Sensitization to human leukocyte antigens (HLAs) remains a daunting barrier to transplantation for patients who have developed anti-HLA antibodies through pregnancy, transfusions, or prior transplants. These patients wait up to twice as long for a compatible graft, and once transplanted, they are at higher risk for acute and chronic rejection episodes and decreased graft survival compared with patients who are not sensitized to HLA (1–5). Addressing the issue of increasing numbers of sensitized patients waiting for compatible grafts in a timely manner is a major challenge for transplant programs. Immunomodulation protocols have been designed to decrease or eliminate the amount of HLA antibody allowing for broadly sensitized patients to be transplanted with compatible grafts. Two major immunomodulation protocols have been developed, one using high-dose intravenous immunoglobulin/rituximab (IVIG/R) (6–11) and the second using plasmapheresis and low-dose IVIG (12–15).
Monitoring antibody levels to determine acceptable levels of donor-specific antibodies (DSAs) before transplantation is critical to the success of immunomodulation (16, 17). The type of monitoring strategy used depends on the desired endpoint of the treatment protocol. For living donor (LD), initial identification of DSAs coupled with monitoring over time is essential. Zachary et al. (15) have reported that the likelihood of successful immunomodulation depends on the titer and specificity of the DSA. For deceased donor (DD) testing, the laboratory must identify all antibody specificities present and monitor for changes in their strength. Compatible DD are identified through a combination of DSA strength and the donor-specific crossmatch (XM), both within an acceptable range.
The combination IVIG/R immunotherapy has become the standard of care at our institution for broadly sensitized patients awaiting LD, if no HLA or ABO compatible donor is available, and DD transplantation, if no LD is available. We have previously reported the acceptable levels of class I-specific antibodies that allow for successful transplantation for both groups after immunotherapy (17). We have expanded this approach to include class II-specific antibodies and the frequencies of the antibodies detected as determined by individual class I or II antigen frequencies and the United Network for Organ Sharing (UNOS) calculated panel reactive antibody (CPRA) method (1). We have also used pronase treatment of donor cells to eliminate the interference of rituximab on B-cell flow XM (FXM) (18, 19). To help guide the appropriate course of treatment, we have expanded our studies to include long-term antibody analyses. Herein, we report our findings to determine the impact of the high-dose IVIG/R immunotherapy on pretransplant HLA-specific antibody levels throughout the course of treatment and our guidelines for acceptable antibody levels, leading to the identification of compatible DDs. This approach gives broadly sensitized patients without LDs the opportunity for transplantation with compatible DD.
We first determined the immunomodulatory effect of the IVIG/R treatments by generating comprehensive antibody profiles using sequential serum samples. Figure 1 shows the antibody profiles from eight patients who received one IVIG/R immunomodulatory treatment. The results are expressed as the mean of all positive beads binding in SFI units (mSFI). The binding intensity more than 20,000 SFI/1000 MFI is considered to correlate with a positive FXM. The results are illustrated with the lowest mSFI obtained for each patient antibody profile indicated at time 0. Toward the left, the first value for each patient profile is the pretreatment value. After this initial level, there is a period of 30 to 120 days when the antibody levels reach the lowest levels for all eight patients. The mean decrease in binding was more prominent for four of the eight patients. After that time period, the mSFI increased or seemed to remain constant, with seven of eight antibody profiles showing strong mean binding more than 100,000 SFI/5,000 MFI. These results indicate that the antibody profiles for all patients do show some impact of the IVIG/R therapy, albeit some more pronounced than others. However, the decrease observed does seem to rebound to higher levels with time.
Figure 2 shows the antibody profile from a patient who received three rounds of immunotherapy and the changes in antibody-binding strength in greater detail. Each round of treatment changed the SFI intensity of each single-antigen bead. The first round of treatment reduced the mSFI of all beads by 68% from 88,963 to 28,372. The binding intensity was greater than 100,000 SFI for 29% of all the beads at day 92, which decreased to 9% of all the beads at day 358. The second round of treatment reduced the mSFI of all beads by 44% from 42,473 (day 361) to 23,907 (day 444). The corresponding percent of beads with binding greater than 100,000 SFI was 20% (day 361), which decreased to 8% (day 444). The third round showed a reduced mSFI at day 561, but overall showed an increase of 8% from 47,190 mSFI at day 454 to 51,074 at day 750. The percent of beads with binding greater than 100,000 SFI was 17% at day 454, which increased to 21% at day 750. The antigen frequency based PRA (aPRA), indicated at the bottom of the figure, showed little variation with time and did not reflect the same changes as seen with the mSFI levels. These results show an oscillation of antibody levels that coincides with each round of therapy. These changes in antibody levels were independent of any changes in the assay system itself in that there is no absolute correlation between class I and II internal positive control bead values with the oscillation of antibody levels (data not shown).
Figure 3(a and b) illustrates the correlation between the pronase-treated B-cell pronase flow crossmatch (BPFXM) by using two different summation methods of DSA levels. Figure 3(a) illustrates the sum of the highest DSA values from class I and II versus BPFXM MCS values. Figure 3(b) illustrates the sum of all the class I and II DSA values versus BPFXM MCS values. Both SFI and MFI values are indicated. Four patients had antibody-mediated rejection (AMR) within 42 days after receiving the DD transplant (red squares). One patient had AMR at 114 day posttransplant (yellow diamond). Four patients had AMR after 200 days (green triangle). In comparison, data for seven patients with no AMR are also illustrated. Both figures showed a 300 MCS cutoff value to differentiate the early AMR group from the late AMR and non-AMR groups. The summation of DSA levels were from 6205 to 22,202 MFI (Fig. 3a) for the highest class I and II DSA beads and 12,865 to 63,065 MFI (Fig. 3b) for the sum of all class I and II DSA beads.
By using the criteria outlined, 74% (80/108) patients were transplanted (Table 1). A total of 26% (28/108) patients remained on the waitlist during this period with mean calculated PRA (mCPRA) 96. Four of these patients were transplanted in 2010. Of the 80 patients transplanted, 42 (53%) were transplanted with a positive donor-specific T- or B-cell flow cytometry XM. A total of 19 (24%) patients experienced AMR±cell mediated rejection (CMR) ranging from 3 to 535 days posttransplant with a median of 109 days. Seven of these patients were diagnosed with AMR±CMR within the first 42 days. Two patients were diagnosed with late AMR±CMR at posttransplant days 144 and 249. One of these patients had cytomegalovirus detected by polymerase chain reaction (PCR) before the diagnosis of AMR, and one patient had cytomegalovirus detected by PCR at the time of AMR diagnosis. Another patient had parvovirus B19-positive PCR reaction before the diagnosis of AMR (posttransplant day 535). One patient was noncompliant (posttransplant day 139) and lost the graft. Twenty-eight patients (mCPRA 84) were transplanted with a negative donor-specific flow cytometry XM. Pretreatment samples were available for 25 of these 28 patients and showed that all pretreatment XMs were negative with the donor samples. Among these 28 patients, none had AMR and 8 had CMR. The CMR occurred between 15 and 773 days posttransplant with the median time at 58 days. A total of 10 patients received zero HLA ABDR-mismatched grafts (mCPRA 90), one had AMR after 101 days posttransplant, and one had CMR at 404 days posttransplant.
We have reported that patients undergoing immunotherapy show a significant decrease in T-cell FXM results, and CDC T-cell PRA results performed on sera obtained before the initiation of therapy and immediately pretransplant (6, 7). These patients were selected for transplant based on achieving acceptable levels of DSA including binding to single-antigen class I beads defining DSA at less than 100,000 SFI/5000 MFI and T-cell donor FXM less than 250 MCS. Our studies have shown that these patients are at low risk for AMR (17). Herein, we report on the DD kidney transplant HLA antibody profiles after high-dose IVIG/R immunotherapy as measured by solid-phase antibody profiling to identify optimal DSA strength and timing. Further, the use of pronase-treated cells allows for monitoring of changes in donor-specific B-cell XMs subsequent to high-dose IVIG/R immunotherapy.
We have reported here on changes of antibody levels subsequent to IVIG/R immunomodulation that allows selection of compatible DDs and optimal timing of transplantation without increasing the risk of AMR. Previously, we have reported the use of an in vitro CDC XM to identify patients who would benefit from the immunomodulatory therapy (10). The question arises whether patients who will benefit sufficiently from immunomodulation to allow for transplant can be identified in the initial workup using current solid-phase testing schematics. To address this question, we determined the effect of IVIG/R on antibody-binding profiles for patients who completed treatment. We used a quantitative Luminex-based single-antigen (L-SA) to measure the level of binding to class I- or class II-coated beads (Fig. 1). Results were also expressed as aPRA for the frequency of the class I or class II antigens and CPRA. We anticipated that quantitative antibody profiles rather than aPRA or UNOS CPRA would reflect more accurately the impact of IVIG/R therapy on antibody binding. We compiled a complete time course of the antibody profiles for eight patients (Fig. 1). The results indicate that all patients have a decrease, some more pronounced than others, in mean antibody levels between 30 and 120 days posttreatment after which the antibody levels rebound and remain high. Despite three rounds of immunomodulation (Fig. 2), there seem to be decreases in antibody levels followed by rebounds. These results suggest that there may be two windows of opportunity for patients to be transplanted. First, a positive response to immunotherapy with the level of antibodies to disparate donor antigens decreasing to less than the 100,000 SFI/5000 MFI and a T-cell XM less than 250 MCS (17) as previously described and a B-cell flow cytometry XM less than 300 MCS. A second opportunity occurs at 30 to 120 days posttreatment when a nadir of DSA is seen. We have optimized the monthly antibody information provided to the transplant team by providing lists of antigens with binding less than 100,000 SFI/5000 MFI (an acceptable XM would be anticipated), antigens with binding greater than 100,000 SFI/5000 MFI (higher XM results would be anticipated), and antigens with binding greater than 200,000 SFI/10,000 MFI (an unacceptable [UA] positive cytotoxicity XM would be anticipated). The later group of antigens is listed as UA in UNOS UNet, thereby eliminating crossmatching with donors with antigens where there is a high likelihood of a highly positive UA XM. We observed no difference in class I or II aPRA or CPRA for IVIG/R-treated patients (Fig. 1). Presumably, these patients were so broadly sensitized with antibodies to high-frequency antigens. Thus for patients undergoing immunotherapy, it is important to monitor changes in DSA levels among serial samples. We also observed (Fig. 2) that the antibody levels changed after IVIG/R treatment showing a cyclic decrease and rebound effect during a 30- to 120-day period.
These results provide insight into the feasibility a patient will receive an acceptable donor and when a patient should receive a second round of immunotherapy. We conclude that all patients do benefit from the IVIG/R immunotherapy as shown by decreases in antibody levels, which seem to be relatively long lived. The longevity of the antibody decrease and slow rebound may be attributable to the combination IVIG/R therapy used. Rituximab reportedly eliminates B cells by a variety of mechanisms including complement-dependent cytotoxicity (20), antibody-dependent cellular cytotoxicity, phagocytosis, and apoptosis (20–23). The B-cell depletion is profound, occurring as early as 24- to 72-hr postadministration (20) and is sustained for approximately 6 months. The half-life of IgG is 21 to 24 days that accounts for the initial lag in decrease of antibody detected by the solid-phase assays. Eventually the antibody levels do start to increase. Sidner et al. (21) reported in the few remaining B cells that after one dose of rituximab, there was an increase in the CD19+ CD27+ memory B cells. Although memory B cells do not secrete antibody, they can convert to antibody-secreting plasma cells on exposure to antigen (24, 25). The B cells that repopulate by 6 months are predominantly the CD19+CD5+ cells (21). No marked increase in viral infection was observed in our patient population. All patients in this study received infection prophylaxis protocols and were monitored by viral PCR assays for 6 months posttransplant (6, 7). Thus, the effect of the IVIG/R to the decrease HLA-specific antibody levels does not seem to result in an increase in viral infections.
We further established the correlation between pronase B-cell FXM and class I and II L-SA values along with the previously published correlation between the T-cell FXM and class I L-SA values (17). We concluded that 6000 MFI (120,000 SFI) achieved by adding the highest DSA values for class I and II beads and 300 MCS pronase B cell XM will increase the risk of immediate AMR (Fig. 3a and b). For the early AMR patient group, there was no major difference between adding both class I and class II highest DSA beads or adding all DSA bead values. Thus, our guidelines for selection of an appropriate compatible DD include a flow cytometry T-cell XM of less than 250 MCS, a pronase B-cell–treated flow cytometry XM less than 300 MCS, and no DSAs in the high-binding range of more than 200,000 SFI/10,000 MFI.
Using these criteria, compatible DD were identified for 80 (74%) of the 108 IVIG/R-treated patients. Forty-two (53%) of these patients were transplanted across positive flow cytometry XMs. Pretreatment and posttreatment single-antigen bead data available for 25 of these patients showed an average decrease in highest DSA-binding strength, 33% for class I and 37% for class II. By combining antibody profiling with strengths and assignment of UA antigens, the virtual XM identified DD with negative XMs for 28 patients (35%). All patients in the positive and negative XM groups had high CPRAs; thus, it was not possible to identify which patients would have DD with negative XMs before treatment. Further, most of these patients had been on the wait list for more than 8 years. The CPRA values of 80% to 96% indicate that 1 in 5 to 1 in 20 donors should be compatible; however, the CPRA does not take into account antibody to HLA-C, DP, or specific relevant alleles. Thus, our results are consistent with some negative XM donors being identified for these patients. The zero mismatched grafts remain an important compatible donor source for these broadly sensitized patients.
In summary, these results indicate that even with one round of IVIG/R therapy, the likelihood that a patient will go on to DD transplant is relatively high. This approach allows for transplantation of patients who otherwise would likely not receive a compatible DD. Rigorous and routine antibody monitoring is essential in guiding transplant timing and therapeutic interventions.
MATERIALS AND METHODS
From January 2006 to December 2009, 108 patients broadly sensitized to HLA who had reached the top of the waitlist and treated with IVIG/R were selected. HLA antibody status was monitored by solid-phase antibody testing methods including flow cytometry, CDC, and L-SA since May 2007. The UAs were based on intensities of L-SA binding expected to produce CDC-positive XMs (>200,000 SFI/10,000 MFI) and entered into the UNOS database generating CPRA. As a part of the final XMs for DD testing, the pretreatment, posttreatment, and current sera were included in the FXM, CDC XM, and L-SA testing protocol. Since February 2008, pronase-treated B-cell XMs have been used to eliminate the interference by the rituximab. AMR or CMR was diagnosed by biopsy. The C4d presented in vessels of the grafts or manifested margination of neutrophils and monocyte or macrophages in glomerular and peritubular capillaries was indicative as markers for AMR diagnosis.
IVIG/R Immunotherapy Protocol
All patients received IVIG (2.0 g/kg on day 1 and 30, maximum dose 140 g/treatment) plus rituximab (1 g administered on day 15). The IVIG doses were infused as described previously (6, 7).
The L-SA was performed according to manufacturer's instructions (One Lambda Canoga Park, CA) as described previously (17). Normalized SFI values were obtained as described previously (17). SFI values are derived from the raw MFI data using the Quantiplex standard beads (26). SFI values more than 20,000 were considered as positive binding that would account for a positive flow cytometry XM. Raw MFI values were obtained from Luminex readout. Rather than using PRA value determined by binding to a given percent of beads tested, we used antigen frequency PRA (aPRA). aPRA is calculated based on the antigen frequencies of the class I or II antigens detected using a panel of 54,458 HLA typed donors in the UNOS database (mTilda version 6.13, mTilda Data Management System, Outland Enterprise, OR).
XMs were performed as described previously (17, 18) with the pronase treatment as follows (19). Pronase is a protease that can remove CD20 from B-cell surfaces, thus eliminating binding by rituximab in the assay. Briefly, 3 to 5×106 cells were used for 1.0 mL (1 mg/mL) pronase treatment. A total of 0.15 to 0.25×106 pronase-treated cells were used for the FXM. For the patient samples and the rituximab control, 10 μL (0.1 mg/mL) of mouse anti-human CD20 (Southern Biotech, AL) were added to block any residual nonspecific binding to CD20, which was not removed by the pronase treatment. FXM results were expressed as mean channel shift over background. T pronase cell XMs were considered positive at more than 70 MCS and B pronase cell FXMs more than 130 MCS.
The authors thank Rosemarie Vadillo for preparation of manuscript and the HLA Laboratory team for their technical expertise and the Transplant Immunotherapy Program team for their excellent patient care and support.
1. The Organ Procurement and Transplantation Network 2007. Available at: http://optn.transplant.hrsa.gov/
2. Jackson AM, Zachary AA. The problem of transplanting the sensitized patient: Whose problem is it? Front Biosci
2008; 13: 1396.
3. Lee PC, Terasaki PI, Takemoto SK, et al. All chronic rejection failures of kidney transplants were preceded by the development of HLA antibodies. Transplantation
2002; 74: 1192.
4. Michaels PJ, Fishbein MC, Colvin RB. Humoral rejection of human organ transplants. Springer Semin Immunopathol
2003; 25: 119.
5. Takemoto SK, Zeevi A, Feng S, et al. National conference to assess antibody-mediated rejection in solid organ transplantation. Am J Transplant
2004; 4: 1033.
6. Vo A, Lukovsky M, Toyoda M, et al. Rituximab
and intravenous immune globulin for desensitization during renal transplantation. N Engl J Med
2008; 359: 242.
7. Vo A, Peng A, Toyoda M, et al. Use of intravenous immune globulin and rituximab
for desensitization of highly HLA-sensitized patients awaiting kidney transplantation. Transplantation
2010; 89: 1095.
8. Jordan SC, Vo AA, Peng A, et al. Intravenous gammablobulin (IVIG): A novel approach to improve transplant rates and outcomes in highly HLA-sensitized patients. Am J Transplant
2006; 6: 459.
9. Jordan SC, Vo A, Bunnapradist S, et al. Intravenous immune globulin treatment inhibits crossmatch positivity and allows for successful transplantation of incompatible organs in living-donor and cadaver recipients. Transplantation
2003; 76: 631.
10. Jordan SC, Tyan DB, Stablein D, et al. Evaluation of intravenous immunoglobulin as an agent to lower allosensitization and improve transplantation in highly sensitized adult patients with end-stage renal disease: Report of the NIH IGO2 trial. J Am Soc Nephrol
2004; 15: 3256.
11. Jordan SC, Vo A, Peng A, et al. Intravenous gammaglobulin (IVIG): A novel approach to improve transplant rates and outcomes in highly HLA-sensitized patients. Am J Transplant
2006; 6: 459.
12. Montgomery RA, Zachary AA. Transplanting patients with a positive donor-specific crossmatch: A single center's perspective. Pediatr Transplant
2004; 8: 535.
13. Zachary AA, Montgomery RA, Leffell MS. Desensitization protocols improving access and outcome in transplantation. Clin Appl Immunol
2005; 5: 373.
14. Crew RJ, Ratner LE. Overcoming immunologic incompatibility: Transplanting the difficult to transplant patient. Semin Dial
2005; 18: 474.
15. Zachary AA, Montgomery RA, Leffell MS. Factors associated with and predictive of persistence of donor-specific antibody after treatment with plasmapheresis and intravenous immunoglobulin. Hum Immunol
2005; 66: 364.
16. Zachary AA, Montgomery RA, Jordan SC, et al. 14th international HLA and immunogenetics workshop: Report on understanding antibodies in transplantation. Tissue Antigens
2007; 69: 160.
17. Reinsmoen NL, Lai CH, Vo A, et al. Acceptable donor-specific antibody levels allowing for successful deceased and living donor kidney transplantation after desensitization therapy. Transplantation
2008; 86: 820.
18. Bray RA, Nickerson PW, Kerman RH, et al. Evolution of HLA antibody detection: Technology emulating biology. Immunol Res
2004; 29: 41.
19. Bearden CM, Agarwal A, Book BK, et al. Pronase treatment
facilitates alloantibody flow cymetric and cytotoxic crossmatching in the presence of rituximab
. Hum Immunol
20004; 65: 803.
20. Maloney DG, Liles TM, Czerwinski DK, et al. Phase I clinical trial using escalating single-dose infusion of chimeric anti-CD20 monoclonal antibody (IDEC-C2B8) in patients with recurrent B-cell lymphoma. Blood
1994; 84: 2457.
21. Sidner RA, Book BK, Agarwal A, et al. In vivo human B-cell subset recovery after in vivo
depletion with rituximab
, anti-human CD20 monoclonal antibody. Hum Antibodies
2004; 13: 55.
22. Maloney DG, Smith B, Appelbaum FR. The anti-tumor effect of monoclonal anti-CD20 antibody (mAb) therapy includes direct anti-proliferative activity and induction of apoptosis in CD20 positive non-Hodgkin's lymphoma (NHL) cell lines. Blood
1996; 88(suppl 1): 637.
23. Manches O, Lui G, Chaperot L, et al. In vitro mechanisms of action of rituximab
on primary non-Hodgkin lymphomas. Blood
2003; 101: 949.
24. Ramos EJ, Pollinger HS, Stegall MD, et al. The effect of desensitization protocols on human splenic B-cell population in vivo. Am J Transplant
2007; 7: 402.
25. Dorshkind K, Rawlings DJ. B-cell development: The hematopoietic hierarchy and B-lineage commitment. In: Hoffman, ed. Hematology: Basic principles and practice, [ed. 4]. Churchill Livingstone Elsevier, Philadelphia 2005.
26. Lai CH, Okhovat J, Ong G, et al. Better dynamic antibody detection range using flow single antigen assay than Luminex single antigen assay. Hum Immunol
2010; 71: S62.