Kidney International, Vol. 63 (2003), pp. 1888–1893

Monocytes and peritubular capillary C4d deposition in acuterenal allograft rejection1

ALEX B. MAGIL and KATHRYN TINCKAM

Departments of Pathology and Laboratory Medicine, St. Paul’s Hospital and the Unversity of British Columbia, Vancouver,BC, Canada; and Division of Nephrology, Department of Medicine, University of British Columbia, Vancouver, BC, Canada

Monocytes and peritubular capillary C4d deposition in acute Complement split factor C4d is generated by the acti-renal allograft rejection. vation of the classical complement pathway by antigen-

Background. Peritubular capillary (PTC) deposition of com- antibody complexes and binds covalently to tissue ele-plement split factor C4d in renal allografts has been shown toments at the site of activation. Deposition of C4d alongbe closely associated with circulating antidonor antibodies and

a marker for relatively poor graft survival. Monocyte/macro- peritubular capillaries (PTC) has been shown to be closelyphage (MO) infiltration of renal allografts has been shown to correlated with circulating donor-specific antibodies inadversely affect graft survival. The purpose of this study was renal allograft recipients experiencing acute rejectionto assess whether the two phenomena are related.

and has been suggested as a marker for AHR [3, 5, 8].Methods. Twenty-three biopsies (from 15 patients) demon-Several studies have addressed the issue of morpho-strated diffuse strong staining of PTC for C4d (C4d� group)

and acute tubular injury with or without significant cellular re- logic markers of AHR [1, 5, 10]. The results of these inves-jection, while 28 biopsies (with acute rejection) but negative tigations have emphasized the association of neutrophilicfor PTC C4d served as controls (C4d� group).

infiltration in biopsies from patients with AHR [5, 10].Results. The C4d� group demonstrated significantly greaterOther lesions that have been shown to be significantlyglomerular and interstitial MO infiltration than did the C4d�

group [3.4 � 2.0 vs. 0.2 � 0.3 MO/glomerulus, P � 0.0001; correlated with AHR include glomerular and arterial12.9 � 9.2 vs. 6.5 � 5.0 MO/high power field (hpf), P � 0.0030]. fibrinoid necrosis, vascular thrombosis, infarction, andNeutrophilic (PMN) infiltration of glomeruli and PTC was also

glomerulitis [5, 10]. In these studies, many of the biopsiessignificantly greater in the C4d� group than in the C4d� one(0.8 � 0.6 vs. 0.3 � 0.3 PMN/glomerulus, P � 0.0003; 0.9 � from patients with AHR also demonstrated the histo-0.8 vs. 0.4 � 0.3 PTC PMN/hpf, P � 0.0035). logic features of acute cellular rejection (ACR), as de-

Conclusion. The results indicate a close association between fined by both the Banff and Cooperative Clinical TrialsPTC C4d deposition and MO infiltration, particularly glomeru-in Transplantation (CCTT) criteria [1, 5, 10].lar, and confirm previous observations regarding the correla-

tion of PTC C4d staining and PMN infiltration. Glomerular and interstitial monocyte/macrophage (MO)infiltration has been observed in biopsies demonstratingACR [11–20] and has been shown to be associated with

There has been considerable interest in acute humoral relatively poor allograft survival [11–16, 21]. As littlerejection (AHR) in recent years. This form of rejection mention has been made of MO in AHR, a detailed studyis often resistant to conventional antirejection therapy of biopsies demonstrating diffuse strong PTC C4d immu-[1–4] and has been shown to carry a relatively poor long- nostaining and the morphologic features associated withterm prognosis [5–8]. Recent studies have suggested that AHR was performed to evaluate this phenomenon. Theunconventional therapeutic measures such as plasma ex-

results indicate that MO infiltration, particularly glomer-change [with tacrolimus-mycophenolate rescue or intra-ular, is a significant feature of these biopsies and confirmvenous immune globulin (IVIG)] or immunoabsorptionprevious observations of significantly increased glomeru-may be of benefit in AHR [2, 9], underlining the impor-lar and PTC neutrophilic infiltration in these biopsiestance of recognizing AHR especially in allograft biopsies.compared to those demonstrating ACR and no PTC C4dreaction [5, 10].

1 See Editorial by Colvin, p. 1953.

Key words: acute humoral rejection, C4d, monocyte.METHODS

Received for publication August 13, 2002 Patientsand in revised form October 4, 2002, and November 15, 2002Accepted for publication December 17, 2002 For the purposes of this study, only those patients who

were biopsied within 6 months of transplantation and 2003 by the International Society of Nephrology

1888

Magil and Tinckam: C4d and monocytes in acute rejection 1889

whose allograft biopsies demonstrated diffuse strong more than 50% of PTC showed strong circumferentialstaining for C4d. For cross-study comparison purposes,PTC staining for C4d and histologic evidence of acute

tubular injury were included in the test group (C4d� biopsies showing focal PTC C4d staining were not in-cluded in this study.group). Twenty-three biopsies from 15 patients (10 fe-

males, 5 males) biopsied serially between January 1, 1999Clinical monitoringand June 30, 2002 met these criteria. Twenty-eight serial

biopsies done between January 1, 1999 and December The patients’ serum panel reactive antibody (PRA)level was determined by standard lymphocytotoxic assay31, 1999 from an unselected group of 24 patients (5 fe-

males, 19 males) with acute rejection by the Banff 97 prior to transplantation. Flow cytometric cross-match(FACSCalibur G3 Flow Cytometer; Becton-Dickinson,criteria (at least grade 1A) [22] and no staining for PTC

C4d served as controls (ACR group). Patient and donor Franklin Lakes, NJ) was performed at the time of trans-plantation in patients at high risk (PRA �30%) or indata were compiled primarily from the British Columbia

Transplant Society renal transplant database. Chart re- those receiving a second or greater transplant and wasused to detect circulating antidonor antibodies at theviews were performed when specific information was not

available from the database. time of rejection.Initial immunosuppression was achieved using a stan-

Histology dard triple therapy regimen of methylprednisolone, 1 mg/kg/day for 3 to 5 days, then prednisone, 0.7 to 1 mg/kg/Renal biopsies were divided into three portions for

light microscopy, electron microscopy, and immunohis- day, tapering to 0.3 mg/kg/day by 2 weeks posttransplant,and 0.15 mg/kg/day by 6 months (all patients); and cyclo-tochemistry. For light microscopy, tissue was fixed in

either B5 fixative (50 biopsies) and embedded in paraffin sporine, 9.0 mg/kg/day adjusted for trough levels 425 to500 ug/mL for the first 30 days, then tapered graduallyor Karnovsky’s fixative (two biopsies) and embedded in

polyglycol methacrylate. Sections embedded in paraffin to levels of 100 to 150 ug/L for long-term maintenance(29 patients), or tacrolimus, 0.12 to 0.15 mg/kg/day ad-were cut at 2 �m and stained with hematoxylin-eosin

(H&E), periodic acid-Schiff (PAS), and periodic acid- justed for trough levels 10 to 15 ug/mL for the first30 days, then tapered gradually to levels of 5 to 8 ug/mLsilver methenamine (PASM). Sections embedded in po-

lyglycol methacrylate were cut at 1 �m and stained with for long-term maintenance (11 patients) and mycophe-nolate mofetil, 2000 mg/day (all patients). Induction ther-H&E and PASM. The biopsies were examined by one

of the authors (A.B.M.). apy was used in four patients at high risk (PRA �30%)and consisted of a 7- to 10-day course of treatment with

Immunohistochemistry antilymphocyte antibodies (OKT3, Ortho Biotech, Rari-tan, NJ, USA), 5 mg/day. All rejection episodes wereThe avidin-biotin peroxidase complex procedure for

antibody localization was used. Acetone-fixed sections of treated with methylprednisolone for 3 to 6 days (total doserange 1500 mg to 4000 mg). Patients with Banff gradesnap-frozen renal tissue were stained with commercially

available mouse monoclonal antibodies specific for com- IIA or greater rejections were treated with OKT3 for7 to 10 days (5 mg/day). One C4d� patient had plasmaplement split factor C4d (Quidel, San Diego, CA, USA).

Snap-frozen sections from biopsies of membranous glo- exchange for two rejection episodes with a total of 10exchanges for each episode and IVIG following themerulonephritis that showed strong glomerular staining

for C4d served as positive controls for C4d. Lymph node plasma exchange for the second episode. Another C4d�patient received IVIG. A full response to therapy wastissue was used as positive controls for CD68 and CD3.

B5-fixed paraffin-embedded sections were stained with a return of renal function to baseline prerejection creati-nine. A partial response was defined as a decrease incommercially available mouse monoclonal antibodies spe-

cific for CD68 (a marker for MO) and CD3 (T-cell marker) creatinine of at least 20% from its peak but not to theprerejection baseline level. Evaluation of graft survival(Dako, Carpinteria, CA, USA). For the two biopsies in

which the histologic portion had been embedded in poly- was not done because of the short follow-up periods insome of the C4d� patients.glycol methacrylate, acetone-fixed sections of snap-

frozen renal tissue were stained for CD68 and CD3 withQuantitative analysiscommercially available mouse monoclonal antibodies

(Dako). Negative controls consisted of cases of thin base- All biopsies were scored according to the Banff 97criteria [22] to determine the type and grade of the rejec-ment membrane disease. Additional control studies were

performed by omitting the primary monoclonal antibody tion reaction. Neutrophils (PMN), MO, and T cells werecounted in all glomeruli in each biopsy and expressedin the staining procedure and by using an irrelevant

mouse monoclonal antibody as the primary antibody. as the number of cells per glomerulus. The number ofcortical PTC PMN and cortical interstitial (CI) MO perThe biopsies were evaluated by one of the authors

(A.B.M.). A biopsy was considered positive for C4d if high power field (hpf) (�40 objective with an object area

Magil and Tinckam: C4d and monocytes in acute rejection1890

Fig. 1. Immunohistochemistry for C4d� Group. (A) There is diffuse strong staining of peritubular capillaries for C4d. (B) The glomerulus containsseveral CD68� cells (monocytes) (arrowheads). (C ) Numerous CD68� cells (monocytes) are present in the interstitium. The glomerulus (G)contains a CD68� cell (arrowhead). (A) Magnification is �160; (B) and (C) Magnification is �260.

Table 1. Comparison of C4d� and acute cellular rejection (ACR)diameter of 0.5 mm and an area of 0.196 mm2) in each(C4d�) groups with respect to biopsy findings

biopsy was determined by counting the number of corti-Variable number C4d� ACR P ccal PTC PMN, CI MO, and cortical hpfs using a highNumber of biopsies 23 28power objective (�40) and dividing the number of cellsGlomerulitis % 13 (57) 3 (11) 0.0014by the number of cortical hpfs. For the quantitative anal-Acute tubular injurya % 22 (96) 11 (39) �0.0001

yses only those biopsies with cortex containing a mini- Peritubular capillarypolymorphonuclear leukocyteb % 23 (100) 28 (100) NSmum of four glomeruli were used. Thus two C4d� biop-

2.0 peritubular capillarysies with three glomeruli were not included in the MOpolymorphonuclear leukocyteb % 2 (9) 0 NS

and T-cell count determinations. Glomerular polymorphonuclearleukocyteb 22 (96) 26 (93) NS

1.0 polymorphonuclear leukocyteb/Statistical analysisglomeruli % 5 (22) 2 (7) 0.1193

Descriptive statistics are presented as mean � stan- Polymorphonuclear leukocyteb

tubulitis % 8 (35) 3 (11) 0.0823dard deviation (SD). Continuous variables were com-Glomerular monocyte/macrophage % 22 (96) 12 (43) 0.0002pared using the t test and categorical variables wereGlomerular T cells % 18 (78) 23 (82) NS

compared using the chi square (2) test. A P value of lessNS is not significant.

than 0.05 for two-sided univariate tests was considered a Acute tubular injury not associated with significant inflammation (i 2) ortubulitis (t 2)significant.

b Neutrophilc C4d� vs. ACR, 2 test

RESULTS

Renal biopsy findings1) with either no, mild, or moderate tubulitis (Banff 97All but two of the 23 biopsies from the 15 patientsscore: t � 0, 1 or 2) or moderate interstitial inflammationwith strong diffuse PTC C4d staining (C4d� group) (Fig.(Banff 97 score: i � 2) with no or mild tubulitis (Banff1) contained more than 10 glomeruli. The other two97 score: t � 0 or 1) and were graded as suspiciousbiopsies had nine glomeruli each. All of the C4d� biop-for cellular rejection according to the Banff 97 gradingsies had at least two interlobular arteries. Twenty of thesystem [22]. Eight of the C4d� biopsies also had acute28 biopsies with no PTC C4d staining (C4d� group)cellular interstitial rejection [Banff 97 grade 1A (N � 5)contained 10 or more glomeruli, while eight had nineor 2A (N � 3)]. The C4d� biopsies (ACR group) allglomeruli. Two of the C4d� biopsies lacked interlobulardemonstrated significant ACR [Banff 97 grades 1A (N �arteries. The mean intervals between the times of trans-13), 1B (N � 5), 2A (N � 9), 3 (N � 1)].plantation and times of biopsy for the C4d� and the

The significant biopsy findings are summarized in Ta-C4d� groups were 22 � 20 days and 31 � 43 days,ble 1. Compared to the C4d� group, the C4d� grouprespectively, the difference not being significant.showed significantly more glomerulitis (P � 0.0003),The C4d� biopsies could be divided into two sub-acute tubular injury not associated with significant tubul-groups based on whether there was a significant cellularitis (t 2) or inflammation (i 2) (P � 0.0001) andcomponent to the acute rejection reaction or not. Fifteenglomerular MO infiltration (Fig 1) (P � 0.0001). Theof the C4d� biopsies demonstrated either mild intersti-

tial mononuclear cellular infiltration (Banff 97 score: i � sensitivity and specificity of glomerular MO infiltration

Magil and Tinckam: C4d and monocytes in acute rejection 1891

Table 2. Results of quantitative analysis of biopsy findings for the C4d� and the acute cellular rejection (ACR) (C4d�) groups

Variable number C4d� ACR P a

Polymorphonuclear leukocyte/glomerulus (range) 0.8�0.6 (0–3.1) 0.3�0.3 (0–1.4) 0.0003Peritubular capillary polymorphonuclear leukocyte/

high power field (�40) (range) 0.9�0.8 (0.1–3.5) 0.4�0.3 (0.1–1.1) 0.0035Monocyte/glomerulus (range) 3.4�2.0 (0.1–8.4) 0.2�0.3 (0–1.0) �0.0001Cortical interstitium monocyte/high power field (�40) (range) 12.9�9.2 (7.0–36.2) 6.5�5.0 (1.0–22.8) 0.0030T cells/glomerulus (range) 1.4�0.9 (0–3.0) 1.1�1.7 (0–6.9)

NS is not significant.a C4d� vs. ACR, t test

for the C4d� group was 0.96 and 0.57, respectively, for merulus or two or more PTC PMNs per hpf than dida mean MO/glomerulus �0 and 0.91 and 0.93, respec- those with ACR (C4d�). In the present study, five (22%)tively, for a mean MO/glomerulus 1.0. Varying num- C4d� biopsies and two (7%) C4d� biopsies had an aver-bers of PMN were observed in the glomeruli and CI age of one or more PMNs per glomerulus, but the differ-PTC in all of the C4d� and most of the C4d� biopsies. ence was not significant. Two (9%) of the C4d� biopsiesCortical interstitial MO were present in all biopsies but and none of the C4d� biopsies had an average of 2.0 orthey appeared more prominent in the C4d� ones (Fig more PMNs per PTC.1). In most of the immunostained sections, it could not A comparison of the C4d� group to one expandedbe determined with confidence whether the CI CD68� by the addition of 10 serially selected allograft biopsiescells were within or outside the PTC. Neutrophilic tubul- (from nine patients) showing acute tubular injury sec-itis was observed in proportionately more C4d� biopsies ondary to ischemia or calcineurin inhibitor toxicity re-(35%) than in C4d� ones (11%) but the difference was vealed no significant differences between the two groupsnot significant. Most of the C4d� (78%) and C4d� with respect to the biopsy variables under study (data(82%) biopsies showed varying numbers of glomerular not shown).T cells. Intimal arterial fibrinoid necrosis of was observedin one C4d� and one C4d� biopsy. None of the biopsies Clinical datashowed glomerular fibrinoid change. One C4d� and one There were significantly more females in the C4d�C4d� biopsy demonstrated focal glomerular capillary group (67%) than in the C4d� one (21%) (2 � 6.371,thrombosis. Although focal interstitial medullary hemor- P � 0.0116). The mean age of the recipients was similarrhage was noted in 4 of 18 (22%) C4d� biopsies in which

in both groups (C4d�, 45.9 � 12.3 years; C4d�, 48.5 �the medulla was present and in one of 18 (6%) C4d�

10.8 years) The number of cadaveric, living-related andbiopsies with medulla, the difference was not significant.

living-unrelated donors was eight (53%), four (27%),Within the C4d� group there were no significant differ-and three (20%), respectively, for the C4d� group andences between the subgroups with and without acute11 (46%), eight (33%), and five (21%), respectively, forcellular interstitial rejection with respect to any of thethe C4d� patients. Significantly more C4d� patients hadvariables assessed (data not shown).a second or third transplant (40%) than did the C4d�

group (8%) (2 � 3.901, P � 0.0483). The C4d� groupQuantitative analysishad a significantly higher proportion (62%) of patientsThe results are summarized in Table 2. The C4d� groupwith a PRA �20% prior to transplantation than did theshowed significantly greater glomerular infiltration byC4d� group (12%) (2 � 8.373, P � 0.0038). An anti-both PMNs (P � 0.0003) and MO (P � 0.0001) than diddonor antibody determination at the time of rejectionthe C4d� group. The mean number of PMNs in PTCswas performed in only one AHR patient and this waswas significantly higher in the C4d� biopsies than in thepositive. Monoclonal antilymphocyte antibody therapyC4d� ones (P � 0.0035). CI MO infiltration was signifi-was used in 53% and 29% of C4d� and C4d� patients,cantly greater in the C4d� group than in the C4d� grouprespectively. One C4d� patient was treated for two re-(P � 0.0030). The level of glomerular T-cell infiltrationjection episodes by plasma exchange with full response.was very similar in both groups. There was no significantThe one C4d� patient treated by IVIG alone did notdifference between the two C4d� subgroups (with andrespond to therapy. Twenty-eight percent of the rejec-without ACR) with respect to any of the above variablestion episodes in the C4d� group failed to respond either(data not shown).fully or partially to antirejection therapy while none ofIn a previous study, Mauiyyedi et al [5] showed thatthe C4d� rejection episodes failed to respond (2 �a significantly higher proportion of cases with AHR

(C4d�) had an average of one or more PMNs per glo- 5.956, P � 0.0147).

Magil and Tinckam: C4d and monocytes in acute rejection1892

DISCUSSION result in a significant correlation is unknown. Furtherstudy is needed to answer these questions.It has been proposed that a definite diagnosis of AHR

Previous biopsy studies have emphasized the signifi-requires the demonstration of circulating antidonor anti-cance of glomerular and PTC PMN infiltration as mark-body and that the term “suspicious for AHR” be re-ers for AHR [5, 10]. In the present investigation, glomer-served for cases in which the allograft biopsies demon-ular and PTC PMNs were present in all of the C4d�strate the histologic and immunopathologic features ofbiopsies and in significantly greater numbers than in theAHR, but circulating antidonor antibody either cannotACR (C4d�) biopsies, in agreement with previous re-be detected or has not been tested for [5]. As antidonorports [5, 10], but the numbers were smaller than thoseantibody assessment was performed in only one C4d�presented in a previous publication [5]. Whereas Mauiy-patient in the present retrospective study, it is uncertainyedi et al [5] detected 2.0 PTC PMNs per hpf in 65%whether the other C4d� cases had definite AHR or wereof biopsies from patients with AHR, only 9% of theonly suspicious for this entity. In view of the strong as-C4d� biopsies in the present investigation showed this.sociation of strong diffuse PTC C4d immunostaining inSimilarly, 1.0 PMN per glomerulus was demonstratedallograft biopsies and circulating antidonor antibodyin 55% of AHR biopsies in the study of Mauiyyedi et al[5, 8], it is likely that most or all of the C4d� patients[5] but in only 22% of our C4d� specimens. Possiblein the present study had AHR.explanations of these discrepancies include differencesThe results of this investigation demonstrate for thein recipient demographics and donor characteristics.first time that glomerular, and to a lesser extent CI, MO

Criteria for the diagnosis of AHR have recently beeninfiltration is closely associated with PTC C4d staining.proposed [5]. These include (1) diffuse and strong PTCTrpkov et al [10] noted that transplant glomerulitis wasstaining for C4d; (2) at least one of an average of 2.0observed significantly more often in patients with AHRPTC PMNs per hpf, arterial fibrinoid necrosis, or acutecompared to those with ACR but did not determine thetubular injury; and (3) circulating antidonor antibody [5].cell type(s) involved. Similarly, Nickeleit et al [23] ob-In view of the results of the present study, we propose

served a significant correlation between PTC C4d stain-that transplant glomerulitis due to MO infiltration be

ing and transplant glomerulitis but did not elaborate on added to the list of histologic criteria described above [5].the cell type(s) infiltrating the glomeruli. A very recentstudy of the pathology of AHR did not evaluate the renal ACKNOWLEDGMENTSbiopsies for MO infiltration [5].

This study was presented in part at the 91st Annual Meeting ofGlomerular and interstitial MO infiltration in renalthe United States-Canadian Academy of Pathology, Chicago, Illinois,

allograft biopsies has been assessed in many previous February 23 to March 1, 2002. We thank Linda Hughes, Jasbir Gill,and Lee Cross for their excellent technical assistance.studies but not specifically in the setting of AHR [11–20].

The MO infiltration has been shown to be associated with Address correspondence and reprint requests to: Dr. Alex Magil,expression of monocyte chemotactic peptide-1 (MCP-1) Laboratory, St. Paul’s Hospital, 1081 Burrard Street, Vancouver, BC,

Canada, V6Z 1Y6, Tel: (604) 806 8700, Fax: (604) 806 8701.[24], granulocyte-macrophage colony-stimulating factorE-mail: amagil@providencehealth.bc.ca(GM-CSF) [25] and macrophage migratory inhibitory

factor (MIF) [26]. Moreover, a number of reports haveREFERENCESdemonstrated that glomerular and/or interstitial MO in-

1. Halloran PF, Schlaut J, Solez K, et al: The significance of thefiltration have prognostic significance and are associatedanti-class I response. II. Clinical and pathologic features of renal

with relatively poor graft survival [11–16, 21]. transplants with anti-class I-like antibody. Transplantation 53:550–555, 1992Several investigations have demonstrated that PTC

2. Pascual M, Saidman S, Tolkoff-Rubin N, et al: Plasma exchangeC4d deposition is an independent prognostic factor andand tacrolimus-mycophenolate rescue for acute humoral rejection

is associated with relatively poor outcome [5–8]. In view in kidney transplantation. Transplantation 66:1460–1464, 19983. Collins AB, Schneeberger EE, Pascual MA, et al: Complementof the relationship between PTC C4d deposition and

activation in acute humoral renal allograft rejection: diagnosticglomerular and CI MO infiltration demonstrated in thesignificance of C4d deposits in peritubular capillaries. J Am Soc

present study and the results of previous investigations Nephrol 10:2208–2214, 19994. Crespo M, Pascual M, Tolkoff-Rubin N, et al: Acute humoralwith respect to the effect of MO on graft survival [11–16,

rejection in renal allograft recipients: I. Incidence, serology and21], it is uncertain whether either of these phenomena clinical characteristics. Transplantation 71:652–658, 2001is an independent prognostic factor. Similarly, interstitial 5. Mauiyyedi S, Crespo M, Collins AB, Schneeberger EE, et al:

Acute humoral rejection in kidney transplantation: II. Morphology,hemorrhage, which has been shown to be associated withimmunopathology, and pathologic classification. J Am Soc Nephrolclinically severe rejection [27] and decreased 1-year graft 13:779–787, 2002

survival [28], occurred more frequently in C4d� biopsies 6. Herzenberg AM, Gill JS, Djurdjev O, et al: C4d deposition inacute rejection: an independent log-term prognostic factor. J Amthan in C4d� ones as demonstrated in the present studySoc Nephrol 13:234–241, 2002and in a previous one [5], but the differences were not 7. Regele H, Exner M, Watschinger B, et al: Endothelial C4d depo-sition is associated with inferior kidney allograft outcome inde-significant. Whether larger numbers of biopsies would

Magil and Tinckam: C4d and monocytes in acute rejection 1893

pendently of cellular rejection. Nephrol Dial Transplant 16:2058– of macrophages and myofibroblasts. A new marker for develop-ment of chronic renal allograft rejection. Transplantation 69:2658–2066, 20012662, 20008. Bohmig GA, Exner M, Habicht A, et al: Capillary C4d deposition

19. Ozdemir BH, Ozdemir FN, Gungen Y, et al: Role of Macrophagesin kidney allografts: a specific marker of alloantibody-dependentand lymphocytes in the induction of neovascularization in renalgraft injury. J Am Soc Nephrol 13:1091–1099, 2002allograft rejection. Am J Kidney Dis 39:347–353, 20029. Bohmig GA, Regele H, Exner M, et al: C4d-positive acute hu-

20. Burkhardt K, Bosnecker A, Hillebrand G, et al: MRP8/14-moral renal allograft rejection: effective treatment by immuno-positive macrophages as early acute cellular rejection markers, andabsorption. J Am Soc Nephrol 12:2482–2489, 2001soluble MRP8/14 and increased expression of adhesion molecules10. Trpkov K, Campbell P, Pazderka F, et al: Pathologic featuresfollowing renal transplantation. Transplant Proc 27:890–891, 1995of acute renal allograft rejection associated with donor-specific

21. Yanagisawa T, Otsubo O, Shimada T, et al: Relationship betweenantibody. Transplantation 61:1586–1592, 1996macrophage infiltration of renal allografts and chronic renal im-11. Hayry P, von Willebrand E: Monitoring of human renal allograftpairment. Tranplant Proc 29:2783–2786, 1997rejection with fine-needle aspiration cytology. Scand J Immunol 13:

22. Racusen LC, Solez K, Colvin RB, et al: The Banff 97 work-87–97, 1981ing classification of renal allograft pathology. Kidney Int 55:713–12. Harry TR, Coles GA, Davies M, et al: The significance of mono-723, 1999cytes in glomeruli of human renal transplants. Transplantation 37:

23. Nickeleit V, Zeiler M, Gudat F, et al: Detection of the comple-70–73, 1984ment degradation product C4d in renal allografts: diagnostic and13. Om A, Baquero A, Raja R, et al: The prognostic significance of thetherapeutic implications. J Am Soc Nephrol 13:242–251, 2002presence of monocytes in glomeruli of renal transplant allografts.

24. Grandaliano G, Gesualdo L, Ranieri E, et al: Monocyte chemo-Transplant Proc 19:1618–1622, 1987tactic peptide-1 expression and monocyte infiltration in acute renal14. Burkhard K, Hofmann GO, Bosnecker A, et al: Early infiltrationtransplant rejection. Transplantation 63:414–420, 1997of renal allografts with 27E10-positive macrophages and graft out- 25. Kajiwara I, Kawamura K, Takebayashi S: An analysis of mono-come. Transpl Int 7(Suppl 1):S577–S579, 1994 cyte/macrophage subsets and granulocyte-macrophage colony-

15. Ramos EL, Barri YM, Croker BP, et al: Thromboxane synthase stimulating factor expression in renal allograft biopsies. Nephronexpression in renal transplant patients with rejection. Transplanta- 73:536–543, 1996tion 59:490–494, 1995 26. Lan HY, Yiang N, Brown FG, et al: Macrophage migration inhibi-

16. Croker BP, Clapp WL, Abu Shamat ARF, et al: Macrophages tory factor expression in human renal allograft rejection. Trans-and chronic renal allograft nephropathy. Kidney Int 50(Suppl 57): plantation 66:1464–1471, 1998S42–S49, 1996 27. Colvin RB, Cohen AH, Saiontz C, et al: Evaluation of pathologic

17. Grimm PC, McKenna R, Nickerson P, et al: Clinical rejection is criteria for acute renal allograft rejection: reproducibility, sensitiv-distinguished from subclinical rejection by increased infiltration ity, and clinical correlation. J Am Soc Nephrol 8:1930–1941, 1997by a population of activated macrophages. J Am Soc Nephrol 10: 28. Nickeleit V, Vamvakas EC, Pascual M, et al: The prognostic1582–1589, 1999 significance of specific arterial lesions in acute renal allograft rejec-

tion. J Am Soc Nephrol 9:1301–1308, 199818. Pilmore HL, Painter DM, Bishop GA, et al: Early up-regulation