidney Dis. 2011;58(2):276-281

Biopsy date post-Tx No. of glomeruli/no. globally sclerosed Glomerular injury pattern Interstitial fibrosis Interstitial inflammation Arteriolar hyaline/arterial intimal thickening No. of glomeruli for IF/no. globally sclerosed IF IgG

2

Am J Kidney Dis. 2011;58(2):276-281

Note: For features graded on a scale of 1-3, 1⫹ is mild; 2⫹, moderate; and 3⫹, severe. Conversion factors for units: SCr in mg/dL to ␮mol/L, ⫻88.4; eGFR in mL/min/1.73 m2 to mL/s/1.73 m2, ⫻0.01667. Abbreviations: EDD, electron-dense deposits; eGFR, estimated glomerular filtration rate (calculated using the Modification of Diet in Renal Diseases [MDRD] Study equation); EM, electron microscopy; ESRD, end-stage renal disease; GN, glomerulonephritis; HBV, hepatitis B virus; HCV, hepatitis C virus; IF, immunofluorescence; IgG4, immunoglobulin G4; MI, myocardial infarction; MPGN, membranoproliferative glomerulonephritis; NA, not applicable; ND, not done; PGNMID, proliferative glomerulonephritis with monoclonal immunoglobulin G deposits; PKD, polycystic kidney disease; PRA, panel reactive antibody; SCr, serum creatinine; T1DM, type 1 diabetes mellitus; Tx, transplant. a Serum and urine immunofixation. b No artery.

Subendothelial, intramembranous, & subepithelial EDD Subendothelial & mesangial EDD; subendothelial electron-lucent widening Subendothelial & mesangial EDD EM

Other relevant findings

1⫹ segmental glomerular capillaries & mesangium 2⫹ C3 granular mesangial & glomerular capillary loops

␭

Subendothelial & mesangial EDD

None 2⫹ C3 smudgy mesangial staining

None

No staining No staining

3 1

2

Patient

Table 1 (Cont’d). Clinical and Kidney Biopsy Findings in Kidney Transplant Recipients With PGNMID

No staining

4

Monoclonal IgG Deposits in Kidney Allografts

Figure 1. Recurrent proliferative glomerulonephritis with monoclonal immunoglobulin G (IgG) deposits (PGNMIGD) in patient 1. (A) Membranoproliferative glomerulonephritis (MPGN) pattern of glomerular injury in the patient’s native kidney biopsy specimen. (Hematoxylin and eosin [H&E]; original magnification, ⫻400.) (B) The same pattern of glomerular injury was observed in the allograft biopsy 1 year posttransplant. (H&E; original magnification, ⫻400.) (C) Immunofluorescence with antibodies to IgG subclasses in the 1-year posttransplant allograft biopsy specimen, just like in the patient’s native kidney biopsy specimen, showed only IgG3 staining. (Original magnification, ⫻400.) (D, E) Subendothelial electron-dense deposits in the (D) native and (E) transplant kidney biopsy specimens. (Uranyl acetate lead citrate; original magnification, ⫻20,000.)

posttransplant in kidney allografts. Patient 1 had obvious PGNMIGD with an MPGN pattern and IgG3 ␬ deposits already 12 months posttransplant (unfortunately, a biopsy at an earlier time was not performed). Morphologic findings were similar to the patient’s native kidney biopsy specimen. We believe that patient 2 also had recurrent disease, although we have insufficient proof for that. However, because glomerulonephritis with an MPGN pattern was diagnosed on the patient’s native kidney biopsy specimen at the age of 57 years is supportive of that conclusion. Unfortunately, this biopsy specimen was small, no immunofluorescence was performed, and only 1 globally sclerotic glomerulus was available for EM. However, because the patient had obvious IgG3 ␬ mesangial and glomerular capillary deposits 2 years posttransplant makes it unlikely that he had de novo disease. In 279

Albawardi et al

patient 3, it is almost certain that PGNMIGD developed de novo in the allograft. This patient had polycystic kidney disease as the cause of ESRD and the biopsy diagnostic of PGNMIGD was performed 13 years posttransplant. In patient 4, it is more difficult to prove the de novo nature of the disease, but it is likely. This patient had a simultaneous kidney-pancreas transplant at the young age of 21 years. He developed ESRD secondary to type 1 diabetes. Although the presence of diabetic nephropathy in the native kidney was not biopsy proven, the clinical history was entirely consistent with progressive diabetic nephropathy. A further argument for de novo disease is that the patient’s kidney allograft biopsy specimens 2 years 6 months and 3 years posttransplant were not truly consistent with PGNMIGD; the diagnostic changes were seen in his allograft biopsy 3 years 6 months posttransplant. In our opinion, recurrent disease would have caused already obvious glomerular deposits in the earlier biopsy specimens. Therefore, our study indicates that PGNMIGD may recur not only in kidney allografts, but also develop de novo. Although a recurrence rate for PGNMIGD cannot be established based on our patients and the 4 patients published by Nasr et al,7 we assume that it is high. At this stage, it is difficult to predict the true natural history of recurrent (or de novo) PGNMIGD in kidney allografts; it probably is similar to the natural history of the native kidney disease, which usually is slowly progressive.2 Proteinuria is present, but may not be nephrotic range. In our patients, comorbid conditions (eg, polyoma virus nephropathy, infection, and cardiovascular disease) or medications probably modified the disease course. Unfortunately, there is no proven effective treatment. Transplant recipients are on continuous immunosuppression therapy, which does not appear to prevent recurrence or de novo development of this disease. Nasr et al7 propose aggressive immunosuppression with rituximab and/or cyclophosphamide, but the effectiveness of this treatment in patients with PGNMIGD remains to be established. Morphologic characteristics of PGNMIGD in kidney allograft biopsy specimens are not different from PGNMIGD in native kidney biopsy specimens and the differential diagnostic approach is the same.2,7 One additional differential diagnostic problem could be transplant glomerulopathy. In transplant glomerulopathy, nonspecific IgG staining along the glomerular capillaries is not unusual and electron densities may be found along glomerular capillaries in the widened subendothelial space. However, IgG staining in transplant glomerulopathy usually is weak; other immunoreactants, including IgM, complement, and ␬ and ␭ light chains; also are present; and IgG subclass staining, if performed, does not indicate monoclonality. Also, by 280

ultrastructural examination, electron-dense deposits in transplant glomerulopathy are less distinct than in PGNMIGD and are localized to the widened subendothelial space rather than the glomerular basement membrane along glomerular capillaries. The clinical presentation and laboratory findings in PGNMIGD are nonspecific. Despite the monoclonal nature of glomerular deposits, only ⬃30% of patients have a detectable monoclonal spike in serum or urine.2 None of our transplant recipients had a detectable monoclonal spike in serum or urine. As mentioned in the introduction, it is likely that monoclonal IgG (usually IgG3 ␬) has a strong affinity to glomerular matrix and even miniscule immeasurable amounts of monoclonal IgG from serum could be absorbed by glomeruli. However, an extensive evaluation for a paraproteinemia/plasma cell dyscrasia that includes serum and urine immunofixation, serum quantitative free light chain measurement, and bone marrow biopsy is warranted. Monoclonal immunoglobulins, including IgG3, can activate the classic complement pathway, and glomerular complement deposition is common in PGNMIGD. However, only ⬃10% of patients with PGNMIGD have low serum complement levels.2 At our institution, serum complement levels in transplant recipients are not measured routinely; only 1 of our 4 patients was tested (this patient had normal serum complement levels) after the diagnosis. Normal serum complement levels in patients with PGNMIGD are explained best by the finding that monoclonal IgG deposits in only glomeruli, and the quantity is not sufficient to cause serum complement depletion. However, measuring serum complement is warranted in these patients. In summary, our study indicates that PGNMIGD may recur rapidly in kidney allografts. De novo PGNMIGD usually appears later after transplant. The natural history of recurrent or de novo PGNMIGD in kidney allografts is difficult to determine based on our limited material, but probably follows a course similar to that in native kidneys.

ACKNOWLEDGEMENTS Portions of these findings were presented in abstract form at the United States Canadian Academy of Pathology Annual Meeting in 2010; Albawardi A et al. Modern Pathol. 2010; 23:337A. Support: None. Financial Disclosure: The authors declare that they have no relevant financial interests.

REFERENCES 1. Nasr SH, Markowitz GS, Stokes MB, et al. Proliferative glomerulonephritis with monoclonal IgG deposits. A distinct entity mimicking immune-complex glomerulonephritis. Kidney Int. 2004; 65(1):85-96. Am J Kidney Dis. 2011;58(2):276-281

Monoclonal IgG Deposits in Kidney Allografts 2. Nasr SH, Satoskar A, Markowitz GS, et al. Proliferative glomerulonephritis with monoclonal IgG deposits. J Am Soc Nephrol. 2009;20(9):2055-2064. 3. Lin J, Markowitz GS, Valeri AM, et al. Renal monoclonal immunoglobulin deposition disease: the disease spectrum. J Am Soc Nephrol. 2001;12(7):1482-1492. 4. Herzenberg AM, Kiaii M, Magil AB. Heavy chain deposition disease: recurrence in a renal transplant and report of IgG2 subtype. Am J Kidney Dis. 2000;35(5):E25.

Am J Kidney Dis. 2011;58(2):276-281

5. Herrera GA. Renal lesions associated with plasma cell dyscrasias practical approach to diagnosis, new concepts, and challenges. Arch Pathol Lab Med. 2009;133(2):249-267. 6. Ivanyi B. A primer on recurrent and de novo glomerulonephritis in renal allografts. Nat Clin Pract Nephrol. 2008;4(8): 446-457. 7. Nasr SH, Sethi S, Cornell LD, et al. Proliferative glomerulonephritis with monoclonal IgG deposits recurs in the renal allograft. Clin J Am J Soc Nephrol. 2007;6(1):122-132.

281