Journal Logo

Original Clinical Science—General

A Virtual Crossmatch-based Strategy Facilitates Sharing of Deceased Donor Kidneys for Highly Sensitized Recipients

Roll, Garrett R. MD1; Webber, Allison B. MD2; Gae, David H. PhD3; Laszik, Zoltan MD4; Tavakol, Mehdi MD1; Mayen, Luis BS5; Cunniffe, Kelly BS3; Syed, Shareef MD1; Hirose, Ryutaro MD1; Freise, Chris MD1; Feng, Sandy MD, PhD1; Roberts, John P. MD1; Ascher, Nancy L. MD, PhD1; Stock, Peter G. MD, PhD1; Rajalingam, Raja PhD3

Author Information
doi: 10.1097/TP.0000000000002924
  • Free



The current kidney allocation system in the United States exposes highly sensitized candidates to deceased donor offers through regional and national sharing to increase the opportunity of finding a compatible kidney.1 As a result, most kidneys allocated to highly sensitized recipients are shipped between regions. Traditionally, a physical crossmatch (PXM) using donor lymphocytes and patient serum has been performed before transplant to establish compatibility for kidney transplantation and prevent hyperacute and acute rejections.2,3 Frequently, the donor blood required for a PXM is sent at the same time as the kidney, and therefore the test cannot be started until the donor's kidney arrives, adding to cold ischemia time. A positive PXM that occurs after the kidney arrives at the recipient’s transplant center requires subsequent reallocation, adding to already lengthy preservation times. Moreover, in addition to these challenging logistical issues, it has recently been reported that 19.2% of kidneys transported between regions to a highly sensitized candidate are eventually transplanted into an unintended recipient, most commonly due to a positive PXM.4

We adopted the virtual crossmatch (VXM) as our predominant pretransplant crossmatch in late 2014. This study explores the impact of our VXM strategy on the ability to reduce the frequency with which a kidney imported to our center for a specific waitlisted recipient is transplanted into an unintended recipient.


Study Population

All kidneys imported to UCSF between December 2014 and September 2017 were included in this analysis. We retrospectively reviewed the final disposition of these imported kidneys and investigated any reason a kidney was not transplanted into the intended recipient. The following data were collected for each recipient: level of pretransplant sensitization (ie, calculated panel reactive antibodies [CPRA]), type of pretransplant crossmatch (virtual or physical) and the results, final organ allocation, reasons for reallocation after kidney arrival, hyperacute rejection episodes, creatinine 1 month, 6 months, and 12 months after transplant, and all kidney biopsies done in the first year after transplant.

Unacceptable Antigen Criterion and Virtual Crossmatch Assessment

Serum samples are obtained every 3 months from candidates who are active on the United Network for Organ Sharing waitlist and estimated to receive an organ offer in 1 year. The estimation of which candidates are likely to draw offers within the next year is done based on the average number of points required to draw organ offers in each blood group in our region, and the specific donor programs such as hepatitis C positive, high kidney donor profile index, and increased Public Health Service risk, to which the waitlisted patient has consented.

Anti-HLA antibodies were measured using the Luminex-based single antigen bead assay (One Lambda Inc., Canoga Park, CA).5 Serum samples are pre-treated with dithiothreitol to prevent aggregation of high titer antibodies to some specific beads (prozone effect) and increase the sensitivity for antigen–antibody binding. Antibody specificity is determined based on known cross-reactivity patterns. The median fluorescence intensity (MFI) is used as an arbitrary unit of antibody quantity. If multiple beads have allelic variants of the same antigen (eg, HLA-A*02:01, *02:03, *02:06—variants of HLA-A2 antigen), then the average MFI of all positive beads is used to quantify HLA-A2 antibody strength. LABXpress Pipettor (One Lambda), a high throughput liquid handling system to aspirate and dispense precise volumes into test wells of a 96-well reaction plate is used to minimize interassay variations. Targets of HLA antibodies identified during the quarterly screening are listed as unacceptable antigens in United Network for Organ Sharing if they are thought to increase the risk of graft rejection. The followings are criterion for listing unacceptable antigens: (1) HLA-A/B/C/DR/DR51/DR52/DR53/DQB1 allotypes are listed as unacceptable antigens if the patient displays antibodies with ≥2000 MFI reactivity against these allotypes; (2) Bw4 or Bw6 are listed as unacceptable antigens if the patient displayed antibodies to these epitopes at any MFI; (3) none of the DQA1, DPA1, and DPB1 allotypes and HLA-A/B/C/DR/DR51/DR52/DR53/DQB1 alleles are listed as unacceptable antigens despite targeted antibodies of any MFI to maximize donor offers. When a kidney is offered by the match run, a VXM is performed by the HLA laboratory director to determine the likelihood of a positive PXM. If the risk of a positive PXM is exceedingly low, we proceeded with transplant based on the VXM assessment only, which is possible about 85% of the time.

Flow Cytometer Crossmatch

If pretransplant PXM has not been performed, a retrospective physical flow cytometry-based crossmatch is always done after transplant even for candidates with a CPRA of 0%. A PXM is required before transplant in 5 situations: (1) presence of multiple DSAs at low levels (<2000 MFI) against the donor, with a low likelihood of getting a more compatible offer in the near future because of broad sensitization; (2) presence of strong allele-specific antibodies (≥5000 MFI) directed against the donor; (3) lack of an up-to-date quarterly antibody test available to perform an adequate VXM; (4) presence of only a single serum sample tested for HLA antibodies when an HLA-mismatched donor is offered; or (5) occurrence of a possible sensitizing event between the most recent antibody profile and the organ offer, such as a blood transfusion or an infection requiring treatment with antibiotics but not vaccine administration. Only in these 5 well-defined situations, or about 15% of organ offers, was a physical T- and B-cell crossmatch using flow cytometry performed on the admission serum sample (as well as the most current sample stored in the laboratory) before transplant.

Three-color flow cytometer crossmatch was performed using FACSCanto flow cytometer (Becton Dickinson, San Jose, CA), which detects the presence or absence of IgG antibodies directed toward donor lymphocyte-specific antigens. Donor lymphocytes are isolated from whole blood using EasySep kits, which yield highly purified lymphocytes. Approximately 0.25 × 106 lymphocytes are incubated with 30 µL of test or control serum for 20 minutes at room temperature followed by 4 washes with cold wash buffer (PBS containing 2% BSA and 0.1% NaN3). An optimized concentration of fluorescein Isothiocyanate (FITC)-conjugated, human-specific anti-IgG monoclonal antibody (clone G18-145, Becton Dickinson) are added and incubated for 20 minutes at 4°C followed by 2 washes with PBS wash buffer. FITC IgG antibodies allow detection of bound recipient antibodies. T and B lymphocytes are analyzed using allophycocyanin-conjugated anti-CD3 (clone SK7) and PE-conjugated anti-CD19 (clone 4G7) monoclonal antibodies (Becton Dickinson). Fluorescence intensity is acquired as logarithmic data and the ratio of the median fluorescent value for each test serum to that of the negative control serum is calculated. The FCXMs are considered positive if the median channel shift for test serum is >50 for T cells and >120 for B cells when compared with the negative control serum. To reduce background reactivity in a B-cell FCXM due to nonspecific immunoglobulin binding by Fc receptors and B-cell surface immunoglobulins, the lymphocytes were treated with pronase.

Management After Transplant

Induction and maintenance immune suppression was based on the level of sensitization and traditionally accepted risk factors for rejection and was not influenced by the type of pretransplant crossmatch. Generally, induction was lymphocyte depletion (Thymoglobulin; Genzyme, Cambridge, MA) for patients with a PRA >40% or in patients with multiple other risks for rejection, while Basiliximab (Simulect; Novartis Pharma AG, Basel, Switzerland) was used for unsensitized patients. A single dose of 70 g of intravenous immune globulin was given the day after transplant to patients with preformed DSA. Protocol biopsies were done at 6 months and again at 12 months if there was inflammation present on the index biopsy.


We imported 254 kidneys from 51 of 58 nonlocal organ procurement organizations. Of the 254 intended recipients, 52.8% were female, 36.3% were retransplant recipients (6.8% were third kidney transplant recipients, 29.5% were second kidney transplant recipients), 33.2% were Hispanic, 21.7% were Asian, and 16.6% were African American. Of these imported kidneys, 215 (84.6%) were transplanted based solely on a VXM. The VXM was used for recipients across all degrees of sensitization (Table 1). In particular, 91 of 215 recipients (42.3%) transplanted without a PXM had a CPRA of 100%.

Pretransplant crossmatch used for imported kidney transplantations

Thirty-nine of the 254 imported kidneys (15.4%) required a pretransplant PXM. The most common reason (20/39; 51.2%) a VXM alone was not adequate to move forward with transplantation was that the recipient did not have antibody testing within the previous 3 months. Of the 39 pretransplant PXMs, only 2 were positive: one was undergoing simultaneous liver-kidney transplant (SLKT) and had a persistently positive PXM after liver transplantation due to strong DQ7 DSA (MFI = 17 612); another patient had strong DP3 DSA (MFI = 22 828) and caused positive B-cell crossmatch by both complement-dependent cytotoxicity and flow cytometry methods. Of the 37 negative pretransplant PXMs, 4 had DSAs (Table 2).

Patients received transplant with preformed DSA

Among the retrospective crossmatches performed for those transplanted based solely on a VXM, only 2 were positive, and both of them were SLKT recipients. For SLKT candidates, we list only the targets of strong DRB1 and DQB1 antibodies with MFI >8500 as unacceptable antigens in the UNet, and therefore the positive crossmatch is expected and did not impact SLKT graft function. Moreover, 27 kidney recipients transplanted based solely on a VXM had preformed DSAs (Table 2). Over the study period, 8 kidneys were declined due to a positive VXM.

The final organ allocation did not change in 93% (236 of 254) of imported kidneys. Of the remaining 18 kidneys that were not transplanted into the intended recipient (Table 3), 9 were reallocated due to a newly identified medical condition in the recipient such as a new lung mass on the pre-operative chest radiograph and 5 were declined for organ quality after inspection on the back table or biopsy findings. Only 4 of 254 imported kidneys (1.6%) required reallocation and were transplanted into an unintended recipient because of a positive PXM or detection of new unacceptable DSA in the admission serum sample (Table 4). One of these 4 patients was undergoing SLKT and had a persistently positive PXM after liver transplantation, so the decision was made to reallocate the kidney rather than move forward with kidney transplantation. The second patient had strong DP3 DSA (MFI = 22 828) which caused positive B-cell crossmatch by both complement-dependent cytotoxicity and flow cytometry methods, and thus the kidney was reallocated. The third patient did not have an up to date serum sample at the time of the organ offer, so a new sample was drawn on admission which had 2 DSAs (A33 with MFI 1197, DQA1*02:01 with MFI 11,167), and thus kidney was reallocated due to multiple DSAs indicating high risk of rejection. The fourth patient developed a new Bw6 DSA in admit serum due to recent blood transfusions, and thus the kidney was reallocated.

Final allocation of 254 imported kidneys
Details of 4 candidates that necessitated reallocation as a result of HLA antibodies

No recipients experienced clinically suspected or biopsy-proven hyperacute rejection, and early acute rejection was very infrequent. Of the 215 recipients transplanted without a PXM, 1 patient experienced cellular rejection (CPRA 100% recipient with no pretransplant DSA), 1 patient experienced humoral rejection (CPRA 100% recipient with low-level pretransplant DSA to DR14 and Cw2), and 2 patients experienced borderline changes within the first month. Of the 39 recipients transplanted with a pretransplant PXM, 1 recipient experienced humoral rejection (CPRA 100% recipient without pretransplant DSA) in the first month. Graft function, as determined by creatinine at 1 month, 6 months, and 12 months after transplant, was not different between patients who had a pretransplant PXM versus VXM (Table 5). For cause biopsies and protocol biopsies done routinely at 6 months, and selectively at 12 months (Table 5), show that rejection rates were low in the first year. Three grafts were lost in the first month, all due to renal vein thrombosis. Pathology review of the explanted kidney did not identify cellular or humoral rejection in any of those cases.

Crossmatch and kidney transplant outcomes


The kidney allocation system in the United States is designed to broadly share kidneys by exposing highly sensitized patients with a CPRA of 99% or 100% to regional and national offers.6 Traditionally a PXM has been used before transplant to predict the risk of hyperacute or acute rejection by detecting preformed antibody. A PXM takes 3 or more hours to complete, requires donor blood to be shipped long distances, and prolongs cold ischemia time.7,8 The time required for a PXM has been shortened by a recently published protocol,9 but this shorter protocol continues to require a donor sample, which may only be available with the shipped kidney. While this shorter solid-phase assay reduces cold ischemia time, it does not reduce the likelihood of a positive result after the kidney has arrived at the recipient center. Frequently, when a positive crossmatch occurs after a kidney has been shipped a long distance, the kidney cannot be returned due to the prolonged preservation time. In this scenario, the kidney is often reallocated locally, which may actually advantage centers that accept kidneys with a high likelihood of having a positive PXM. In addition, several studies have shown that the presence of pretransplant DSAs, even at low levels, with a negative PXM, adversely affects long-term graft survival substantiating the notion that a VXM is the most accurate characterization of immunologic risk.10-13 Finally, there is a risk of a false-positive PXM, which is as high as 20%.14,15 For these same reasons, VXM is being used more commonly in heart and lung transplant, where the demand for short cold ischemia time prohibits late reallocation.

Using a VXM as the default pretransplant crossmatch has a number of advantages. It is not a laboratory test, but rather it is the interpretation of existing data that can predict the risk of hyperacute rejection and be performed in 5–30 minutes. The single antigen bead assay has greatly enhanced our ability to precisely identify the type and strength of HLA antibodies.16 With advances in the sensitivity and specificity of single antigen assays for HLA antibody testing combined with the knowledge of cross-reactive groups and outcomes of correlative studies with the strength of donor-specific HLA antibodies and flow crossmatch outcomes gave us confidence that our HLA laboratory could accurately and reliably predict the result of PXMs by virtual assessment in the majority of clinical scenarios. As a result, we were convinced that pretransplant PXMs lacked specificity, were extraneous to our process of safe donor and recipient matching, and were only required in a small number of well-defined clinical scenarios. With the application of broader sharing in the kidney allocation changes, we began to rely on a VXM in a vast majority of cases. A VXM takes into account the current and all the historic antibodies a particular recipient has against a specific donor and is therefore a more comprehensive assessment of immunologic compatibility based on the patient’s alloantibody profile compared with the donor’s histocompatibility antigens, whereas a PXM represents a single point in time.

There are a growing number of descriptions of VXM-based practices.15,17-23 These existing descriptions concentrate on patients of low immunologic risk and do not include a large proportion of highly sensitized patients. In fact, the need to confirm the efficacy of a VXM strategy in highly sensitized recipients was recently highlighted.15 The composition of our cohort is 62.9% of patients with a CPRA ≥ 80%, and 138 patients with a CPRA of ≥98%, of whom 86.2% were transplanted without a pretransplant PXM.

It is not known how many patients are transplanted nationally based on a VXM, but a recent survey of HLA laboratories suggests that an overwhelming majority of recipients still have a PXM before kidney transplantation in the United States.22 The variation in crossmatch practice has been amplified by the recent kidney allocation changes.24 Modeling suggests that a VXM is a more effective way to get highly sensitized candidates transplanted when compared with reliance on a PXM,25 and long-term outcomes in patients with a negative VXM and positive PXM have been equivalent to outcomes in patients transplanted with a negative PXM.15 The expanding use of VXM in heart and lung transplant, where prolonged cold ischemia time and the need to avoid hyperacute rejection are paramount underscores how a national VXM-based practice could enhance sharing of kidneys for highly sensitized patients.

A national allocation system based on the VXM is within reach. A centralized VXM-based allocation system would improve organ allocation efficiency, reduce cold ischemia time, and reduce cost. To raise all centers to the same level of expertise, specific MFI cutoffs could be suggested by a consensus conference of HLA experts based on the presumed risk of rejection from each antibody, and centers could have the option to raise and lower the MFI cutoffs for individual patients as desired. The fact that the VXM strategy has been slow to penetrate much of the community likely means that some are uncomfortable setting the avoids based the single antigen testing so the match run can act as the VXM, or interpreting the results of the VXM. To assist those centers, a centralized HLA expert could interpret testing and help set MFI cut offs for centers if they do not feel confident with this practice. A centralized VXM-based system would then reduce the experience gap between centers and level the field.

Our center’s waiting list has almost 5000 patients, which presents unique challenges but also magnifies the value of a VXM-based strategy. While 11% of our waitlist is broadly sensitized with a CPRA >85, approximately 70% of our waitlist candidates have no detectable HLA antibodies at all. For these patients, a VXM alone is clearly sufficient, and a PXM only risks a false-positive result. The unsensitized patient is a perfect place to start to incorporate a VXM-based strategy at the center or national level. But our goal has been to design a crossmatch strategy that would be accurate and efficient for all candidates and would not limit access to organ offers for highly sensitized patients. Our results show that over the study period, we have been able to transplant a large number of highly sensitized candidates using this VXM strategy, as 137 of 254 (54%) imported kidneys were for intended recipients with a CPRA of 99%–100%.

Our study has several weaknesses. Most importantly, we do not have a control group to directly compare the rate of transplant into an unintended recipient, preservation time, and rejection rate. The reallocation data for this analysis was supplied by our organ procurement organization, and they did not precisely track late reallocation before the study period. Moreover, our center moved to the VXM-based strategy as the kidney allocation changes that brought about broader sharing were implemented in 2014. As a result, we cannot provide a representative control group but can show a late reallocation rate of 1.6% that compares very favorably to the existing national rate of 19.2%, with infrequent rejection in a population of recipients of whom many were highly sensitized. Long-term allograft function and the risk of development of post-transplant DSA and rejection after VXM have been described elsewhere; the data presented here address the acute need for organ allocation efficiency improvement.

There are clearly areas for process improvement at our center, influenced by the size of our waiting list. Over half of the required pretransplant PXMs could have been avoided if the waitlisted candidates had up-to-date HLA antibody testing. In addition, 8 of the 17 cases of kidney reallocation after arrival occurred due to a new medical condition in the intended recipient, which could possibly be improved by more frequent contact with waitlisted patients.

In summary, sharing which requires transportation over long distances is possible, and late reallocations can be almost entirely avoided with a strategy that relies heavily on VXM, without limiting access for highly sensitized patients on the waitlist. A conversation about a systematic national VXM strategy is warranted to facilitate broad kidney sharing.



1. Formica RN Jr. A critical assessment on kidney allocation systems. Transplant Rev (Orlando). 2017; 31:61–67
2. Patel R, Terasaki PI. Significance of the positive crossmatch test in kidney transplantation. N Engl J Med. 1969; 280:735–739
3. O’Rourke RW, Osorio RW, Freise CE, et al. Flow cytometry crossmatching as a predictor of acute rejection in sensitized recipients of cadaveric renal transplants. Clin Transplant. 2000; 14:167–173
4. Paramesh AS, Neidlinger N, Salvatore M, et al. OPO strategies to prevent unintended use of kidneys exported for high PRA (>98% cPRA) recipients. Am J Transplant. 2017; 17:2139–2143
5. Pei R, Lee JH, Shih NJ, et al. Single human leukocyte antigen flow cytometry beads for accurate identification of human leukocyte antigen antibody specificities. Transplantation. 2003; 75:43–49
6. Hart A, Gustafson SK, Skeans MA, et al. OPTN/SRTR 2015 Annual Data Report: early effects of the new kidney allocation system. Am J Transplant. 2017; 17Suppl 1)543–564
7. Shrestha S, Bradbury L, Boal M, et al. Logistical factors influencing cold ischemia times in deceased donor kidney transplants. Transplantation. 2016; 100:422–428
8. Eby BC, Redfield RR, Ellis TM, et al. Virtual HLA crossmatching as a means to safely expedite transplantation of imported pancreata. Transplantation. 2016; 100:1103–1110
9. Liwski RS, Greenshields AL, Conrad DM, et al. Rapid optimized flow cytometric crossmatch (FCXM) assays: the Halifax and Halifaster protocols. Hum Immunol. 2018; 79:28–38
10. Tian H, Chen GH, Xu Y, et al. Impact of pre-transplant disease burden on the outcome of allogeneic hematopoietic stem cell transplant in refractory and relapsed acute myeloid leukemia: a single-center study. Leuk Lymphoma. 2015; 56:1353–1361
11. Schwaiger E, Eskandary F, Kozakowski N, et al. Deceased donor kidney transplantation across donor-specific antibody barriers: predictors of antibody-mediated rejection. Nephrol Dial Transplant. 2016; 31:1342–1351
12. 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; 16:3458–3467
13. Wu P, Jin J, Everly MJ, et al. Impact of alloantibody strength in crossmatch negative DSA positive kidney transplantation. Clin Biochem. 2013; 46:1389–1393
14. Sullivan HC, Dean CL, Liwski RS, et al. (F)utility of the physical crossmatch for living donor evaluations in the age of the virtual crossmatch. Hum Immunol. 2018; 79:711–715
15. Johnson CP, Schiller JJ, Zhu YR, et al. Renal transplantation with final allocation based on the virtual crossmatch. Am J Transplant. 2016; 16:1503–1515
16. Tait BD, Süsal C, Gebel HM, et al. Consensus guidelines on the testing and clinical management issues associated with HLA and non-HLA antibodies in transplantation. Transplantation. 2013; 95:19–47
17. Gebel HM, Bray RA. The evolution and clinical impact of human leukocyte antigen technology. Curr Opin Nephrol Hypertens. 2010; 19:598–602
18. Bielmann D, Hönger G, Lutz D, et al. Pretransplant risk assessment in renal allograft recipients using virtual crossmatching. Am J Transplant. 2007; 7:626–632
19. Baxter-Lowe LA, Cecka M, Kamoun M, et al. Center-defined unacceptable HLA antigens facilitate transplants for sensitized patients in a multi-center kidney exchange program. Am J Transplant. 2014; 14:1592–1598
20. Cecka JM, Kucheryavaya AY, Reinsmoen NL, et al. Calculated PRA: initial results show benefits for sensitized patients and a reduction in positive crossmatches. Am J Transplant. 2011; 11:719–724
21. Taylor CJ, Kosmoliaptsis V, Sharples LD, et al. Ten-year experience of selective omission of the pretransplant crossmatch test in deceased donor kidney transplantation. Transplantation. 2010; 89:185–193
22. Kamoun M, Phelan D, Noreen H, et al. HLA compatibility assessment and management of highly sensitized patients under the new kidney allocation system (KAS): a 2016 status report from twelve HLA laboratories across the U.S. Hum Immunol. 2017; 78:19–23
23. Ferrari P, Fidler S, Holdsworth R, et al. High transplant rates of highly sensitized recipients with virtual crossmatching in kidney paired donation. Transplantation. 2012; 94:744–749
24. Parsons RF, Locke JE, Redfield RR 3rd, et al. Kidney transplantation of highly sensitized recipients under the new kidney allocation system: a reflection from five different transplant centers across the United States. Hum Immunol. 2017; 78:30–36
25. Gebel HM, Kasiske BL, Gustafson SK, et al. Allocating deceased donor kidneys to candidates with high panel-reactive antibodies. Clin J Am Soc Nephrol. 2016; 11:505–511
Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.