Human leukocyte antigen (HLA) antibodies are associated with antibody-mediated rejection, detected by histology and C4d staining, transplant glomerulopathy, and graft failure (1, 2). Development of de novo donor-specific antibodies (DSA) after transplant portends a poor prognosis (3, 4). In some patients, however, the presence of circulating antibodies does not result in graft dysfunction (5, 6).
For many years, cell-based assays were the standard approach for detection of HLA antibodies, namely complement-dependent cytotoxicity and flow cytometry crossmatching. The recent advent of more sensitive and specific solid phase tests, such as HLA single antigen bead (SAB) assays, allows for detection of HLA antibodies specific to individual donor antigens (7, 8). Yet, it is still not possible to distinguish which HLA antibodies are detrimental, especially antibodies at low but detectable levels that bind but do not activate complement.
Considering the important role of complement fixation as detected by C4d deposition in defining antibody-mediated rejection (9), one way to distinguish which HLA antibodies are harmful is to determine which HLA antibodies can activate the classical complement pathway. This has been done recently using a C4d two-step screening and identification approach (9). However, this assay has modest sensitivity and did not correlate with antibody-mediated rejection and C4d deposition. We have developed a one-step C1q-SAB assay that identifies complement-fixing HLA antibodies with high sensitivity and specificity. C1q is the first step in the classical complement cascade activated by antibody and precedes C4d deposition.
In this study, we sought to determine whether the development of posttransplant de novo complement-fixing DSA as detected by the C1q assay could identify clinically relevant DSA. We hypothesized that C1q-positive DSA would be associated with C4d deposition and with transplant glomerulopathy and, ultimately, graft loss.
Patient characteristics and clinical data of the 31 patients are detailed in Table 1. Although there were more women in the C1q-positive group, the difference was not statistically significant. There were no significant differences in calculated panel reactive antibody levels, HLA mismatching, living versus deceased donor transplants, or previous transplants. Immunosuppression was not significantly different between the two groups in the eras in which they were transplanted. During this period, most patients received antithymocyte globulin, basilixumab, or daclizumab for induction therapy and a calcineurin inhibitor (tacrolimus or cyclosporine), mycophenolate mofetil, and corticosteroid maintenance or withdrawal. Two patients received an investigative immunosuppression medication but converted to a tacrolimus-based regimen 1-year after transplant.
Patients who had C1q-fixing DSA experienced more acute rejection (83% vs. 42%, P=0.03), C4d deposition (67% vs. 26%, P=0.06), transplant glomerulopathy (50% vs. 0%, P=0.001), and graft loss (67% vs. 16%, P=0.007). Serum creatinine concentrations up to 1-year posttransplant and 1- and 5-year graft survival were not significantly different among patients with and without C1q-fixing DSA although 5-year graft survival was 9% lower in the C1q-positive group.
Association of DSA by IgG and C1q With C4d Deposition
To assess the association between the presence of DSA and C4d deposition in the allograft, we tested sera pretransplant and at the time of biopsy from all 31 patients for DSA by IgG and C1q (Table 2). Twelve of 13 C4d-positive biopsies (92%) were associated with the presence of DSA by IgG, giving a strong correlation between the presence of IgG-positive DSA and the presence of C4d deposition (P=0.0007) with a high level of sensitivity (92%, 95% confidence interval [CI] 0.62–1). Five of 18 C4d-negative biopsies (28%) also had DSA detected by IgG, giving a specificity of 72% (95% CI 0.46–0.89).
Similar analysis of C1q-positive DSA and C4d deposition was performed. Eight of 13 C4d-positive biopsies (62%) were associated with the presence of DSA by C1q, yielding a sensitivity of 62% (95% CI 0.32–0.85). Four of 18 C4d-negative biopsies (22%) had DSA detected by C1q, yielding a specificity of 78% (95% CI 0.52–0.93). Six samples were positive for DSA by both IgG and C1q. DSA that was positive by C1q always tested positive by IgG but not vice versa.
Association of De Novo IgG-Positive and C1q-Positive DSA With C4d Deposition, Transplant Glomerulopathy, and Graft Loss
To assess the association of the development of de novo DSA present at the time of biopsy with C4d deposition, transplant glomerulopathy, and graft loss, we identified recipients without DSA pretransplant who developed de novo DSA after transplantation. From the 31 original patients in the study, we eliminated four patients who had DSA detected before transplantation (two patients with DSA by IgG only and two patients with DSA by IgG and C1q).
We identified 13 patients (eight from the C4d-positive group and five from the C4d-negative group) who developed de novo DSA detected by IgG only (three patients) or by IgG and C1q (10 patients) and compared them with the 14 patients who did not have DSA pre- or posttransplant (Tables 3 and 4). We found no significant association between the presence or the mean fluorescence intensity (MFI) strength (Table 5) of de novo DSA by IgG or C1q and the presence of C4d deposition, transplant glomerulopathy, or graft loss. However, the absence of de novo DSA by IgG and C1q had a high negative predictive value for the absence of C4d deposition (IgG: 100%, 95% CI 0.73–1; C1q: 88%, 95% CI 0.62–0.98), transplant glomerulopathy (IgG: 100%, 95% CI 0.73–1; C1q: 100%, 95% CI 0.77–1), and graft failure (IgG: 86%, 95% CI 0.56–0.97; C1q: 88%, 95% CI 0.62–0.98). Testing for DSA by IgG was more sensitive than C1q for C4d deposition (IgG: 100%, 95% CI 0.60–1; C1q: 75%, 95% CI 0.36–0.96). C1q testing was more specific for transplant glomerulopathy (C1q: 81%, 95% CI 0.57–0.94; IgG: 67%, 95% CI 0.43–0.85) and graft loss (C1q: 79%, 95% CI 0.54–0.93; IgG: 63%, 95% CI 0.39–0.83). Patients developed an average of three de novo DSA (range, 1–6; class I, 3; class II, 4; classes I and II, 6). All patients with transplant glomerulopathy and most patients with graft loss (89%) developed HLA class II de novo DSA.
Finally, when comparing C1q-positive versus C1q-negative groups for death-censored graft survival (Fig. 1), we found lower nominal graft survival rates for C1q-positive patients, but the difference was not statistically significant (P=0.20). Five of the 12 patients in the C1q-positive group (42%) lost their grafts beyond 5 years posttransplant, whereas the three patients who had DSA by IgG but not C1q did not lose their grafts within the 2.5 to 5 posttransplant years included in the study period. There were no deaths with functioning grafts in either group during the study period.
This pilot study demonstrates for the first time in adult kidney transplant recipients that the C1q SAB assay detects clinically relevant antibodies. Detection of C1q-positive DSA is specific for outcomes associated with late graft failure, such as transplant glomerulopathy and graft loss. Our results also show that absence of de novo DSA by IgG and C1q has a high negative predictive value for absence of C4d deposition, transplant glomerulopathy, and graft failure. Other investigators have reported the use of complement-fixing assays to predict graft outcomes (9–16). Wahrmann et al. (13) and Smith et al. (15) showed that C4d-fixing HLA antibodies correlated with the development of antibody-mediated rejection and decreased graft survival in kidney and heart transplant patients, respectively. However, the C4d-fixing SAB assay showed discordance between serum complement-fixing antibodies and C4d deposition, findings consistent with our data (9). There are important differences to note between the C1q and C4d assays. First, whereas the C1q assay uses purified human C1q at the same concentration for each sample, the C4d assay uses sera from healthy volunteers as a complement source, thereby introducing variability (9–16). Furthermore, individual sera can inhibit HLA antibodies similar to that seen with pooled intravenous immunoglobulin (our unpublished observations), and reported MFI values in the C4d assay were reduced compared with normal binding by 15% to 50% (15). Second, the C1q-SAB assay detects complement-fixing antibodies at low levels due to its high sensitivity, and MFIs up to 30,000 are routinely seen. The C4d assay uses a first step on relatively insensitive “screening beads” before reflexing positives to the SABs, which could lead to missing complement-fixing antibodies. MFIs in the C4d assay seem to be in the range of 500 to 3500 according to published data (15). Therefore, we believe that as compared with the C4d assay, the C1q assay has higher specificity and sensitivity and is able to detect low-level complement-fixing antibodies missed by both the complement-dependent cytotoxicity and C4d assays. In addition, the C1q assay can detect IgM antibodies that can fix complement. Studies have suggested associations of IgM HLA antibodies with transplant rejection (17, 18).
Our data show that C4d deposition may be evident in the absence of circulating C1q-fixing DSA, and, conversely, some patients who have C1q-fixing DSA have C4d-negative biopsies. The C4d assay also yielded results with C4d-positive and IgG-negative reactions (13, 16). There are several explanations for the differences between in vitro serum results as detected by C1q and in vivo complement fixation as detected by C4d staining. First, the kidney allograft has the ability to absorb antibodies, thereby leaving no detectable circulating antibody (19, 20). Second, C4d deposition could reflect other non-HLA antibody-mediated pathways that activate complement and lead to deposition of C4d such as the lectin pathway (21–23). Third, as in the case of ABO-incompatible transplantation, C4d may be a marker of accommodation, or a protective response of the graft (5, 24–26). Fourth, treatment for rejection may have effectively suppressed or inhibited C1q-positive antibodies while the sequelae of complement fixation—namely C4d deposition—remains.
In the cases of C1q-positive DSA, negative C4d staining, the variability in the immunoperoxidase technique may lead to lower sensitivity and lack of detection (27). Although immunoperoxidase methods are less sensitive than immunofluorescence for detecting C4d (28), the concordance rate between the two methods in our hands is 97% with 87.5% sensitivity by immunohistochemistry (29). We assessed biopsies for C4d using immunohistochemistry because the ease of using paraffin tissue outweighs the slightly greater sensitivity of immunofluorescence on frozen tissue. Therefore, the decreased sensitivity of the immunoperoxidase method may have altered the results minimally. Alternatively, low levels of circulating antibody may indicate an earlier stage of disease preceding overt graft injury (30). As another possibility, perhaps some C1q-positive antibodies cause injury by mechanisms unrelated to complement activation such as immunoglobulin clustering and antibody-dependent cell-mediated cytotoxicity through natural killer cell-mediated mechanisms (31, 32).
Our findings show that the presence of C1q-fixing antibodies has a high level of specificity with transplant glomerulopathy, which is often associated with late graft failure. All patients who developed transplant glomerulopathy tested positive for C1q-fixing antibodies. Consistent with other studies, there was a predominance of DSA against HLA class II antigens (33–35). This, in itself, is an interesting finding because many HLA class II antibodies (more than class I), which are strongly positive by IgG are negative by C1q (our unpublished observations). Recent studies of the mechanisms that lead to late graft failure have focused on ongoing immunologic injury. Interestingly, previous reports show that circulating de novo antibody formed after transplant may precede graft loss by many years (3). Our data support the mechanism that a major cause of late kidney transplant failure is antibody-mediated microcirculation injury (35, 36). Moreover, the histologic diagnosis of transplant glomerulopathy, which remains poorly understood, is often seen in patients with previous episodes of antibody-mediated rejection (34). Thus, the presence of C1q-fixing DSA may be a marker for ongoing immune activation and complement-mediated graft injury that lead to eventual graft failure. It is also possible that it is the C1q-negative DSA that is involved in chronic injury, whereas the C1q-positive DSA is involved in earlier and more acute adverse reactions. The availability of the C1q assay gives us a tool for addressing this question.
The main goals of this pilot study were to evaluate the utility of the C1q assay in predicting C4d deposition in kidney biopsies and in identifying patients at risk of future graft failure. Despite the small sample size, patients studied were relatively diverse in terms of age, sex, and race. All biopsies and assays were analyzed blinded to the clinical outcomes. All patients received relatively uniform immunosuppression and received their long-term follow-up at our institution. On the other hand, we lacked the power to determine small to moderate-sized differences in graft survival by C1q antibody status, in part because of the small sample size and in part because of overall excellent graft survival at our institution. In summary, this study in adult kidney transplant recipients describes for the first time the utility of the C1q assay, a novel and sensitive method for detecting complement-fixing DSA. The C1q assay is specific for long-term adverse clinical sequelae, such as transplant glomerulopathy and graft loss. Larger, prospective studies are required to define the role of C1q antibody testing after kidney transplantation and to elucidate the mechanism by which it might lead to poorer graft outcome.
MATERIALS AND METHODS
We conducted a retrospective single center study of 274 adult kidney transplant recipients at Stanford University Medical Center transplanted between 1993 and 2009 who had at least one biopsy performed for clinical cause or by protocol between 1995 and 2009. Only patients with stored serum samples (n=341) pretransplant and within 7 days from the time of biopsy were included. We reviewed and categorized all biopsies (n=559) for the presence or absence of C4d staining by immunohistochemistry using formalin-fixed, paraffin-embedded tissue in all but one instance. Positive cases showed peritubular capillary endothelial staining and, for these, we estimated the percentage of capillaries stained. Glomerular endothelial cell staining was noted but was not sufficient for a case to be scored as “positive.” Of the 36 patients identified with C4d-positive staining, 20 did not have serum samples both pretransplant and at the time of biopsy and three were younger than 18 years, leaving 13 patients for analysis. If patients had multiple biopsies, the first biopsy with serum available at the time of biopsy was used for analysis. From the 238 recipients with C4d-negative staining, we used a random number generator to select 40 patients with C4d-negative staining. Twenty-two patients did not have serum samples available, leaving 18 patients for comparative analysis. We obtained clinical and laboratory data of donors and recipients from electronic databases and medical records. The Institutional Review Board at Stanford University approved the study.
Assessment of HLA Antibodies Using the (Standard) IgG SAB Assay
Antibody screening was performed blinded to biopsy and clinical data. We analyzed 31 recipients' sera both pretransplant and at the time of biopsy (n=62) to identify HLA classes I and II IgG(positive) DSA using a Luminex platform (LABScan 100) and commercially available SAB kits (LABScreen, One Lambda, Inc., Canoga Park, CA) according to the manufacturer's instructions. This assay uses pooled luminescent beads, each uniquely distinguishable and coated with a different purified, single HLA classes I or II antigen. Collectively, there are beads representing 97 class I (A, B, C) and 97 class II (DR, DQ, DP) antigens/alleles. Normalized MFI values from Fusion software (One Lambda, Inc.) were used to assign positive (“true”; MFI >1000) and “possible” (MFI=500–999) antibodies. This assay detects all IgG-binding antibodies irrespective of their complement-fixing ability.
Assessment of HLA Antibodies Using the C1q SAB Assay
The C1q assay was performed in parallel with the IgG SAB method. Serum (5 μL) spiked with 150 μg/mL PE-labeled purified human C1q (hC1q; Sigma, St. Louis, MO) to introduce uniformity and adequate C1q concentrations into all specimens was incubated with 2.5 μL of LABScreen classes I or II SAB and 5 μL of PE-conjugated hC1q (custom-labeled) for 40 min at room temperature with gentle shaking, washed twice, resuspended in 60 μL wash buffer and analyzed on the Luminex. MFI values were analyzed as for the IgG SAB assay with the exception that standard cutoff values were not used. Negative background values range from 0 to 50 MFI with most less than 10 MFI. Antibodies were assigned as “possible” when the first increase more than 33% (but at least ∼300 MFI) over the prior lower MFI bead was observed. Antibodies showing more than 50% increment over the last “possible” value or a log increase in MFI were assigned as “true” antibodies. Antibodies assigned as “possible” have values 300 to 1000 MFI depending on the background values, and positives frequently give a 100% to a log difference in MFI over the highest “possible.” This technique detects the subset of IgG antibodies that can bind human C1q without the need for further functional complement activation. The assay also detects IgM antibodies, which are IgG(−)/C1q(positive) in parallel testing. This assay routinely detects MFI values up to approximately 30,000.
All biopsies performed by protocol or for evaluation of allograft dysfunction were formalin-fixed, paraffin-embedded specimens stained with hematoxylin-eosin and periodic acid-Schiff. A single pathologist blinded to the results of the serum studies and clinical outcomes read the 31 biopsies using lymphocytic glomerulitis with capillary wall double contouring (34) to define transplant glomerulopathy. We excluded biopsies with double contouring of the glomerular basement membrane due to recurrent disease. C4d staining was performed using the immunoperoxidase technique in all cases but one, which was performed by immunofluorescence on frozen tissue. Biopsies were considered C4d positive if there was linear endothelial C4d staining in at least 20% of cortical peritubular capillaries. Biopsies were scored according to the Banff '97 classification.
The Fisher's exact test with two-tailed P values was used for comparisons between groups. Sensitivity, specificity, positive predictive value, negative predictive value, and the respective 95% CI were calculated according to standard formulas. The Kaplan-Meier product limit method was used to calculate graft survival. Graph Pad and SAS software were used to assist with randomization and analysis.
The authors thank Ling Ling Lin and Calvin Lou for excellent technical and database assistance, and Glenn Chertow and Basit Javaid for review and critique of the manuscript.
1. Patel R, Terasaki PI. Significance of the positive crossmatch test in kidney transplantation. N Engl J Med
1969; 280: 735.
2. Worthington JE, McEwen A, McWilliam LJ, et al. Association between C4d
staining in renal transplant biopsies, production of donor-specific HLA antibodies, and graft outcome. Transplantation
2007; 83: 398.
3. Terasaki P, Lachmann N, Cai J. Summary of the effect of de novo HLA antibodies on chronic kidney graft failure. Clin Transpl
4. Li X, Ishida H, Yamaguchi Y, et al. Poor graft outcome in recipients with de novo donor-specific anti-HLA antibodies after living related kidney transplantation. Transpl Int
2008; 21: 1145.
5. Salama AD, Delikouras A, Pusey CD, et al. Transplant accommodation in highly sensitized patients: A potential role for Bcl-xL and alloantibody. Am J Transplant
2001; 1: 260.
6. Zhang X, Reed EF. Effect of antibodies on endothelium. Am J Transplant
2009; 9: 2459.
7. Pei R, Lee J, Chen T, et al. Flow cytometric detection of HLA antibodies using a spectrum of microbeads. Hum Immunol
1999; 60: 1293.
8. Gibney EM, Cagle LR, Freed B, et al. Detection of donor-specific antibodies using HLA-coated microspheres: Another tool for kidney transplant risk stratification. Nephrol Dial Transplant
2006; 21: 2625.
9. Honger GWM, Amico P, Hopfer H, et al. C4d
-fixing capability of low-level donor-specific HLA antibodies is not predictive for early antibody-mediated rejection. Transplantation
2010; 89: 1471.
10. Wahrmann M, Exner M, Regele H, et al. Flow cytometry based detection of HLA alloantibody mediated classical complement activation. J Immunol Methods
2003; 275: 149.
11. Bartel G, Wahrmann M, Exner M, et al. In vitro detection of C4d
-fixing HLA alloantibodies: Associations with capillary C4d
deposition in kidney allografts. Am J Transplant
2008; 8: 41.
12. Bartel G, Wahrmann M, Exner M, et al. Determinants of the complement-fixing ability of recipient presensitization against HLA antigens. Transplantation
2007; 83: 727.
13. Wahrmann M, Bartel G, Exner M, et al. Clinical relevance of preformed C4d
-fixing and non-C4d
-fixing HLA single antigen reactivity in renal allograft recipients. Transpl Int
2009; 22: 982.
14. Wahrmann M, Exner M, Haidbauer B, et al. [C4d
]FlowPRA screening—A specific assay for selective detection of complement-activating anti-HLA alloantibodies. Hum Immunol
2005; 66: 526.
15. Smith JD, Hamour IM, Banner NR, et al. C4d
fixing, luminex binding antibodies—A new tool for prediction of graft failure after heart transplantation. Am J Transplant
2007; 7: 2809.
16. Wahrmann M, Exner M, Schillinger M, et al. Pivotal role of complement-fixing HLA alloantibodies in presensitized kidney allograft recipients. Am J Transplant
2006; 6(5 pt 1): 1033.
17. Stastny P, Ring S, Lu C, et al. Role of immunoglobulin (Ig)-G and IgM antibodies against donor human leukocyte antigens in organ transplant recipients. Hum Immunol
2009; 70: 600.
18. Mizutani K, Terasaki P, Rosen A, et al. Serial ten-year follow-up of HLA and MICA antibody production prior to kidney graft failure. Am J Transplant
2005; 5: 2265.
19. Martin L, Guignier F, Bocrie O, et al. Detection of anti-HLA antibodies with flow cytometry in needle core biopsies of renal transplants recipients with chronic allograft nephropathy. Transplantation
2005; 79: 1459.
20. Martin L, Guignier F, Mousson C, et al. Detection of donor-specific anti-HLA antibodies with flow cytometry in eluates and sera from renal transplant recipients with chronic allograft nephropathy. Transplantation
2003; 76: 395.
21. Sun Q, Liu Z, Yin G, et al. Detectable circulating antiendothelial cell antibodies in renal allograft recipients with C4d
-positive acute rejection: A report of three cases. Transplantation
2005; 79: 1759.
22. Sumitran-Holgersson S. Relevance of MICA and other non-HLA antibodies in clinical transplantation. Curr Opin Immunol
2008; 20: 607.
23. Tinckam KJ, Chandraker A. Mechanisms and role of HLA and non-HLA alloantibodies. Clin J Am Soc Nephrol
2006; 1: 404.
24. Lynch RJ, Platt JL. Accommodation in renal transplantation: Unanswered questions. Curr Opin Organ Transplant
2010; 15: 481.
25. Aikawa A, Ohara T, Arai K, et al. Clinical outcome and accommodation in ABO incompatible kidney transplantation. Clin Transpl
26. Narayanan K, Jendrisak MD, Phelan DL, et al. HLA class I antibody mediated accommodation of endothelial cells via the activation of PI3K/cAMP dependent PKA pathway. Transpl Immunol
2006; 15: 187.
27. Ludovico-Martins H, Silva C, Teodoro WR, et al. Analysis of different staining techniques for c4d
detection in renal allograft biopsies. Transplant Proc
2009; 41: 862.
28. Seemayer CA, Gaspert A, Nickeleit V, et al. C4d
staining of renal allograft biopsies: A comparative analysis of different staining techniques. Nephrol Dial Transplant
2007; 22: 568.
29. Troxell ML, Weintraub LA, Higgins JP, et al. Comparison of C4d
immunostaining methods in renal allograft biopsies. Clin J Am Soc Nephrol
2006; 1: 583.
30. Cai J, Terasaki PI. Humoral theory of transplantation: Mechanism, prevention, and treatment. Hum Immunol
2005; 66: 334.
31. Hidalgo LG, Sis B, Sellares J, et al. NK cell transcripts and NK cells in kidney biopsies from patients with donor-specific antibodies: Evidence for NK cell involvement in antibody-mediated rejection. Am J Transplant
2010; 10: 1812.
32. Kirby JA, Givan AL, Shenton BK, et al. Renal allograft rejection. Possible involvement of antibody-dependent cell-mediated cytotoxicity. Transplantation
1990; 50: 225.
33. Gloor JM, Sethi S, Stegall MD, et al. Transplant glomerulopathy: Subclinical incidence and association with alloantibody. Am J Transplant
2007; 7: 2124.
34. Cosio FG, Gloor JM, Sethi S, et al. Transplant glomerulopathy. Am J Transplant
2008; 8: 492.
35. Hidalgo LG, Campbell PM, Sis B, et al. De novo donor-specific antibody at the time of kidney transplant biopsy associates with microvascular pathology and late graft failure. Am J Transplant
2009; 9: 2532.
36. Einecke G, Sis B, Reeve J, et al. Antibody-mediated microcirculation injury is the major cause of late kidney transplant failure. Am J Transplant
2009; 9: 2520.