1. 0041-1337/03/7509-1490/0
TRANSPLANTATION Vol. 75, 1490–1495, No. 9, May 15, 2003
Copyright © 2003 by Lippincott Williams & Wilkins, Inc. Printed in U.S.A.
BENEFICIAL EFFECT OF PLASMAPHERESIS AND INTRAVENOUS
IMMUNOGLOBULIN ON RENAL ALLOGRAFT SURVIVAL OF
PATIENTS WITH ACUTE HUMORAL REJECTION1
PAULO N. ROCHA,2,5 DAVID W. BUTTERLY,2 ARTHUR GREENBERG,2 DONAL N. REDDAN,2
JANET TUTTLE-NEWHALL,3 BRADLEY H. COLLINS,3 PAUL C. KUO,3 NANCY REINSMOEN,4
TIMOTHY FIELDS,4 DAVID N. HOWELL,4 AND STEPHEN R. SMITH2
Background. Acute humoral rejection (AHR) has immunosuppression protocols directly target T lymphocytes
been associated with enhanced graft loss. Our study (2). Conversely, the role of alloantibodies in acute rejection is
compared the renal allograft survival of patients with far less clear. During the last decade, however, a substantial
AHR treated with plasmapheresis (PP) and intrave- body of evidence has indicated a central role for antibodies in
nous immunoglobulin (IVIG) with allograft survival in the pathogenesis of acute rejection in a subset of renal trans-
patients with acute cellular rejection (ACR).
plant recipients (3–11). These patients tend to present soon
Methods. We retrospectively analyzed all kidney
transplants performed at our institution between Jan- after transplant with severe allograft dysfunction that is usu-
uary 1999 and August 2001 (n 286). Recipients were ally unresponsive to conventional anti-T– cell therapy (11).
classified into three groups according to biopsy re- This type of accelerated acute rejection, also termed de-
ports: AHR, ACR, or no rejection. The ACR group was layed antibody-mediated rejection or acute humoral rejection
further divided into early and late rejection (<90 and (AHR), may represent an anamnestic antibody response to
>90 days posttransplant, respectively). donor alloantigens (11, 12) or, alternatively, reflect the pres-
Results. After a mean follow-up of 569 19 days, the ence of low levels of preformed antibodies not detected by the
incidence of AHR was 5.6% (n 16). Recipient presen- usual crossmatch techniques. AHR is not to be confused with
sitization, delayed graft function, early rejection, and the immediate type of antibody-mediated rejection, hyper-
higher creatinine at diagnosis were characteristic of
acute rejection, in which high amounts of preformed antibod-
AHR. Most AHR patients (14/16) were treated with PP
and IVIG. One patient received only IVIG, whereas ies destroy the allograft within minutes of vascular anasto-
another received only PP. All AHR patients were given mosis. Earlier studies using plasmapheresis (PP) to remove
steroid pulses, but only four received antilymphocyte donor-specific alloantibodies (DSAs) from patients with AHR
therapy because of concomitant severe ACR. The ACR yielded disappointing results (13–15). More recently, various
group comprised 43 patients (15%). One patient with groups have successfully treated AHR with PP or intrave-
mild rejection received no therapy, 20 improved with nous immunoglobulin (IVIG) or both in combination (9, 16,
steroids alone, and 22 required additional antilympho- 17). In most of these studies, the patients received antilym-
cyte therapy. One-year graft survival by Kaplan Meier phocyte therapy in addition to PP and IVIG.
analysis was 81% and 84% in the AHR and ACR groups, At our institution, we have noted an increase in the inci-
respectively (P NS). Outcomes remained similar
dence of AHR during the last few years. It has been our
when AHR patients were compared with those with
early ACR. practice to initiate PP and IVIG as soon as the diagnosis of
Conclusions. We conclude that AHR, when diag- AHR is made, as opposed to reserving this therapy for cases
nosed early and treated aggressively with PP and unresponsive to steroids and antilymphocyte preparations.
IVIG, carries a short-term prognosis that is similar to This report focuses on the renal allograft outcome of 16
ACR. patients diagnosed with AHR and treated with PP and IVIG
during the study period (January 1999 –August 2001).
The immunologic response to a renal transplant involves
cellular and humoral components. Cell-mediated immunity
MATERIALS AND METHODS
in transplantation has been extensively studied and long
considered to be the main effector during acute allograft Study Group
rejection (1). In fact, current induction and maintenance We conducted a retrospective analysis of all kidney transplants
performed at Duke University Medical Center from January 1999 to
1
This work was presented in abstract format at the American August 2001. A total of 286 recipients were identified and followed
Transplant Congress, Washington, DC, April 29, 2002. until July 2002. For the purposes of this study, combined kidney-
2
Department of Medicine, Duke University Medical Center, Duke pancreas transplants were analyzed as cadaveric kidney trans-
University Medical Center, Durham, North Carolina. plants. All pertinent patient information was obtained from elec-
3
Department of Surgery, Duke University Medical Center, Duke tronic medical records. The renal biopsy reports were used to classify
University Medical Center, Durham, North Carolina. recipients into three groups: AHR, ACR, or no rejection (NO REJ).
4
Department of Pathology, Duke University Medical Center, The ACR group was further divided into early and late rejection ( 90
Duke University Medical Center, Durham, North Carolina. and 90 days posttransplant, respectively). When features of both
5
Address correspondence to: Paulo N. Rocha, Box 3014, Duke ACR and AHR were present in the same biopsy specimen, the patient
University Medical Center, Durham, NC 27710. E-mail: rocha001@ was included in the AHR group. The remaining subjects who did not
mc.duke.edu. undergo renal biopsy or whose biopsy reports revealed conditions
Received 4 November 2002. Revision requested 23 December 2002. other than acute rejection were included in the NO REJ group. One
Accepted 7 January 2003. patient was empirically treated for rejection without a renal biopsy;
DOI: 10.1097/01.TP.0000060252.57111.AC 1490

2. May 15, 2003 ROCHA ET AL. 1491
he received only steroids and was included in the ACR group. Renal Treatment of Acute Rejection
biopsies were performed only to evaluate clinically significant graft The majority (14/16) of our patients with AHR received the combined
dysfunction (delayed graft function [DGF] or rise in serum creatinine regimen of PP and IVIG. For reasons that were not stated in the
after initial function). We defined DGF as a need for renal replace- medical records, one patient was treated with only PP, whereas another
ment therapy within the first week of transplant. received only IVIG. A typical PP regimen consisted of four daily sessions
(range 3– 6) with 5% human albumin replacement. The number of PP
Antibody Detection, Specificity Analysis, and Crossmatch Techniques sessions was based on the clinical response to therapy as measured by
urine output and serum creatinine. Alloantibody titers were not rou-
The conventional complement-dependent cytotoxicity (CDC) tech-
tinely measured and, therefore, not used to guide the duration of PP.
nique with and without antihuman globulin (AHG) was used to
IVIG was administered after the last PP session. The average patient
define antibody specificity and to test donor-recipient crossmatches. received a total of 215 53 g (approximately 2g/kg), but there was a wide
Antibodies were detected and specificity was defined using a panel of variation in dosage (range 30– 420 g in 1–3 divided doses). In patients
cells (56 for T cells and 26 for B cells, Pel Freez Inc., Brown Deer, WI) who received more than one dose of IVIG, the subsequent doses were
selected to represent all major human leukocyte antigens (HLAs). usually given in the outpatient setting (range 3–15 days after first
The crossmatches were performed using purified T and B cells, dose). All patients with AHR were treated with intravenous pulses of
undiluted serum, and a series of three serum dilutions. As controls, methylprednisolone (500 mg) to treat a concomitant component of ACR
autologous crossmatches were performed. In patients known to have or empirically before the final diagnosis was made. Antilymphocyte
immunoglobulin-M antibodies, sera were treated with dithiothreitol antibody therapy was not routinely used to treat AHR; however, when
before testing. All patients included in this study demonstrated AHR was accompanied by severe ACR, OKT3 (three cases) or anti-
negative crossmatches by the CDC-AHG technique. We also used the thymocyte globulin (one case) was given to treat the cellular component.
flow cytometry technique (Flow-Panel Reactive Antibody [PRA], One Of the 43 patients with ACR, 20 improved with steroids alone, whereas
Lambda, Inc., Canoga Park, CA) to detect anti-HLA antibodies as 22 required additional antilymphocyte antibody therapy with OKT3 (20
previously described (18). cases) or anti-thymocyte globulin (two cases) because of severe rejection
(Banff grade II or greater) or steroid resistance.
Pathology
Statistics
A portion of each biopsy specimen was routinely processed and We used the two-tailed, unpaired t test to compare continuous
stained by hematoxylin-eosin, periodic acid-Schiff, methenamine sil- variables and the chi-square test to compare dichotomous variables
ver, and Masson trichrome methods. ACR was diagnosed and graded between groups; when sample sizes were small, the Mann-Whitney
according to the Banff 1997 schema (12). Clear histologic criteria for U test was used for continuous variables to avoid distributional
the diagnosis of AHR have yet to be established. Provisionally, this assumptions. All calculations were performed using MINITAB for
diagnosis was rendered when a biopsy contained an interstitial in- Windows version 10.2 (Minitab Ltd., State College, PA). Graphs were
filtrate of inflammatory cells (1) involving more than 25% of the created with SigmaPlot 2000 for Windows version 6.00 (SPSS, Chi-
biopsy tissue, (2) focused on peritubular capillaries, (3) composed at cago, IL). The results were expressed as mean standard error of the
least in part of neutrophils, and (4) usually accompanied by only mean or median and interquartile range. Statistical significance was
minimal or mild lymphocytic tubulitis. All biopsies initially classified defined as a P value less than 0.05. Survival analysis was performed
as AHR were re-reviewed by an experienced renal pathologist with SAS software version 8.2 (SAS Institute Inc., Cary, NC) using
(D. N. H.) and confirmed to be consistent with humoral rejection the Kaplan-Meier method, and comparisons between survival curves
before inclusion in this study. Routine staining for C4d on all biop- were made with the log-rank test.
sies with histologic evidence of AHR began in February 2001. The
staining was performed on a portion of fresh-frozen tissue by the RESULTS
immunofluorescence method using a mouse monoclonal anti-C4d
We identified 286 kidney transplant recipients between
primary antibody (clone 10 –11, Biogenesis, Sandown, NH). For the
January 1999 and August 2001 and followed them for a mean
AHR biopsies received before February 2001 (n 12), C4d staining
was retrospectively performed on stored frozen tissue.
Induction and Maintenance Immunosuppression
Recipients of pediatric en bloc kidneys, simultaneous kidney-pan-
creas allografts, and those at high immunologic risk (African Amer-
icans, recipients of living unrelated donors, those with lupus, and
those with a high risk of acute tubular necrosis) received induction
therapy with daclizumab (two doses 1 mg/kg). Presensitized pa-
tients, such as those with a peak PRA greater than 40, recipients of
a second transplant after early graft loss, and recipients of a third or
greater transplant, were given Thymoglobulin (five doses 1.5 mg/kg).
All other patients received no induction. Maintenance immunosup-
pression consisted of a calcineurin inhibitor, mycophenolate, and
prednisone. Tacrolimus was the calcineurin inhibitor used in 80% of
our patients; the target trough level was 10 to 15 ng/mL during the
first month and 5 to 10 ng/mL afterward. Mycophenolate mofetil was
started at 1 g twice daily and dose-adjusted, if needed, for gastroin-
testinal or bone marrow toxicity. Methylprednisolone was given as a
250 mg intravenous bolus in the operating room and then tapered to FIGURE 1. Retrospective analysis of kidney and kidney-pan-
30 mg oral prednisone by postoperative day 4. The prednisone dose creas transplants performed at Duke University Medical
was tapered by 5 mg every 2 weeks until a maintenance dose of 5 to Center. Data were collected on transplants from January
10 mg per day was achieved. 1999 to August 2001. Follow-up was extended until July 2002.

3. 1492 TRANSPLANTATION Vol. 75, No. 9
TABLE 1. Pretransplant characteristics tory infiltrate, and in many, they represented the main pop-
AHR ACR NO REJ ulation. The baseline characteristics of patients according to
Parameters
(n 16) (n 43) (n 227) rejection status are shown in Table 1. Although most demo-
Age 44.9 2.7 43.6 2.0 46.8 0.9 graphic values were not significantly different between rejec-
Gender tion groups, there was a trend toward a predominance of
Male 38% (6)a 56% (24) 63% (142) women and African Americans in patients with AHR; specif-
Female 63% (10)a 44% (19) 37% (85) ically, only one white man was diagnosed with AHR. In
Race addition, patients with AHR were more likely to have re-
White 38% (6) 49% (21) 59% (133)
ceived a cadaveric kidney and to have a positive PRA. Mean
Black 63% (10) 51% (22) 37% (85)
Others 0% (0) 0% (0) 4% (9)
peak historic PRAs were significantly higher in patients who
Transplant type went on to develop AHR than in those who developed ACR or
Cadaveric 88% (14) 72% (31) 65% (148) no rejection. In response to this higher degree of presensiti-
Living donor 13% (2) 28% (12) 35% (79) zation, 81% of patients in the AHR group received induction
Positive PRA ( 10%) 56% (9)b 28% (12) 20% (44) antibody therapy before transplantation. This percentage
Peak historic PRA T 46.7 11.3c,d 17.0 5.1 11.9 1.9 was greater than that seen in the ACR (56%) and NO REJ
Peak historic PRA B 30.6 10.7d,e 9.9 4.0 7.3 1.6 (54%) groups but did not reach statistical significance. Age
Induction therapy 81% (13) 56% (24) 54% (122) and cold ischemia time were similar between groups.
a
P 0.047 vs. NO REJ. The overall incidence of DGF was 25%. DGF was more
b
P 0.014 vs. NO REJ. common in the AHR group (69%) when compared with the
c
P 0.01 vs. ACR. ACR (30%) and NO REJ (21%) groups (Fig. 2a). The serum
d
P 0.001 vs. NO REJ.
e
creatinine at the time of diagnosis of rejection was signifi-
P 0.05 vs. ACR.
cantly higher in patients with AHR (AHR 8.3 1.0 vs. ACR
PRA, panel reactive antibody; AHR, acute humoral rejection;
ACR, acute cellular rejection; NO REJ, no rejection. 5.3 0.4, P 0.003). The initial hospital stay was significantly
shorter if patients did not have an episode of rejection
(7.0 0.2 days, P 0.001 vs. AHR and ACR), but it was not
of 569 19 days (Fig. 1). At the end of follow-up, at least one different between rejection groups (AHR 15.4 1.1 vs. ACR
episode of acute allograft rejection had occurred in 59 pa- 12.1 1.7, P NS).
tients; in 43 patients (73%), the rejection was classified as The treatment of AHR consisted of PP (one case), IVIG (one
cellular, which corresponds to an ACR incidence of 15% (43/ case), or both PP and IVIG in combination (14 cases). Fifteen
286). AHR was diagnosed histologically in 16 patients, for an patients received treatment with one of these modalities as
incidence of 5.6% of all transplants (16/286) and 27% of all soon as the diagnosis of AHR was made; all such patients
acute rejection episodes (16/59). Nine patients demonstrated demonstrated a positive initial response to treatment as ev-
biopsy findings of both AHR and ACR; the remaining seven idenced by a decrease in serum creatinine and, for those
patients demonstrated “pure AHR.” In all cases classified as requiring renal replacement therapy, the ability to discon-
AHR, neutrophils constituted at least 10% of the inflamma- tinue dialysis. In one case, AHR was initially treated with
FIGURE 2. High incidence of delayed graft function (DGF) and early rejection in acute humoral rejection (AHR). (a) Incidence
of DGF according to rejection status (†P 0.008 vs. acute cellular rejection [ACR], ‡P<0.001 vs. no rejection [NO REJ], 2 test).
(b) Time to diagnosis of acute rejection. The 25th percentile (boundary of the box closest to zero); median (line within the box);
75th percentile (boundary of the box farthest from zero); 90th and 10th percentiles (whiskers above and below the box); and
outliers (filled circles beyond the whiskers). Data on the y-axis are expressed on a logarithm scale.

4. May 15, 2003 ROCHA ET AL. 1493
only steroids, and PP and IVIG were withheld until a second nosis, we further divided the ACR group into early ( 90 days,
biopsy (1 week later) showed persistent humoral rejection. In 25 patients) and late ( 90 days, 18 patients) cellular rejec-
this case, AHR could not be reversed, and the patient expe- tion. When we compared the AHR group with the early ACR
rienced primary nonfunction. Two patients in the AHR group group, the percentage reaching ESRD and the last creatinine
experienced graft loss after an initial response to therapy. of surviving grafts was not significantly different (Table 2).
The reasons for graft loss were pyelonephritis involving the To refine the diagnosis of AHR, we stained all AHR biop-
allograft and chronic rejection, 2 and 17 months posttrans- sies with anti-C4d antibody (Table 3). We were able to doc-
plant, respectively. A fourth patient in the AHR group died of ument strong staining in peritubular capillaries in 9 of 16
sepsis 1 month posttransplant in the setting of pancytopenia patients (56%). Staining was equivocal in two patients
attributed to mycophenolate mofetil. (present but with only faint staining of peritubular capillar-
In patients without episodes of acute rejection, graft sur- ies) and clearly negative in five. Further, we used the flow
vival at 1 year was 94%, which is significantly better than in cytometry technique to detect anti-HLA antibodies in the
both acute rejection groups (P 0.01 for NO REJ vs. AHR, stored pretransplant sera of all AHR patients. Despite a
P 0.007 for NO REJ vs. ACR). The 1-year graft survival in negative crossmatch by the CDC-AHG technique, we found
patients with AHR was 81%; this is no different than what that seven patients showed circulating donor antigen-specific
was observed in patients with ACR (AHR 81% vs. ACR 84%, anti-HLA antibodies at the time of transplant; six were anti-
P NS) (Fig. 3). major histocompatibility complex (MHC) class I and one was
In general, AHR occurred soon after transplant, and on anti-MHC class II. In three patients, posttransplant serum
only one occasion was the diagnosis made more than 15 days was available for testing. We detected de novo anti-MHC
postoperatively (Fig. 2b). In contrast, many cases of ACR class II alloantibodies against donor antigens at the time of
happened late after transplant (median 6 and 49 days for allograft rejection in two of these patients. In the other, a new
AHR and ACR, respectively, P 0.001). Because late episodes class II antibody was also found, but it was not directed at
of rejection have been associated with a worse allograft prog- donor HLAs. Taken together, these data substantiate the
diagnosis of AHR by at least two methods (histology and C4d
or flow cytometry) in 12 of 16 patients. One-year graft sur-
vival for these 12 patients was 75%. When only the nine
C4d-positive patients are considered, 1-year graft survival is
78%. These survival rates are not significantly different than
those observed in the larger AHR and ACR groups.
DISCUSSION
The main findings of this single-center retrospective anal-
ysis were that AHR occurs soon after transplant in a sub-
group of highly sensitized recipients, is associated with DGF,
and responds well to treatment with PP and IVIG. The 1-year
graft survival of AHR patients treated with this regimen was
81%; this is superior to what has been previously reported
and, in fact, no different than what was observed in patients
with ACR.
FIGURE 3. Kaplan Meier survival according to rejection sta- On the basis of histologic criteria alone, the incidence of
tus. NO REJ group (dashed line), ACR group (solid line), and AHR during the entire study was 5.6%, corresponding to 27%
AHR group (dotted line). Values on top of horizontal axis of all acute rejection episodes. However, in only 12 of the
represent the number of patients at risk. Death with graft initial 16 patients with AHR was the diagnosis confirmed by
function was included as allograft loss. One-year graft sur-
at least two techniques (histology and C4d or flow cytom-
vival was significantly better in the NO REJ group when
compared with the ACR and AHR groups (P<0.001, log rank
etry). Thus, the true incidence of AHR at our institution was
test). Survival was not different between the ACR and AHR between 4.2% (12/286) and 5.6% (16/286). This is consonant
groups (P 0.69). with other reports that the overall incidence of AHR ranges
TABLE 2. Renal and patient outcomes at 1 year and at last follow-up
AHR ACR NO REJ
(n 16) Early (n 25) Late (n 18) (n 227)
1-yr graft survival 81% 83% 85% 94%a
Last creatinine, median (IQR) 1.6 (1.2–2.2) 1.4 (1.1–1.8) 1.9b,c (1.6–2.8) 1.4 (1.1–1.7)
ESRD at last follow-up, % (n) 19% (3) 20% (5) 44% (8) 5% (12)
Death with functioning graft, % (n) 6% (1) 0% (0) 0% (0) 2% (4)
Follow-up in days, mean SEM 457 76 641 75 633 69 564 21
a
P 0.001 vs. all groups.
b
P 0.02 vs. early ACR.
c
P 0.009 vs. NO REJ.
ESRD, end-stage renal disease; SEM, standard error of mean; IQR, interquartile range.

5. 1494 TRANSPLANTATION Vol. 75, No. 9
TABLE 3. C4d staining and flow crossmatch in patients with AHR
Pre-TX alloantibody New post-TX
Neutrophils alloantibody
Patients C4d PTC
in PTC Class I/II
Class I/II (%) DSA (%)
a
1 Yes Yes 0/0 N/A Not tested
2 Yes No 99/39 Yesb N/A
3 Yes Yes 84/59 Yesb N/A
4 Yes No 0/0 N/A Not tested
5 Yes Yes 97/98 Yesb N/A
6 Yes No 0/0 N/A 0/22d
7 Yes Yes 0/0 N/A Not tested
8 Yes Yes 10/0 No 0/44c
9 Yes Equivocal 97/0 Yesb N/A
10 Yes Yes 99/98 Yesb N/A
11 Yes Yes 0/81 Yesc N/A
12 Yes Yes 93/84 Yesb N/A
13 Yes Equivocal 0/0 N/A 0/21c
14 Yes No 0/0 N/A Not tested
15 Yes No 31/0 No Not tested
16 Yes Yes 0/0 N/A Not tested
a
Historic PRA B 35% (CDC/AHG).
b
DSA MHC class I.
c
DSA MHC class II.
d
This antibody is not donor-specific.
PTC, peritubular capillary; TX, transplant; DSA, donor-specific alloantibody; N/A, not applicable; MHC, major histocompatibility complex;
CDC, complement-dependent cytotoxicity; AHG, antihuman globulin.
from 3% to 10%, and that approximately 20% to 30% of acute successfully reversed AHR in seven of seven renal transplant
rejections have a humoral component (19). recipients using a regimen of 2 g/kg IVIG; notably, five of
The current study reveals a potential association between seven patients were also given antilymphocyte therapy. Al-
DGF and AHR, with a significantly higher proportion of lografts remained functional in all four patients for whom
patients in the AHR group (compared with the ACR and NO long-term follow-up ( 1 year) was available (16). In a recent
REJ groups) requiring renal replacement therapy within the report, Montgomery et al. showed that a combined regimen of
first week posttransplant. In the study by Crespo et al., the PP followed by IVIG was effective in reversing AHR in three
percentage of patients with DGF was also significantly patients; this regimen was also successful in desensitizing
higher in the AHR group when compared with patients with four patients before living-donor kidney transplantation (17).
mild ACR but not when compared with patients with severe The precise mechanism of action of IVIG is still unknown.
ACR (11). The present study does not offer a definite expla- Proposed mechanisms include neutralization by anti-idio-
nation for the higher incidence of DGF in patients with AHR. typic antibodies, altered complement or Fc-receptor binding,
However, one can speculate that, in such cases, DGF actually and suppression of de novo immunoglobulin synthesis (16,
reflects ongoing immunologic injury caused by low levels of 23). When given after PP, IVIG has the added benefit of
DSA present at the time of transplant. An alternative hy- replenishing gamma globulins and, thereby, decreasing the
pothesis is that DGF, presumably caused by ischemic injury, risk of infections.
triggers humoral immunity. It has been documented that Unlike prior reports, the majority of our patients with AHR
ischemic injury may increase the immunogenicity of the graft (12/16) did not receive antilymphocyte therapy. Therefore,
by inducing the class II transcriptional activator and, conse- their excellent outcome attests to the efficacy of PP and IVIG
quently, MHC class II expression (20). in the treatment of AHR. However, because all AHR patients
In the pre-cyclosporine era, attempts to treat AHR by re- received methylprednisolone boluses, we cannot rule out a
moving DSA with PP were largely unsuccessful (13–15). Al- potential benefit from steroids. Because many cases of AHR
though PP is effective in removing DSA from the circulation, are mixed (9/16 in our experience), steroids may indeed be
it does not suppress antibody synthesis, and a rebound in needed to treat the cellular component. Moreover, steroids
circulating DSA after PP has been documented (21). More may also down-regulate humoral responses by interfering
recently, Pascual et al. reported success in treating AHR by with T-cell help.
combining PP with tacrolimus-mycophenolate mofetil rescue We used renal allograft biopsy reports to classify patients
(9). This regimen reversed AHR in 14 of 15 patients (11). The into the three groups; therefore, a potential criticism of our
authors attribute the success of this regimen to the effective study is that the diagnosis of AHR rested mainly on histol-
suppression of antibody resynthesis by tacrolimus and my- ogy. Mauiyyedi et al. suggest that use of histology alone may
cophenolate mofetil, but because all patients also received lead to overdiagnosis of AHR (24). However, the incidence of
low-dose IVIG (0.4 g/kg) and antilymphocyte therapy, it is AHR in our study was similar to that reported by Mauiyyedi
difficult to sort out the relative contributions of each agent. using the complement split-product C4d as the diagnostic
Polyclonal IVIG preparations are widely used in the treat- criterion (9, 24). Although several studies indicate an impor-
ment of antibody-mediated diseases (22, 23). Jordan et al. tant role for C4d in the diagnosis of AHR (11, 25–29), its

6. May 15, 2003 ROCHA ET AL. 1495
specificity has been recently questioned (30). We used a com- response. II. Clinical and pathologic features of renal transplants with
bination of histology, C4d, and flow data to further substan- anti-class I-like antibody. Transplantation 1992; 53: 550.
8. Lobo PI, Spencer CE, Stevenson WC, et al. Evidence demonstrating poor
tiate the diagnosis of AHR in 12 of the initial 16 patients. kidney graft survival when acute rejections are associated with IgG
One-year graft survival for this subgroup was 75%, not sig- donor-specific lymphocytotoxin. Transplantation 1995; 59: 357.
nificantly different than that observed for the larger AHR 9. Pascual M, Saidman S, Tolkoff-Rubin N, et al. Plasma exchange and
group. For the four patients in which the diagnosis rested tacrolimus-mycophenolate rescue for acute humoral rejection in kidney
transplantation [see comments] [erratum appears in Transplantation
solely on histology, we unfortunately did not have posttrans-
1999 Feb 15; 67(3): 495]. Transplantation 1998; 66: 1460.
plant sera in three and therefore cannot exclude the possi- 10. Bohmig GA, Regele H, Saemann MD, et al. Role of humoral immune
bility that they developed de novo DSA. In the fourth patient, reactions as target for antirejection therapy in recipients of a spousal-
a new class II antibody was found posttransplant, but it was donor kidney graft. Am J Kidney Dis 2000; 35: 667.
not directed at donor targets. We hypothesize that antibodies 11. Crespo M, Pascual M, Tolkoff-Rubin N, et al. Acute humoral rejection in
renal allograft recipients: I. Incidence, serology and clinical character-
against non-HLAs may have been responsible for this case.
istics. Transplantation 2001; 71: 652.
12. Racusen LC, Solez K, Colvin RB, et al. The Banff 97 working classification
CONCLUSION of renal allograft pathology. Kidney Int 1999; 55: 713.
This retrospective analysis shows that the early diagnosis 13. Power D, Nicholls A, Muirhead N, et al. Plasma exchange in acute renal
and treatment of AHR of renal allografts with a combined allograft rejection: is a controlled trial really necessary? Transplanta-
tion 1981; 32: 162.
regimen of PP and IVIG is associated with excellent results. 14. Allen NH, Dyer P, Geoghegan T, et al. Plasma exchange in acute renal
Prospective trials comparing this with other regimens are allograft rejection. A controlled trial. Transplantation 1983; 35: 425.
needed. Given the low incidence of humoral rejection, a mul- 15. Soulillou JP, Guyot C, Guimbretiere J, et al. Plasma exchange in early
ticenter approach will likely be required to recruit the nec- kidney graft rejection associated with anti-donor antibodies. Nephron
1983; 35: 158.
essary number of patients for such studies. In the meantime,
16. Jordan SC, Quartel AW, Czer LS, et al. Posttransplant therapy using
our recommendation is to perform early biopsy (3–5 days high-dose human immunoglobulin (intravenous gammaglobulin) to con-
posttransplant) in patients at risk for humoral rejection (fe- trol acute humoral rejection in renal and cardiac allograft recipients
male, African American, and high PRA) who experience and potential mechanism of action. Transplantation 1998; 66: 800.
DGF. If the pathologist suspects a humoral component, a 17. Montgomery RA, Zachary AA, Racusen LC, et al. Plasmapheresis and
intravenous immune globulin provides effective rescue therapy for re-
repeat crossmatch should be performed and therapy with PP
fractory humoral rejection and allows kidneys to be successfully trans-
and IVIG initiated without delay. Our data indicate that the planted into cross-match-positive recipients. Transplantation 2000; 70:
use of antilymphocyte agents may be reserved for those cases 887.
with a concomitant moderate to severe cellular rejection. PP 18. Pei R, Wang G, Tarsitani C, et al. Simultaneous HLA class I and class II
should be performed until there is clinical improvement in antibodies screening with flow cytometry. Hum Immunol 1998; 59: 313.
19. Watschinger B, Pascual M. Capillary c4d deposition as a marker of hu-
graft function; in our experience, this is usually achieved moral immunity in renal allograft rejection. J Am Soc Nephrol 2002; 13:
after three to six sessions. The optimal dose of IVIG remains 2420.
unclear. Although the average dose used in our series was 2 20. Sims TN, Goes NB, Ramassar V, et al. In vivo class II transactivator
g/kg, there are data supporting the use of low-dose IVIG (0.4 expression in mice is induced by a non-interferon-gamma mechanism in
response to local injury. Transplantation 1997; 64: 1657.
g/kg) in this setting (9).
21. Vangelista A, Frasca GM, Nanni CA, et al. Value of plasma exchange in
renal transplant rejection induced by specific anti-HLA antibodies.
Acknowledgments. The authors thank Dr. Mary Foster for her
Trans Am Soc Artif Intern Organs 1982; 28: 599.
suggestions and critical review of this manuscript and Mr. Jeffrey 22. Dwyer JM. Manipulating the immune system with immune globulin [re-
Crews for performance of C4d staining. view]. New Engl J Med 1992; 326: 107.
23. Kazatchkine MD, Kaveri SV. Immunomodulation of autoimmune and
REFERENCES inflammatory diseases with intravenous immune globulin. N Engl
1. Krensky AM. Immune response to allografts. In: Norman J, Turka LA, eds. J Med 2001; 345: 747.
Primer on transplantation. Mt. Laurel, NJ: American Society of Trans- 24. Mauiyyedi S, Crespo M, Collins AB, et al. Acute humoral rejection in
plantation 2001, p 16. kidney transplantation: II. Morphology, immunopathology, and patho-
2. Gaston RS. Maintenance immunosuppression in the renal transplant re- logic classification. J Am Soc Nephrol 2002; 13: 779.
cipient: an overview [review]. Am J Kidney Dis 2001; 38(6 Suppl 6): 25. Feucht HE, Felber E, Gokel MJ, et al. Vascular deposition of complement-
S25-S35. split products in kidney allografts with cell-mediated rejection. Clin Exp
3. Gailiunas P, Jr., Suthanthiran M, Busch GJ, et al. Role of humoral pre- Immunol 1991; 86: 464.
sensitization in human renal transplant rejection. Kidney Int 1980; 17: 26. Feucht HE, Schneeberger H, Hillebrand G, et al. Capillary deposition of
638. C4d complement fragment and early renal graft loss. Kidney Int 1993;
4. Lordon RE, Wilson RL, Shield CF III, et al. Humoral immunity in renal 43: 1333.
transplantation detection of cytotoxic antibodies to lymphocyte anti- 27. Collins AB, Schneeberger EE, Pascual MA, et al. Complement activation
gens other than HLA-A, B, C and Ia-like antigens in patients with in acute humoral renal allograft rejection: diagnostic significance of C4d
vascular graft rejection. Transplantation 1981; 32: 286. deposits in peritubular capillaries. J Am Soc Nephrol 1999; 10: 2208.
5. Majoor GD, van de Gaar MJ, Vlek LF, et al. The role of antibody in rat 28. Nickeleit V, Zeiler M, Gudat F, et al. Detection of the complement degra-
renal allograft rejection. I. Absence of antibody to erythrocyte-associ- dation product C4d in renal allografts: diagnostic and therapeutic im-
ated antigens is associated with enhanced renal allograft survival. plications. J Am Soc Nephrol 2002; 13: 242.
Transplantation 1981; 31: 369. 29. Bohmig GA, Exner M, Habicht A, et al. Capillary C4d deposition in kidney
6. Halloran PF, Wadgymar A, Ritchie S, et al. The significance of the anti- allografts: a specific marker of alloantibody-dependent graft injury.
class I antibody response. I. Clinical and pathologic features of anti- J Am Soc Nephrol 2002; 13: 1091.
class I-mediated rejection. Transplantation 1990; 49: 85. 30. Platt JL. C4d and the fate of organ allografts. J Am Soc Nephrol 2002; 13:
7. Halloran PF, Schlaut J, Solez K, et al. The significance of the anti-class I 2417.