Abstract

Objectives: A T-cell-positive and B-cell-negative flow cytometry crossmatch result remains a conundrum since HLA class I antigens are expressed on both T and B cells. We investigated the frequency, donor HLA specificity of the antibodies, and mechanisms for these crossmatch results.
Materials and Methods: We analyzed 3073 clinical flow cytometry crossmatch tests performed in an American Society of Histocompatibility and Immunogenetics-accredited histocompatibility laboratory. The sera associated with the T-cell positive and B-cell negative flow cytometry crossmatches were also tested for donor HLA immunoglobulin G antibodies using LABScreen single antigen assays.
Results: Among the 3073 test results, 1963 were T-cell negative and B-cell negative, 811 were T-cell negative and B-cell positive, 274 were T-cell positive and B-cell positive, and 25 were T-cell positive and B-cell negative. The LABScreen single antigen assay detected HLA class I immunoglobulin G donor-specific antibodies in 23 of 25 sera associated with a T-cell positive and B-cell negative flow cytometry crossmatch result, and donor-specific antibodies directed at not only HLA-Cw but also at HLA-A or HLA-B were observed. In addition, we identified that the B-cell channel shift threshold used to classify a B-cell flow cytometry crossmatch was a potential contributor to a T-cell-positive and B-cell-negative flow cytometry crossmatch result.
Conclusions: Our analysis of 3073 flow cytometry crossmatches, in addition to demonstrating that HLA antibodies directed at the HLA-A, -B, or -Cw locus were associated with a T-cell-positive and B-cell-negative result, identified mechanisms for the surprising T-cell-positive and B-cell-negative flow cytometry crossmatch result.

Key words : Donor-specific antibody, Transplantation

Introduction

Kidney transplantation is the treatment of choice for most patients with irreversible kidney failure, but presensitization to donor human leukocyte antigens (HLA) and antibody-mediated rejection undermine its full benefits. In a landmark publication, Patel and Terasaki reported that 24 of 30 kidney transplants performed across a positive complement-dependent cytotoxicity crossmatch (CDC-XM) failed in an accelerated fashion.¹ Test systems to detect circulating antibodies at a higher sensitivity than CDC-XM have been developed, and flow cytometry crossmatch (FCXM) is in widespread use to identify donor-specific immunoglobulin G (IgG) antibodies (DSA).^2-5 Pretransplant FCXM results are used in many centers to optimize donor selection, and pretransplant T- and B-cell FCXM results have been associated with allograft outcomes.^6-10 In addition to detection of HLA IgG DSA, FCXM also detects binding of any type of antibodies to T cells and B cells, including autoantibodies and non-HLA antibodies. HLA class I antigens are expressed on T cells and B cells, and some but not all studies suggest that the cell surface expression of HLA class I antigens is higher on B than on T cells.^11,12 Thus, a T-cell-positive, B-cell-positive (T+B+) FCXM, rather than a T-cell-positive, B-cell-negative (T+B-) FCXM, is the expected result in the presence of HLA class I antibodies. A T-cell-negative, B-cell-positive (T-B+) FCXM is the anticipated result in the presence of HLA class II antibodies only since HLA class II antigens are expressed on B cells but not on resting T cells.¹³

We aimed to determine the frequency, HLA specificity of the antibodies, and potential mech-anisms responsible for the T+B- FCXM. We analyzed 3073 clinical FCXM tests performed using a standardized protocol in an American Society of Histocompatibility and Immunogenetics (ASHI)-accredited histocompatibility laboratory and report here the frequency, HLA-specificity of the antibodies, and mechanisms responsible for the T+B- FCXM.

Materials and Methods

Isolation of mononuclear cells for the flow cytometry crossmatch and complement-dependent cytotoxicity crossmatch tests
Peripheral blood collected in acid citrate dextrose tubes was used for isolation of peripheral blood mononuclear cells (PBMC) using Ficoll-Paque Premium solution (GE Healthcare Biosciences) density gradient centrifugation. Mononuclear cells from the lymph node and spleens from deceased donors were also isolated using Ficoll-Paque Premium solution density gradient centrifugation. Mononuclear cells from the interface were washed 3 times with phosphate-buffered saline, and red blood cells were lysed with Tris ammonium chloride. Mononuclear cells were counted using either the Neubauer counting chamber or the LUNA automated cell counter (model L20001; Logos Biosystems). Cell viability was assessed with the Trypan blue exclusion test or the acridine orange/propidium iodide assay. In the assays reported here, viability of mononuclear cells exceeded 80%.

Separation of serum
Peripheral blood collected in serum separation tubes was centrifuged at 2300 revolutions/min for 20 minutes. The serum was collected and treated with 0.3% sodium azide prior to storage.

Flow cytometry crossmatch
Flow cytometry crossmatch was performed in a 96-well V-bottom plate (catalog no. 3894, Corning, Inc). We incubated 300 000 PBMC or mononuclear cells, isolated from the spleen or the lymph node, with 30 μL of negative control serum (NCS; normal male human AB serum free of HLA class I and II antibodies), recipient serum (RS), or positive control serum (PCS; pooled serum from highly sensitized patients with broad HLA reactivity) for 20 minutes at 22 °C. Before testing, the NCS, RS, and PCS were centrifuged at 42 030 relative centrifugal force to remove any aggregates or cryoprecipitate.

The mononuclear cells used as target cells in the FCXM were washed twice with wash buffer (phosphate-buffered saline with 0.1% fetal bovine serum and 0.1% sodium azide) and stained with 10 μL each of anti-CD3 PerCP mAb (catalog no. 340663; BD Biosciences), anti-CD19 PE mAb (catalog no. 340720; BD Biosciences), and fluorescein isothiocyanate (FITC) affinity-purified Fab’ 2 fragment goat anti-human IgG, Fcγ fragment specific (catalog no. 109-096-098; Jackson ImmunoResearch Laboratories; 1:200 diluted) for 30 minutes at 4 °C. Cells were then washed once and fixed with 1% paraformaldehyde, and FCXM was performed using the BD FACS Canto II instrument, with analysis of data using FACSDiva software (versions 7.1.3 and 8.0.1). Lymphocytes were gated using forward scatter versus side scatter plot. Median channel fluorescence (MCF) of goat-anti-human IgG FITC fluorescence on T cells and on B cells was measured from the histograms. Channel shift (CS) was calculated by subtracting the MCF observed with the NCS from the MCF observed with the RS or PCS.

A T-cell FCXM test was reported as positive if the MCF of T cells incubated with the RS showed a CS of 40 or higher. A B-cell FCXM test was reported as positive if the MCF of B cells incubated with the RS showed a CS of 50 or higher. The CS thresholds were based on the interquartile range (IQR; calculated by subtracting the 25th percentile value from the 75th percentile value) observed in our laboratory following treatment of target mononuclear cells with NCS. In 2020 consecutive T-cell FCXM performed at our center, the MCF of T cells treated with NCS was 229 and the IQR was 39; in 2019 consecutive B-cell FCXM, the MCF of B cells treated with NCS was 252 and the IQR was 47. In 8 consecutive ASHI proficiency challenges involving 66 unique sera and 16 unique HLA-typed cells, the use of these CS thresholds resulted in 65 of 66 T-cell FCXM tests and 64 of 66 B-cell FCXM tests being concordant with the 80% ASHI consensus result (ground truth).

Complement-dependent cytotoxicity crossmatch
The CDC?XM tests were performed with donor T and B lymphocytes isolated from mononuclear cells by using immunomagnetic beads (Dynabeads CD19 pan B for B lymphocytes and Dynabeads CD2 for the isolation of T lymphocytes; Life Technologies, Thermo Fisher Scientific). Recipient serum was used untreated or was heat-inactivated for 6 minutes at 63 °C. In a microtiter tray, 1000 cells in 1 μL volume per well were treated with 2 μL of serum and incubated at 22 °C for 30 minutes. For T-cell anti-human globulin (AHG) CDC-XM tests, the trays were washed 3 times with wash buffer (2.5% fetal calf serum in McCoy’s media), 1 μL of appropriate dilution of AHG was added to each well, and the tray was incubated at 22 °C for 2 minutes. Next, 5 μL of rabbit complement (catalog no. 310245; Invitrogen Life Technologies) was added to both the T-cell-AHG CDC-XM tray and the B-cell CDC-XM tray and incubated at 22 °C for 1 hour. Trays were then stained and quenched with 2 μL of ethidium bromide/acridine orange (FlouroQuench Plus; catalog no. FQZAE100; One Lambda Inc). Percent cell death was ascertained by fluorescence microscopy, and the results were scored as per the ASHI scoring method.

Detection of anti-HLA antibodies
Sera, untreated or treated with EDTA (0.025 M final concentration), were tested for the presence of anti-HLA IgG antibodies leveraging Luminex bead-based multiplexing technology. Anti-HLA class I (HLA-A, -B, -Cw) IgG antibodies and anti-HLA class II (HLA-DR, -DQ, -DP) IgG antibodies were identified using the LABScreen single antigen (LSA) class I assay (LS1A04) and the LSA class II assay (LS2A01), respectively (One Lambda). For these assays, 10 µL of serum samples were incubated with 2.5 μL of microbeads coated with HLA class I or HLA class II antigens, respectively, at 22 °C for 30 minutes. The specimens were then washed 3 times in 1× wash buffer (One Lambda). The microbeads were treated with PE-labeled goat anti-human IgG polyclonal antibody (catalog no. LS-AB2; One Lambda) at 22 °C for 30 minutes in dark conditions and washed 3 times in wash buffer. We used LABScan 100/200 (Luminex Corporation) for acquisition of microbead fluorescence and data analyses. Data were then exported to HLA fusion software (One Lambda) for analyses.

Data analyses
From February 2014 to April 2018, 3609 FCXM tests were performed in our laboratory. We reported data from 3073 FCXM tests using potential organ donor’s cells as target cells and after exclusion of recipient auto-FCXM and repeat donor FCXM with the same results.

For the LSA assay, any single antigen bead with a trimmed mean fluorescence intensity (MFI) of ≥2000 was scored as positive for the presence of anti-HLA IgG DSA in accordance with our threshold used in the ASHI proficiency challenges. When there was more than 1 bead in the test panel for a given antigen and the donor HLA antigen’s high-resolution typing was not available, the highest bead’s MFI value for the given antigen was used as DSA MFI. We summed all DSA MFIs (≥0 MFI) of class I and of class I and II for the correlational analyses between summed MFI and CS of T cells and B cells treated with RS, respectively.

Statistical analyses
We performed statistical analyses using Prism 9 software (GraphPad Software). We summarized the data as median and IQR and used Mann-Whitney tests and Wilcoxon matched-pairs signed-rank tests to compare the groups. A 2-sided P < .05 was considered significant. Data analysis was performed under a Weill Cornell Medicine institutional review board-approved protocol.

Results

Frequency of T-cell-positive, B-cell-negative FCXM results
A total of 3609 clinical FCXM tests were performed in our laboratory from February 2014 to April 2018, and data from 3073 FCXM tests were included in downstream data analysis (Figure 1). Among the 3073 tests, 1963 FCXM tests were negative (64%) and 1110 FCXM tests were positive (36%). Among the positive tests, 811 were T-B+ (73%), 274 were T+B+ (24.7%), and 25 FCXM were T+B- (2.3%).

Twenty-four unique patients awaiting kidney transplant contributed the 25 sera associated with the T+B-; serum from 1 patient was tested against 2 different donors, and the results were T+B- in both instances. In 12 FCXM, a recent serum and a historical serum from the same patient were tested against the same donor. Data analysis showed that 10 of the 12 historical sera were also T+B- (with 2 sera being borderline T positive with CS of 31 and 39) and 2 were T+B+.

Eleven of the 25 T+B- FCXM tests (44%) were performed using deceased donor cells as targets, and 14 (56%) were performed using living donor cells as targets. We found that PBMCs were the targets in 22 of 25 T+B- FCXM and that mononuclear cells isolated from spleen (n = 2) or lymph node (n = 1) were the targets in 3.

Donor HLA antibodies associated with the T-cell-positive, B-cell-negative flow cytometry crossmatch
Donor HLA-specific IgG antibodies directed at HLA class I antigens (>2000 MFI) were detected in 23 of 25 sera associated with the T+B- FCXM. For case 24, the MFI of IgG DSA directed at donor HLA-A1 was 1403 and the MFI of DSA directed at HLA-B57 was 1741. For case 25, the highest MFI of DSA was only 212. Analysis for HLA class I locus specificity showed that the antibodies were often directed at more than 1 locus (Table 1). Among the 23 serum samples, antibodies directed at HLA-A were identified in 12, antibodies at HLA-B were identified in 10, and antibodies at HLA-Cw were identified in 10. Antibodies directed at HLA class II antigens were also detected in 6 serum samples.

We performed CDC-XM tests for 21 of the 25 T+B- FCXM cases (Table 1). Among 21 CDC-XM tests, 2 were T+B+ CDC-XM, 3 were T-B+ CDC-XM, and 16 were T-B- CDC-XM. The CDC-XM test is less sensitive in detecting antibodies compared with the FCXM test but does detect complement-fixing IgM and IgG antibodies, whereas FCXM detects both complement-fixing and complement-nonfixing IgG antibodies. Because of limited sensitivity of the CDC-XM test and because the CDC-XM test detects only complement-fixing antibodies, we expected fewer cases to result in a positive CDC-XM result. Interestingly, case 3 with highest cumulative HLA IgG DSA of 53 187 MFI resulted in a higher ASHI score for T-cell AHG CDC-XM compared with B-cell CDC-XM (data not shown), reflective of a higher T-cell FCXM CS compared with B-cell FCXM CS.

Characteristics of T-cell-positive, B-cell-negative flow cytometry crossmatch
Table 2 lists the characteristics of the 25 T+B- FCXM tests. The median (IQR) of MCF was 230 (36) with the potential donor T cells treated with NCS and 250 (65) for the corresponding B cells (P < .05, Wilcoxon signed-rank test for paired samples). This difference remained significant even after exclusion of case 5, which had a low T-cell MCF and an unusually high B-cell MCF. A B-cell MCF higher than T-cell MCF was also observed in the 3048 FCXM tests that were not T+B- (B-cell MCF = 253 [45] vs T-cell MCF = 233 [38]; P < .001).

We observed that the MCF of T cells and the MCF of B cells treated with RS in the 25 T+B- FCXM texts were not different. The median (IQR) MCF was 287 (33) with the T cells treated with the RS and 283 (74) for the corresponding B cells treated with the RS (P = .81).

The CS, calculated by subtracting the MCF of cells treated with NCS from the MCF of cells treated with the RS, differed significantly between the T cells and the B cells in the 25 T+B- FCXM. The median (IQR) CS with T cells was 49 (20), which was significantly higher than the median (IQR) CS with B cells of 38 (15) (P < .001).

There was no significant association between summed MFI of HLA class I DSA and the CS of T cells treated with RS (Figure 2; Spearman rank correlation coefficient rs = 0.34, P = .09), summed MFI of HLA class I DSA and CS of B cells treated with RS (Figure 2; Spearman rank correlation coefficient rs = 0.08, P = .69), and summed MFI of HLA class I and II DSA and the CS of B cells treated with RS (Figure 2; Spearman rank correlation coefficient rs = 0.04, P = .83) in the 25 T+B- FCXM.

Potential mechanisms for the T-cell-positive, B-cell-negative flow cytometry crossmatch result
A mechanism for the T+B- FCXM is that the MCF of B cells treated with the NCS was uniquely higher when the result is T+B- FCXM compared with the MCF of B cells when the result is T+B+ FCXM or T-B+ FCXM. To investigate this possibility, we compared the MCF of B cells treated with NCS among the FCXM classified as T+B-, T+B+, or T-B+ FCXM. The median (IQR) of B-cell MCF was 250 (65) in the 25 T+B- FCXM tests, 248 (49) in the 274 T+B+ FCXM, and 254 (45) in the 811 T-B+ FCXM (P = .2, Kruskal-Wallis test; Figure 3A). Thus, a uniquely high B-cell MCF was not responsible for the observed T+B- test result.

Another mechanism for the T+B- FCXM result is that the MCF of T cells treated with NCS was uniquely lower in T+B- FCXM compared with T+B+ or T-B+ FCXM. To investigate this possibility, we compared the MCF of T cells treated with NCS among FCXM scored as T+B-, T+B+, or T-B+. The median (IQR) of T-cell MCF was 230 (36) in the 25 T+B- FCXM tests, 229 (41) in the 274 T+B+ FCXM, and 234 (43) in the 811 T-B+ FCXM (P = .06; Figure 3B). Thus, a uniquely low T-cell MCF was an unlikely contributor to the T+B- FCXM.

A CS of 50 is used as the threshold to classify a B-cell FCXM as positive, and a CS of 40 is used to classify a T-cell FCXM as positive in our laboratory. This differential cutpoint is a potential contributor to the T+B- FCXM result. Our data analysis showed that the CS of 50 applied to score a B-cell FCXM as positive may have contributed to 12 of 25 T+B- FCXM results (Table 1).

Figure 4 shows the flow cytometry acquisition setup and representative T+B- FCXM tests to illustrate potential mechanisms. The lymphocyte subpopulation was gated using forward scatter versus side scatter plot (Figure 4A). The T cells were gated using CD3 PerCP and the B cells using CD19 PE, and the MCF of goat anti-human IgG FITC levels on T cells and on B cells were measured using histograms (Figure 4B). Figure 4C is a representative T+B- FCXM in which the CS of 50 threshold used to score a B-cell FCXM as positive contributed to the T+B-. Figure 4D is a representative T+B- FCXM in which the CS of 50 threshold used to score a B-cell FCXM as positive did not contribute to the T+B- FCXM.

Clinical outcomes
Three patients (cases 16, 23, and 24) with a T+B- FCXM received kidney transplants at our center. All 3 received antithymocyte globulin as induction therapy, and 2 (cases 16 and 23) received additional treatment with rituximab and intravenous immunoglobulin as per our transplant center policy for patients with DSA by LSA and a positive FCXM. The third patient, case 24 without DSA by LSA assay but a T+B- FCXM, did not receive additional therapy with rituximab or intravenous immunoglobulin. All 3 received maintenance therapy consisting of tacrolimus, mycophenolate mofetil, and prednisone. None had an episode of acute rejection within the first year posttransplant and have excellent kidney graft function with serum creatinine levels of 0.88 mg/dL (case 16), 1.0 mg/dL (case 23), and 1.36 mg/dL (case 24) at their last follow-up visit.

Monitoring of DSAs showed that the DSA directed at HLA-B50 in case 16 decreased from pretransplant MFI of 9349 to MFI of 1911 at 1-year posttransplant. In case 23, pretransplant DSA MFI of 12 350 remained at the same level, with MFI of 12 418 at 1-year posttransplant. In addition, pretransplant DSA MFI of 3673 directed at HLA-Cw17 increased to MFI of 5768 at 1-year posttransplant. This patient also developed de novo DSA directed at HLA-B42 during this time period, and the MFI was 4050. There were no pretransplant DSAs above MFI >2000 in case 24, and this patient did not develop de novo DSA during the first 3 months of monitoring posttransplant.

Discussion

A systematic analysis of 3073 clinical FCXM performed using a standardized protocol in an ASHI-accredited histocompatibility laboratory identified that 8.4% of the T-cell positive FCXM are T+B-. We found that antibodies directed not only at HLA-Cw but also antibodies directed HLA-A and antibodies directed at HLA-B were associated with a T+B- FCXM.

A mechanism for the T+B- FCXM test result is the differential expression of HLA class I antigens on T cells and B cells. HLA-Cw antigen expression, as detected using mAb DT9, has been found to be higher on T cells than on B cells¹²; in this study, 7 of 10 FCXM T+B- FCXM were observed with sera containing HLA-Cw DSA. Sera with anti-HLA-Cw IgG have also been shown by others to cause T+B- FCXM.^14,15 Our findings are consistent with the earlier findings that antibodies directed at HLA-Cw antigens may contribute to a T+B- FCXM. Importantly, we reported that antibodies directed at HLA-A and antibodies directed at HLA-B were also associated with a T+B- FCXM.

Our correlational analyses did not identify a significant association between the summed MFI of class I DSA and the CS of T cells or B cells treated with the RS. The lack of a linear relationship may be related to the LSA assay being a qualitative assay rather than a quantitative assay¹⁶ and possible confounding due to the presence of non-HLA antibodies. Our findings provided an explanation for the observation of Higgins and colleagues¹⁷ that serum with an MFI of 591 was associated with a positive FCXM, whereas serum with an MFI of 4773 was associated with a negative FCXM.

We reported that the MCF of B cells treated with NCS is significantly higher than the MCF of T cells treated with NCS.¹⁸ The higher B-cell MCF with NCS would reduce the B-cell CS observed with RS and could have potentially caused the T+B- FCXM result. However, the higher B-cell MCF is unlikely to be the major mechanism since the MCF of B cells did not differ among FCXM that were classified as T+B-, T-B+, or T+B+ tests.

It is a common practice to use a larger CS to classify a B-cell FCXM as positive compared with classifying a T-cell FCXM as positive.^19-22 Had we applied the CS of 40 threshold to classify not only T-cell FCXM but also B-cell FCXM, 12 of 25 T+B- FCXM would be reclassified as a T+B+ FCXM. Among these 12 FCXM tests, 7 of 12 would still be classified as T+ FCXM if we had applied the CS of 50 used for the B-cell FCXM. Two of the 12 T+B- FCXM tests with B-cell CS of 49 and CS of 47 resulted in T+B+ CDC-XM, and 1 of the 12 T+B- FCXM tests with B-cell CS of 43 resulted in T-B+ CDC-XM. Thus, the higher CS used to classify a B-cell FCXM test is a potential contributor to 12 of the 25 T+B- results. In this regard, the CS threshold that should be applied to classify a FCXM test as positive or negative is the Achilles’s heel of this assay. In the FCXM tests, serum from a male donor of AB blood group and Rh negative and negative for anti-HLA-class I and class II IgG antibodies were used to establish baseline MCF of T cells and B cells and this serum was used as the NCS. The target cells were from heathy volunteers or potential donors. A predefined ratio of the fluorescence of the T cells and B cells treated with RS to the fluorescence of the corresponding cells treated with NCS is one approach to classify a FCXM.6 A fluorescence value that is 2 or 3 standard deviations (SD) above the mean value observed with the NCS is also used to classify a FCXM as a positive or a negative.^19-23 A value that is 2 SD higher than the mean would place the observed value outside the 95% probability, and a value that is 3 SD higher would place the observed value outside the 99.7% probability. However, this approach assumes normal distribution of fluorescence values. Fluorescence intensity is acquired using a logarithmic scale and is best estimated using medians rather than means, and the distribution of fluorescence values of T cells and B cells treated with NCS is seldom normally distributed. Indeed, the MCF values in the 3073 FCXM analyzed here were not distributed normally. These considerations led to the use of IQR values to derive CS in this study. Finally, had we used 2 SD or 3 SD instead of IQR to develop the CS, none of the 25 T+B- results would change.

In our study, we used the same CS for classifying FCXM irrespective of whether the cells were from a living donor or a deceased donor and regardless of whether the cells were isolated from peripheral blood, spleen, or lymph node. The use of same CS irrespective of the donor cell type may be of concern since Badders and colleagues found that HLA class I expression is lower on B cells isolated from deceased donor blood compared with B cells isolated from spleen, lymph node, or B cells isolated from living donor blood.²⁴ The source of cells used as target cells in our study was not a significant issue since 22 of 25 T+B- were performed using PBMC.

Autoantibodies directed at T cells could result in a T+B- result. In our analysis of 1475 recipient auto-FCXM tests performed in our laboratory during 2014 to 2018 (the same vintage as the 25 T+B- results reported here), not even a single auto-FCXM was T+B- (0%) and 5 were T+B+ (0.34%), 119 were T-B+ (8.17%), and 1333 of 1475 were T-B- (91.49%). Thus, a T-cell autoantibody being responsible for the observed T+B- FCXM appears extremely unlikely.

There are limitations to our study. The donor cells used in our FCXM tests were not treated with pronase, and our results may not be generalizable to tests performed using pronase-treated cells. Pronase treatment has been reported to “improve flow cytometric detection of HLA antibodies”²⁵ and to “improve flow cytometry crossmatch results.”²⁶ However, pronase may also unmask cryptic epitopes on T cells,²⁷ elicit new reactivity to donor B cells and autologous B cells,²⁸ and reduce HLA density on target cells, leading to incorrect FCXM results,²⁹ including false-positive T-cell FCXM tests.³⁰

The donors in our study were not HLA typed at high resolution; therefore, the detected antibody may not be donor HLA allele specific. The LSA assay used in this study may not have included donor-specific alleles. Because the highest bead’s MFI for the given antigen was used as DSA MFI when there was more than 1 bead in the panel, it is possible that the summed MFI may be higher here than what it really was. Our reanalysis of data using lowest bead’s MFI for the given HLA antigen, when there was more than 1 bead in the panel, lowered the HLA class II MFI in 5 of 6 cases with HLA class II DSA; 4 of 8 HLA class II DSA specificities became negative (<2000 MFI). However, summed HLA class I and II MFI for any of the cases with >10 000 MFI using highest bead’s MFI did not reduce to <10 000 MFI using lowest bead’s MFI, suggesting that the use of highest bead MFI did not influence interpretation of the results. These 2 limitations, however, apply to the lack of association between the summed MFI of DSA and MCF but not to the T+B- FCXM result since the B and T cells were from the same donor.

Conclusions

Our data analyses involving 3073 clinical FCXM tests performed using a standardized protocol in an ASHI-accredited histocompatibility laboratory identified that almost 10% of FCXM tests that were T+ FCXM were T+B- FCXM. We demonstrated that IgG antibodies directed at HLA- A, -B and/or -C locus-determined antigens are associated with a T+B- result. Differential expression of HLA on T cells and B cells and/or differential binding of antibodies to T cells and B cells and CS used to qualify a FCXM test as positive or negative are potential contributors to the surprising T+B- FCXM result.

References:

1. Patel R, Terasaki PI. Significance of the positive crossmatch test in kidney transplantation. N Engl J Med. 1969;280(14):735-739. doi:10.1056/NEJM196904032801401
   CrossRef - PubMed
   
2. Lee PP, Garovoy MR. Flow cytometry crossmatching: The first 10 years. Transplant Rev. 1994;8(1):1-4.
   CrossRef - PubMed
   
3. Bray RA, Tarsitani C, Gebel HM, Lee JH. Clinical cytometry and progress in HLA antibody detection. Methods Cell Biol. 2011;103:285-310. doi:10.1016/B978-0-12-385493-3.00012-7
   CrossRef - PubMed
   
4. Maguire O, Tario JD, Jr., Shanahan TC, Wallace PK, Minderman H. Flow cytometry and solid organ transplantation: a perfect match. Immunol Invest. 2014;43(8):756-774. doi:10.3109/08820139.2014.910022
   CrossRef - PubMed
   
5. Guillaume N. Improved flow cytometry crossmatching in kidney transplantation. HLA. 2018;92(6):375-383. doi:10.1111/tan.13403
   CrossRef - PubMed
   
6. Scornik JC, Clapp W, Patton PR, et al. Outcome of kidney transplants in patients known to be flow cytometry crossmatch positive. Transplantation. 2001;71(8):1098-1102. doi:10.1097/00007890-200104270-00015
   CrossRef - PubMed
   
7. Lentine KL, Graff RJ, Xiao H, et al. Flow cytometry crossmatch before kidney transplantation in contemporary practice: target cell utilization, results patterns, and associated long-term graft survival. Clin Transpl. 2008:253-266.
   CrossRef - PubMed
   
8. Graff RJ, Xiao H, Schnitzler MA, et al. The role of positive flow cytometry crossmatch in late renal allograft loss. Hum Immunol. 2009;70(7):502-505. doi:10.1016/j.humimm.2009.04.002
   CrossRef - PubMed
   
9. Limaye S, O'Kelly P, Harmon G, et al. Improved graft survival in highly sensitized patients undergoing renal transplantation after the introduction of a clinically validated flow cytometry crossmatch. Transplantation. 2009;87(7):1052-1056. doi:10.1097/TP.0b013e31819d17b0
   CrossRef - PubMed
   
10. Norin AJ, Mondragon-Escorpizo MO, Brar A, et al. Poor kidney allograft survival associated with positive B cell - Only flow cytometry cross matches: a ten year single center study. Hum Immunol. 2013;74(10):1304-1312. doi:10.1016/j.humimm.2013.06.027
    CrossRef - PubMed
    
11. Pellegrino MA, Belvedere M, Pellegrino AG, Ferrone S. B peripheral lymphocytes express more HLA antigens than T peripheral lymphocytes. Transplantation. 1978;25(2):93-95.
    CrossRef - PubMed
    
12. Honger G, Krahenbuhl N, Dimeloe S, Stern M, Schaub S, Hess C. Inter-individual differences in HLA expression can impact the CDC crossmatch. Tissue Antigens. 2015;85(4):260-266. doi:10.1111/tan.12537
    CrossRef - PubMed
    
13. Daar AS, Fuggle SV, Fabre JW, Ting A, Morris PJ. The detailed distribution of MHC Class II antigens in normal human organs. Transplantation. 1984;38(3):293-298. doi:10.1097/00007890-198409000-00019
    CrossRef - PubMed
    
14. Good DJ, Zhang A, Kemesky J, Stopczynski N. Previously believed to be nonsensical crossmatch results, explained by ANTI-HLA-C antibodies (P030). Hum Immunol. 2017;78:75.
    CrossRef - PubMed
    
15. Inlow J, Steller P, Eltayeb A, et al. Case study: positive T-cell flow crossmatch (TFXM)/negative B-cell flow crossmatch (BFXM) can be mediated by a Class I antibody (220-P). Hum Immunol. 2011;1(72):S154.
    CrossRef - PubMed
    
16. Sullivan HC, Liwski RS, Bray RA, Gebel HM. The road to HLA antibody evaluation: do not rely on MFI. Am J Transplant. 2017;17(6):1455-1461. doi:10.1111/ajt.14229
    CrossRef - PubMed
    
17. Higgins R, Hathaway M, Lowe D, et al. Blood levels of donor-specific human leukocyte antigen antibodies after renal transplantation: resolution of rejection in the presence of circulating donor-specific antibody. Transplantation. 2007;84(7):876-884. doi:10.1097/01.tp.0000284729.39137.6e
    CrossRef - PubMed
    
18. Putheti P, Sharma VK. Determining channel shift cutoff in T-cell and B-cell flow cytometry crossmatch. Exp Clin Transplant. 2020. doi:10.6002/ect.2020.0021
    CrossRef - PubMed
    
19. Kotb M, Russell WC, Hathaway DK, Gaber LW, Gaber AO. The use of positive B cell flow cytometry crossmatch in predicting rejection among renal transplant recipients. Clin Transplant. 1999;13(1 Pt 2):83-89. doi:10.1034/j.1399-0012.1999.130104.x
    CrossRef - PubMed
    
20. Lobashevsky AL, Senkbeil RW, Shoaf J, et al. Specificity of preformed alloantibodies causing B cell positive flow crossmatch in renal transplantation. Clin Transplant. 2000;14(6):533-542. doi:10.1034/j.1399-0012.2000.140604.x
    CrossRef - PubMed
    
21. Karpinski M, Rush D, Jeffery J, et al. Flow cytometric crossmatching in primary renal transplant recipients with a negative anti-human globulin enhanced cytotoxicity crossmatch. J Am Soc Nephrol. 2001;12(12):2807-2814. doi:10.1681/ASN.V12122807
    CrossRef - PubMed
    
22. de Moraes P, Fagundes I, Cardone JM, et al. Accuracy of the median channel shift in the flow cytometry for predicting complement dependent cytotoxicity crossmatching in kidney transplant candidates. Transpl Immunol. 2019;52:27-31. doi:10.1016/j.trim.2018.10.005
    CrossRef - PubMed
    
23. Adebiyi OO, Gralla J, Klem P, et al. Clinical significance of pretransplant donor-specific antibodies in the setting of negative cell-based flow cytometry crossmatching in kidney transplant recipients. Am J Transplant. 2016;16(12):3458-3467. doi:10.1111/ajt.13848
    CrossRef - PubMed
    
24. Badders JL, Jones JA, Jeresano ME, Schillinger KP, Jackson AM. Variable HLA expression on deceased donor lymphocytes: Not all crossmatches are created equal. Hum Immunol. 2015;76(11):795-800. doi:10.1016/j.humimm.2015.09.029
    CrossRef - PubMed
    
25. Vaidya S, Cooper TY, Avandsalehi J, et al. Improved flow cytometric detection of HLA alloantibodies using pronase: potential implications in renal transplantation. Transplantation. 2001;71(3):422-428. doi:10.1097/00007890-200102150-00015
    CrossRef - PubMed
    
26. Apithy MJ, Desoutter J, Gicquel A, et al. Pronase treatment improves flow cytometry crossmatching results. HLA. 2017;90(3):157-164. doi:10.1111/tan.13073
    CrossRef - PubMed
    
27. Szewczyk K, Barrios K, Magas D, et al. Flow cytometry crossmatch reactivity with pronase-treated T cells induced by non-HLA autoantibodies in human immunodeficiency virus-infected patients. Hum Immunol. 2016;77(6):449-455. doi:10.1016/j.humimm.2016.04.014
    CrossRef - PubMed
    
28. Hart JD, Lutz CT, Jennings CD, et al. Falsely incompatible B-cell flow cytometry crossmatch after pronase treatment: a case report. Transplant Proc. 2015;47(3):831-833. doi:10.1016/j.transproceed.2014.12.022
    CrossRef - PubMed
    
29. Hetrick SJ, Schillinger KP, Zachary AA, Jackson AM. Impact of pronase on flow cytometric crossmatch outcome. Hum Immunol. 2011;72(4):330-336. doi:10.1016/j.humimm.2011.01.005
    CrossRef - PubMed
    
30. Park H, Lim YM, Han BY, Hyun J, Song EY, Park MH. Frequent false-positive reactions in pronase-treated T-cell flow cytometric cross-match tests. Transplant Proc. 2012;44(1):87-90. doi:10.1016/j.transproceed.2011.12.048
    CrossRef - PubMed