Immunoglobulin A–dominant postinfectiousglomerulonephritis: frequent occurrence in nondiabeticpatients with Staphylococcus aureus infectionSuchin Worawichawong MDa, Louis Girard MDb, Kiril Trpkov MDa,James C. Gough MDa, Daniel B. Gregson MDa, Hallgrimur Benediktsson MDa,⁎
aDepartment of Pathology and Laboratory Medicine and Calgary Laboratory Services, Foothills Medical Centre,University of Calgary, Alberta, Canada T2T 1K6bDepartment of Internal Medicine, University of Calgary, Alberta, Canada T2T 1K6
Received 12 May 2010; revised 26 July 2010; accepted 28 July 2010
Postinfectious glomerulonephritis (PIGN) typicallyoccurs in children and most often follows a respiratorytract infection with Streptococcus. IgG, IgM, and C3immunoreactants are typically detected by immunofluores-cence (IF). Electron microscopy (EM) shows immune-complex deposits as characteristic humps in the subepithelialareas of the glomerular basement membranes. The outcomeis typically excellent in most patients, and the diagnosis maynot require renal biopsy. Less commonly, PIGN may occurin adults and in association with other microorganisms. In1980, Spector et al [1] reported a PIGN-like GN in 3 patients,with dominant and codominant IgA deposits, demonstratedby IF, in association with staphylococcal infection. Sincethen, additional cases have been reported. Koyama et al [2]published 10 cases associated with methicillin-resistantStaphylococcus aureus (MRSA) infection and suggestedthat staphylococcal superantigen may play an important rolein this condition. The association of MRSA superantigensand PIGN with IgA was also documented in other studiesfrom Japan [3-5]. Nasr et al [6] designated the entity IgA-dominant acute poststaphylococcal GN; they described thisentity in 5 patients with diabetic nephropathy and with poorclinical outcome. A strong association of Staphylococcusinfection and IgA-dominant or codominant PIGN was foundin 8 patients by Satoskar et al [7]; only 2 of the 8 patientswere diabetic. Their preferred terminology was Staphylo-coccus infection–associated GN mimicking IgA nephropa-thy. More recently, Haas et al [8] reported 13 patients withIgA-dominant PIGN, including 5 patients with diabetes,showing a good short-term outcome in the nondiabeticpatient group presenting without severe renal insufficiency.
In the present study, we report 7 cases of PIGN withdominant or codominant IgA (IgA-PIGN), which wereencountered during 2½ years in a large renal-biopsy practice,serving the population of Southern Alberta, Canada (∼2million). All cases were examined by light microscopy (LM),IF, and EM, and we used strict criteria for the diagnosis ofIgA-PIGN, using all 3 modalities.
2. Materials and methods
A total of 905 native kidney biopsies were performed inour institution between July 1, 2007, and March 31, 2010.Twenty biopsies (2.2%) were diagnosed as PIGN, and ofthese, 7 (0.8%) biopsies from 7 patients fulfilled the clinicaland pathologic criteria for PIGN with dominant or codom-inant IgA immunoglobulin (IgA-PIGN), demonstratedby IgA IF staining intensity at least 2+ and by subepithelialhumps on EM. Patient information included age, sex, race,underlying diseases, history of infections, and laboratorydata, including baseline serum creatinine (SCr), peak SCr,urine analysis, and serum complement (C3 and C4) levels.
We also investigated the modalities of patient treatment, andwe documented the last SCr retrieved from our informationsystem. MRSA isolates were identified at a central regionallaboratory using standard methods and were typed via pulsedfield gel electrophoresis [9].
All biopsies were studied by LM, IF, and EM. Tissuesamples for LMwere embedded in paraffin aswell as in glycolmethacrylate resin (Leica Historesin Embedding Kit; LeicaMicrosystems Nussloch GmbH, Heidelberg, Germany).Twenty-seven slides (20 plastic-embedded, 7 paraffin-embedded) for each biopsy were stained with hematoxylinand eosin, periodic acid–Schiff, toluidine blue, periodic acidsilver methenamine, Masson trichrome, and Congo red. IFwas performed on cryostat sections, and slides were stainedfor IgG, IgA, IgM, C3c, C1q, fibrinogen, and κ and λ lightchains. We applied intensity grading scale from 0 to 3+ for IF.Hitachi H-7650 transmission EM was used for ultrastructuralstudies, and on average, 16 (range, 12-28) electron micro-graphs were sampled per case.
3. Results
3.1. Clinical features
A summary of patient demographic and clinical features isprovided in Table 1. Median patient age was 64 years (range,46-86 years). All patients (5 males, 2 females) presented withhematuria, proteinuria, and hypertension. Six patients hadsevere renal insufficiency (anuria in 2 and oliguria in 4). Meanbaseline SCr of nondiabetic patients was 70.5 μmol/L (range,65-77 μmol/L) and increased 7 to 14 times at the time of thebiopsy (all patients had peak SCr N450 μmol/L). One patienthad underlying diabetes mellitus but without biopsy-provendiabetic nephropathy; he had chronic kidney disease withbaseline SCr of 180 μmol/L that increased to 681 μmol/L atthe time of the clinical presentation. Serum complement levels(C3 and C4) were available in 6 patients. In 4 patients,complement levels were normal, and in 2 patients, they werereduced (1 C3 and 1 C4).
3.2. Pathology of IgA-PIGN
3.2.1. Light microscopyA mean of 37 glomeruli (range, 8-51) were examined in
each biopsy. All biopsies showed diffuse proliferative GN(DPGN) and exudative GN (Fig. 1A). One patient showed amembranoproliferative GN (MPGN) pattern (Fig. 1B), and 1patient's biopsy showed crescentic GN with 75% crescents.No crescents were seen in the remaining biopsies. Focalintraglomerular hyaline thrombi were seen in 2 patients, andfocal and segmental endocapillary necrosis was seen in 2patients. Mild tubular atrophy was found in 5 of 6nondiabetic patients, and severe atrophy was seen in thediabetic patient who also had mild interstitial fibrosis. We
Organism SCr (μmol/L) SerumcomplementHT DM Other Baseline Peak New baseline
(1) 46/M ARF Y N Alcoholism Skin lesions MSSA NA 527 99 Low C3(2) 59/M ARF Y N DVT Bloody diarrhea No growth 77 926 146 Normal(3) 74/M ARF Y Y COPD Diabetic foot ulcer No growth 180 681 Dead Normal(4) 64/M NS Y N Asthma Pneumonia No growth 69 N/A 155 Low C4(5) 52/F ARF Y N IVDU Endocarditis CMRSA-10 65 588 71 Normal(6) 86/F ARF Y N – Cellulitis MRSA 71 1086 97 Normal(7) 84/M ARF/NS N N COPD Pneumonia CMRSA-10 108 483 Dead Normal
Abbreviations: ARF, acute renal failure; NS, nephrotic syndrome; DVT, deep vein thrombosis; COPD, chronic obstructive pulmonary disease; IVDU,intravenous drug abuse; M, male; F, female; NA, not applicable.
found mild to moderate hypertensive nephrosclerosis witharteriosclerosis and hyaline arteriolosclerosis in all nondia-betic patients, whereas the diabetic patient had moderate tosevere vascular disease.
3.2.2. Immunofluorescence microscopyDominant mesangial IgA and C3 deposits were seen in all
patients (Table 2 and Fig. 2A and B). In addition,membranous granular IgA and C3 deposits were also presentin 3 biopsies in nondiabetic patients. Coexpression of IgG (1-2+) was seen in 4 biopsies (Fig. 2C). In 2 biopsies, theintensity of IgG was the same as IgA (2+), whereas in 2biopsies, it was weaker than IgA (1+ versus 2+ and 2+ versus3+, respectively). Six biopsies showed mesangial κ and λlight-chain deposits, and 4 biopsies also showed granularmembranous deposits of both light chains. C1q was negativein all biopsies.
3.2.3. Electron microscopyElectron-dense deposits forming prominent subepithelial
humps were documented in all patients on EM (Fig. 3).Subepithelial deposits measured from 360 to 4590 nm in
Fig. 1 Histologic patterns of IgA-PIGN. A, DPGN and exudative GN (h(periodic acid–Schiff stain; magnification ×400).
width. Mesangial deposits were also documented in allpatients. Subendothelial and intramembranous deposits wereseen in 5 and 3 patients, respectively.
3.3. Microbiology
Clinical infections were documented in all patients at thetime of presentation, including skin ulcers in 2, infectiousdiarrhea in 1, pneumonia in 2, and 1 case each of endocarditisand cellulitis. S aureus was documented in 4 patientsincluding methicillin-sensitive S aureus (MSSA) in 1 andMRSA in 3. Community-wide specific clone community–associated MRSA-10 (CMRSA-10; USA300) was identifiedin 2 patients with Staphylococcus infection.
3.4. Treatment and follow-up
All patients were treated with antibiotics, and 4 of themreceived prednisolone, as shown in Table 3. Hemodialysiswas performed in 5 patients (range, 1-201 days). During amean follow-up of 7.7 months (median, 7 months; range, 1-18 months), SCr of 5 patients (71-155 μmol/L) remained
ematoxylin and eosin stain; magnification ×400). B, MPGN pattern
Table 2 Summary of pathologic features
Patient LM IF EM
IgG IgA IgM C3 C1q κ λ Mes Subendo Memb Subepi
1 Crescentic DPGN 1+ 2+ 1+ 3+ 0 1+ 1+ Y Y N Y2 DPGN 0 2+ Trace 2+ 0 1+ 1+ Y N N Y3 DPGN 2+ 2+ 1+ 3+ 0 1+ 1+ Y Y Y Y4 DPGN 0 3+ Trace 2+ 0 Trace 2+ Y Y N Y5 DPGN 2+ 3+ 0 3+ 0 2+ 2+ Y Y Y Y6 MPGN-like DPGN 2+ 2+ 0 2+ 0 1+ 2+ Y Y Y Y7 DPGN 1+ 2+ Trace 3+ 0 0 Trace Y N N Y
higher than the previous baseline range of 65 to 77 μmol/L.Two patients died, including the diabetic patient who becamedialysis-dependent for 8 months and died of cardiopulmo-nary arrest secondary to pneumonia. Another patient died ofrenal failure, after having refused hemodialysis. Autopsieswere not performed on these patients.
4. Discussion
Acute PIGN with dominant IgA deposition (IgA-PIGN)has been increasingly recognized as a sequel of staphylo-coccal infection, including MRSA [1-8]. This entity can bedistinguished from typical PIGN and from IgA nephropathyby IF and EM findings [5,7]. Studies from Japan suggested arole for staphylococcal superantigen in the pathogenesis ofIgA-PIGN [2-4], but pathologic features were not clearlyoutlined in these studies.
All patients in this study presented with rising SCr,hematuria, and proteinuria, and 6 of 7 patients had severerenal insufficiency. All patients had history of infection at thetime of the biopsy. We identified S aureus in 4 (57%) of 7cases in this study: MSSA in 1 and MRSA in 3. We havedocumented for the first time an association between
Fig. 2 Immunofluorescence features of IgA-PIGN. A, Marked IgA gMarked C3 deposits (3+). C, Weaker IgG granular mesangial and memb
CMRSA-10 (USA300) infection and PIGN. This is impor-tant because the incidence of CMRSA-10 infections hasincreased to comprise up to 25% of community acquired Saureus infections in Calgary, since this strain's arrival in2004 [10,11]. Complement levels were not consistentlydepressed in our study; this has also been documented byothers [12].
The spectrum of light microscopic patterns of GNassociated with Staphylococcus infection includes mesan-gial proliferative GN, MPGN, DPGN, and crescentic GN[1,6-8]. Spector et al [1] reported 3 patients withstaphylococcal infection who had moderate mesangialproliferative GN (2 patients) and DPGN (1 patient), notingthat they exhibited either dominant or codominant IgAdeposits on IF. Nasr et al [6] also found that IgA was thedominant immunoglobulin in 5 patients with staphylococcalinfection. All patients in their study had diabetic nephrop-athy. On morphology, 4 patients demonstrated PIGN and 1showed MPGN pattern, and on EM, humps were docu-mented in all patients. Satoskar et al [7] described 8 patientswith Staphylococcus infection–associated GN mimickingIgA nephropathy. LM revealed variable degree of mesangialand intracapillary hypercellularity, and 1 of the patients inthis study had crescents. Of note, no humps weredocumented on EM in their study. Haas et al [8] used
strict criteria to define IgA-dominant PIGN based on IgApredominance on IF and presence of humps on EM, and wefollowed the same criteria for case inclusion in this study.An association with staphylococcal infection was docu-mented in 6 (46%) of 13 patients, 3 with MRSA and 3 withMSSA. They described 13 patients with acute, subacute,and resolving/persistent GN patterns with variable degree ofglomerular hypercellularity, including diffuse or focalendocapillary and/or mesangial hypercellularity. Crescentswere documented in 4 (31%) of 13 in their study, typicallyin less than 10% of the glomeruli. In the previous studies,patients who presented with severe renal insufficiency ordiabetic nephropathy had poor outcome [1,6-8].
In the current study, we documented DPGN andexudative GN in 6 patients, which resembled typicalPIGN, and MPGN pattern was seen in 1 patient. CrescenticGN (crescents affecting N50% of glomeruli) superimposedon DPGN was seen in 1 biopsy; however, no crescents wereseen in the remaining patients. Two patients in our studyshowed intraglomerular capillary hyaline thrombi, resem-bling those reported by Satoskar et al [7]. We foundsubepithelial electron-dense deposits (humps) characteristic
Table 3 Treatment and follow-up
Patient SCr (μmol/L) Treatment
Baseline Peak Antibiotics Steroid He
1 NA 527 Y Y Y2 77 926 Y Y Y3 180 681 Y Y Y (4 69 NA Y N N5 65 588 Y N Y6 71 1086 Y Y Y7 108 483 Y N N
Abbreviations: Y, yes; N, no; NA, not applicable.NOTE. Causes of death are as follows:
a Cardiopulmonary arrest due to pneumonia.b Renal failure—refusal of hemodialysis.
of PIGN in all patients. All patients in our study also haddocumented infections at the time of the presentation; and in4 (57%) of 7 patients, we isolated S aureus (1 MSSA and 3MRSA). In 2 of 3 patients with MRSA, we found CMRSA-10 (equivalent to USA300), a clone responsible for theongoing community-wide MRSA epidemic. Although wecould not specifically document Staphylococcus in 3patients, prior undocumented staphylococcal infectionscannot be excluded in these patients. In addition, CMRSA-10 was isolated in another patient who was not included inthis study because, although he presented with typical PIGNand humps, he showed dominant IgG immune deposits anddid not show any IgA on IF. This patient died 4 months afterthe presentation. In the remaining 12 PIGN cases withdominant IgG immune deposits seen in our institution duringthe study period, only weak IgA was documented in 5 cases(1+ in 4 biopsies, trace in 1), and no staphylococcal infectionwas documented in any case.
The etiology of IgA-PIGN is unknown. A role forsuperantigens has been suggested [2-4]. Superantigens areproteins that may be detected in several pathogenic microbes(including Staphylococcus). They are powerful activators ofthe immune system and stimulate the production of largenumbers of T cells along with a variety of cytokines(including TNF-α). They also induce antibody productionincluding IgA. Staphylococcal enterotoxins such as entero-toxin B and C can act as superantigens and have beenidentified in staphylococcal-associated GN [2-4].
Several reports have suggested an association withdiabetes mellitus, which was observed either in all patients[6] or in 25% to 38% of the patients in some studies [7,8]. Inour study, only 1 (14%) patient was diabetic, and he died8 months after the biopsy. Associated conditions in ourpatients included alcoholism, deep venous thrombosis,chronic obstructive lung disease, allergic asthma, intrave-nous drug abuse, and staphylococcal endocarditis withdocumented CMRSA-10.
In summary, we report 7 patients with IgA-PIGN, only 1of whom had preexisting diabetes mellitus. All patientspresented with hematuria and proteinuria, and 6 of 7 patients
Follow-up
modialysis (d) New baseline SCr (μmol/L) Time (mo)
had severe renal insufficiency (1 presented only withnephrotic syndrome). All patients had documented infectionsat the time of presentation, and S aureus was isolated in 4 of7 patients (1 MSSA and 3 MRSA). In 2 patients, we foundCMRSA-10 (USA300), which has not previously beenlinked to GN. Although the follow-up in this study wasrelatively short, 2 of 7 patients (1 diabetic) died within 1 and8 months after presentation. Although the renal functionimproved after therapy in 5 nondiabetic patients, fullrecovery was not seen during the limited follow-up. Anawareness and an accurate diagnosis of this entity shouldfacilitate the appropriate treatment in these patients.
References
[1] Spector DA, Millan J, Zauber N, Burton J. Glomerulonephritis andStaphylococcal aureus infections. Clin Nephrol 1980;14:256-61.
[2] Koyama A, Kobauashi M, Yamaguchi N, et al. Glomerulonephritisassociated with MRSA infection: a possible role of bacterial super-antigen. Kidney Int 1995;47:207-16.
[3] Yoh K, Kobayashi M, Hirayama A, et al. A case of superantigen-related glomerulonephritis after methicillin-resistant Staphylococcusaureus (MRSA) infection. Clin Nephrol 1997;48:311-6.
[4] Kobayashi M, Koyama A. Methicillin-resistant Staphylococcus aureus(MRSA) infection in glomerulonephritis—a novel hazard emerging onthe horizon. Nephrol Dial Transplant 1998;13:2999-3001.
[5] Handa T, Ono T, Watanabe H, et al. Glomerulonephritis induced bymethicillin-sensitive Staphylococcus aureus infection. Clin ExpNephrol 2003;7:7247-9.
[6] Nasr SH, Markowitz GS, Whelan JD, et al. IgA-dominantpoststaphylococcal glomerulonephritis complicating diabetic nephropathy.HUM PATHOL 2003;34:1235-41.
[7] Satoskar AA, Nadasdy G, Plaza JA, et al. Staphylococcus infection-associated glomerulonephritis mimicking IgA nephropathy. Clin J AmSoc Nephrol 2006;1:1179-86.
[8] Haas M, Racusen LC, Bagnasco SM. IgA-dominant postinfectiousglomerulonephritis: a report of 13 cases with common ultrastructuralfeatures. HUM PATHOL 2008;39:1309-16.
[9] Mulvey MR, Chui L, Ismail J, et al. Development of a Canadianstandardized protocol for subtyping methicillin-resistant Staphylococcusaureus using pulsed-field gel electrophoresis. J Clin Microbiol 2001;39:3481-5.
[10] Gilbert M, MacDonald J, Gregson D, et al. Outbreak in Alberta ofcommunity-acquired (USA300) methicillin-resistant Staphylococcusaureus in people with a history of drug use, homelessness orincarceration. CMAJ 2006;175:149-54.
[11] Al-Rawahi GN, Reynolds S, Porter SD, et al. Community-associated CMRSA-10 (USA-300) is the predominant strainamong methicillin-resistant Staphylococcus aureus strains causingskin and soft tissue infections in patients presenting to the emergencydepartment of a Canadian tertiary care hospital. J Emerg Med 2010;38:6-11.
[12] Nadasdy T, Silva FG. Acute postinfectious glomerulonephritis andglomerulonephritis caused by persistent bacterial infection. In:Jennette JC, Olson JL, Schwartz MM, Silva FG, editors. Heptinstall'spathology of the kidney, 6th ed.Philadelphia: Lippincott Williams &Wilkins; 2007. p. 321-96.
