Collaborators: Jerome Olagne and Sophie Caillard have equally contributed to this work. Marcellin L: Pathology Department, Parissiadis A: Etablissement Français du sang, Strasbourg, France. Thierry A, Goujon JM, Jollet I: Poitiers University Hospital, France. Ducloux D, Félix S, Chabot J: Besancon University Hospital, France. Essig M, Paraf F: Limoges University Hospital, France. Frimat L, Champigneulle J, Perrier P: Nancy University Hospital, France. Etienne I, François A: Rouen University Hospital, France. Legris T, Moal V, Daniel L: Marseille University Hospital, France. Toupance O, Diebold D, Tabarry T: Reims University Hospital, France. Moal MC, Doucet L: Brest University Hospital, France. Heng AE, Kemeny JL, Quainon F: Clermont-Ferrand University Hospital, France. Rivalan J, Rioux-Leclerc N: Rennes University Hospital, France. Rondeau E, Callard P: Tenon Hospital, Paris, France.
Post-transplant Lymphoproliferative Disorders: Determination of Donor/Recipient Origin in a Large Cohort of Kidney Recipients
Article first published online: 12 MAY 2011
©2011 The Authors Journal compilation©2011 The American Society of Transplantation and the American Society of Transplant Surgeons
American Journal of Transplantation
Volume 11, Issue 6, pages 1260–1269, June 2011
How to Cite
Olagne, J., Caillard, S., Gaub, M. P., Chenard, M. P. and Moulin, B. (2011), Post-transplant Lymphoproliferative Disorders: Determination of Donor/Recipient Origin in a Large Cohort of Kidney Recipients. American Journal of Transplantation, 11: 1260–1269. doi: 10.1111/j.1600-6143.2011.03544.x
- Issue published online: 6 JUN 2011
- Article first published online: 12 MAY 2011
- Received 27 October 2010, revised 06 march 2011 and accepted for publication 06 march 2011
- kidney transplantation;
- post-transplant lymphoproliferative disorders;
- tumoral cell origin
- Top of page
Although in previous studies most post-transplant lymphoproliferative disorders (PTLD) were reported to derive from recipient cells, some cases derived from donor lymphocytes have been reported. To provide a better description of the features and outcome of PTLD according to the origin of the lymphoma, we performed histologic and molecular studies of PTLD in kidney recipients. Forty-three specimens were analyzed by histochemistry, fluorescent hybridization of the Y chromosome and analysis of multiple short tandem repeat microsatellite loci. Sixteen tumors were shown to be of donor origin and 27 of recipient origin. Time to PTLD was shorter in donor-derived PTLDs (20±27 vs. 69±67 months, p = 0.013). Ten-year patient survival was similar among patients with recipient- and donor-derived PTLD, but when PTLD-related mortality was analyzed, there was a trend to better survival in patients with donor lymphomas. Among the 21 PTLDs localized in the allograft, 14 lymphomas were of donor origin and seven of recipient origin. No difference was found between the two groups. Our analysis of the origin of PTLDs in the largest cohort studied to date with a description of the clinical and histological characteristics of donor and recipient PTLDs should lead to a better understanding of lymphomagenesis.
post-transplant lymphoproliferative disorders
fluorescence in situ hybridization
Epstein Barr Virus
- Top of page
Post-transplant lymphoproliferative disorders (PTLDs) are a well-known complication after organ transplantation. Their incidence is relatively low, ranging from 1 to 10% depending on which organ is transplanted and on the population concerned (adults or children) (1,2). Some risk factors are well known including the Epstein–Barr virus (EBV), and potent immunosuppression (2–7). PTLD, typically of B-cell origin, displays some particularities compared to lymphomas in immunocompetent patients: implication of EBV in two-thirds of the cases, the frequency of extra-nodal sites, localization of PTLD in the graft itself, frequent localization in the brain and possible response to tapering of immunosuppression (8). The clinical features of PTLD vary depending on the time to occurrence and/or on the main localization of the lymphoma. From the histological point of view, a spectrum of lesions ranging from infectious mononucleosis to a malignant lymphoma comprises the clinical entity known as PTLD (9). A number of small-scale studies and case reports have shown that PTLDs localized in the graft organ itself may originate from donor cells without being donor transmitted (10–31). Nevertheless, contrary to bone marrow transplant where PTLDs typically derive from donor lymphocytes, the majority of PTLDs that develop after organ transplantation derive from recipient lymphocytes (32–34). To improve our knowledge of this rare complication, using analysis of multiple short tandem repeat microsatellite (STR) loci and fluorescence in situ hybridization of the Y chromosome, we analyzed the origin of lymphoma cells in a large cohort of adult kidney recipients with PTLD, half of which were graft lymphomas. We studied the features of lymphomas and patient prognosis according to the origin of the tumoral cells.
- Top of page
We analyzed 43 specimens of PTLD diagnosed in 43 kidney transplant recipients from 11 French transplant centers. Inclusion in the study was based on the availability of frozen or paraffin-embedded tissues from the PTLD together with a control tissue from the donor or the recipient or both (for example, from the recipient, a previous skin or gastric biopsy, or nephrectomy, from the donor, previous graft biopsies or lymphocytes stored in the Histocompatibility Laboratory). Among the 11 participating centers, 139 cases of lymphomas were reported in the French Registry during the period concerned, 27 were localized in the graft and 112 outside the graft. Six cases of graft PTLD were excluded from our study because specimens of the tumor or control tissues were no longer available. We received 22 cases of extra-graft lymphomas from the participating centers. We performed a statistical analysis to compare the cases included in our study with the cases that were not included (n = 90). No statistical differences in patient or tumor characteristics were observed in the two groups except that fewer patients were treated with ATG in the cases we analyzed than in the cases not analyzed in our study (62 vs. 86%, p = 0.03). Clinical data concerning the patients included in our study were extracted from the Registry database. Specimens were centralized in Strasbourg University Hospital, and processed in the Pathology Department for histological and in situ hybridization investigations and in the Molecular Biochemistry Laboratory for analysis of multiple STR microsatellite loci.
Histological sections of formalin-fixed, paraffin-embedded tissues from lymphomatous tumors were examined after routine staining with hematoxylin and eosin. The B-cell nature of the lymphoid proliferation was evaluated using an anti-CD20 antibody (L26, Lab. Dako, Glostrup, Denmark); EBV was detected and localized in formalin-fixed, paraffin-embedded tissue by nonisotopic in situ hybridization using the EBV probe BamH1 W (Ventana, Tucson, AZ, USA, 85755). PTLDs were classified according to the 2008 WHO classification (35) as early lesions, polymorphic PTLD, monomorphic lymphoma or classical Hodgkin's lymphoma-type PTLD (Table 1).
|Infection mononucleosis-like PTLD|
|Florid follicular hyperplasia|
|Diffuse large B-cell lymphoma|
|Plasma cell myeloma|
|Peripheral T-cell lymphoma|
|Hepatosplenic T-cell lymphoma|
|Classical Hodgkin lymphoma-type PTLD|
Analysis of immunoglobulin gene rearrangement
Immunoglobulin gene rearrangement was analyzed on paraffin-embedded lymphoma tissue. DNA was extracted from paraffin-embedded tissue by heating at 95°C for 10 min in lysis buffer including 10 mM Tris-HCl at pH 8, 100 mM KCl, 2.5 mM MgCl2, 0.45% Tween 20, followed by Proteinase K (4 μg/μl) digestion overnight at 65°C and extraction with phenol chloroform. After fluorescent PCR amplification of IgH VDJ or IgKappa Kde junctions using a set of primers (IGK and IGH Gene Clonality Assay kits) supplied by In Vivo Scribe Technologies (Carlsbad, CA92009, USA), PCR products were identified by capillary electrophoresis on an ABI PRISM 3130 (Applied Biosystems, Carlsbad, CA, USA).
Fluorescence in situ hybridization (FISH): Fluorescent hybridization was used in cases of sex mismatch (n = 19), Figure 1. In situ hybridization was performed on 4 μm thick formalin-fixed, paraffin-embedded sections of PTLD using CEP Y (satellite III) SpectrumGreen probe (Abbott, Wiesbaden, Germany) that hybridizes with the Y chromosome (bandYq12, locus DYZ1). Hybridization with 5 μL of probe occurred at 37°C after 20 h in Hybridizer (Dako) following a denaturing cycle at 73°C for 5 min. The glass slide was fixed with rubber cement during hybridization. The section was placed in 0.4× SSC/0.3% Igepal at 73°C for 2 min then in 2× SSC/0.1% Igepal at room temperature for 1 min before dehydration in alcohol. Finally, 15 μL Iodure of Propidium was added on the section.
HLA immunostaining: The origin of the PTLD was determined by HLA immunostaining in two previously published cases (36). Briefly, staining was performed with HLA A24 (Ab H41, One-Lambda) or HLA DQ3 (kindly provided by Marie-Marthe Tongio, Etablissement Francais du Sang, Strasbourg, France) mouse monoclonal IgG antibodies. Immunohistochemical staining of formalin-fixed, paraffin-embedded sections was performed using a peroxidase-antiperoxidase reaction.
The donor versus recipient origin of tumor tissues was evaluated by analyzing DNA at multiple polymorphic microsatellite loci and comparing the results with those obtained using recipient or donor tissues (sequencer ABI 3130, Applied Biosystems), Figure 2. PTLD and control cells were microdissected from formalin-fixed paraffin-embedded tissues. Kidney allograft rejection samples, samples displaying inflammation, and allograft samples with PTLD but with low tumor cell populations (< 60–70%) were not used. The tumor samples analyzed contained at least 80–90% of tumor cells. DNA samples were subjected to microsatellite analysis as previously described (18). Samples were examined at 9–24 highly polymorphic loci, depending on the case concerned. STRs were chosen for their high rate of heterozygosity. Loci were considered to be informative if the tumor and the control DNA (recipient or donor or both) produced two bands that were clearly distinguishable. Analyses were considered to be in concordance when all the STRs were identical. Discordance was assumed when one or more STR differed.
- Top of page
Univariate analyses were performed using χ2 for categorical variables and Student's two-sided t-test for continuous variables, values of p < 0.05 were considered significant. Kaplan–Meier analyses were used to construct survival plots of time to death after PTLD diagnosis. Groups were compared using a log rank test. Statistical analyses were performed using SPSS 11.5 TM (SPSS, Inc., Chicago, IL, USA).
- Top of page
We studied 43 tumors from 43 kidney recipients (15 women and 28 men). Patients received transplants between 1989 and 2008. Age at transplant ranged from 22 to 69 years (mean 49 ± 12 years). One patient received the graft from a living donor, all the others from deceased donors. Five patients had their second transplant and one the third. PTLD occurred after a mean interval of 56 ± 64 months (3–240 months), 15 (35%) before month 12. Three patients were EBV negative before transplantation and two received a kidney from an EBV-positive donor. The lymphoma was located in the graft in 21 patients, and the graft was the only site in 14 patients (67%). Other PTLDs (n = 22) involved the central nervous system in two patients, gastro-intestinal tract in eight patients, lymph nodes or hematopoietic organs in five patients, skin in four patients, lung, amygdale or parotid gland each in one patient. Out of the 22 patients investigated, 16 had EBV viremia at the time of PTLD diagnosis. One PTLD was classified as early lesion, 10 as polymorphic PTLD, 31 as monomorphic B-cell lymphoma and one as classical Hodgkin's lymphoma. All tumors were B type and CD20 positive. PTLD was positive for EBV staining in 29 cases (68%). Twenty-four cases out of 37 cases tested were monoclonal (65%).
Analysis of donor or recipient origin
Results of analyses of the origin of tumor cells are listed in Table 2. Sixteen tumors were of donor origin and 27 of recipient origin. All donor PTLDs except two were localized in the graft itself. The two tumors of donor origin located outside the graft occurred a long time after transplantation and involved skin in one case and the gastro-intestinal tract in the other. These two cases will be described in detail in a forthcoming paper. We decided to not include these two cases in the following statistical analyses because we think they are an exception and probably have a particular pathophysiology.
|Sex and age of recipient at diagnosis||Time to PTLD (months)||PTLD loc1||PTLD loc2||EBV tumor||Histology||Clonality||Microsatellite||FISH||PTLD origin|
Table 3 summarizes the characteristics of the patients and lymphoma according to donor or recipient origin. The 14 tumors that developed from donor cells were confined to the graft in 10 cases, and were localized in the kidney graft plus in other organs in four cases (lymph nodes in two cases—one with additional localization in bone,spleen and liver in one case, and skin in one case). The tumors developed 3–99 months after transplantation (median 7 months). PTLDs of donor origin that were confined to the graft (n = 9) were diagnosed sooner (3–28 months, including seven during the first year after transplant). The 27 lymphomas of recipient origin were localized in the graft in seven cases (four at an single location and three as part of a multifocal disease) and were localized outside the graft in the 20 other cases. They developed 3–240 months after transplant (median 44 months).
|Donor PTLD (n = 14)||Recipient PTLD (n = 27)||p|
|Age at PTLD (years)||53 ± 11 [35–69]||55 ± 12 [25–69]||NS|
|CMV donor (% pos)||21%||48%||0.03|
|CMV recipient (% pos)||64%||52%||NS|
|MM CMV (%)||7%||18%||0.15|
|EBV donor (% pos)||72%||94%||0.17|
|EBV recipient (n,% neg)||0||3 (13%)||0.07|
|MM EBV (n)||0||2||NS|
|Early onset PTLD (< 12 months) (n)||9 (64%)||6 (22%)||0.01|
|Time to PTLD (months ± SD)||20 ± 28 [3–99]||69 ± 68 [3–240]||0.01|
|LDH (mean ± SD, UI/L)||612 ± 372 [222–1300]||598 ± 415 [100–1835]||NS|
|Single site (n,%)||10 (71%)||12 (44%)||0.09|
|Monomorphic PTLD (n, %)||8 (57%)||23 (85%)||0.057|
|PTLD EBV+ (n, %)||11 (78%)||17 (63%)||NS|
|Monoclonal PTLD (n, %)||8/13 (61%)||15/22 (68%)||NS|
|IS tapering only (n, %)||3 (21%)||2 (7.5%)||0.2|
|Antiviral therapy (n, %)||2 (14%)||7 (28%)||NS|
|Rituximab (n, %)||6 (43%)||21 (78%)||0.03|
|Chemotherapy (n, %)||8 (57%)||14 (52%)||NS|
|Chemotherapy + Rituximab (n, %)||3 (21%)||11 (40%)||NS|
|Transplantectomy (n, %)||7 (50%)||3 (11%)||0.01|
|Graft failure||10 (71%)||7 (26%)||0.007|
|Complete or partial remission (%, n)||85% (11/14)||70% (19/27)||NS|
Differences in clinical features and outcome according to the origin of the lymphoma
Donor lymphoma displayed the following features: CMV-positive donors (21 vs. 48%, p = 0.032) and a trend toward fewer CMV mismatches (7 vs. 18%, p = 0.1), more early-onset PTLDs (64 vs. 22%, p = 0.01) with a shorter time to PTLD diagnosis: 20 ± 27 versus 69 ± 67 months, p = 0.01, an increasing trend toward single localization (71 vs. 44%, p = 0.09), particularly within the graft itself (p = 0.001). Lymphomas of donor origin were more frequently polymorphic than were tumors of recipient origin (43 vs. 15%, p = 0.057). EBV was identified in 78% of donor tumors and 63% of recipient tumors (NS). Patients with donor-derived PTLD were more frequently transplantectomized (50 vs. 11%, p = 0.01) and returned more frequently to dialysis (71 vs. 26%, p = 0.007). No long-term impact of the origin of PTLD cells on patients’ global survival rate was observed (57% in patients with donor tumors vs. 58% in patients with recipient tumors 5 years after diagnosis) (see Figure 3a). However, when we analyzed death related to PTLD, fatal events were more frequent in patients with recipient PTLD (8 out of 27 patients, 68% survival 5 years post-transplant) than in patients with donor lymphoma (2 out of 14, 85% survival 5 years post-transplant) but the difference was not statistically significant (p = 0.3 by log rank test), Figure 3b.
Graft PTLD (n = 21): The characteristics of patients with graft lymphomas are summarized in Table 4. We found no difference between donor (n = 14) and recipient PTLDs (n = 7) in terms of clinical features or prognosis. Of note, there was no difference in time to PTLD (20 ± 27 and 23 ± 36 months respectively) nor in the site of the PTLD (respectively 29% and 43% of PTLD spread beyond the graft, p = 0.4). There was no difference in histological features (monomorphic PTLD in 57% of donor PTLDs vs. 71% of recipient PTLDs and EBV-positive tumors in 78% of donor PTLDs vs. 100% of recipient tumors, NS). In this subgroup, there was no difference in the 10-year patient global or PTLD-related survival rate between donor and recipient PTLDs (57% survival 5 years post-transplant in both, Figures 4 and 5).
|Graft donor PTLD (n = 14)||Graft recipient PTLD (n =7)||p|
|Age at PTLD (years)||53 ± 11 [35–69]||56 ± 11 [40–69]||NS|
|LDH (mean ± SD, UI/L)||612 ± 372 [222–1300]||741 ± 730 [325–1835]||NS|
|Single site (n, %)||10 (71%)||4 (57%)||NS|
|Early onset PTLD (< 12 months) (n)||9 (64%)||4 (57%)||NS|
|Time to PTLD (months ± SD)||20 ± 28 [3–99]||23 ± 36 [3–102]||NS|
|Monomorphic PTLD (n)||8 (57%)||5 (71%)||NS|
|PTLD EBV+ (n)||11 (78%)||7 (100%)||NS|
|Monoclonal PTLD (n)||8/13 (61%)||6/7 (85%)||NS|
|Complete or partial remission (%)||85% (12/14)||71% (5/7)||NS|
- Top of page
In this paper we describe the largest series used for the determination of donor/recipient origin of post-transplant lymphoproliferative disorders to date. Among the 43 tumors analyzed, we found 16 tumors originating from donor cells and 27 from recipient lymphocytes. Of note, 21 of the lymphomas studied were localized in the graft itself. PTLD pathogeny is specific because PTLDs are related to EBV and to the immunosuppressive condition of the recipient. Among the particularities of PTLD, the possibility of lymphomas developing within the graft is one of the most intriguing. Several studies have shown that PTLD developed in the graft itself in 12% of kidney recipients (French Registry, personal data), in the liver in 15–30% of liver recipients (37) and even more frequently in lung and gastro-intestinal tract patients (50–70%) (2,38,39), probably because many lymphocytes infiltrate lung and intestinal tissues. Moreover, it has been shown that lymphomas occurring within the graft often derive from donor cells in liver, lung and heart lung and in kidney transplant recipients (Table 5) (10,12,14–17,36). It is assumed that these tumors are not transmitted by the donor but, after transplantation, grow from donor passenger lymphocytes present in the organ or in nearby lymph nodes transplanted together with the organ (40). Moreover, donor-origin hematopoietic cells have been identified in native tissues in recipients of allografts without PTLD (41).
|Transplanted organ||Cases (n)||PTLD of recipient origin (n)||PTLD of donor origin (n)||PTLD of donor origin localized in the allograft (n)|
|Kidney and kidney/pancreas||51||32||19||18|
|Lung and heart lung||17||13||4||4|
We showed that the great majority of donor PTLDs arose in the graft itself (14/16). In the literature, 18 donor PTLDs were localized in or near the graft (10–31,42). Moreover, in these published reports of donor lymphomas, only one tumor was known to disseminate outside the graft (32). The only widespread donor PTLD, which presented as a pararenal mass with multifocal marrow involvement, was a multiple myeloma (43). Surprisingly, we describe two donor-derived PTLDs that developed only outside the graft (one in the skin and one in the gastrointestinal tract) and occurred a long time after transplantation. This is an unexpected finding since no other case has been described in the literature until now. Nevertheless, we can hypothesize that the donor lymphocytes survived for a long time in these two patients due to tolerance, as the patients had been transplanted since 1985 and 1992 respectively, had a perfectly functioning graft and no rejection episode before the onset of PTLD. Very long-term chimerism (up to 30 years) is reported in the literature (44). What actually triggered the development of the lymphomas in these patients remains unclear.
Next we performed a head-to-head comparison between donor and recipient-derived PTLDs. According to a report by Petit et al. (10), donor-derived tumors develop earlier than recipient lymphomas. Analyzing 12 PTLDs by multiple STR loci analysis and HLA immunostaining, Petit identified four tumors derived from donor cells (all in the allograft) and eight tumors originating from recipients. All recipient lymphomas occurred after the first post-transplant year, whereas the four donor-derived tumors were confined to the graft and occurred during the first post-transplant year. Similarly, Capello (45) analyzed 15 PTLDs after liver transplantation and showed that time to diagnosis differed in donor and recipient lymphomas (median 7 vs. 63 months respectively) and in the localization of the tumor (9/9 tumors of donor origin were localized in the hepatic hilum vs. 5/6 recipient tumors were localized outside the liver). In our series, donor PTLDs were more often polymorphic lymphomas, probably because early-onset PTLDs are more likely to be less aggressive than late-onset tumors (46). Capello made the same observation in his cohort of liver transplant patients: 7/9 polymorphic lymphomas in donor PTLDs versus 5/5 diffuse large B-cell lymphomas in recipient PTLDs (47).
Finally, we showed that prognosis for patients with tumors of donor or recipient origin is comparable with a similar 10-year survival rate. This is in accordance with the Capello series where the outcome did not differ significantly between liver patients with a donor or recipient PTLD (47). Nevertheless, when we analyzed patient survival related to PTLD, we found a trend toward more deaths in patients with recipient PTLD, consistent with a probably slightly better prognosis for patients with a donor PTLD, especially in the case of early-onset PTLD and if the lymphoma was strictly confined to the graft (one out of eight patients in our series and the patient died from complications of chemotherapy whereas he was in partial remission). Isolated cases of donor PTLD reported in the literature appeared to have a good prognosis. This could be explained by the fact that these tumors could be more sensitive to restoration of recipient cytotoxicity when immunosuppression is reduced. Another explanation could be that direct involvement of the graft may favor early detection of the tumor because the graft is the object of increased surveillance after transplantation. The apparently better prognosis of donor tumors in the literature could also be due to a selection bias for publication. Finally, the published series was too small to allow statistical analysis of patient survival (10,11). In the report by Petit et al., one patient with donor PTLD out of four died while five out of eight patients with recipient lymphoma died. For the first time, the present series enabled us to present statistics on the survival of patients with PTLD according to tumor origin.
In the subgroups of patients with graft PTLD, 14 had donor tumors and 7 recipient lymphomas. This in accordance with a review of published cases in which 18 donor tumors and 11 recipient tumors were reported (Table 5). It is well known that allorecognition brings recipients’ lymphoid cells into the graft, where they can give rise to PTLD. We found no difference in terms of clinical features and prognosis between donor and recipient-derived tumors, but our number of patients was probably too small to reveal statistical differences. The observation that four patients with a donor lymphoma presented multifocal disease caused by donor lymphoma cells spreading outside the graft is surprising. On the other hand, we showed that 40% of lymphomas in the graft derived from recipient lymphocytes. Gulley analyzed the origin of 20 PTLD tumors, of which three occurred in the graft itself. Of these three, only one tumor originated from the donor and the other two from recipient cells (34). The same observations were made in Peterson's series in which three out of four lung PTLDs in lung transplant patients were of recipient origin (48).
Microsatellite techniques can facilitate the identification of the origin of the lymphoma in most cases of PTLD. However, when PTLDs occur in the graft, the origin of the tumor is sometimes difficult to determine because of the presence of recipient lymphocytes in the donor graft resulting in chimerism. In these cases, the use of an immunohistochemical analysis or FISH can help identify the exact origin of the tumor by enabling simultaneous analysis of tumoral cell morphology and fluorescent staining, as already pointed out by Petit et al. using HLA immunostaining (10). However, fluorescence hybridization is only possible for the Y chromosome in cases of sex mismatch (40% of the cases in our series) (42).
- Top of page
In conclusion, using a large series, we showed that the great majority of donor PTLDs developed in the allograft, with a shorter time to diagnosis than for recipient lymphomas. The pathogeny of these two kinds of PTLD requires further study to identify the factors leading to donor or recipient lymphomas. There appears to be a trend toward better prognosis in donor PTLDs but further investigation is required. Determination of the origin of the tumor cells not only provides clues for lymphomagenesis, but also prognostic guidance in the clinical management of patients, for example for the use of cytotoxic T-cell infusions.
- Top of page
We are indebted to all collaborators who declared cases to the French Registry, and to those who prepared and sent specimens. We thank Christine Meyer, Eliane Supper, Fabienne Reymann and Martine Muckensturm for their technical assistance.
- Top of page
The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.
- Top of page
- 30Epstein–Barr virus-associated posttransplant lymphoproliferative disorder of donor origin after simultaneous pancreas-kidney transplantation limited to pancreas allograft: A case report. Am J Transplant 2006; 6: 2506–2511., , et al.
- 35Post-transplant lymphoproliferative disorders. In: Swerdlow SH, Campo E, Harris NL et al, eds. WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues Lyon , France : ARC, 2008., , , .