Human Pathology (2015) 46, 1521–1528

www.elsevier.com/locate/humpath

Original contribution

Non–immunoglobulin A mesangial immune

complex glomerulonephritis in kidney
transplants☆
Giovanna A. Giannico MD a , Shanna Arnold PhD, MSCI a,b , Anthony Langone MD c ,
Heidi Schaefer MD c , J. Harold Helderman MD c , David Shaffer MD, FACS c ,
Agnes B. Fogo MD a,⁎
a
Department of Pathology, Microbiology and Immunology, Vanderbilt University Medical Center, Nashville, TN 37232
b
Department of Veterans Affairs, Nashville, TN 37212
c
Division of Kidney and Pancreas Transplantation, Vanderbilt University Medical Center, Nashville, TN 37232

Received 24 April 2015; revised 29 May 2015; accepted 10 June 2015

Keywords:
Summary We have observed a predominantly mesangial non–immunoglobulin A immune complex
Renal biopsy;
mesangial glomerulopathy (MG) in renal transplants with mesangial deposits by immunofluorescence
Transplant biopsy;
and electron microscopy. Clinicopathological features of 28 patients with MG were analyzed and
Mesangial
compared with 28 transplant controls, matched for age, sex, ethnicity, donor type, estimated glomerular
glomerulonephritis;
filtration rate, and interval from transplant to biopsy. Indications for biopsy in the MG group were
Immune complex
allograft dysfunction in 64%, allograft dysfunction/proteinuria in 29%, and proteinuria in 7%. Biopsy
glomerulonephritis;
indications in controls were allograft dysfunction (61%), allograft dysfunction/proteinuria (18%),
Transplant
proteinuria (14%), and delayed graft function (7%). Most MG cases had mild mesangial hypercellularity
glomerulonephritis
with endocapillary proliferation in 2 and crescents in 2 without fibrinoid necrosis. Immunoglobulin
M–dominant deposits were present in 83%, and immunoglobulin G was dominant in 17% with
mesangial deposits in 93% of cases by electron microscopy. Compared with controls, MG had higher
Banff interstitial inflammation score (i) (P = .036) and was associated with concurrent acute T-cell–
mediated rejection (P = .023), but not with acute or chronic antibody-mediated rejection. MG patients
and controls had similar prevalence of polyomavirus nephropathy and Epstein-Barr virus infection. At
follow-up, most MG patients had stable estimated glomerular filtration rate with no or stable proteinuria.
Disease-specific graft survival was not different in MG versus controls. We conclude that, in view of the
apparent self-limited nature of this lesion, additional treatment may not be required in these patients.
Awareness of this lesion may thus spare patients unwarranted further intervention.
© 2015 Elsevier Inc. All rights reserved.

☆
Disclosures: None.
⁎ Corresponding author at: Department of Pathology, Microbiology and Immunology, Vanderbilt University Medical Center, C-3310 MCN, Nashville, TN
37232-2561.
E-mail addresses: giovanna.giannico@vanderbilt.edu (G. A. Giannico), Shanna.arnold@vanderbilt.edu (S. Arnold), anthony.langone@vanderbilt.edu
(A. Langone), heidi.schaefer@vanderbilt.edu (H. Schaefer), hal.helderman@vanderbilt.edu (J. H. Helderman), david.shaffer@vanderbilt.edu (D. Shaffer),
agnes.fogo@vanderbilt.edu (A. B. Fogo).

http://dx.doi.org/10.1016/j.humpath.2015.06.012
0046-8177/© 2015 Elsevier Inc. All rights reserved.
1522
1. Introduction
Recurrent and de novo disease may significantly affect
long-term kidney allograft survival. The Renal Allograft
Disease Registry reported a prevalence of 3.4% to 13% with
doubled relative risk of allograft loss and significantly
increased graft failure in patients with recurrent versus de
novo disease [1,2]. Recurrent glomerular disease is the third
most frequent cause of late kidney allograft loss at 10 years
[3]. Most common forms of glomerular disease can recur in
the transplant. However, accurate estimation of the exact
prevalence of posttransplant glomerulonephritis (GN),
whether de novo or recurrent, is hampered by incomplete
renal biopsy evaluation by immunofluorescence (IF) and
electron microscopy (EM) in most centers.
In a recent study on recurrent and de novo disease in renal
transplant biopsies, 9 of 23 biopsies showed immunoglobulin
M (IgM) mesangial deposits with or without C3 and mesangial
deposits by EM. Concurrent or subsequent infection with
hepatitis B virus (HBV), hepatitis C virus (HCV), polyoma
(BK) virus, or cytomegalovirus (CMV) occurred in 3 of these
patients [4]. The authors concluded that the etiology of IgM
deposits is not clear and speculated on possible etiologies
including T-cell deficiency, mesangial overloading by immune
complexes (ICs) containing unidentified microorganism
antigens, or rejection-related immune events. In this study,
we characterized the clinicopathological features and signif-
icance of non–immunoglobulin A (IgA) mesangial glomeru-
lopathy (MG) in the renal transplant.
2. Materials and Methods
Renal allograft biopsies performed for cause at Vanderbilt
University Medical Center from 2005 to 2013 were
reviewed. Twenty-eight patients with diagnosis of MG,
without known history of IC GN, or with known GN in the
native kidney unrelated to MG were selected. For controls,
28 transplant patients without MG were randomly selected
after matching for age, sex, ethnicity, donor type, estimated
glomerular filtration rate (eGFR) at presentation, and time
interval from renal transplant to biopsy.
Allograft biopsies were performed under ultrasound
guidance using a 16-gauge automated biopsy instrument
and were allocated for light microscopy, IF, and EM studies.
Renal biopsies were processed by standard techniques for
light microscopy with multiple serial sections stained with
hematoxylin and eosin, periodic acid–Schiff (PAS), and
PAS–methenamine silver, IF on frozen sections (stained for
immunoglobulin G [IgG], 1:8; IgM, 1:8; IgA, 1:8; C3, 1:8;
C1q, 1:12; kappa, 1:8; lambda, 1:8 [Dako, Carpentaria, CA];
and C4d, 1:10 [AbD Serotec-MorphoSys, Deutschland,
Germany), and EM. Kappa and lambda were performed in
selected cases before 2011 based on clinical indication and in
all cases after 2011. IF intensity was semiquantitatively
G. A. Giannico et al.
graded as follows: negative, 0; mild, 1+; moderate, 2+; and
severe, 3+, with specification of staining localization and
pattern. EM was done on all initial transplant biopsies with
available tissue. In repeat biopsies, IF and EM were
selectively repeated depending on clinical setting and
pathologic findings in the previous biopsy. All biopsies
were interpreted by experienced renal pathologists.
Acute allograft dysfunction was defined as a rise in
creatinine of 0.4 mg/dL or greater over the baseline during a
1-month period. Delayed graft function was defined as
hemodialysis requirement in the first week after transplant.
Glomerular filtration rate was estimated using the Modifi-
cation of Diet in Renal Disease Study equation [5].
Proteinuria was evaluated by spot urine protein/creatinine
ratio (uPCR; in milligrams per milligram) with greater than
0.2 mg/mg considered positive, and 3.5 mg/mg or greater for
nephrotic proteinuria. Histologic findings were graded by the
National Institutes of Health Cooperative Clinical Trials in
Transplantation (CCTT) classification [6] and subsequently
scored based on the 2007 Banff classification [7] with 2013
update [8].
Statistical analysis was performed using SPSS 22 for
Windows (IBM SPSS, Chicago, IL). Categorical variables
were described with relative frequencies and analyzed by χ2
test. Continuous variables were reported as mean ± SD and/
or median and were analyzed with Mann-Whitney U test, as
applicable. Survival curves were calculated by the Kaplan-
Meier method and compared with log-rank test. Multivariate
Cox proportional hazards models were used to evaluate the
effect of risk factors on renal survival. Two-tailed P values less
than .05 were considered statistically significant. The study
was approved by the Vanderbilt Institutional Review Board.
3. Results
3.1. Clinical findings
A total of 791 renal transplant biopsies were performed at
Vanderbilt University Medical Center from 2005 to 2013, of
which 28 were diagnosed with MG (3.5%). The clinical and
demographic characteristics of MG patients and controls are
shown in Table 1. MG patients and controls were matched
for age, sex, ethnicity, donor type, eGFR at presentation, and
time from transplant to biopsy and did not differ significantly
for number of HLA mismatches, indications for biopsy, and
proteinuria. In MG, 1 patient was highly sensitized with
positive cross-match and treated with desensitization proto-
cols before transplant. The induction treatment history was not
available in 1 patient receiving his third transplant. In controls,
2 highly sensitized patients had positive cross-match and were
treated with preoperative intravenous immunoglobulin. None
were ABO blood group incompatible.
Clinical and laboratory data are shown in Table 1 and
primary kidney diseases in Table 2. In MG, the interval from
Mesangial glomerulopathy in kidney allografts 1523

Table 1 Clinical characteristics at time of renal biopsy Table 2 Primary disease

MG Controls P MG (n = 28), Controls (n = 28),
(n = 28) (n = 28) n (%) n (%)
Age (y) (range) 42 (6-62) 42 (18-64) 1.0 Glomerular 7 (25) 7 (25)
Sex, M/F (%) 20/8 (71/29) 18/10 (64/36) .8 MN with anti-TBM nephritis 1 0
Race, C/AA (%) 18/10 (64/36) 19/9 (68/32) .8 Alport's 1 0
Donor type FSGS 2 3
D/LR/LU (%) 14/5/9 14/6/8 1.0/.7/1.0 Chronic postinfectious 1 0
(50/18/32) (50/21/29) mesangial GN
HLA compatibility a 1/26 2/26 .6 Anti-GBM GN 1 1
(incompatible/compatible) HUS 0 1
HLA mismatches LN 0 1
A 1.3 ± 0.7 1.2 ± 0.6 .3 IgA nephropathy 1 1
B 1.5 ± 0.6 1.4 ± 0.6 .3 Hypertension/Diabetes 13 (47) 7 (25)
DR 1.1 ± 0.7 1.1 ± 0.8 .6 Congenital 6 (21) 3 (11)
Time between transplant 53 53 .8 Hereditary 2 (7) 4 (14)
and biopsy (mo) Other 0 5 (18)
Indication for biopsy, n (%) Unknown 0 2 (7)
Increased creatinine 19 (68) 17 (60) .8 Abbreviations: MG, mesangial glomerulopathy; GN, glomerulonephri-
Proteinuria 6 (21) 4 (14) .7 tis; MN, membranous nephropathy; TBM, tubular basement membrane;
Increased creatinine/ 3 (10) 5 (18) .7 GBM, glomerular basement membrane; HUS, hemolytic-uremic
proteinuria syndrome; LN, lupus nephritis.
DGF 0 2 (7) .5
eGFR (mL/min per 37.9 34.8 .9
1.73 m2) uPCR, 3.5-12), including 2 with concomitant findings of
Creatinine (mg/dL) 2.5 2.6 .9 transplant glomerulopathy, 2 with diabetic nephropathy, and
Proteinuria, n (%) b 1 with extensive foot process effacement 5 years after
Negative 5 (18) 8 (30) .5
transplant with native disease of focal segmental glomerulo-
Subnephrotic 17 (64) 14 (52) .6
sclerosis (FSGS; Mann-Whitney U test, P = .016, MG only
Nephrotic 5 (18) 5 (18) 1.0
Hematuria 5/28 (18) 4/28 (14) 1.00 versus MG with other diagnosis). Proteinuria at biopsy was
HCV 2/28 (7) 0/28 1.00 assessed in 27 (96%) of 28 controls and was nephrotic in 5
HBV 0/28 0/28 1.00 (18%; uPCR, 6.5-11), with diabetic nephropathy in 2 and
HIV c 0/9 0/11 membranous nephropathy, IgA nephropathy, and transplant
CMV c 1/15 (6) 6/13 (46) .02 glomerulopathy in 3 each, respectively. C3, C4, rheumatoid
EBV c 0/10 1/3 (33) .2 factor, cryoglobulin, anti–double-stranded DNA, and/or
BK c 5/20 (25) 4/16 (25) 1.00 antineutrophil cytoplasmic antibody were assessed in a
Hypocomplementemia c 0/11 0/3 subset of MG patients and were unremarkable. Four patients
RF/cryoglobulin c 0/9 0/1 were antinuclear antibody (ANA) positive without clinical or
ANA c 4/8 (50) 0/3 .1
other serologic evidence of systemic lupus erythematosus (SLE)
ANCA c 0/5 0/3
and had no lupus nephritis in the native biopsy. MG patients
Anti-dsDNA c 0/5 0/0
DSA c 7/13 (54) 2/10 (20) .2 showed HCV positivity in 2, CMV positivity in 1, and blood
BK viremia in 5, of which 3 had polyomavirus nephropathy
Abbreviations: MG, mesangiopathic glomerulopathy; M, male; F, female;
C, Caucasian; AA, African American; D, deceased; LR, living related; LU,
(PVN) on renal biopsy. HBV, Epstein-Barr virus (EBV), and
living unrelated; HLA, human leukocyte antigen; HCV, hepatitis C virus; human immunodeficiency virus (HIV) were negative in all
HBV, hepatitis B virus; HIV, human immunodeficiency virus; CMV, tested MG patients. There was no clinical evidence of bacterial
cytomegalovirus; EBV, Epstein-Barr virus; BK, BK polyomavirus; RF, infection at the time of biopsy in MG patients.
rheumatoid factor; ANA, anti-nuclear antibody; DGF, delayed graft All tested control patients were negative for HCV, HBV,
function; ANCA, antineutrophil cytoplasmic antibody; anti-dsDNA, anti–
double-stranded DNA antibody; DSA, donor-specific antibody.
and HIV, whereas 6 were CMV positive, 1 was EBV
a
HLA compatibility information at time of transplant was not positive, and 4 were BK positive.
available in 1 patient. At biopsy, 25 MG patients (89%) were receiving triple
b
Proteinuria at biopsy was not available in 1 patient in both groups. immunosuppression (calcineurin inhibitor, mycophenolate
c
Serologies were available in limited subsets of patients in both groups. mofetil and prednisone, or calcineurin inhibitor, sirolimus
and prednisone), 2 were on a calcineurin inhibitor–sparing
regimen with sirolimus and mycophenolate mofetil, and 1
transplant to MG diagnosis ranged from 2 to 222 months was on double immunosuppression (calcineurin inhibitor
(average 52 months). Proteinuria was present in 22 (81%) of and prednisone). In controls, 26 patients (92%) were
27 assessed MG patients and was nephrotic in 5 (18%; receiving triple immunosuppression, 1 was on a calcineurin
1524 G. A. Giannico et al.

inhibitor–sparing regimen with sirolimus and prednisone, Table 3 Histopathologic findings

and 1 was on double immunosuppression (calcineurin MG Controls P
inhibitor and mycophenolate mofetil). (n = 28) (n = 28)
Mesangial hypercellularity, 23 (82) 14 (50) .02 a
3.2. Histopathologic findings n (%)
Additional pathologic diagnosis, n (%)
All cases showed mild, albeit variable, degree of TCMR or Borderline 10 (36) 2 (7) .02 a
mesangial expansion/proliferation, with matrix increase ABMR 7 (54) 1 (10) .2 b
CAMR 4 (30) 1 (10) .6 b
exceeding mesangial cell proliferation (Table 3 and Fig. 1).
CNIT 7 (25) 8 (28) 1.0
Focal and segmental endocapillary proliferation was present
DN 3 (10) 5 (18) .6
in 2. Two cases had crescents (1 cellular and 1 fibrocellular) ATI 9 (32) 6 (21) .5
without fibrinoid necrosis. Six glomeruli showed usual-type PVN 3 (10) 4 (14) 1.0
segmental sclerosis with adhesions to Bowman capsule, and IFTA 3 (10) 4 (14) 1.0
1 case had collapsing features, attributed to calcineurin- GN a 28 (100) 4 (14) b.001 a
inhibitor toxicity in the absence of FSGS as primary disease Banff scores, mean ± SD
or other possible causes of collapsing FSGS [9]. IF (Table 4) Glomerulitis (g) 2.11 ± 0.83 1.70 ± 0.84 .2
revealed mesangial staining in all cases with dominant IgM Mononuclear interstitial 2.25 ± 1.03 1.27 ± 0.46 .04 a
staining in 23 MG cases, largely in a mesangial pattern with inflammation (i)
rare segmental capillary loop extension (Fig. 2), and Tubulitis (t) 1.33 ± 0.61 1.25 ± 0.45 .9
Intimal arteritis (v) 1.00 ± 0.0 1.00 ± 0.0 N/A
IgG-dominant mesangial staining in 5. Mesangial IgA was
Allograft glomerulopathy (cg) 1.84 ± 0.37 1.64 ± 0.49 .2
present in 13 cases, with lower intensity than IgG or IgM, by
Interstitial fibrosis (ci) 1.57 ± 0.81 1.58 ± 0.77 .9
case definition. C3, C1q, or both were positive in 17 cases. Tubular atrophy (ct) 1.65 ± 0.87 1.58 ± 0.77 .9
C3 was present in 14 cases, of which 9 also had C1q. In 2 cases, Vascular fibrous intimal 1.33 ± 0.48 1.60 ± 0.63 .2
C3 was stronger than Ig, had similar intensity in 3, and was thickening (cv)
weaker than Ig in all remaining cases. C1q was co-dominant in 2 Arteriolar hyaline (ah) 1.38 ± 0.76 1.52 ± 0.71 .4
cases with weak intensity, b2+, and dominant in 2, of which 1 Mesangial matrix (mm) 1.13 ± 0.34 1.14 ± 0.36 .9
showed short fibrillary substructure of deposits by EM. All 10 Peritubular capillary 1.70 ± 0.58 1.56 ± 0.81 .5
cases tested for kappa and lambda light chains were polyclonal. margination (ptc)
EM showed IC deposits in 26 of 28 MG cases. Deposits Total interstitial 1.91 ± 0.97 1.91 ± 0.83 .9
were mesangial in 17 cases (Fig. 3), subepithelial in 1, inflammation (ti)
C4d staining (C4d) 2.25 ± 1.03 2.25 ± 0.95 .9
subendothelial in 1, mesangial and subendothelial in 6, and
ABMR scores (g + ptc + v) 2.42 ± 2.16 1.96 ± 2.08 .4
mesangial and subepithelial in 3. Cases without mesangial
ABMR + CAMR scores 3.28 ± 2.86 2.78 ± 2.68 .5
deposits by EM had strong mesangial staining by IF (1 with (g + ptc + v + cg)
IgA and stronger IgM, 1 with full-house staining with
Abbreviations: MG, mesangiopathic glomerulopathy; TCMR,T-cell
dominant IgM). Vague substructure, from indistinct to short mediated rejection; ABMR, antibody-mediated rejection; CAMR, chronic
fibrillary, was present in 4 cases, of which 2 had IgG and antibody-mediated rejection; CNIT, calcineurin-inhibitor toxicity; DN,
IgM, 1 had IgA and IgM (2 IgM dominant, 1 IgG dominant), diabetic nephropathy; ATI, acute tubular injury; IFTA, interstitial
and 1 had only IgM. Tubuloreticular aggregates were seen in 1 fibrosis-tubular injury; PVN, polyomavirus nephropathy; GN, glomeru-
patient who had FSGS as primary disease. In this patient, lonephritis; N/A, not applicable.
a
GN included MG in the MG group by definition, and membranous
serology was negative for ANA, whereas IF showed IgM with nephropathy, IgA nephropathy, and recurrent lupus nephropathy in the
lesser IgG, C3, and C1q, and SLE was clinically excluded. control group.
One patient with primary disease of antitubular basement b
Evaluated only in patients tested for donor-specific antibodies.
membrane antibody with membranous nephropathy had
mesangial IgG-dominant deposits with less IgM and C3 with
only mesangial and subendothelial deposits by EM, thus not I, and 1 both classes I and II. Concurrent T-cell–mediated
representing recurrent membranous nephropathy. Foot process rejection (TCMR) was also present in 1 of these patients.
effacement was less than 50% in 16 cases and at least 50% in Nine biopsies were suspicious for ABMR (DSA negative in
11. One patient with primary disease of FSGS and uPCR 10 3, not tested in 6), and 2 of these had concomitant TCMR.
had at least 80% foot process effacement without segmental Overall, 3 patients had diagnostic TCMR and 7 had
sclerosis, raising the possibility of recurrent unsampled FSGS borderline rejection by Banff criteria. Of these, 4 had
in addition to MG, but felt to be unlikely in view of the long coexisting ABMR and/or chronic antibody-mediated rejec-
interval from transplant (5 years). tion (CAMR), all with positive DSA (3 class II and 1 class I).
Diagnoses rendered in addition to MG (Table 3) included In the remaining 6 patients with TCMR/borderline rejection,
antibody mediated-rejection (ABMR) in 7 patients, of which DSA was negative in 2 and was not tested in 4. Thirteen
5 had had class II donor-specific antibody (DSA), 1 had class cases showed transplant glomerulopathy, of which 4 had
Mesangial glomerulopathy in kidney allografts 1525

Fig. 1 Varying degrees of mesangial expansion/proliferation in MG, from absent (A) to global (B) to segmental (C). PAS, original
magnification, ×400.

diagnostic CAMR, all with class II DSA, and 1 also with and CAMR with class I and II DSA in 1, suspicious for
class I, and 7 were suspicious for CAMR (negative DSA in 3 ABMR in 11, and suspicious for CAMR in 8. Three patients
or nontested in 4). Two were diagnosed as isolated transplant had IgA nephropathy and 1 had recurrent lupus nephritis.
glomerulopathy. DSAs were evaluated in 10 patients, of which 2 were positive
Repeat biopsy was done in 8 MG patients, 4 of which had and 8 negative.
increased creatinine, 3 had increased creatinine and persis-
tent proteinuria, and 1 had increased proteinuria only, on 3.3. Clinical outcome
average after 9.6 months. MG persisted in 4 cases and was
absent in 4 cases. Cases with persistent MG had more Clinical follow-up was available in all patients and
mesangial deposits and new subendothelial deposits in 1 controls. Mean MG follow-up was 24.4 months (range, 4-94
case, new focal endocapillary proliferation in another, and months; median, 13.5 months). ABMR was treated with a
new focal segmental sclerosis and slight increase in combination of antithymocyte globulin, plasmapheresis,
interstitial fibrosis from 10% to 20% in a third case, also intravenous immunoglobulin, and rituximab. TCMR was
showing coexistent diabetic nephropathy. A fourth patient treated with pulse corticosteroid. Cases suspicious but not
had borderline rejection and no residual mesangial expan- diagnostic for rejection did not receive increased immuno-
sion, but IgM and C3 persisted by IF. Repeat EM was not suppression. In PVN, mycophenolate mofetil was discon-
done due to short time interval after the initial diagnostic tinued and tacrolimus was switched to cyclosporine. Diabetic
biopsy. Treatment after repeat biopsy in patients with nephropathy was treated with angiotensin-converting en-
persistent MG included increased immunosuppression in zyme inhibitors. No specific treatment was introduced after
the presence of concomitant rejection and potentiation of MG diagnosis.
angiotensin-converting enzyme inhibitors in cases with Five MG cases had high BK viremia (mean, 238 549
increased proteinuria. copies/mL; range, 810 000-33 839 copies/mL), of which 3
In controls, biopsies showed multiple lesions, with had PVN on renal biopsy. Four of these 5 patients (2 with
TCMR or borderline changes in 2 cases, diagnostic ABMR PVN and 2 without) had variably rapid serologic resolution
of active infection with absent or persistent low-level (b1 g)
Table 4 IF microscopy findings in MG proteinuria in 3 and persistent proteinuria with progression of
Positive, n (%) Intensity a renal dysfunction in 1, in the absence of active infection. One
patient with PVN on initial biopsy showed persistent MG but
IgG 17 (60) 0.7 ± 0.4
no PVN on repeat biopsy, despite persistent BK serologic
IgA 13 (46) 0.5 ± 0.2
IgM 27 (96) 0.9 ± 0.4
positivity with stable renal function and no proteinuria.
C3 14 (50) 0.7 ± 0.5 Five MG patients progressed to end-stage renal disease,
C1q 12 (43) 0.8 ± 0.2 including 3 with MG remission at repeat biopsy. Death
Kappa 9 (90) 0.3 ± 0.1 occurred in 3 patients, including 1 with persistent MG, due to
Lambda 10 (100) 0.5 ± 0.2 cardiovascular complications. Mean eGFR at follow-up in
C4d (peritubular capillaries) remaining patients was 34.8 ± 16.1 mL/min per 1.73 m2
Diffuse or focal C4d 5 (18) N/A (mean creatinine, 2.5 ± 1.3 mg/dL). Proteinuria at follow-up
(Banff C4d3/C4d2) was nephrotic in 4, including 3 patients with persistent MG,
Minimal or negative C4d 3 (11) N/A subnephrotic in 13, and absent in the remaining 9 (not
(Banff C4d1/C4d0) available for 2).
Abbreviations: IF, immunofluorescence; MG, mesangial glomerulopa- In controls, mean follow-up was 22.3 months (range, 1-90
thy; N/A, Not applicable. months). TCMR and calcineurin-inhibitor toxicity were
a
Mean ± SD; scale 0-3+.
treated as described for MG. Seven controls progressed to
1526 G. A. Giannico et al.

Fig. 2 There is moderate global mesangial staining for IgM (A) and C3 (B) with segmental capillary wall staining. IF, ×400.

end-stage renal disease, and cardiac death with stable renal
function occurred in 1 patient. Mean eGFR at follow-up in
remaining controls was 38.8 ± 20.8 mL/min per 1.73 m2
(mean creatinine 2.2 ± 1.0 mg/dL). Proteinuria was nephrotic
in 2, subnephrotic in 8 and absent in the remaining 11 (not
available in 7).
MG did not affect disease-specific graft survival com-
pared with the controls (Fig. 4; log-rank test, P = .67). MG
had significantly higher Banff interstitial inflammation score
(i) compared with controls (Mann-Whitney U test, P = .036)
and was significantly associated with TCMR/borderline
rejection (Fisher exact test, 36% versus 7%; P = .023).
However, the Banff tubulitis score (t) did not differ
significantly between the 2 groups (P = .873). MG was not
significantly associated with ABMR or CAMR compared
with controls (ABMR: Fisher exact test, 54% versus 10%
[P = .20]; CAMR: 30% versus 10% [P = .61]). When
adjusting for sex, race, donor type, age, and eGFR at biopsy,
the presence of concurrent ABMR, CAMR, or TCMR did
not increase the risk of graft loss (Cox regression, P = .467,
.898, and .721). There was no significant difference in PVN/
EBV infection prevalence between MG and controls (PVN:
Fisher exact test, 25% versus 25% [P = 1.00]; EBV: Fisher
exact test, 25% versus 25% [P = .238]). CMV was higher in
controls compared with MG (Fisher exact test, 6% versus
46%, P = .023).
4. Discussion
We investigated the clinicopathological features of post-
transplant MG and examined possible factors associated with
its development, as well as impact on graft loss. We showed
that in most cases, MG is an IgM-dominant IC process with

MG
Control Group

P = .67

Number at risk
MG 24 16 12 6 2 1 1 1 0 0 0
Control Group 23 12 7 4 4 4 2 1 1 0 0
Fig. 3 Numerous mesangial IC deposits with mild mesangial
matrix increase. EM, ×8900. Fig. 4 Kaplan-Meier analysis of renal survival.
Mesangial glomerulopathy in kidney allografts
mild mesangial proliferation and presence of discrete mesan-
gial and occasional subendothelial IC deposits by EM, rarely
with vague substructure, and occurring at a median of 38
months after transplantation. Mesangial deposits were present
in all cases by IF and confirmed in most cases by EM, with lack
of EM deposits in a few cases, most likely reflecting the
scattered focal and segmental distribution of deposits.
Mesangial GN encompasses a heterogeneous group of
primary and secondary glomerular IC diseases. Mesangial
expansion with an increase in cellularity and/or matrix
represents a common response to glomerular injury, either
immune mediated or not. We ruled out the possibility of
nonspecific glomerular injury causing MG by establishing
the presence of IC deposits by IF/EM. The differential
diagnosis of this mesangial IC process would include
primary and secondary GN. IgA nephropathy is one of the
most common primary mesangial GN. The absence of
significant IgA staining by IF in our MG cohort ruled out this
possibility. Another form of mesangial GN is IgM
nephropathy [10]. Our cohort did not show clinicopatholog-
ical features of IgM nephropathy, in that only a minority of
cases presented with nephrotic proteinuria, which was
associated with concomitant lesions of, for example,
transplant glomerulopathy or diabetic nephropathy. Most
of our cases showed dominant IgM staining. One MG patient
with non–biopsy-proven presumed arterionephrosclerosis as
the primary disease had IgG-dominant full-house staining
with positive ANA. Thus, the possibility that the post-
transplant MG could have represented recurrent mesangial
lupus nephritis could not be ruled out, but felt to be unlikely
in the absence of clinical and serologic evidence of SLE and
lack of tubuloreticular aggregates on biopsy.
De novo C1q nephropathy in the transplant is also
characterized by mesangial hypercellularity and is not
associated with poor graft survival in most patients [11].
This entity could also be entertained in the differential
diagnosis. In our cohort, 2 cases had co-dominant weak C1q
staining, b2+, and 2 cases had dominant C1q by IF, of which
1 showed short fibrillary substructure of deposits by EM,
suggesting cryoglobulinemic GN, and 1 showed limited
proteinuria and foot process effacement by EM, which could
possibly be considered an atypical form of C1 nephropathy,
although not precisely fitting in this category. We thus
considered these cases within our cohort of MG.
Mesangial IgG primary GN, characterized by exclusive or
predominant mesangial deposits, staining dominantly with
IgG in patients without evidence of SLE has been described
[12–14]; however, whether this represents a unique clinico-
pathological entity is not established. Six of 14 cases described
in the largest series [14] demonstrated evidence of humps
by EM, supporting the possibility of resolving infection-
associated GN. In our MG cohort, there was no serologic
evidence or clinical history to specifically indicate recurrent
disease or a known secondary etiology, such as de novo
autoimmune disease. Thus, we hypothesized underlying
subclinical chronic/smoldering infection as a possible etiology
1527
of this MG. In MG cases demonstrating deposit substructure
by EM, the differential diagnosis of cryoglobulinemic GN was
entertained. However, none showed clinical or serologic
evidence of cryoglobulinemia.
Scattered subendothelial and mesangial IC deposits have
been associated with viral infections. In a study of 9 patients
with mesangial expansion and IgM-positive mesangial
deposits [4], 3 demonstrated or subsequently developed
viral infection (BK, CMV, HCV, or HBV), raising the
possibility of antibody/virus antigen IC deposits. Seven of
these cases occurred less than 1 year after transplantation,
and 4 had proteinuria greater than 500 mg/L. BK virus has
also been hypothesized to cause IC deposition. Glomerular
changes, including mesangial IgM and C3 deposition with
BK virion material within hump-type deposits, have rarely
been described [15,16], and tubular basement membrane
deposits were present in 16 of 30 patients in one PVN case
series [17]. In our cohort, 5 MG cases had serologic BK
infection, of which 3 had PVN on renal biopsy. Most of these
5 patients had rapid serologic resolution of active infection
with little or no proteinuria in 3, persistent proteinuria with
progression of renal dysfunction in 1 despite resolution of
infection, and resolution of PVN with persistent MG in 1.
Thus, with the limitations imposed by the small number of
patients, progression of renal disease and/or proteinuria did not
appear related to persistent BK infection. Furthermore, active
BK infection did not differ in MG compared with controls.
CMV has also been associated with glomerular changes
[18–20]. In our series CMV active infection was signifi-
cantly lower in MG than in controls.
Glomerular changes similar to MG have also been
reported in transplant glomerulopathy [21,22]. However, in
our cohort, transplant glomerulopathy was similar in MG and
controls, arguing against the possibility of a direct
relationship of MG with transplant glomerulopathy.
De novo membranous nephropathy may occur due to
ABMR. In our series, ABMR or CAMR in MG patients was
not statistically different versus controls, making an
association of MG with humoral mechanisms of rejection
unlikely. However, MG was more frequently associated with
TCMR/borderline rejection compared with controls. Other
lesions of acute interstitial nephritis, including eosinophils,
granulomas, plasma cells, crystals, or tubular basement
membrane deposits were not found in our MG cohort. The
significance of this association is unclear. We speculate that
TCMR could initiate glomerular injury by generating
pathogen- and danger-associated molecular patterns stimu-
lating Toll-like receptors and complement and, thus, initiate
transient deposition of mesangial IgM ICs by innate and
adaptative immune response. Conversely, transient glomer-
ular IC deposition elicited by an unknown cause, thereby
triggering an immunologic response leading to TCMR,
cannot entirely be ruled out. In our series, patients were
treated accordingly to concomitant diagnoses of rejection
and/or PVN, and no specific treatment was initiated for the
diagnosis of MG. Furthermore, MG was not statistically
1528
significantly associated with change in disease-specific
survival versus control.
In conclusion, we describe the clinicopathological
features of posttransplant MG and demonstrate that this
entity is frequently associated with concurrent TCMR on
kidney biopsy. Our study has several limitations, including
the retrospective nature of the analysis and the limited
number of patients. Additional studies with larger cohort size
and longer follow-up would be required to more specifically
draw conclusions about the possible pathogenesis of this
entity in kidney transplants. However, we suggest that, given
the relative good prognosis and transient nature of these
changes, additional treatment may not be required due to the
apparent self-limited nature of this lesion.
Acknowledgment
The authors would like to thank Mr John Bobbitt for his
technical assistance with digital images and Dr Samih Nasr
for helpful discussions.
References
[1] Hariharan S, Peddi VR, Savin VJ, et al. Recurrent and de novo renal
diseases after renal transplantation: a report from the renal allograft
disease registry. Am J Kidney Dis 1998;31:928-31.
[2] Hariharan S, Adams MB, Brennan DC, et al. Recurrent and de novo
glomerular disease after renal transplantation: a report from Renal
Allograft Disease Registry (RADR). Transplantation 1999;68:
635-41.
[3] Magee CC, Pascual M. Update in renal transplantation. Arch Intern
Med 2004;164:1373-88.
[4] Gough J, Yilmaz A, Yilmaz S, Benediktsson H. Recurrent and de novo
glomerular immune-complex deposits in renal transplant biopsies.
Arch Pathol Lab Med 2005;129:231-3.
[5] Levey AS, Coresh J, Greene T, et al. Using standardized serum
creatinine values in the Modification of Diet in Renal Disease Study
equation for estimating glomerular filtration rate. Ann Intern Med
2006;145:247-54.
G. A. Giannico et al.
[6] Colvin RB, Cohen AH, Saiontz C, et al. Evaluation of pathologic
criteria for acute renal allograft rejection: reproducibility, sensitivity,
and clinical correlation. J Am Soc Nephrol 1997;8:1930-41.
[7] Solez K, Colvin RB, Racusen LC, et al. Banff 07 classification of renal
allograft pathology: updates and future directions. Am J Transplant
2008;8:753-60.
[8] Haas M, Sis B, Racusen LC, et al. Banff 2013 meeting report:
inclusion of c4d-negative antibody-mediated rejection and antibody-
associated arterial lesions. Am J Transplant 2014;14:272-83.
[9] Cosio FG, Frankel WL, Pelletier RP, Pesavento TE, Henry ML,
Ferguson RM. Focal segmental glomerulosclerosis in renal allografts
with chronic nephropathy: implications for graft survival. Am J
Kidney Dis 1999;34:731-8.
[10] Cohen AH, Border WA, Glassock RJ. Nehprotic syndrome with
glomerular mesangial IgM deposits. Lab Invest 1978;38:610-9.
[11] Said SM, Cornell LD, Valeri AM, et al. C1q deposition in the renal
allograft: a report of 24 cases. Mod Pathol 2010;23:1080-8.
[12] Sato M, Kojima H, Nabeshima K, Nakajima Y, Koshikawa S. Primary
glomerulonephritis with predominant mesangial immunoglobulin G
deposits–a distinct entity? Nephron 1993;64:122-8.
[13] Yoshikawa N, Iijima K, Shimomura M, Nakamura H, Ito H. IgG-associated
primary glomerulonephritis in children. Clin Nephrol 1994;42:281-7.
[14] Fakhouri F, Darre S, Droz D, et al. Mesangial IgG glomerulonephritis:
a distinct type of primary glomerulonephritis. J Am Soc Nephrol 2002;
13:379-87.
[15] Celik B, Randhawa PS. Glomerular changes in BK virus nephropathy.
HUM PATHOL 2004;35:367-70.
[16] Brealey JK. Ultrastructural observations in a case of BK virus
nephropathy with viruses in glomerular subepithelial humps. Ultra-
struct Pathol 2007;31:1-7.
[17] Bracamonte E, Leca N, Smith KD, et al. Tubular basement membrane
immune deposits in association with BK polyomavirus nephropathy.
Am J Transplant 2007;7:1552-60.
[18] Richardson WP, Colvin RB, Cheeseman SH, et al. Glomerulopathy
associated with cytomegalovirus viremia in renal allografts. N Engl J
Med 1981;305:57-63.
[19] Vichot AA, Formica Jr RN, Moeckel GW. Cytomegalovirus
glomerulopathy and cytomegalovirus interstitial nephritis on sequen-
tial transplant kidney biopsies. Am J Kidney Dis 2014;63:536-9.
[20] Herrera GA, Alexander RW, Cooley CF, et al. Cytomegalovirus
glomerulopathy: a controversial lesion. Kidney Int 1986;29:725-33.
[21] Freese PM, Svalander CT, Molne J, Nyberg G. Renal allograft
glomerulopathy and the value of immunohistochemistry. Clin Nephrol
2004;62:279-86.
[22] Petersen VP, Olsen TS, Kissmeyer-Nielsen F, et al. Late failure or
human renal transplants. An analysis of transplant disease and graft
failure among 125 recipients surviving for one to eight years. Medicine
(Baltimore) 1975;54:45-71.