Secondary Logo

Journal Logo

Red Blood Cell Transfusions and the Risk of Allosensitization in Patients Awaiting Primary Kidney Transplantation

Leffell, Mary S.1,4; Kim, Deborah2; Vega, Renato M.1; Zachary, Andrea A.1; Petersen, Jeffrey3; Hart, John M.1; Rossert, Jerome3; Bradbury, Brian D.2

doi: 10.1097/01.tp.0000437435.19980.8f
Clinical and Translational Research
Free
SDC

Background Most studies of HLA sensitization after red blood cell transfusion in transplant candidates were done before widespread use of leukoreduced blood and based on relatively insensitive, nonspecific antibody assays. We evaluated the effect of transfusion on the breadth and magnitude of HLA antibody formation using current, sensitive, HLA-specific immunoassays.

Methods Serial HLA antibody data were merged with transfusion data from the US Renal Data System for 1324 patients on the kidney transplant waitlist (2004–2010). Two study groups were identified: a matched cohort consisting of 89 patients who received transfusion and 251 patients who did not receive transfusion and a crossover cohort consisting of 69 patients. Changes in antibody levels and calculated panel-reactive antibody (CPRA) were compared using χ2 and Sign tests, respectively. Logistic regression was used to estimate the relative risk of antibody responses.

Results Among the matched cohort, 20% of those who received transfusion compared to 3% of those who did not receive transfusion exhibited an antibody response (P=0.001), whereas in the crossover cohort, 19% exhibited a response in those who received transfusion compared to 1% of those who did not receive transfusion (P=0.0001). Moreover, 26.3% of those who received transfusion had increased CPRA compared to 5.8% of those who did not receive transfusion . These effects were greater in women and blacks compared to men and whites, respectively. Importantly, patients who received transfusion were at an increased risk of a potentially crossmatch positive response (odds ratio=9.6, 95% confidence interval=3.0–30.7).

Conclusions Sensitization from transfusion can occur in up to 20% of transplant candidates, resulting in higher antibody levels and CPRA values that adversely impact access to transplantation. These results support transfusion avoidance whenever possible.

Supplemental digital content is available in the text.

1 Johns Hopkins University Immunogenetics Laboratory, Johns Hopkins University School of Medicine, Baltimore, MD.

2 Center for Observational Research, Amgen, Inc., Thousand Oaks, CA.

3 Clinical Development, Amgen, Inc., Thousand Oaks, CA.

4 Address correspondence to: Mary S. Leffell, Ph.D., John Hopkins University Immunogenetics Laboratory, The Johns Hopkins University School of Medicine, Baltimore, MD.

Support for this study was supplied by a contract from Amgen, Inc.

The opinions and conclusions reported in this investigation are those of the authors and do not represent the NIDDK or the USRDS.

The authors declare conflicts of interest.

E-mail: msl@jhmi.edu

M.S.L., D.K., A.A.Z., J.P., J.R., and B.D.B. participated in the research design, writing of the article, performance of the research, and analysis of data. R.M.V. participated in the research design, performance of the research, and analysis of data. J.M.H. participated in the research design and writing of the article.

Received 20 June 2013. Revision requested 21 August 2013.

Accepted 7 September 2013.

Supplemental digital content (SDC) is available for this article. Direct URL citations appear in the printed text, and links to the digital files are provided in the HTML text of this article on the journal’s Web site (www.transplantjournal.com).

Accepted December 3, 2013

In the dialysis population, anemia is a common and debilitating comorbidity affecting more than 90% of patients (1, 2) because of insufficient endogenous erythropoietin production (3). From 1989 through 2010, the use of erythropoiesis-stimulating agents (ESAs) for the management of anemia increased substantially and contributed to a marked decline in red blood cell (RBC) transfusions and the percentage of patients on the kidney transplant waitlist with elevated panel-reactive antibody (PRA) levels (2). In the mid-2000s, results from clinical trials in patients with chronic kidney disease showed that targeting higher hemoglobin concentrations with ESA treatment increased the risk of cardiovascular events and mortality, which ultimately led to revisions in ESA labeling that were implemented in 2011 (4, 5). As this new safety information became more widely known, the use of ESAs began to decrease and hemoglobin levels declined (6, 7).

In 2011, in addition to the revised labeling, there were important changes to anemia management in patients on dialysis, which occurred after the implementation of the end-stage renal disease (ESRD) prospective payment system (PPS) (8–10). The ESRD PPS bundles payment for separately billable dialysis services including injectable medications (e.g., ESAs, iron), laboratory orders. and the dialysis procedure (8). Early reports indicate that PPS and the revisions to ESA labeling contributed to large declines in ESAs and hemoglobin levels and to an increase in the use of RBC transfusions (2, 9, 10).

The role of RBC transfusion in the development of HLA antibodies and its potential impact on transplant access have been well recognized by the transplant and nephrology communities, although much of the supporting data are decades old (11, 12). In 1981, Opelz et al. (13) reported a 28% incidence of HLA antibodies in patients who had transfusion, with greater risk among multiparous females and patients who previously had transplantation. Sensitization declined in the 1990s subsequent to the introduction of ESAs and the growing use of leukodepleted products (14, 15). Given that sensitized patients wait for a kidney longer and, if transplanted, experience shortened graft survival, with the use of RBC transfusions rising among patients on dialysis, transfusion avoidance has again become a topic of discussion (16, 17).

Contemporary assays for the detection and semiquantification of HLA-specific antibodies are highly sensitive and specific and are a far superior method for evaluating the risk of HLA sensitization after transfusion than was available in older studies (18). Accordingly, we merged serial HLA antibody data from patients on dialysis awaiting primary transplant with US Renal Data System (USRDS) data that capture reliable data on RBC transfusion administration. We then evaluated the risk of apparent de novo or increased allosensitization after a single transfusion event.

Back to Top | Article Outline

RESULTS

The demographic and clinical characteristics of the overall waitlist population are shown in Table 1, and the derivation of the matched and crossover cohorts is shown in Figure 1. The patients included in both cohorts were generally comparable to the overall population. Patients who received and those who did not receive transfusion had a mean age of 56 years, 39% were female, 51% were white, and 46% were black, with a mean time on dialysis of 1 year. Diabetes was the primary cause of ESRD in 38% of patients, 1% had a history of receiving a transfusion, and the prevalence of most major cardiovascular comorbidities was less than 25%. The crossover cohort was slightly younger (53 years), on average, and had a longer time on dialysis (∼2 years).

TABLE 1

TABLE 1

FIGURE 1

FIGURE 1

Back to Top | Article Outline

Changes in Antibody Strength

At the index time point, adjusted HLA class I and class II antibody values were not different between those who received and did not receive transfusion (class I, P=0.96; class II, P=0.38). The proportion of patients with any change in class I and/or II antibody levels was substantively higher for those who received transfusion compared to those who did not receive transfusion (20.4% vs. 2.4%, P<0.001) (Fig. 2). Substantial increases in antibody strength, i.e., potentially crossmatch positive increases, were observed among 13.5% of those who received transfusion compared to 1.6% of those who did not receive transfusion. In the crossover cohort, the proportion of patients with any response was substantively higher in the group who received transfusion than in the group who did not receive transfusion (20.3% vs. 1.4%, P<0.001). Substantial increases in antibody strength were seen among 11.6% of those who received transfusion compared to none of those who did not receive transfusion in the crossover cohort (Fig. 2). The results were nearly identical when the analysis was limited to patients with additional specificity testing performed by the phenotype and/or single-antigen assays (see Figure S1, SDC, http://links.lww.com/TP/A896), when stratified by whether patients experienced any proinflammatory events (see Figure S2, SDC, http://links.lww.com/TP/A896) and by time between the transfusion event and the follow-up HLA antibody measurement (see Figure S3, SDC, http://links.lww.com/TP/A896). Aside from sex, the stratified analyses mirrored the overall population. Among females, 44.7% experienced an increase in antibody strength compared to 6.1% of males (see Figure S4, SDC, http://links.lww.com/TP/A896).

FIGURE 2

FIGURE 2

In the regression analyses, patients who received transfusion were at an elevated relative risk of having any HLA antibody response at the lower thresholds (odds ratio [OR]=10.4, 95% confidence interval [CI]=4.0–27.1). Similar estimates were observed for a potentially crossmatch positive antibody at the higher thresholds (OR =9.6, 95% CI=3.0–30.7) as shown in the Forrest plot in Figure 3 (estimates also provided in Table S1; see SDC, http://links.lww.com/TP/A896). The OR estimates remained unchanged after case-mix adjustment and were consistent across all subgroups where ORs could be estimated. The OR estimates for any antibody response were numerically much higher for black (OR=14.5, 95% CI=3.0–70.4) than for white patients (OR=8.5, 95% CI=2.5–29.4). The OR for any antibody response among females (OR=12.4, 95% CI=1.4–35.3) was numerically higher than for the overall population; no male patients who did not receive transfusion showed any antibody response so an OR could not be estimated. Although antibody levels declined gradually over time, analysis of posttransfusion samples for 20 patients with, on average, 5 additional measurements (standard deviation=4.1, interquartile range [IQR]=1–8) showed antibody persistence over the following year (see Figure S5, SDC, http://links.lww.com/TP/A896).

FIGURE 3

FIGURE 3

Back to Top | Article Outline

Changes in Antibody Breadth

The proportion of patients with an increase in changes in antibody breadth (CPRA) was greater among those who received transfusion than among the matched patients who did not receive transfusion (26.3% vs. 5.8%) and during the transfusion period (compared to the control period [38% vs. 2%]) in the crossover cohort (Fig. 4). Among patients who received transfusion, 15.2% (matched cohort) and 22% (crossover) experienced an increase in CPRA greater than 20 points; approximately 6% in both cohorts had a 50-point change or higher in CPRA. Increases in CPRA greater than 90 were observed among 4.7% of all who received transfusion (data not shown). Female patients who received transfusion, patients with some degree of sensitization before transfusions (i.e., non-zero CPRA at baseline), were more susceptible to increases in CPRA. Greater increases in CPRA were observed with increased time between the transfusion event and the follow-up HLA antibody measurement (see Figures S6–S8, SDC, http://links.lww.com/TP/A896).

FIGURE 4

FIGURE 4

Back to Top | Article Outline

DISCUSSION

We evaluated the risk of transfusion-induced allosensitization among patients who were receiving chronic dialysis and awaiting their first transplant and found that nearly 14% of patients who received transfusion developed a biologically relevant increase in antibody strength compared to less than 2% of matched controls. Using a broader definition of antibody response at the lower assay thresholds, 20% of patients who received transfusion compared to less than 2% of patients who did not receive transfusion exhibited an antibody response. These results were nearly identical in our crossover population. These findings corresponded to a nearly 10-fold increased relative risk of broad sensitization and a 32-point mean increase in CPRA. These effects were more pronounced among females and blacks and were not explained by proinflammatory events such as surgeries or infections. There was a modest decline in antibody strength for both class I and II antibodies over the subsequent year among those patients with serial posttransfusion samples. These results provide strong evidence using contemporary antibody detection methods in the era of leukoreduced blood for a causal relationship between RBC transfusions and the development of allosensitization.

While it was recognized as early as 1954 that blood transfusions induce antibodies reactive with leukocytes (19), most studies were conducted with relatively insensitive and nonspecific cell-based assays (14, 20, 21) and results were equivocal, with some reporting significant antibody formation after a transfusion while others did not (22–24). Given the declining use of ESAs to treat anemia in patients receiving dialysis and the concomitant increase in transfusions, we sought to quantify the effect of transfusion on sensitization to HLA antigens using contemporary antibody assays. Our study is one of three to our knowledge that have examined transfusion-related sensitization using today’s antibody assays. In 2012, Balasubramaniam et al. (25) reported a 16.7% sensitization rate tested by single-antigen bead assays after receipt of leukodepleted blood transfusions among men awaiting primary kidney transplant, and this rate was four-fold higher than among patients who did not receive transfusion. More recently, Yabu et al. (26) observed a 20% sensitization rate for patients on dialysis awaiting primary transplant who received a single RBC transfusion relative to matched and self-controls. Our study used a similar design but included a demographically and regionally distinct population that contained a substantially higher percentage of black patients with results that essentially replicate the finding of a 20% sensitization rate and additionally provide estimates suggesting a nearly 10-fold increased RR of antibody response after a single RBC transfusion.

In our analyses, we used phenotype panels on the Luminex platform to evaluate changes in antibody strength as we have found that the MFI values from these assays yield superior correlations with actual crossmatch results than do single-antigen panels (27). There are many reasons why MFI values from single-antigen assays are frequently difficult to interpret when taken out of context: the values vary appreciably among the different antibody specificities; there is considerable variability between runs because of the high sensitivity of the assay; non-HLA antibodies can react with denatured antigens on the test beads, which has been shown to affect as much as 60% of some antigens; and the high sensitivity of the single-antigen assays has made it difficult to establish a threshold above which one can be sure that the reactivity is due to HLA-specific antibody (28, 29). A recent multicenter study in which all participating centers rigidly adhered to the same technique using test results from the same lots found that MFI values varied according to manufacturer, bead type, kit, and lot. While ROC analyses showed excellent consistency in antibody assignments between manufacturers (area under the curve>0.9), global normalization reduced MFI variation to levels near 20% (30). This 20% variability under very strict conditions substantiates our concern with using single-antigen data in this present study. We established a correlation for changes in antibody levels between the pooled antigen and phenotype tests (discussed in the Methods section) and determined that changes at the higher pooled antigen thresholds correlate with antibodies consistent with or approaching positive crossmatch levels. Recognizing that many centers use only single-antigen assays, we note that comparable single-antigen results yielded higher MFI values and varied according to different antibody specificities.

Like Yabu et al. (26), we also found that changes in HLA antibody levels after a transfusion corresponded to significant increases in CPRA, with nearly 40% of patients who received transfusion exhibiting a positive change and half were 20% or greater, translating to a sizable reduction in the potential number of available donor organs. Also notable was the absence of any change in CPRA among the comparator groups who did not receive transfusion, particularly during the self-control period in the crossover cohort. This effect on CPRA highlights the clinical and economic consequences of allosensitization because sensitized patients receive substantially fewer offers for a kidney translating into much longer wait times; have greater likelihood of dying while waiting; and, if transplanted, have reduced graft survival (2, 13, 31–33). Our findings also show that patients with some degree of sensitization are more likely to experience antibody increases after transfusion. Because 30% or more kidney transplant waitlist candidates have PRA/CPRAs greater than 20% (34), the clinical and economic consequences of transfusion may be even greater for these candidates.

Our results corroborate previous findings regarding sensitization within demographic subpopulations. Data from the OPTN/SRTR have shown that sensitization is more common among female and black patients (35), and when transfused, their antibodies increase both in level and breadth (26, 36, 37). We found that sensitization after transfusion was substantially higher in female and black patients. This racial disparity has implications considering the recent evidence of increased transfusion particularly among black patients after implementation of the ESRD PPS (9).

Recent evidence also suggests that proinflammatory events (e.g., surgeries, infections) can provoke an antibody response (38–40). In the current analysis, a large proportion of the patients experienced a proinflammatory event proximate to the transfusion event, raising the possibility that these events might have confounded the observed antibody responses. However, our results remained unchanged when those patients and their matched controls were eliminated from the analysis. In addition, the issue of whether leukoreduction reduces the risk of sensitization remains an area of controversy. Our study, which was conducted from 2004 onward, provides indirect evidence that leukoreduction does not mitigate the risk of transfusion-associated sensitization. Our findings are consistent with those reported by Karpinski et al. (41) who showed that, even after a Canadian province uniformly adopted leukoreduction, there was no decrease in the level of allosensitization after RBC transfusion among kidney transplant candidates. Finally, given the available serial antibody measurements, we found that the antibody response persisted up to 1 year later.

The importance of minimizing the use of transfusion in patients with kidney disease, particularly potential transplant candidates, has long been recognized by the nephrology community and has been reinforced by recent updates to clinical practice guidelines for managing anemia of chronic kidney disease (42). The significant decline in transfusion among patients on dialysis over the past two decades coincident with the introduction and widespread adoption of ESAs has meaningfully affected sensitization in this population (2). However, the safety concerns raised by the high-hemoglobin target randomized controlled trials, the changes in ESA labeling, and the implementation of the ESRD PPS have all contributed to downward shifts in the use of ESAs (6, 7, 9, 43). This change in the management of anemia has resulted in falling hemoglobin levels and an increase in the use of transfusion (44). In this context, our findings reinforce the importance of weighing the benefits and risks of alternative therapies like ESAs when deciding on the course of anemia treatment for potential transplant candidates.

This study should be evaluated in light of the following limitations. Although our study population was from a relatively large single-center waitlist, it represents only a fraction of the entire waitlisted population and, therefore, may not generalize to all waitlist candidates. CPRA data were not available for all patients because only selected patient sera received further antibody specificity testing. Although we replicated the incidence of antibody responses limited to those with available CPRA data, we could not fully assess increases in antibody breadth in our entire study population. Lastly, we obtained transfusion data from Medicare claims, which are expected to have high specificity (low false positives) but only moderate sensitivity (modest number of false negatives). Hence, we may have misclassified some patients who did not receive transfusion; this would, under expectation, underestimate the true excess risk of sensitization after transfusion.

In conclusion, our findings indicate a causal relationship between RBC transfusions and clinically relevant HLA antibody development. The observed changes in HLA antibodies translated into a clinically meaningful increase in CPRA consistent with a ∼25% reduction in the pool of available donor organs. Importantly, the sensitizing effects of transfusion may differentially affect female and black patients, further limiting their access to transplantation. These findings should reinforce the importance of transfusion avoidance when clinically possible for patients on the transplant waitlist.

Back to Top | Article Outline

MATERIALS AND METHODS

Data Sources

Antibody data from Immunogenetics Laboratory at the Johns Hopkins University Comprehensive Transplant Center were merged with data from the USRDS. Sera are received monthly for antibody testing, with mean compliance rates of 65% to 80% among patients active on the renal waitlist. USRDS data capture demographic information, comorbid conditions, medication utilization, and encounters with the health care system including hospitalizations, surgeries, and general dialysis care. Deidentified patient antibody data were merged with the USRDS records, and subsequent data analyses were conducted after approval by the Johns Hopkins University Institutional Review Board and the USRDS.

Back to Top | Article Outline

Study Population

From an initial sample of 2324 patients, 1342 adult patients were included who 1) were candidates for a primary transplant on the Johns Hopkins renal waitlist (2004–2010), 2) were receiving chronic dialysis, and 3) had at least two measurements for HLA antibodies separated by a mean of 30 days (median=30 days, interquartile range [IQR]=25–42 days). We limited our analysis to patient data from 2004 to 2010, a period during which most of transfused blood was leukoreduced (45). Of these, 110 received one or more RBC transfusions between two consecutive HLA antibody measurements. Each two consecutive HLA antibody measurements were considered a pair (e.g., measurements 1 and 2, 2 and 3, etc.). We identified all patients with one or more pairs with an intervening transfusion event; pairs without any intervening transfusion events were eligible for selection into the unexposed group.

Back to Top | Article Outline

Matched and Crossover Cohorts

Two cohorts of patients were identified: a matched cohort and a crossover cohort. Of the 110 patients who received transfusion, 89 were matched to 251 patients who did not receive transfusion; 69 were included in the crossover cohort (Fig. 1). The matched cohort consisted of all patients with a pair of HLA antibody measurements with an intervening transfusion event matched to individuals who had a pair of antibody measurements but without an intervening transfusion event. Each patient who received transfusion was individually matched to up to four patients who did not receive transfusion based on age (±2 years), sex, race (white, black, other), history of transfusion (yes/no), time on dialysis (±6 months), and calendar time (year of index HLA antibody measurement). The distribution of our matching ratio was 1:4 (45%), 1:3 (15%), 1:2 (18%), and 1:1 (22%). Twenty-one patients who received transfusion could not be matched owing to their lengthy/brief time on dialysis, older/younger age, and/or race. The first antibody measurement in any pair was considered the index (pre) and the second was considered the follow-up (post). For patients who received transfusion, the index was the antibody measurement immediately preceding the transfusion event, and for the patients who did not receive transfusion, the index was the first available antibody measurement as of the match date for the patient who received transfusion. The post sample was the HLA antibody with the maximum adjusted value during the follow-up period (median=77 days, IQR=32–166 days).

From the patients who received transfusion identified in the matched cohort, those who had at least one HLA antibody measurement preceding their index measurement were included in the crossover cohort. These patients had three consecutive HLA antibody measurements with a transfusion event occurring between the second and third (transfused period), but none occurring between the first and the second (control period).

Back to Top | Article Outline

Exposure

The exposure in this investigation was an RBC transfusion event occurring in the inpatient or outpatient setting. Transfusion events were identified using codes defined by USRDS and previous studies (44, 46) (for description of codes used, see SDC, http://links.lww.com/TP/A896). Transfusion events were defined as one or more transfusion administrations occurring on a single day; information on the units of blood transfused was not available.

Back to Top | Article Outline

HLA Antibody Measurements

Patient sera were tested for HLA-specific antibodies using solid-phase immunoassays against pooled HLA class I and class II antigens, single HLA class I and class II phenotypes, and single HLA antigens on the Luminex platform (Lifecodes Immuncor, San Diego, CA; One Lambda, Thermo Fisher Scientific, Canoga Park, CA). All patient sera were routinely tested using the pooled antigen assays, and the results were used as the primary measure of changes in antibody strength and breadth. Interpretation of HLA antibody reactivity was made using an adjusted value that normalizes the antibody measurement to control values. The adjusted, normalized value was calculated using software provided by the manufacturer. When substantial changes were noted in serial patient samples, sera were then tested with the single HLA phenotype and/or single HLA antigen assays. As these were retrospective analyses, only a subset of patient sera had additional antibody testing performed with the phenotype or single antigen assays at the relevant time points. These data were used to evaluate changes in antibody specificity and to confirm changes in antibody strength. The breadth of the antibody response was measured as a calculated panel-reactive antibody value (CPRA) that represents the percentage of donors having HLA antigens corresponding to patient antibodies (47).

Criteria were established for changes in antibody strength in the pooled antigen assays based on correlations of the adjusted, normalized values from the pooled antigen assays with normalized MFI values from the HLA phenotype assays. Briefly, 67 and 72 sera were tested for class I and class II specific antibodies, respectively, on both the pooled antigen and phenotype platforms. The phenotype assay was used to establish the relative antibody strength based on prior correlations with crossmatch results, with MFI values greater than 6000 and greater than 10,000 for class I specific antibodies corresponding to positive flow cytometric and cytotoxicity crossmatch tests, respectively (27). The phenotype assay was used to evaluate antibody strength because it provides good correlation with crossmatch reactivity. It should be noted that comparable MFI values from single antigen assays correlating with positive crossmatch tests are substantially higher than those obtained with phenotype panels and the single antigen assay results yielded much lower correlations with crossmatch results (27).

Two conservative threshold levels for assessing changes in antibody strength were defined from the above correlations. Changes in the adjusted values of 25 to 49 and 30 to 59, respectively, for class I and class II specific antibodies, were considered as moderate changes in antibody strength that would correspond to MFI values below a positive crossmatch level. This level of reactivity was sufficiently high to rule out any changes due to day-to-day variation in test results, which has been established in our laboratory as between 4.9% and 6.6% using a robotic liquid handling system (29). Because the moderate changes were well above the range that might be due to inherent test variability, such changes were indicative of some increase and/or broadening of patient sensitization. Changes in the adjusted values for class I and class II specific antibodies greater than 50 and greater than 60, respectively, were considered substantial and potentially “biologically significant” because changes in these ranges correlated with MFI values that approached or were consistent with positive crossmatch assays.

Back to Top | Article Outline

Other Measurements

Demographic characteristics including age, sex, race, and primary cause of ESRD were ascertained from the Medical Evidence form (CMS 2728). Comorbidities including heart failure, atherosclerotic heart disease, cancer, and cerebrovascular disease were identified from inpatient and outpatient Medicare claims in the 1-year period preceding the index HLA antibody measurement using established methods (48). Proinflammatory events, including major surgeries and infections, were identified before and after the index measurement.

Back to Top | Article Outline

Statistical Analysis

Summary statistics for continuous variables (mean, standard deviation, median) and categorical variables (count [n], percent [%], 25th/75th percentile) were used to characterize patients and changes in HLA antibody levels from the index to follow-up measurement. Baseline patient characteristics were compared using Pearson χ2 analysis and Student t-test, as appropriate. Changes in HLA class I and II antibody levels from index to the follow-up measurements were compared using Pearson χ2 analysis in the matched cohort and Wilcoxon sign rank test in the crossover cohort.

Logistic regression analysis was used to estimate ORs and 95% CIs for the effect of a single transfusion event on the development of an antibody response. The primary analysis evaluated “any” response defined as an increase at either the lower or the higher pooled antigen assay cutoff values for HLA class I or II antibodies evaluated together and separately. A secondary analysis evaluated increases only at the higher thresholds, which were more likely clinically relevant. Patients who did not receive transfusion served as the reference group in all analyses. Statistical adjustment was made for residual imbalances in patient characteristics between patients who received and did not receive transfusion. Because not all sera received additional specificity testing, analyses were replicated in the subset of patients for whom antibody specificities were determined for the CPRA calculation. Subgroup analyses were conducted within strata of age (<40, 40 to <50, 50 years), sex (male/female), and race/ethnicity (white/black/other) and according to whether patients experienced a proinflammatory event (yes/no) between the antibody measurements because proinflammatory events may stimulate an antibody response (38).

Changes in CPRA for patients who received and did not receive transfusion in the matched cohort and during the transfusion and control periods in the crossover cohort were compared using Pearson χ2 analysis and the Wilcoxon sign rank test, respectively. Analyses were conducted for the overall population and according to demographic characteristics, the presence of concurrent proinflammatory events, and baseline level of sensitizations (CPRA=0 vs. CPRA >0).

Lastly, among those patients who experienced an increase in HLA antibodies after a transfusion, we examined changes in HLA class I and II antibody levels for up to 1 year using generalized estimating equations (see more detailed description in the SDC, http://links.lww.com/TP/A896). For all analyses, P<0.05 was considered statistically different. All analyses were done using SAS, version 9.2 (Cary, NC), and all analyses were performed on limited data sets in compliance with the provisions of the Health Insurance Portability and Accountability Act.

Back to Top | Article Outline

ACKNOWLEDGMENTS

The authors thank Fangfei Chen and Corina Bigham for their assistance with the statistical programming for this study.

Back to Top | Article Outline

REFERENCES

1. Pisoni RL, Bragg-Gresham JL, Young EW, et al. Anemia management and outcomes from 12 countries in the Dialysis Outcomes and Practice Patterns Study (DOPPS). Am J Kidney Dis 2004; 44: 94.
2. United States Renal Data System. USRDS 2012 Annual Data Report: Atlas of End-Stage Renal Disease in the United States. Volume 2, Chapter 10. Bethesda, MD: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases; 2012.
3. Eschbach JW, Haley NR, Adamson JW. The anemia of chronic renal failure: pathophysiology and effects of recombinant erythropoietin. Contrib Nephrol 1990; 78: 24; discussion 37.
4. Pfeffer MA, Burdmann EA, Chen CY, et al. A trial of darbepoetin alfa in type 2 diabetes and chronic kidney disease. N Engl J Med 2009; 361: 2019.
5. Singh AK, Szczech L, Tang KL, et al. Correction of anemia with epoetin alfa in chronic kidney disease. N Engl J Med 2006; 355: 2085.
6. Regidor D, McClellan WM, Kewalramani R, et al. Changes in erythropoiesis-stimulating agent (ESA) dosing and haemoglobin levels in US non-dialysis chronic kidney disease patients between 2005 and 2009. Nephrol Dial Transplant 2011; 26: 1583.
7. Spiegel DM, Khan I, Krishnan M, et al. Changes in hemoglobin level distribution in US dialysis patients from June 2006 to November 2008. Am J Kidney Dis 2010; 55: 113.
8. Medicare Program: End-stage renal disease prospective payment system. Final rule. Federal Register 2010; 75: 49029.
9. Brunelli SM, Monda KL, Burkart JM, et al. Early trends from the Study to Evaluate the Prospective Payment System impact on small dialysis organizations (STEPPS). Am J Kidney Dis 2013; 61: 947.
10. Collins AJ. ESRD Payment Policy Changes: The New “Bundled” Dialysis Prospective Payment System (PPS) in the United States. 2012. Available from: http://www.usrds.org/2012/pres/USDialysisBundle_impact_NKFCM2012.pdf. Accessed June 21, 2012.
11. Locatelli F, Nissenson AR, Barrett BJ, et al. Clinical practice guidelines for anemia in chronic kidney disease: problems and solutions. A position statement from Kidney Disease: Improving Global Outcomes (KDIGO). Kidney Int 2008; 74: 1237.
12. KDOQI Clinical Practice Guideline and Clinical Practice Recommendations for anemia in chronic kidney disease: 2007 update of hemoglobin target. Am J Kidney Dis 2007; 50: 471.
13. Opelz G, Graver B, Mickey MR, et al. Lymphocytotoxic antibody responses to transfusions in potential kidney transplant recipients. Transplantation 1981; 32: 177.
14. Vella JP, O’Neill D, Atkins N, et al. Sensitization to human leukocyte antigen before and after the introduction of erythropoietin. Nephrol Dial Transplant 1998; 13: 2027.
15. Grimm PC, Sinai-Trieman L, Sekiya NM, et al. Effects of recombinant human erythropoietin on HLA sensitization and cell mediated immunity. Kidney Int 1990; 38: 12.
16. Scornik JC, Meier-Kriesche HU. Blood transfusions in organ transplant patients: mechanisms of sensitization and implications for prevention. Am J Transplant 2011; 11: 1785.
17. Macdougall IC, Obrador GT. How important is transfusion avoidance in 2013? Nephrol Dial Transplant 2013; 28: 1092.
18. Zachary AA, Leffell MS. Detecting and monitoring human leukocyte antigen-specific antibodies. Human immunology 2008; 69: 591.
19. Dausset J. [Indications for transfusion with concentrated or washed erythrocytes]. Concours Med 1954; 76: 2975.
20. Opelz G, Terasaki PI. Improvement of kidney-graft survival with increased numbers of blood transfusions. N Engl J Med 1978; 299: 799.
21. Lagaaij EL, Hennemann IP, Ruigrok M, et al. Effect of one-HLA-DR-antigen-matched and completely HLA-DR–mismatched blood transfusions on survival of heart and kidney allografts. N Engl J Med 1989; 321: 701.
22. United States Renal Data System. USRDS 2010 Annual Data Report: Atlas of Chronic Kidney Disease and End-Stage Renal Disease in the United States. Bethesda, MD: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases; 2010.
23. Opelz G, Vanrenterghem Y, Kirste G, et al. Prospective evaluation of pretransplant blood transfusions in cadaver kidney recipients. Transplantation 1997; 63: 964.
24. Vanrenterghem Y, Waer M, Roels L, et al. A prospective, randomized trial of pretransplant blood transfusions in cadaver kidney transplant candidates. Leuven Collaborative Group for Transplantation. Transplant Int 1994; 7: S243.
25. Balasubramaniam GS, Morris M, Gupta A, et al. Allosensitization rate of male patients awaiting first kidney grafts after leuko-depleted blood transfusion. Transplantation 2012; 93: 418.
26. Yabu JM, Anderson MW, Kim D, et al. Sensitization from transfusion in patients awaiting primary kidney transplant. Nephrol Dial Transplant 2013; 28: 2908.
Zachary AA, Sholander JT, Houp JA, et al. Using real data for a virtual crossmatch. Hum Immunol 2009; 70: 574.
El-Awar N, Terasaki PI, Nguyen A, et al. Epitopes of human leukocyte antigen class I antibodies found in sera of normal healthy males and cord blood. Hum Immunol 2009; 70: 844.
Zachary AA, Vega RM, Lucas DP, et al. HLA antibody detection and characterization by solid phase immunoassays: methods and pitfalls. Methods Mol Biol 2012; 882: 289.
Reed EF, Rao P, Zhang Z, et al. Comprehensive assessment and standardization of solid phase multiplex-bead arrays for the detection of antibodies to HLA. Am J Transplant 2013; 13: 1859.
Organ Procurement and Transplantation Network (OPTN). Proposal to Substantially Revise the National Kidney Allocation System. 2012. Available from: http://optn.transplant.hrsa.gov/PublicComment/pubcommentPropSub_311.pdf. Accessed June 17, 2013.
Network OPTN. Scientific Registry of Transplant Recipients: OPTN: data. 2011. Available from: http://optn.transplant.hrsa.gov/. Accessed September 2012.
Ibrahim HN, Skeans MA, Li Q, et al. Blood transfusions in kidney transplant candidates are common and associated with adverse outcomes. Clin Transplant 2011; 25: 653.
Organ Procurement and Transplantation Network (OPTN) and Scientific Registry of Transplant Recipients (SRTR). OPTN/SRTR 2011 Annual Data Report. Rockville, MD: Department of Health and Human Services, Health Resources and Services Administration, Healthcare Systems Bureau, Division of Transplantation; 2012.
Organ Procurement and Transplantation Network (OPTN) and Scientific Registry of Transplant Recipients (SRTR). OPTN/SRTR 2010 Annual Data Report. Rockville, MD: Department of Health and Human Services, Health Resources and Services Administration, Healthcare Systems Bureau, Division of Transplantation; 2011.
Scornik JC, Ireland JE, Howard RJ, et al. Assessment of the risk for broad sensitization by blood transfusions. Transplantation 1984; 37: 249.
Triulzi DJ, Kleinman S, Kakaiya RM, et al. The effect of previous pregnancy and transfusion on HLA alloimmunization in blood donors: implications for a transfusion-related acute lung injury risk reduction strategy. Transfusion 2009; 49: 1825.
Locke JE, Zachary AA, Warren DS, et al. Proinflammatory events are associated with significant increases in breadth and strength of HLA-specific antibody. Am J Transplant 2009; 9: 2136.
Alberu J, Morales-Buenrostro LE, de Leo C, et al. A non-allogeneic stimulus triggers the production of de novo HLA antibodies in healthy adults. Transplant Immunol 2007; 18: 166.
Troidle L, Eisen T, Pacelli L, et al. Complications associated with the development of bacteremia with Staphylococcus aureus. Hemodial Int 2007; 11: 72.
Karpinski M, Pochinco D, Dembinski I, et al. Leukocyte reduction of red blood cell transfusions does not decrease allosensitization rates in potential kidney transplant candidates. J Am Soc Nephrol 2004; 15: 818.
Kliger AS, Foley RN, Goldfarb DS, et al. KDOQI US Commentary on the 2012 KDIGO Clinical Practice Guideline for Anemia in CKD. Am J Kidney Dis 2013; 62: 849.
United States Renal Data System. USRDS 2011 Annual Data Report: Atlas of Chronic Kidney Disease and End-Stage Renal Disease in the United States. Bethesda, MD: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases; 2011.
Gill KS, Muntner P, Lafayette RA, et al. Red blood cell transfusion use in patients with chronic kidney disease. Nephrol Dial Transplant 2013; 28: 1504.
Food and Drug Administration Update on: Leukocyte Reduction of Blood and Blood Components Public Workshop. 2005. Available from: http://www.fda.gov/downloads/biologicsbloodvaccines/newsevents/workshopsmeetingsconferences/transcriptsminutes/ucm054436.pdf, 2013.
Ibrahim HN, Ishani A, Foley RN, et al. Temporal trends in red blood transfusion among US dialysis patients, 1992–2005. Am J Kidney Dis 2008; 52: 1115.
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.
Hebert PL, Geiss LS, Tierney EF, et al. Identifying persons with diabetes using Medicare claims data. Am J Med Qual 1999; 14: 270.
Keywords:

Transfusion; Sensitization; HLA antibody; Calculated panel-reactive antibody; Kidney transplantation; Dialysis

Supplemental Digital Content

Back to Top | Article Outline
© 2014 by Lippincott Williams & Wilkins