Low-level donor-specific human leukocyte antigen (HLA) antibodies (HLA-DSA)—defined by positive single-antigen flow beads (SAFB) and negative complement-dependent cytotoxicity crossmatch (CDC-XM)—represent a risk factor for early antibody-mediated rejection (AMR) and decreased allograft survival (1, 2). However, in the past years, it became evident that not all low-level HLA-DSA are detrimental. Several studies reported that low-level HLA-DSA are associated with different clinical outcomes ranging from acute AMR with subsequent allograft loss to uneventful and rejection-free courses (1–4).
Because the clinical outcome of low-level HLA-DSA is not yet predictable, reliable parameters that allow distinction between harmful from presumably irrelevant HLA-DSA are required. Unfortunately, readily available parameters pretransplant such as HLA-DSA characteristics (i.e., number, class, and strength) and the route of sensitization are not sufficiently predictive in this regard (5, 6).
Complement activation triggered by HLA-DSA plays an important role in allograft rejection (7). Because the four IgG subclasses have different capabilities to activate the complement cascade, assessing the IgG subclass pattern in pretransplant sera might be useful to define the clinical relevance of HLA-DSA. So far, only few studies investigated the pretransplant IgG subclass distribution of HLA-DSA in transplant recipients, even fewer correlated the results with clinical outcomes. All these studies have important limitations, which preclude drawing firm conclusions. First, the analyzed HLA antibodies were not declared as being donor-specific (8, 9) or if so, they were detected by a less sensitive and specific method (10). Second, only a few patients with a limited rejection event rate were investigated (19 patients with three rejections and three patients with one rejection) (11, 12). Third, an adequate control group with patients having HLA-DSA but not experiencing AMR was missing (13).
The aims of this study were (i) to establish a reliable assay for detection of IgG subclasses of HLA-DSA; and (ii) to investigate whether pretransplant IgG subclasses of HLA-DSA are predictive for the occurrence and phenotype of AMR in a reasonably large and well-defined patient population (n=74).
The study population consisted of 74 patients with pretransplant HLA-DSA. Forty of 74 patients (54%) experienced AMR within the first 6 months posttransplant (AMR group), whereas 34 patients (46%) did not experience AMR (no AMR group). Patient characteristics of the AMR and no AMR group are detailed in Table 1. There were no differences regarding recipients and donor parameters, HLA-matching, and number, HLA class (including HLA class I and II specificities), and cumulative strength of HLA-DSA (P≥0.07). In addition, previous transplants and pregnancies were equally often observed in both groups.
The two groups were statistically not different regarding the rate of clinical and surveillance biopsies performed within the first 6 months posttransplant. Importantly, only three patients of the no AMR group had no allograft biopsy within the first 6 months posttransplant. Graft survival at 5 years was only 59% in the AMR group, but still 86% in the no AMR group (P=0.02). In addition, death-censored allograft survival at 5 years was lower in the AMR than in the no AMR group (69% vs. 90%; P=0.05).
Specificity of Reporter Antibodies
The IgG subclass SAFB assay was developed and optimized as detailed in the Materials and Methods section. Reporter antibodies for IgG1, IgG3, and IgG4 had a high specificity. However, the reporter antibody for IgG2 showed some cross-reactivity with IgG1,3,4 (approximately 13%–15%) (see Table, Supplemental Digital Content 1,http://links.lww.com/TP/A420).
Frequency of IgG Subclasses, Their Patterns, and Biological Groups
The investigated 74 patients had in total 141 HLA-DSA. IgG1 was the predominant subclass (111/141; 78%), followed by IgG2 (69/141; 49%), IgG3 (51/141; 36%), and IgG4 (28/141; 20%). There are 16 possible combinations of the different IgG subclasses. Thirteen of these 16 patterns were observed (Fig. 1). From a biological point of view, the subclass patterns could be divided into three groups according to their complement-activating capability: (i) strong complement-activating group (i.e., IgG1 and IgG3); (ii) weak/no complement- activating group (i.e., IgG2 and IgG4); and (iii) mixture group (i.e., IgG1 and IgG3; IgG2 and IgG4). The majority of HLA-DSA appeared in the mixture group (67/141; 48%), followed by the strong complement-activating group (48/141; 34%) and the weak/no complement-activating group (9/141; 6%) (Fig. 1). Seventeen of 141 HLA-DSA (12%), which were detectable by the standard SAFB assay using the generic IgGPAN reporter antibody, were negative in all IgG subclass analyses. These HLA-DSA revealed low mean fluorescence intensity (MFI) by the standard SAFB assay (median, 1524; range, 1083–3584). The three IgG subclass groups (i.e., strong complement-activating, weak/no complement-activating, and mixture) showed no statistically significant difference regarding the distribution of the targeted HLA loci (HLA-A- [11/1/11], HLA-B- [8/2/14], HLA-Cw- [2/0/3], HLA-DR- [16/4/17], HLA-DQ- [10/2/18], and HLA-DP-DSA [1/0/4]; P=0.91).
IgG Subclass Groups and Clinical Outcomes
As mentioned above, the investigated 74 patients had in total 141 HLA-DSA (33 patients had 1, 24 patients had 2, 11 patients had 3, three patients had 4, and three patients had 5 HLA-DSA). To assess the effect of the IgG subclasses of HLA-DSA on clinical outcomes, the 74 patients were classified according to the overall complement-activating capability of all their HLA-DSA: (i) strong complement-activating group (i.e., all HLA-DSA were IgG1 and/or IgG3; n=21 [28%]); (ii) weak/no complement-activating group G (i.e., all HLA-DSA were IgG2 and/or IgG4; n=4 [5%]); and (iii) mixture group (i.e., HLA-DSA contained IgG1 and/or IgG3 and IgG2 and/or IgG4; n=46 [62%]). Three patients had HLA-DSA that were negative for all IgG subclasses. They were excluded, leaving 71 patients for the following analyses.
The incidence of AMR within the first 6 months posttransplant was statistically not different between the three groups, but the incidence was numerically lower in the weak/no complement-activating group than in the other two groups (25% vs. 54% vs. 57%; P≥0.23) (Fig. 2A). The histologic phenotypes of AMR were statistically not different among the three groups (P≥0.22; Table 2). Remarkably, the only patient in the weak/no complement-activating group experiencing AMR demonstrated a subclinical C4d-negative phenotype. Time to AMR was not different among the strong complement-activating (median, 27 days; range, 5–180 days) and the mixture group (median, 28 days; range, 5–180 days) (P=0.97). Death-censored allograft survival at 5 years was 78% in both the strong complement-activating and the mixture group, whereas no allograft loss occurred yet in the weak/no complement-activating group (P≥0.40; Fig. 2B).
Next, we compared the amount and relative percentage of IgG subclasses between patients experiencing AMR (n=38) and patients not experiencing AMR (n=33). There was no difference between these two groups regarding the amount of IgG2 and IgG4 determined by MFI ratios (Table 3). In the AMR group, the amount of IgG1 was higher, whereas the amount of IgG3 was lower compared with the no AMR group (P=0.06 and P=0.02, respectively). Strong complement-activating subclasses (i.e., IgG1 and IgG3) accounted for 97.3% and 90.2% of the total amount of HLA-DSA in the AMR and no AMR group, respectively (P=0.24; Table 3).
Subgroup Analysis in Patients With Only One HLA-DSA
To investigate whether a single IgG subclass of HLA-DSA was predictive for AMR, we performed a subgroup analysis in the 33 patients with only 1 HLA-DSA. In these patients, IgG1 was the predominant subclass (32/33; 97%), followed by IgG3 (21/33; 64%), IgG2 (20/33; 61%), and IgG4 (6/33; 18%). The presence or absence of a single IgG subclass of HLA-DSA was neither predictive for the occurrence (P≥0.10) nor for the phenotype of AMR (P≥0.11; data not shown).
In this study, we evaluated the utility of pretransplant IgG subclass analysis of HLA-DSA to identify patients at risk for development of AMR or reduced allograft survival. The key observations were (i) most patients have pretransplant HLA-DSA composed of isolated strong complement activating (28%) or a mixture of strong and weak/no complement-activating antibodies (62%); and (ii) no difference among these two groups regarding clinical outcomes.
The first observation is in line with earlier analyses demonstrating that the complement-activating IgG1 is the predominant subclass of HLA-DSA, which is variably accompanied by other IgG subclasses (13–15). The diversity of the IgG subclass patterns provides evidence that the humoral immune response to foreign HLA antigen often involves the production of more than one IgG subclass. Because of the heterogeneous IgG subclass response of every individual, identifying a predictive subclass pattern might thus be difficult.
To the best of our knowledge, this is the first study investigating the association between pretransplant IgG subclasses of HLA-DSA and major clinical outcomes in a reasonably large and well-defined population. There are two main possibilities why the pretransplant IgG subclass pattern of HLA-DSA was not predictive for development of AMR. First, the ability to activate complement in vivo may not depend on the IgG subclass alone but is largely determined by several additional parameters (e.g., synergistic effect of several antibodies targeting different epitopes on the same HLA molecule, antibody-binding strength to the target HLA alloepitope, and complement-regulating and protective mechanisms on endothelial cells) (14–16). Second, it is conceivable that the IgG subclass pattern may change from pre- to posttransplant depending on the magnitude of the humoral memory response and the antigenicity of the target HLA alloepitope (6). Clearly, further studies are needed to investigate whether a specific change of IgG subclasses from pre- to posttransplant is predictive for development of AMR and allograft survival.
The developed IgG subclass assay deserves some critical comments. In general, the MFI values of the IgG subclass assays—especially IgG2,3,4—were significantly lower compared with those of the standard SAFB assay. Thus, instead of using an arbitrary MFI value as a cutoff for all assays, we determined a positive result for each IgG subclass and every individual bead by using four negative control sera. Seventeen of 141 HLA-DSA (12%) detected by the standard SAFB assay (median MFI, 1524; range, 1083–3584) were negative by all IgG subclass assays. This suggests a lower sensitivity of the IgG subclass assays compared with the standard SAFB test. This could be due to different binding properties of the reporter antibodies, which target different epitopes on the IgG molecule. In addition, it is important to note that the reporter antibody for IgG2 has some cross-reactivity with the other subclasses. Therefore, the frequency of the IgG2 subclass might be overestimated in sera with high amounts of IgG1. This could theoretically lead to some misclassifications of isolated strong complement-activating HLA-DSA into the mixture group. However, this potential confounder is likely clinically not relevant because the outcomes were similar in patients with isolated strong complement-activating HLA-DSA and patients in the mixture group.
Our study has two important advantages, which enforce the reliability of the results and conclusions. First, the IgG subclass assay was developed using well-characterized chimeric human/murine antibodies as positive controls. This allowed to select IgG subclass-specific reporter antibodies with high to very high specificity and to optimize the assay conditions (i.e., amount of IgG subclass-specific reporter antibodies giving the best signal-to-noise ratio). Second, the study population consists of two well-characterized clinicopathologic groups (AMR and no AMR group) that did not differ regarding major immunologic parameters and HLA-DSA characteristics.
A limitation of our study is the sample size, which is not large enough to explore the clinical impact of pretransplant isolated weak/no complement-activating HLA-DSA. Indeed, isolated weak/no complement-activating HLA-DSA seem to be less harmful, but this observation has to be interpreted with caution because it is derived from a few patients. Notably, only 4 of 74 patients (5%) had pretransplant isolated weak/no complement-activating HLA-DSA in our population. Even if isolated weak/no complement-activating HLA-DSA turn out to be of lower pathogenicity in a larger study, the overall predictive value of pretransplant IgG subclass analysis will likely not significantly increase because of the low prevalence of isolated weak/no complement-activating HLA-DSA.
In addition, quantification of HLA-DSA and IgG subclasses by means of cumulative MFI measured on SAFB has important limitations. First, most HLA-DSA are directed against epitopes shared by several HLA antigens and thus are distributed across several SAFB. Second, the amount and condition of HLA molecules may vary among the different SAFB. Clearly, the additional performance of flow cytometric crossmatches would have been helpful to better assess antibody amounts (17), but unfortunately, donor cells were not available in 80% of patients. However, we would like to point out that IgG subclass analyses of HLA-DSA are primarily reported as qualitative results (i.e., positive or negative), and we believe that these binary results are robust and reliable.
In conclusion, in 90% of patients pretransplant HLA-DSA are composed of isolated strong or a mixture of strong and weak/no complement-activating subclasses. Because the clinical outcomes in these two groups are similar, pretransplant IgG subclass analysis is likely not providing substantial value beyond the standard IgG SAFB assay for pretransplant risk stratification.
MATERIALS AND METHODS
The patients for this study are derived from a previously reported population, which consists of 104 patients who have been transplanted across low-level HLA-DSA (i.e., positive by SAFB, but negative by T-cell and B-cell CDC-XM) at the University Hospital Basel between 1999 and 2008 (5). Thirty of 104 patients (29%) were excluded from the current study: (i) patients with cumulative HLA-DSA MFI less than 2000, because initial experiments had demonstrated that HLA-DSA with less than 2000 MFI were often negative in the IgG subclass analysis (n=25); and (ii) patients without sufficient pretransplant serum for all analyses (n=5). The final study population consisted of 74 patients. Forty patients (54%) experienced AMR within the first 6 months posttransplant (AMR group), whereas 34 patients (46%) did not experience AMR (no AMR group). All retrospective analyses were performed with approval from the ethics committee of the University of Basel.
Diagnosis and Treatment of AMR
AMR was diagnosed by allograft biopsies in all cases. Clinically indicated allograft biopsies were performed when serum creatinine increased by more than 20%. Surveillance biopsies at 3 and 6 months posttransplant were routinely performed since 2001. AMR was defined as C4d positivity by immunofluorescence in peritubular capillaries alone or with transplant-glomerulitis and peritubular capillaritis and arteritis and thrombotic microangiopathy in glomeruli. Morphologic features of AMR (i.e., transplant glomerulitis and peritubular capillaritis) without C4d positivity were also considered as AMR (18, 19). Accordingly, AMR phenotypes were subdivided as previously reported (5): (i) clinical/subclinical C4d-positive AMR; (ii) clinical/subclinical C4d-negative AMR; and (iii) clinical/subclinical C4d positivity only.
Treatment for AMR changed over time (5). From 1999 to 2004, most AMR episodes were treated with polyclonal anti-T lymphocyte globulin and steroid pulses. Since 2004, steroid pulses, intravenous immunoglobulins, plasmapheresis, and rituximab (Mabthera, Roche, Switzerland) were used according to the severity of AMR.
Detection and Assignment of HLA-DSA by Standard SAFB
All sera have previously been tested for class I (i.e., HLA-A/B/Cw) and class II (i.e., HLA-DR/DQ/DP) HLA antibodies using SAFB on a Luminex platform (LabScreen® Single Antigen LS1A04 and LS2A01; OneLambda, Canoga Park) (2, 5). A positive result was defined as a baseline normalized MFI more than 500. Donor specificity of HLA antibodies was determined by comparison of the HLA antibody specificities with the HLA typing of the donor, including high-resolution and HLA-Cw and HLA-DP typing, if necessary (2, 5). For every individual HLA-DSA, the reported strength was based on the MFI of one SAFB. In case of more than one HLA-DSA against different HLA antigens, the cumulative strength was calculated by adding the individual MFI values.
Development of the Modified SAFB Assay Detecting IgG Subclasses
The standard SAFB was modified by replacing the generic reporter antibody IgGPAN (OneLambda) by monoclonal reporter antibodies specific for IgG1–4 subclasses conjugated with phycoerythrin (PE) (IgG1 clone HP6001, IgG2 clone 31-7-4, IgG3 clone HP6050, IgG4 clone HP6025; Southern Biotech, Birmingham, AL). We established the assay on HLA class II SAFB using four chimeric human/murine IgG subclass-specific antibodies directed against HLA class II antigens (pan HLA II IgG1–4) as positive controls (20, 21). The applied controls differ only regarding their human heavy chains (i.e., IgG1, IgG2, IgG3, and IgG4), but all target the same epitope existing on HLA class II antigens. Complete saturation of 1.5 μL SAFB suspension was achieved with 28 ng chimeric pan HLA II IgG1–4 antibodies. Next, the specificity of the IgG1–4 subclass reporter antibodies was determined on the saturated SAFB. Finally, we determined the amount of each IgG1–4 subclass reporter antibody giving the best signal-to-noise ratio (noise=appropriate control mouse PE-labeled IgG1, clone 15H6; Southern Biotech).
IgG Subclass Analysis by the Modified SAFB Assay
The same pretransplant sera evaluated for IgG HLA-DSA as detailed earlier were used for the IgG subclass SAFB assay. The assay was performed in light-protected black V-bottom trays. In brief, 6 μL of patient serum and 1.5 μL of SAFB suspension (LabScreen® SA LS1A04 Lot5 and LS2A01 Lot7) were supplemented with 20 μL wash buffer (OneLambda) and gently agitated for 30 min at room temperature. Next, 170 μL wash buffer (OneLambda) was added, and the plate was centrifuged for 5 min at 1300g. After discarding the supernatant, two more washing steps were performed using 200 μL wash buffer. Then, 25 μL of appropriately diluted PE-labeled IgG1–4 subclass reporter antibody (concentration: anti-IgG1=1.3 μg/mL; anti-IgG2=1.3 μg/mL; anti-IgG3=10.6 μg/mL; anti-IgG4=0.68 μg/mL) was added and incubated for 30 min. After two washing steps, SAFB were solved in 80 μL phosphate-buffered saline and data acquired on the Luminex100™ analyzer. A positive result was defined by a MFI value above a cutoff that was generated for each IgG subclass and for every individual bead by using four negative control sera (NC1–4) from healthy, nonsensitized, and HLA antibody–negative men: Cutoff MFI=mean NC1–4+3 standard deviations NC1–4. To determine the amount of IgG subclasses, we used the ratio above the corresponding cutoff (i.e., ratio=MFI IgGsubclass divided by MFI cutoff).
We used JMP software version 8.0 (SAS Institute Inc., Cary, NC) for statistical analysis. For categorical data, Fisher's exact test or Pearson's chi-square test were used. Parametric continuous data were analyzed by Student's t tests. For nonparametric continuous data, the Wilcoxon rank-sum test was used. Survival analysis was performed by the Kaplan-Meier method and groups compared using the log-rank test. A P value less than 0.05 was considered to indicate statistical significance.
1. Gupta A, Iveson V, Varagunam M, et al. Pretransplant donor-specific antibodies in cytotoxic negative crossmatch kidney transplants: Are they relevant? Transplantation
2008; 85: 1200.
2. Amico P, Hönger G, Mayr M, et al. Clinical relevance of pretransplant donor-specific HLA antibodies
detected by single-antigen flow-beads. Transplantation
2009; 87: 1681.
3. van den Berg-Loonen EM, Billen EV, Voorter CE, et al. Clinical relevance of pretransplant donor-directed antibodies detected by single antigen beads in highly sensitized renal transplant patients. Transplantation
2008; 85: 1086.
4. Patel AM, Pancoska C, Mulgaonkar S, et al. Renal transplantation
in patients with pre-transplant donor-specific antibodies and negative flow cytometry crossmatches. Am J Transplant
2007; 7: 2371.
5. Bächler K, Amico P, Hönger G, et al. Efficacy of induction therapy with ATG and intravenous immunoglobulins in patients with low-level donor-specific HLA-antibodies. Am J Transplant
2010; 10: 1254.
6. Burns JM, Cornell LD, Perry DK, et al. Alloantibody levels and acute humoral rejection early after positive crossmatch kidney transplantation. Am J Transplant
2008; 8: 2684.
7. Feucht HE, Schneeberger H, Hillebrand G, et al. Capillary deposition of C4d complement fragment and early renal graft loss. Kidney Int
1993; 43: 1333.
8. Regan J, Monteiro F, Speiser D, et al. Pretransplant rejection risk assessment through enzyme-linked immunosorbent assay analysis of anti-HLA class I antibodies. Am J Kidney Dis
1996; 28: 92.
9. Griffiths EJ, Nelson RE, Dupont PJ, et al. Skewing of pretransplant anti-HLA class I antibodies of immunoglobulin G isotype solely toward immunoglobulin G1 subclass is associated with poorer renal allograft survival. Transplantation
2004; 77: 1771.
10. Monteiro F, Mineiro C, Rodrigues H, et al. Pretransplant and posttransplant monitoring of anti-HLA class I IgG1 antibodies by ELISA identifies patients at high risk of graft loss. Transplant Proc
1997; 29: 1433.
11. Gao ZH, McAlister VC, Wright JR Jr, et al. Immunoglobulin-G subclass antidonor reactivity in transplant recipients. Liver Transpl
2004; 10: 1055.
12. Lobashevsky A, Rosner K, Goggins W, et al. Subtypes of immunoglobulin (Ig)-G antibodies against donor class II HLA and cross-match results in three kidney transplant candidates. Transpl Immunol
2010; 23: 81.
13. Karuppan SS, Ohlman S, Moller E. The occurrence of cytotoxic and non-complement-fixing antibodies in the crossmatch serum of patients with early acute rejection episodes. Transplantation
1992; 54: 839.
14. Kushihata F, Watanabe J, Mulder A, et al. Human leukocyte antigen antibodies and human complement activation
: Role of IgG subclass, specificity, and cytotoxic potential. Transplantation
2004; 78: 995.
15. Bartel G, Wahrmann M, Exner M, et al. Determinants of the complement-fixing ability of recipient presensitization against HLA antigens. Transplantation
2007; 83: 727.
16. 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.
17. Zachary AA, Sholander JT, Houp JA, et al. Using real data for a virtual crossmatch. Hum Immunol
2009; 70: 574.
18. Loupy A, Suberbielle-Boissel C, Hill GS, et al. Outcome of subclinical antibody-mediated rejection
in kidney transplant recipients with preformed donor-specific antibodies. Am J Transplant
2009; 9: 2561.
19. Sis B, Jhangri GS, Bunnag S, et al. Endothelial gene expression in kidney transplants with alloantibody indicates antibody-mediated damage despite lack of C4d staining. Am J Transplant
2009; 9: 2312.
20. Arnold ML, Zacher T, Dechant M, et al. Detection and specification of noncomplement binding anti-HLA alloantibodies. Hum Immunol
2004; 65: 1288.
21. Arnold ML, Dechant M, Doxiadis II, et al. Prevalence and specificity of immunoglobulin G and immunoglobulin A non-complement-binding anti-HLA alloantibodies in retransplant candidates. Tissue Antigens
2008; 72: 60.