Abstract
The association between HLA polymorphisms and PTLD was investigated in a case-control study, comparing 110 predominantly adult solid-organ transplant recipients who developed PTLD to 5601 who did not. Donor and recipient HLA were analyzed. We detected a significant association between recipient HLA-A26 and the development of PTLD (OR 2.74; P=0.0007). In Caucasian recipients, both recipient and donor HLA-A26 were independently associated with development of PTLD (recipient A26 OR 2.99; P=0.0004, donor A26 OR 2.81; P=0.002). Analysis of HLA-A and -B haplotypes revealed that recipient HLA-A26, B38 haplotype was strongly correlated with a higher incidence of EBV-positive PTLD (OR 3.99; p=0.001). The common ancestral haplotype HLA-A1, B8, DR3, when carried by the donor, was protective against PTLD (OR 0.41; p=0.05). Several other HLA specificities demonstrated associations with clinical and pathological characteristics as well as survival. These findings demonstrate the importance of HLA polymorphisms in modulating the risk for PTLD, and may be useful in risk stratification and development of monitoring and prophylaxis strategies.
Introduction
Solid-organ transplant recipients have a well-described increased risk of malignancy, particularly of lymphoid tumors (1–4). Post-transplant lymphoproliferative disorder (PTLD) represents a range of abnormal lymphoid proliferations, most commonly of B-cell origin, that occur in the context of transplant-related immunosuppression. The clinical severity of the condition ranges from a benign mononucleosis-like illness to aggressive monoclonal non-Hodgkin lymphoma (5, 6), which can be rapidly progressive and fatal.
Epstein-Barr virus (EBV) is associated with a number of malignancies including Burkitt lymphoma, Hodgkin lymphoma, and nasopharyngeal carcinoma (7, 8). EBV is detected in 70–90% of PTLD cases, where it induces uncontrolled proliferation of B-cells in the absence of properly functioning CD8+ cytotoxic T-lymphocytes (CTL).
The highly polymorphic Human Leukocyte Antigen (HLA) loci are associated with a number of autoimmune disorders as well as susceptibility to certain infectious diseases (9). HLA molecules are responsible for antigen processing and presentation to the immune system and therefore have the potential to modulate the adaptive immune response to different infectious pathogens as well as to predispose to immune dysregulation. Several reports suggested associations between HLA polymorphisms and the occurrence of PTLD and other lymphoid proliferations, but limitations of sample size and ethnic diversity limit the wide applicability of these findings (10–13). As the field of PTLD research moves to develop better methods of monitoring and pre-emptive therapy, the identification of genetic risk factors such as HLA polymorphisms becomes important. In this study, we used a large cohort of predominantly adult solid-organ transplant donor/recipient pairs and applied rigorous statistical methods to investigate the association between HLA polymorphisms and the incidence of PTLD, its clinical and pathological characteristics, and its outcome.
Materials and Methods
Patients
We performed a retrospective case-control association analysis using data collected in our center and by the United Network for Organ Sharing (UNOS). The Institutional Review Board at the University of Pennsylvania approved the study. Data provided by UNOS was obtained via the Organ Procurement and Transplantation Network (OPTN) and was accurate as of June 6, 2008.
The UNOS database was used to retrieve donor and recipient HLA data for all solid-organ transplants at the University of Pennsylvania from January 1986 to June 2008. For completeness and verification, HLA data were also collected from the University of Pennsylvania clinical immunology lab. In keeping with the standard HLA-typing for solid organ transplants, antigen-level typing was recorded for donors and recipients. For HLA class I, HLA-A and -B were analyzed. For HLA class II, HLA-DR was analyzed. Subject race was also recorded. HLA data were available for 68% and 76% of transplant recipients with and without PTLD respectively. If a patient underwent more than one transplant, data relevant to the transplant immediately preceding the development of PTLD were collected.
During the study period, 162 adult solid organ recipients were diagnosed with PTLD at our institution. We obtained data on HLA typing, race, clinical and pathological characteristics, and outcome. The WHO classification of hematopoietic tumors was used for pathological classification of PTLD (6). Presence of CD20 surface marker was ascertained by immunohistochemistry or flow cytometry where appropriate. EBV positivity was defined as either a positive Epstein Barr-encoded RNA (EBER) in-situ hybridization or a positive latent membrane protein (LMP) stain in accordance with WHO guidelines (6). Clinical stage at diagnosis was defined using the Ann Arbor staging criteria (14). Tumor responses to therapy were graded according to the Response Evaluation Criteria in Solid Tumors (RECIST) guidelines (15). Survival was calculated from the time of PTLD diagnosis until death from any cause or last follow-up.
Statistical Analysis
Fisher's exact test was used to assess the association of each HLA specificity with the occurrence of PTLD. The analysis was conducted for the aggregate population and for the Caucasian subset to diminish the diversity in HLA specificities and increase the power of the study. The association of HLA specificities with clinical and pathological parameters was tested using logistic regression. These parameters included pathological classification at diagnosis according to WHO criteria, CD20 positivity, EBV positivity, stage at diagnosis, graft involvement, extra-nodal disease, time from transplant to PTLD diagnosis and response to initial treatment. The correlation between HLA specificities and survival was tested with the Cox proportional hazards model. Where indicated, we adjusted for multiplicity of tests using the Bonferroni method within each HLA locus separately. A similar approach was used to analyze the correlation between common HLA haplotypes and the occurrence of PTLD. The analysis was conducted in SAS Release 9.1.3 (SAS Institute, Cary, NC).
Results
The study design is summarized in Figure 1. According to the UNOS database, 5929 patients underwent solid organ transplantation at the University of Pennsylvania from January 1986 to June 2008 (kidney 46.2%, liver 26.8%, heart 12.9%, lung 10.1%, other 4%). From that group, we excluded patients who were diagnosed with PTLD, and patients for whom no HLA data were available. This resulted in a cohort of 5601 consecutive transplant recipients without PTLD, of whom 4084 were Caucasian. During the same study period, 162 solid organ transplant recipients were diagnosed with PTLD in our center. HLA data were available for 110 of these patients, including 93 Caucasians. In a minority of patients, HLA data were incomplete; a list of available HLA data for each locus appears in Supplementary Table 1.
Figure 1.
Study design.
The characteristics of 110 evaluable PTLD patients and Caucasian subset are summarized in Table 1. The majority of patients were Caucasian (84.5%) and the male to female ratio was approximately 2:1. Heart, lung, kidney and kidney/pancreas recipients were represented in this cohort. Liver recipients with PTLD were unlikely to have sufficient recipient HLA data for analysis but donor HLA data were usually available through UNOS. Most of our subjects were adult transplant recipients but 5 patients were transplanted at age 17 and later diagnosed with PTLD as adults. One cardiac recipient was transplanted at 14 and diagnosed with PTLD at 15. Patients presented with polymorphic (39.8%), monomorphic (49.1%) or other less common subtypes of PTLD. EBV-negative disease was present in 21.8%.
Table 1.
PTLD Patient characteristics
| Race | All (n=110) | Caucasian (n=93) |
|---|---|---|
| Caucasian | 84.5% | |
| African American | 11.8% | |
| Asian | 0.9% | |
| Hispanic | 0.9% | |
| Other | 1.9% | |
|
| ||
| Sex | ||
|
| ||
| Male | 63.0% | 63.7% |
| Female | 37.0% | 36.3% |
|
| ||
| Organ-type | ||
|
| ||
| Heart | 18.5% | 17.6% |
| Lung | 30.6% | 30.4% |
| Kidney | 41.7% | 34.3% |
| Kidney/Pancreas | 7.4% | 7.8% |
| Liver | 1.8% | 9.8% |
|
| ||
| Mean age at transplant (range) | 44.6 (14–76) | 45.5 (14–70) |
|
| ||
| Mean age at PTLD diagnosis (range) | 49 (15–76) | 50 (15–74) |
|
| ||
| Mean time ± SD in days from transplant to diagnosis (range) | 1399 ± 1621 (6–8402) | 1355 ± 1585 (6–6771) |
| Median | 714 | 694 |
|
| ||
| Pathology - WHO classification | ||
|
| ||
| Benign PTLD | 1.9% | 2.0% |
| Polymorphic | 39.8% | 40.2% |
| Monomorphic | 49.1% | 49.0% |
| Plasmacytoma-like | 2.8% | 3.9% |
| Not otherwise specified | 6.4% | 4.9% |
|
| ||
| CD20 (+) | ||
|
| ||
| Yes | 53.7% | 52.0% |
| No | 12.0% | 13.7% |
| Unknown | 34.3% | 34.3% |
|
| ||
| EBV (+) | ||
|
| ||
| Yes | 58.2% | 54.9% |
| No | 21.8% | 20.6% |
| Unknown | 20% | 24.5% |
|
| ||
| Stage (Ann Arbor Staging) | ||
|
| ||
| I | 40.7% | 38.2% |
| II | 18.5% | 18.6% |
| III | 10.2% | 13.7% |
| IV | 27.8% | 27.5% |
| Unknown | 2.8% | 2.0% |
Association between Expression of HLA-A26 and Development of PTLD
The significant associations between recipient HLA specificities and the development of PTLD are summarized in Table 2A. An association between the expression of HLA-A26 in the recipient and the development of PTLD was observed in both the aggregate (OR 2.74, 95% CI [1.62–4.63], P=0.0007) and Caucasian recipient populations (OR 2.99, 95% CI [1.71–5.22], P=0.0004). These associations remained significant after adjustment for multiple testing using the Bonferroni method (all patients, P=0.015; Caucasians, P=0.008). The frequencies of HLA-A26 carriers among organ transplant recipients were 8.51% and 3.29% in patients with and without PTLD, respectively. The association was more prominent among Caucasian recipients, where HLA-A26 carriers comprised 17.58% and 6.66% of patients with and without PTLD, respectively.
Table 2.
Associations between expression of HLA specificities and the development of PTLD. A. Recipient HLA. B. Donor HLA. P-values represent unadjusted Fisher's exact test. Associations with significance at the 0.05 value are presented. P values that remained significant after Bonferroni correction are highlighted in bold and the adjusted P value is shown in parentheses. NA= Not applicable (not represented in this group). Odds ratios not calculated where the contingency table contained zeros but the direction of the association is presented.
| Table 2A. Recipient HLA Associations | ||||
|---|---|---|---|---|
| All Patients | Caucasian Patients | |||
| HLA Specificity | Odds Ratio (95% CI) | P value | Odds Ratio (95% CI) | P value |
| A26 | 2.74 (1.62, 4.63) | 0.0007 (0.015) | 2.99 (1.71, 5.22) | 0.0004 (0.008) |
| A25 | 2.26 (1.04, 4.90) | 0.044 | 2.078 (0.94, 4.59) | 0.05 |
| Bw6 | >1 | 0.024 | >1 | 0.029 |
| B65 | <1 | 0.052 | <1 | 0.05 |
| DR12 | <1 | 0.037 | <1 | 0.174 |
| Table 2B. Donor HLA Associations | ||||
|---|---|---|---|---|
| All Patients | Caucasian Patients | |||
| HLA Specificity | Odds Ratio (95% CI) | P value | Odds Ratio (95% CI) | P value |
| A26 | 2.17 (1.20, 3.94) | 0.017 | 2.81 (1.54, 5.13) | 0.002 (0.042) |
| B38 | 2.874 (1.54, 5.36) | 0.003 | 3.113 (1.58, 6.12) | 0.003 |
| B8 | 0.509 (0.27, 0.97) | 0.035 | 0.481 (0.25, 0.93) | 0.025 |
| DR3 * | 0.565 (0.3, 0.98) | 0.043 | 0.527 (0.27, 0.95) | 0.033 |
| DR6 | 2.766 (1.48, 5.16) | 0.004 | NA | NA |
| DR5 | 4.092 (1.47, 11.41) | 0.02 | NA | NA |
| DR9 | 2.51 (1.16, 5.42) | 0.028 | 2.786 (1.19, 6.53) | 0.028 |
DR3 represents a combined analysis of DR17 & DR18.
The significant associations between donor HLA specificities and the development of PTLD are summarized in Table 2B. An association between the expression of HLA-A26 in the donor and the development of PTLD in the recipient was observed in the Caucasian population (OR 2.81, 95% CI [1.54–5.13], P=0.002, adjusted P=0.042). There was a similar trend observed in the aggregate population that did not maintain significance after Bonferroni correction. Among Caucasian transplant recipients, 14% of those who developed PTLD and 5.46% of those who did not develop PTLD received an organ from a donor carrying the HLA-A26 specificity. The racial characteristics of donors were unknown.
Among 16 HLA-A26+ recipients who developed PTLD, only two donor-recipient pairs (both kidney transplants) were matched at the A26 specificity, implying that both donor and recipient HLA-A26 independently confer an increased risk for development of PTLD. The increased risk for PTLD among HLA-A26 carriers was not limited to a single allograft type. HLA-A26+ recipients who developed PTLD consisted of heart, lung, kidney, and kidney/pancreas recipients in similar ratios to our entire PTLD cohort. HLA data were not available for the majority of liver recipients.
PTLD patients who carried the HLA-A26 specificity (n=16) were likely to present with PTLD in the first post-transplant year (63% as opposed to 42% in non-A26 carriers, p=0.17). They presented in various disease stages (54% Stage I-II, 46% Stage III-IV) and were mostly EBV positive (69% positive, 25% negative, 6% unknown). 62% had polymorphic disease while the remaining had monomorphic disease (38%). These and other characteristics, including baseline immunosuppressive regimen and serostatus of CMV, HBV and HCV, did not show any significant differences in comparison with HLA-A26-negative PTLD patients. PTLD patients who received an organ from a donor carrying the HLA-A26 specificity (n=13) again showed a trend towards early PTLD within the first post-transplant year (69% as opposed to 41% in non-A26 donors; p=0.07). Otherwise they exhibited typical patterns of disease stage (69% stage I–II, 31% stage III–IV), EBV positivity (62% positive, 23% negative, 15% unknown) and histology (38% polymorphic, 56% monomorphic, 6% unclassified).
Recipient Haplotype HLA-A26, B38 predisposes to PTLD
We investigated the impact of combined expression of HLA-A and -B specificities on the occurrence of PTLD in order to estimate the relative risk in carriers of different haplotypes that contain HLA-A26. A significant association between recipient haplotype HLA-A26, B38 and the development of PTLD was observed in Caucasians (OR 3.99, 95% CI [1.94–8.21], P=0.001). This association maintained significance after Bonferroni correction (adjusted P=0.043) and was unique to recipients and not observed among donors, likely due to the extensive correction for multiple testing required for haplotype analysis. HLA-A26, B38 was common in the Caucasian patient population. 9.9% (9/91) of Caucasian PTLD patients and 2.7% (82/3060) of Caucasian recipients without PTLD carried this haplotype. Notably, all HLA-A26, B38 patients with PTLD who had a known EBV status had EBV-positive PTLD (77.8% vs. 58.5% in non-HLA-A26, B38 patients; p=0.18). Other characteristics of the HLA-A26, B38 population were not significantly different from non-HLA-26, B38 recipients with PTLD (Table 3).
Table 3.
Characteristics of PTLD patients with and without the HLA A26, B38 haplotype
| Caucasian HLA A26, B38 recipients | Caucasian non-HLAA26, B38 recipients | |
|---|---|---|
| N | 9 | 82 |
|
| ||
| Sex | ||
| Male | 55.6% | 64.6% |
| Female | 44.4% | 35.4% |
|
| ||
| Organ-type | ||
| Heart | 22.2% | 19.5% |
| Lung | 11.1% | 36.6% |
| Kidney | 55.6% | 34.1% |
| Kidney/Pancreas | 11.1% | 8.5% |
| Liver | 0% | 1.2% |
|
| ||
| Mean age at transplant (yrs) | 38.7 (17–63) | 45.4 (14–70) |
|
| ||
| Mean age at PTLD diagnosis (yrs) | 42 (9–69) | 50 (15–74) |
|
| ||
| Mean time from transplant to diagnosis (days) | 1404 (6–4985) | 1392 (11–6771) |
|
| ||
| WHO class | ||
| Benign PTLD | 0% | 2.4% |
| Polymorphic | 33.3% | 41.5% |
| Monomorphic | 66.7% | 46.3% |
| Plasmacytoma-like | 0% | 3.7% |
| Unknown | 0% | 6.1% |
|
| ||
| EBV (+) PTLD | ||
| Yes | 77.8% | 58.5% |
| No | 0% | 24.4% |
| Unknown | 22.2% | 15.9% |
|
| ||
| Stage | ||
| I–II | 77.8% | 57.3% |
| III–IV | 22.2% | 41.5% |
| Unknown | 0% | 1.2% |
|
| ||
| Previous rejection episodes – | ||
| Yes | 66.7% | 47.6% |
| No | 22.2% | 39% |
| Unknown | 11.1% | 13.4% |
|
| ||
| Immunosuppressive regimen at diagnosis: | ||
| Tacrolim us-based | 33.3% | 30.5% |
| Cyclosporine-based | 55.6% | 56.1% |
| No CNI or Unknown | 11.1% | 13.4% |
|
| ||
| Azathioprine at diagnosis | 55.6% | 56.1% |
|
| ||
| MMF at diagnosis | 33.3% | 26.8% |
|
| ||
| Steroids at diagnosis | 88.9% | 82.9% |
|
| ||
| CMV serostatus – Positive | 44.4% | 36.6% |
| Negative | 11.1% | 54.9% |
| Unknown | 44.5% | 8.5% |
|
| ||
| HCV - Positive | 0% | 1.2% |
| Negative | 44.4% | 47.6% |
| Unknown | 55.6% | 51.2% |
|
| ||
| HBV – Positive | 0% | 0% |
| Negative | 77.8% | 62.2% |
| Unknown | 22.2% | 37.8% |
We examined common HLA-A, B, DR haplotypes in the context of HLA-A26, B38. PTLD risk was not unique to a specific combination, but the strongest associations were with HLA-A26, B38, DR14 (OR 16.9, p=0.01) and HLA-A26, B38, DR11 (OR 11.5, p=0.0036).
Protective effect of donor common ancestral haplotype (A1, B8, DR3) against PTLD
As displayed in Table 2B, analysis of individual HLA loci identified associations between the expression of donor HLA-B8 and HLA-DR3 and a reduced risk for PTLD. Donor HLA-A1 showed a similar trend, which almost reached statistical significance (OR 0.64; 95% CI [0.39, 1.04], p=0.073). This led us to investigate the common haplotype HLA-A1, B8, DR3, also known as the European ancestral haplotype. A comparison between organ transplant recipients with and without PTLD revealed that donors expressing the HLA-A1, B8, DR3 haplotype produced a significantly lower risk of PTLD (OR 0.41; 95% CI [0.13, 0.99], p=0.05). A similar association was observed when looking at the HLA-B8, DR3 haplotype alone (OR 0.41; 95% CI [0.22, 0.73], p=0.001). These associations were unique to donors and were not identified in organ recipients who carried this haplotype.
Association of Clinical and Pathological Parameters with PTLD
We explored possible correlations between HLA specificities and a number of clinical and pathological characteristics of PTLD (Table 4). When recipient HLA specificities were analyzed, HLA-A1 had a positive association with an earlier stage at diagnosis and HLA-A24 was positively associated with early PTLD (<1 year from transplant to diagnosis). Donor HLA specificities were similarly studied, revealing an association between donor HLA-A2 and polymorphic PTLD (vs. monomorphic PTLD). HLA-B57 was associated with an increased risk for graft involvement and HLA-DR15 with early PTLD. We report unadjusted p-values for these associations but they may represent trends that should be explored further. We did not identify any correlation between HLA specificities and response to reduction of immunosuppression or rituximab.
Table 4.
Associations between HLA specificities and PTLD characteristics
| Recipient/Donor | OR (95% CI) | P-value | ||
|---|---|---|---|---|
| Stage at diagnosis | A1 | Recipient | 0.31 (0.12, 0.77) | 0.013 |
| Monomorphic/Polymorphic PTLD | A2 | Donor | 0.35 (0.17, 0.73) | 0.006 |
| Time from transplant to PTLD | A24 | Recipient | 0.31 (0.12, 0.81) | 0.013 |
| DR15 | Donor | 0.28 (0.1, 0.78) | 0.02 | |
| Graft involvement | B57 | Donor | 11 (1.25, 96.9) | 0.016 |
| HR (95% CI) | ||||
| Survival | B44 | Recipient | 1.93 (1.09, 3.41) | 0.021 |
| Survival | A1 | Donor | 0.43 (0.19, 0.94)* | 0.035 |
This association was significant only among Caucasian recipients.
A survival analysis for individual HLA specificities was conducted using the Cox proportional hazards model (Table 4; Figure 2). Recipient HLA B-44 was associated with shorter survival (HR 1.93, 95% CI [1.09, 3.41]; unadjusted p-value=0.021). In a subgroup analysis of Caucasian patients, donor HLA-A1 was associated with improved survival (HR 0.43, 95% CI [0.19, 0.94]; unadjusted p-value=0.035). These associations did not maintain significance at the 5% level after Bonferroni correction.
Figure 2.
Survival analysis of PTLD patients. A. Recipient HLA-B44 is associated with shortened survival and B. donor HLA-A1 (in Caucasians) is associated with improved survival after the development of PTLD. P-values represent an unadjusted log-rank test.
Discussion
EBV-related malignancies such as PTLD and EBV-associated Hodgkin lymphoma are thought to arise when the immune system no longer effectively suppresses the latent viral infection allowing virally infected cells to proliferate (13, 16). Pharmacologic immunosuppression is therefore the cause for PTLD, but not all immunosuppressed patients develop this complication; genetic and other acquired factors predispose patients to PTLD.
It has been hypothesized that particular HLA specificities at donor or recipient HLA loci may affect the success of T-cell surveillance for EBV and thereby influence a transplant recipient's predisposition to PTLD. This theory derives in part from the knowledge that EBV-specific CTLs target EBV epitopes in an HLA-restricted manner (17, 18). Class I and class II HLA polymorphisms have been associated with other virus-associated cancers including HPV-driven cervical cancer, HIV-associated HHV-8-driven Kaposi's sarcoma, and EBV-associated Hodgkin lymphoma (13, 19–32).
We found that HLA-A26 in both the transplant donor and recipient predisposes to the development of PTLD. We showed that the HLA-A26, B38 haplotype is mainly responsible for this predisposition, at least among Caucasian recipients. Among PTLD patients, the predisposing haplotype did not significantly correlate with a specific immunosuppressive regimen, viral serologies or previous rejection episodes, implying that it is an independent risk factor. Almost a fifth of our Caucasian PTLD patients carried the HLA-A26 antigen, which is normally found in only 3% of European Americans, 1.4% of African Americans and 2.9% of people of Hispanic background (33). The increased incidence of HLA-A26 cannot be explained by a predisposition for HLA-A26-associated diseases that are treated with organ transplantation since HLA-A26 was prevalent across allograft types without predilection to certain disease conditions. In fact, the incidence of HLA-A26 in our transplant recipients without PTLD was similar to its incidence in the general population. The exception is liver transplant recipients for whom recipient HLA data are not routinely available. This population is under-represented in this study and our conclusions may not be generalizable to it. The HLA-A26, B38 haplotype has been reported in a familial case of EBV-associated Hodgkin lymphoma (34), which supports the susceptibility of carriers of this haplotype to EBV-driven B-cell proliferation. Furthermore, a recent report using genome-wide association methods indicated that HLA-A26 itself is a primary susceptibility allele involved in the development of Behçet disease, another immune-centered disease that has an association with a Herpes virus (35).
One could conjecture that individuals possessing the HLA-A26 specificity or a linked allele are less able to present EBV-derived peptides, a deficiency exacerbated by iatrogenic immunosuppression or additional genetic hits. This notion is supported by the fact that among our HLA-A26, B38 carriers with a known EBV status, only cases of EBV-positive PTLD were identified. A larger cohort will be needed to establish a specific association between this haplotype and EBV-positive PTLD.
Our second finding establishes a protective effect of the ancestral haplotype (HLA-A1, B8, DR3) against PTLD; only 5/90 patients (5.5%) developed PTLD from donors carrying this haplotype as opposed to its 10.7% prevalence in non-PTLD donors. This common haplotype is associated with autoimmune disorders and is thought to represent a genetically determined “hyperactive” immune state. We describe here, for the first time, a protective association between this haplotype and a viral-induced lymphoid proliferation. HLA-A1 has been previously associated with an increased risk for EBV-associated Hodgkin lymphoma (13), possibly due to an inefficient EBV-specific CTL response (36). Anti-EBV immunity, which confers protection from PTLD, is likely different from Hodgkin lymphoma because of a superimposed alloimmune response, driven by HLA mismatches, which may enhance the T-cell response to PTLD.
Surprisingly, the association of the ancestral haplotype was found only in donor HLA specificities. Similarly, donor HLA-A26 conferred an increased risk for PTLD independent of recipient match at the A locus. How does the donor HLA haplotype determine the risk for PTLD, a disease thought to derive from the delicate interplay between the recipient's B and T-cells? This interesting finding may be explained by the possible underestimation of donor-derived PTLD. In such cases, EBV-derived peptides presented by cell surface HLA molecules expressed by malignant B-cells of donor origin may enhance the recipient T-cell alloresponse against PTLD. Donor-derived PTLD is thought to be rare (<5%) based on small case series (37, 38), but suggested to be more common in other reports (39–41). In support of this hypothesis, donor HLA-A26 demonstrated a trend towards early onset of PTLD within the first year in our study (69% vs. 41%, p=0.07), which is typical for donor-derived PTLD. We did not have sufficient access to original biopsies or statistical power to ascertain whether these cases were truly donor derived.
It is not clear how HLA specificities at different loci work together to modulate the risk for PTLD. It is possible that the risk is based on the cumulative efficiency of peptide presentation by several HLA molecules as well as other HLA-linked genes. EBV can alter HLA class I peptide loading through TAP-1 and TAP-2 (42–43); polymorphisms in these genes may regulate the immune response towards EBV.
Another plausible mechanism for the association between HLA haplotypes and PTLD involves NK cell interactions with HLA molecules and escape mechanisms from CTL as previously reported for leukemias (44–45) and HIV infection (46). The importance of NK cells in controlling EBV-associated lymphoproliferation is undetermined but recently pointed out by studies of the SH2D1A gene in X-linked lymphoproliferative disorder (47).
Other groups have reported on associations between HLA specificities and PTLD. A group from Germany reported that HLA-B18 and B21 were positively associated with PTLD while HLA-A3 and HLA-DR7 had a negative association (10). An Iranian study demonstrated a higher frequency of HLA-B22 in PTLD patients (11). Finally, a group from the Netherlands reported that mismatches at the HLA-B locus, but not at the HLA-A and -DR loci, were associated with PTLD (12). These studies included mainly kidney recipients and were limited by small sample sizes and lack of adjustment for multiple testing. Inference for the American population is also problematic due to demographic differences.
The findings of this study allow us to identify distinct groups of patients who have a three-fold increased risk (with HLA-A26) or a 39% risk reduction (with HLAA1, B8, DR3) for PTLD based on their immunologic make-up alone. The HLA-A26 specificity is a common antigen in all characterized US populations, with a prevalence of 1.4–3.9%, and the ancestral haplotype is the most common haplotype in Caucasians. These findings can be clinically useful as HLA specificities are readily available at the time of transplant and prevention of EBV-driven PTLD is feasible through enhanced viral load surveillance, minimization of immunosuppression when clinically appropriate and perhaps pre-emptive rituximab, which has been explored and reported (48–49). The optimal management of patients felt to be at high risk for the development of PTLD based on their immunologic background must be the focus of prospective investigation.
A number of statistical trends have been identified between HLA specificities and pathological characteristics, clinical characteristics, and outcome of PTLD, including associations with overall survival. These results, while unadjusted for multiple testing and requiring validation in a larger cohort, have the potential to be useful in predicting the outcome of PTLD.
Our study was designed to encompass an entire transplant population across 20 years of experience at a single center. Its main strength is by representing the largest North American study to look at HLA polymorphisms as a risk factor for PTLD. Still, certain limitations are inherent to this study – our data do not represent allele level typing; some of the data were derived from typing done prior to the implementation of molecular methodologies and do not include HLA-Cw and -DQ typing information. In addition, the identification of some HLA split specificities may be inaccurate or incomplete, in particular in non-Caucasian individuals. Our transplant population does not represent the entire HLA genetic diversity of the human population; and recipient HLA data were missing for most liver transplants. This study also does not account for the evolving trends in immunosuppressive regimens in the last 20 years. Data on these regimens as well as other established risk factors for PTLD, such as CMV and hepatitis serologies, are missing for the control group due to limitations of data retrieval from UNOS. The risk for PTLD varies with different immunosuppressive regimens; tacrolimus was reported to induce more cases of PTLD compared to cyclosporine (50–52), although these findings have been disputed by a Cochrane meta-analysis (53). The use of muromonab-CD3 (Orthoclone OKT3) and anti-thymocyte globulin (ATG) has also been associated with PTLD (54). Belatacept, a T-cell costimulation blocker, has been associated with PTLD of the CNS (55). Larger case control studies will be needed to account for all possible combinations of immunosuppression that modulate the risk for PTLD.
With these limitations in mind, our findings imply a set of protective and predisposing factors for the development of PTLD that are embedded in the genome and can be used for screening and risk stratification prior to organ transplantation as well as post-transplant surveillance. These findings can be used in the future in the planning of prophylactic and preemptive strategies for PTLD.
Supplementary Material
Acknowledgements
This work was supported in part by Health Resources and Services administration contract 234-2005-370011C. The content is the responsibility of the authors alone and does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.
Funding sources: National Institutes of Health grants CA16520 (S.V. and D.F.H.), CA117879 (D.L.P.) & HL069286 (E.A.S.). R.R. is a fellow of the Institute for Translational Medicine and Therapeutics, University of Pennsylvania.
Footnotes
Presented in part in the 2009 American Transplant Congress, Boston, MA.
Disclosure The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.
References
- 1.Murray JE, Wilson RE, Tilney NI, Merrill JP, Cooper WC, Birtch AG, et al. Five years' experience in renal transplantation with immunosuppressive drugs: survival, function, complications, and the role of lymphocyte depletion by thoracic duct fistula. Ann Surg. 1968;168(3):416–35. doi: 10.1097/00000658-196809000-00010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Penn I, Hammond W, Brettschneider L, Starzl TE. Malignant lymphomas in transplantation patients. Transplant Proc. 1969;1:106–112. [PMC free article] [PubMed] [Google Scholar]
- 3.Opelz G, Dohler B. Lymphomas after solid organ transplantation: a collaborative transplant study report. Am J Transplant. 2004;4(2):222–30. doi: 10.1046/j.1600-6143.2003.00325.x. [DOI] [PubMed] [Google Scholar]
- 4.Opelz G, Henderson R. Incidence of non-Hodgkin lymphoma in kidney and heart transplant recipients. Lancet. 1993;342:1514–16. doi: 10.1016/s0140-6736(05)80084-4. [DOI] [PubMed] [Google Scholar]
- 5.Nalesnik MA. Posttransplantation lymphoproliferative disorders (PTLD): Current perspectives. Semin Thorac Cardiovasc Surg. 1996;8:139–48. [PubMed] [Google Scholar]
- 6.Swerdlow SH, International Agency for Research on Cancer, World Health Organization . WHO classification of tumours of haematopoietic and lymphoid tissues. ed 4th International Agency for Research on Cancer; Lyon, France: 2008. [Google Scholar]
- 7.Epstein MA, Achong BG, Barr YM. Virus Particles in Cultured Lymphoblasts from Burkitt's Lymphoma. Lancet. 1964;1(7335):702–3. doi: 10.1016/s0140-6736(64)91524-7. [DOI] [PubMed] [Google Scholar]
- 8.Williams H, Crawford DH. Epstein-Barr virus: the impact of scientific advances on clinical practice. Blood. 2006;107(3):862–9. doi: 10.1182/blood-2005-07-2702. [DOI] [PubMed] [Google Scholar]
- 9.Klein J, Sato A. The HLA System—First of Two Parts. N Engl J Med. 2000;343(10):702–709. doi: 10.1056/NEJM200009073431006. [DOI] [PubMed] [Google Scholar]
- 10.Subklewe M, Marquis R, Choquet S, Leblond V, Garnier JL, Hetzer R, et al. Association of Human Leukocyte Antigen Haplotypes with Posttransplant Lymphoproliferative Disease After Solid Organ Transplantation. Transplantation. 2006;82(8):1093–1100. doi: 10.1097/01.tp.0000235889.05171.12. [DOI] [PubMed] [Google Scholar]
- 11.Pourfarziani V, Einollahi B, Taheri S, Nemati E, Nafar M, Kalantar E. Associations of Human Leukocyte Antigen (HLA) haplotypes with risk of developing lymphoproliferative disorders after renal transplantation. Ann Transplant. 2007;12(4):16–22. [PubMed] [Google Scholar]
- 12.Bakker NA, van Imhoff GW, Verschuuren EA, van Son WJ, van der Heide JJ, Lems SP. HLA Antigens and Post Renal Transplant Lymphoproliferative Disease: HLA-B Matching is Critical. Transplantation. 2005;80(5):595–599. doi: 10.1097/01.tp.0000173793.03228.bd. [DOI] [PubMed] [Google Scholar]
- 13.Hjalgrim H, Rostgaard K, Johnson PC, Lake A, Shield L, Little AM, et al. HLAA alleles and infectious mononucleosis suggest a critical role for cytotoxic T-cell response to EBV-related Hodgkin lymphoma. Proc Natl Acad Sci USA. 2010;107(14):6400–6405. doi: 10.1073/pnas.0915054107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Carbone PP, Kaplan HS, Musshoff K, Smithers DW, Tubiana M. Report of the Committee on Hodgkin's Disease Staging Classification. Cancer Res. 1971;31(11):1860–1. [PubMed] [Google Scholar]
- 15.Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent D, Ford R, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1) Eur J Cancer. 2009;45(2):228–47. doi: 10.1016/j.ejca.2008.10.026. [DOI] [PubMed] [Google Scholar]
- 16.Thorley-Lawson DA, Gross A. Persistence of the Epstein-Barr virus and the origins of associated lymphomas. N Engl J Med. 2004;350(13):1328–37. doi: 10.1056/NEJMra032015. [DOI] [PubMed] [Google Scholar]
- 17.Richinson AB, Moss DJ. Human cytotoxic T lymphocyte responses to Epstein-Barr virus infection. Annu Rev Immunol. 1997;15:405–431. doi: 10.1146/annurev.immunol.15.1.405. [DOI] [PubMed] [Google Scholar]
- 18.Subklewe M, Chahroudi A, Bickham K, Larsson M, Kurilla MG, Bhardwaj N, et al. Presentation of Epstein-Barr virus latency antigens to CD8 (+), interferon-gamma-secreting, T lymphocytes. Eur J Immunol. 1999;29(12):3995–4001. doi: 10.1002/(SICI)1521-4141(199912)29:12<3995::AID-IMMU3995>3.0.CO;2-E. [DOI] [PubMed] [Google Scholar]
- 19.Madeleine MM, Johnson LG, Smith AG, Hansen JA, Nisperos BB, Li S, et al. Comprehensive analysis of HLA-A, HLA-B, HLA-C, HLA-DRB1, and HLA-DQB1 loci and squamous cell cervical cancer risk. Cancer Res. 2008;68(9):3532–9. doi: 10.1158/0008-5472.CAN-07-6471. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Madeleine MM, Brumback B, Cushing-Haugen KL, Schwartz SM, Daling JR, Smith AG, et al. Human leukocyte antigen class II and cervical cancer risk: a population-based study. J. Infect Dis. 2002;186(11):1565–74. doi: 10.1086/345285. [DOI] [PubMed] [Google Scholar]
- 21.Wu Y, Liu B, Lin W, Xu Y, Li L, Zhang Y, et al. Human leukocyte antigen class II alleles and risk of cervical cancer in China. Hum Immunol. 2007;68(3):192–200. doi: 10.1016/j.humimm.2006.07.005. [DOI] [PubMed] [Google Scholar]
- 22.Sanjeevi CB, Hjelmström P, Hallmans G, Wiklund F, Lenner P, Angström T, et al. Different HLA-DR-DQ haplotypes are associated with cervical intraepithelial neoplasia among human papillomavirus type-16 seropositive and seronegative Swedish women. Int J Cancer. 1996;68(4):409–14. doi: 10.1002/(SICI)1097-0215(19961115)68:4<409::AID-IJC1>3.0.CO;2-4. [DOI] [PubMed] [Google Scholar]
- 23.Wang SS, Wheeler CM, Hildesheim A, Schiffman M, Herrero R, Bratti MC, et al. Human leukocyte antigen class I and II alleles and risk of cervical neoplasia: results from a population-based study in Costa Rica. J Infect Dis. 2001;184(10):1310–4. doi: 10.1086/324209. [DOI] [PubMed] [Google Scholar]
- 24.Mann DL, Murray C, O'Donnell M, Blattner WA, Goedert JJ. HLA antigen frequencies in HIV-1-related Kaposi's sarcoma. J Acquir Immune Defic Syndr. 1990;3(Suppl 1):S51–5. [PubMed] [Google Scholar]
- 25.Dorak MT, Yee LJ, Tang J, Shao W, Lobashevsky ES, Jacobson LP, et al. HLA-B, -DRB1/3/4/5, and -DQB1 gene polymorphisms in human immunodeficiency virus-related Kaposi's sarcoma. J Med Virol. 2005;76(3):302–10. doi: 10.1002/jmv.20361. [DOI] [PubMed] [Google Scholar]
- 26.Gayà A, Esteve A, Casabona J, McCarthy JJ, Martorell J, Schulz TF, et al. Amino acid residue at position 13 in HLA-DR beta chain plays a critical role in the development of Kaposi's sarcoma in AIDS patients. AIDS. 2004;18(2):199–204. doi: 10.1097/00002030-200401230-00008. [DOI] [PubMed] [Google Scholar]
- 27.Hildesheim A, Apple RJ, Chen CJ, Wang SS, Cheng YJ, Klitz W, et al. Association of HLA class I and II alleles and extended haplotypes with nasopharyngeal carcinoma in Taiwan. J Natl Cancer Inst. 2002;94(23):1780–9. doi: 10.1093/jnci/94.23.1780. [DOI] [PubMed] [Google Scholar]
- 28.Li X, Ghandri N, Piancatelli D, Adams S, Chen D, Robbins FM, et al. Associations between HLA class I alleles and the prevalence of nasopharyngeal carcinoma (NPC) among Tunisians. J Transl Med. 2007;5:22. doi: 10.1186/1479-5876-5-22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Dardari R, Khyatti M, Jouhadi H, Benider A, Ettayebi H, Kahlain A. Study of human leukocyte antigen class I phenotypes in Moroccan patients with nasopharyngeal carcinoma. Int J Cancer. 2001;92(2):294–7. doi: 10.1002/1097-0215(200102)9999:9999<::aid-ijc1177>3.0.co;2-9. [DOI] [PubMed] [Google Scholar]
- 30.David AL, Taylor GM, Gokhale D, Aplin JD, Seif MW, Tindall VR. HLADQB1*03 and cervical intraepithelial neoplasia type III. Lancet. 1992;340(8810):52. doi: 10.1016/0140-6736(92)92464-q. [DOI] [PubMed] [Google Scholar]
- 31.Vandenvelde C, De Foor M, van Beers D. HLA-DOB1*03 and cervical intraepithelial neoplasia grades I-III. Lancet. 1993;341(8842):442. doi: 10.1016/0140-6736(93)93044-2. [DOI] [PubMed] [Google Scholar]
- 32.Ioannidis JP, Skolnik PR, Chalmers TC, Lau J. Human leukocyte antigen associations of epidemic Kaposi's sarcoma. AIDS. 1995;9(6):649–651. doi: 10.1097/00002030-199506000-00019. [DOI] [PubMed] [Google Scholar]
- 33.Maiers M, Gragert L, Klitz W. High resolution HLA alleles and haplotypes in the US population. Hum Immunol. 2007;68:779–788. doi: 10.1016/j.humimm.2007.04.005. [DOI] [PubMed] [Google Scholar]
- 34.Kamper PM, Kjeldsen E, Clausen N, Bendix K, Hamilton-Dutoit S, d'Amore F. Epstein-Barr virus-associated familial Hodgkin lymphoma: paediatric onset in three of five siblings. Br J Haematol. 2005;129(5):615–7. doi: 10.1111/j.1365-2141.2005.05499.x. [DOI] [PubMed] [Google Scholar]
- 35.Meguro A, Inoko H, Ota M, Katsuyama Y, Oka A, Okada E, et al. Genetics of Behçet disease inside and outside the MHC. Ann Rheum Dis. 2010;69(4):747–54. doi: 10.1136/ard.2009.108571. [DOI] [PubMed] [Google Scholar]
- 36.Brennan RM, Burrows SR. A mechanism for the HLA-A*01-associated risk for EBV+ Hodgkin lymphoma and infectious mononucleosis. Blood. 2008;112(6):2589–90. doi: 10.1182/blood-2008-06-162883. [DOI] [PubMed] [Google Scholar]
- 37.Gulley ML, Swinnen LJ, Plaisance KT, Jr, Schnell C, Grogan TM, Schneider BG. Tumor origin and CD 20 expression in posttransplant lymphoproliferative disorder occurring in solid organ transplant recipients: implications for immune-based therapy. Transplantation. 2003;76(6):959–64. doi: 10.1097/01.TP.0000079832.00991.EE. [DOI] [PubMed] [Google Scholar]
- 38.Chadburn A, Suciu-Foca N, Cesarman E, Reed E, Michler RE, Knowles DM. Post-transplantation lymphoproliferative disorders arising in solid organ transplant recipients are usually of recipient origin. Am J Pathol. 1995;147(6):1862–70. [PMC free article] [PubMed] [Google Scholar]
- 39.Cherqui D, Duvoux C, Plassa F, Gaulard P, Julien M, Fagniez PL, et al. Lymphoproliferative disorder of donor origin in a liver transplant recipient: complete remission after drastic reduction of immunosuppression without graft loss. Transplantation. 1993;56(4):1023–6. [PubMed] [Google Scholar]
- 40.Spiro IJ, Yandell DW, Li C, Saini S, Ferry J, Powelson J, et al. Brief report: lymphoma of donor origin occurring in the porta hepatic of a transplanted liver. N Engl J Med. 1993;329(1):27–9. doi: 10.1056/NEJM199307013290105. [DOI] [PubMed] [Google Scholar]
- 41.Armes JE, Angus P, Southey MC, Battaglia SE, Ross BC, Jones RM, et al. Lymphoproliferative disease of donor origin arising in patients after orthotopic liver transplantation. Cancer. 1994;74(9):2436–41. doi: 10.1002/1097-0142(19941101)74:9<2436::aid-cncr2820740908>3.0.co;2-s. [DOI] [PubMed] [Google Scholar]
- 42.Horst D, van Leeuwen D, Croft NP, Garstka MA, Hislop AD, Kremmer E, et al. Specific targeting of the EBV lytic phase protein BNLF2a to the transporter associated with antigen processing results in impairment of HLA class I-restricted antigen presentation. J Immunol. 2009;182(4):2313–24. doi: 10.4049/jimmunol.0803218. [DOI] [PubMed] [Google Scholar]
- 43.Zeidler R, Eissner G, Meissner P, Uebel S, Tampé R, Lazis S, et al. Downregulation of TAP1 in B lymphocytes by cellular and Epstein-Barr virus-encoded interleukin-10. Blood. 1997;90(6):2390–7. [PubMed] [Google Scholar]
- 44.Demanet C, Mulder A, Deneys V, Worsham MJ, Maes P, Claas FH, et al. Down-regulation of HLA-A and HLA-Bw6, but not HLA-Bw4, allospecificities in leukemic cells: an escape mechanism from CTL and NK attack? Blood. 2004;103(8):3122–30. doi: 10.1182/blood-2003-07-2500. [DOI] [PubMed] [Google Scholar]
- 45.Middleton D, Diler AS, Meenagh A, Sleator C, Gourraud PA. Killer immunoglobulin-like receptors (KIR2DL2 and/or KIR2DS2) in presence of their ligand (HLA-C1 group) protect against chronic myeloid leukaemia. Tissue Antigens. 2009;73(6):553–60. doi: 10.1111/j.1399-0039.2009.01235.x. [DOI] [PubMed] [Google Scholar]
- 46.Flores-Villanueva PO, Yunis EJ, Delgado JC, Vittinghoff E, Buchbinder S, Leung JY, et al. Control of HIV-1 viremia and protection from AIDS are associated with HLA-Bw4 homozygosity. Proc Natl Acad Sci USA. 2001;98(9):5140–5. doi: 10.1073/pnas.071548198. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Hislop AD, Palendira U, Leese AM, Arkwright PD, Rohrlich PS, Tangye SG, et al. Impaired Epstein-Barr virus-specific CD8+ T cell function in X-linked lymphoproliferative disease is restricted to SLAM family positive B cell targets. Blood. 2010;116(17):3249–57. doi: 10.1182/blood-2009-09-238832. [DOI] [PubMed] [Google Scholar]
- 48.Blaes AH, Cao Q, Wagner JE, Young JA, Weisdorf DJ, Brunstein CG. Monitoring and preemptive rituximab therapy for Epstein-Barr virus reactivation after antithymocyte globulin containing nonmyeloablative conditioning for umbilical cord blood transplantation. Biol Blood Marrow Transplant. 2010;16:287–91. doi: 10.1016/j.bbmt.2009.10.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Ahmad I, Cau NV, Kwan J, Maaroufi Y, Meuleman N, Aoun M, et al. Preemptive management of Epstein-Barr virus reactivation after hematopoietic stem-cell transplantation. Transplantation. 2009;87:1240–45. doi: 10.1097/TP.0b013e31819f1c49. [DOI] [PubMed] [Google Scholar]
- 50.Cox KL, Lawrence-Miyasaki LS, Garcia-Kennedy R, Lennette ET, Martinez OM, Krams SM, et al. An increased incidence of Epstein-Barr virus infection and lymphoproliferative disorder in young children on FK506 after liver transplantation. Transplantation. 1995;59:524–9. [PubMed] [Google Scholar]
- 51.Sokal EM, Antunes H, Beguin C, Bodeus M, Wallemacq P, de Ville de Goyet J, et al. Early signs and risk factors for the increased incidence of Epstein-Barr virus-related posttransplant lymphoproliferative diseases in pediatric liver transplant recipients treated with tacrolimus. Transplantation. 1997;64:1438–42. doi: 10.1097/00007890-199711270-00011. [DOI] [PubMed] [Google Scholar]
- 52.Dharnidharka VR, Sullivan EK, Stablein DM, Tejani AH, Harmon WE. Risk factors for posttransplant lymphoproliferative disorder (PTLD) in pediatric kidney transplantation: a report of the North American Pediatric Renal Transplant Cooperative Study (NAPRTCS) Transplantation. 2001;71:1065–8. doi: 10.1097/00007890-200104270-00010. [DOI] [PubMed] [Google Scholar]
- 53.Webster AC, Woodroffe RC, Taylor RS, Chapman JR, Craig JC. Tacrolimus versus ciclosporin as primary immunosuppression for kidney transplant recipients: meta-analysis and meta-regression of randomised trial data. BMJ. 2005;331(7520):810. doi: 10.1136/bmj.38569.471007.AE. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Swinnen LJ, Costanzo-Nordin MR, Fisher SG, O'Sullivan EJ, Johnson MR, Heroux AL, et al. Increased incidence of lymphoproliferative disorder after immunosuppression with the monoclonal antibody OKT3 in cardiac-transplant recipients. N Engl J Med. 1990;323(25):1723–8. doi: 10.1056/NEJM199012203232502. [DOI] [PubMed] [Google Scholar]
- 55.Durrbach A, Pestana JM, Pearson T, Vincenti F, Garcia VD, Campistol J, et al. A phase III study of belatacept vs cyclosporine in kidney transplants from extended criteria donors (benefit-ext study) Am J Transplant. 2010;10:571–581. doi: 10.1111/j.1600-6143.2010.03016.x. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.