Post-Transplant Lymphoproliferative Disorder in the United States: Young Caucasian Males are at Highest Risk
Version of Record online: 6 NOV 2002
American Journal of Transplantation
Volume 2, Issue 10, pages 993–998, November 2002
How to Cite
Dharnidharka, V. R., Tejani, A. H., Ho, P.-L. and Harmon, W. E. (2002), Post-Transplant Lymphoproliferative Disorder in the United States: Young Caucasian Males are at Highest Risk. American Journal of Transplantation, 2: 993–998. doi: 10.1034/j.1600-6143.2002.21019.x
- Issue online: 6 NOV 2002
- Version of Record online: 6 NOV 2002
- Received 8 February 2002, revised and accepted for publication 6 June 2002
- Post-transplant lymphoproliferative disorder;
We have previously documented Caucasian race and cadaver donor source as risk factors for post-transplant lymphoproliferative disorder (PTLD) development in recipients registered in the North American Pediatric Renal Transplant Cooperative Study (NAPRTCS). We analyzed data from the Scientific Registry of the United Network of Organ Sharing (UNOS) (from January 1988 to December 1999) to determine risk factors for the development of PTLD in all organ systems and its frequency, and we compared these factors to the risk factors in the most recent NAPRTCS database (1987–2000). In the UNOS database, PTLD was reported in 2365 of 205 114 organ-transplant recipients (1.2%). PTLD was reported in 3% or more of all intestinal and thoracic organ recipients, but in less than 1% of other abdominal organ recipients. Recipient age < 18 years, Caucasian race and male gender were independent risk factors [Odds Ratios (OR) 2.81, 2.22 and 1.40, respectively, p = 0.0001], but not cadaver donor source. The combination of all three risk factors increased the OR to 8.78. The occurrence of PTLD showed a significant rise per year for heart-lung, kidney, kidney-pancreas and liver transplants, but decreased significantly for heart transplants (p < 0.001). Similar frequencies of PTLD were found in smaller organ-specific registries of heart, intestine, pediatric liver and pediatric kidney transplants. The PTLD incidence per year and incidence density have increased in recent years. Young Caucasian males are at highest risk for PTLD development among solid-organ-transplant recipients. The incidence of PTLD is increasing.
Post-transplant lymphoproliferative disorder (PTLD) is an unusual and troublesome complication of solid-organ transplantation. This disorder is characterized by abnormal proliferation of immune cells (typically host-derived B cells) in the setting of post-transplant immunosuppression (1,2). PTLD was extremely rare till the advent of potent immunosuppressive agents, but has emerged in the last decade as a major complication of transplantation (3,4). Previous reports from single-center series have delineated risk factors in small patient sets, such as pretransplant Epstein-Barr virus (EBV) seronegativity in the recipient (5–7), coexisting cytomegalovirus (CMV) infection (8,9), the use of the immunosuppressive agents OKT3 (10,11) or tacrolimus (12–14). In an analysis of multicentered registry data from pediatric renal transplant recipients, we have demonstrated that PTLD was more frequent in Caucasian race and cadaver donor recipients (15). We wished to determine if these risk factors were specific to pediatric kidney transplantation or were active in other organ transplants as well. In addition, the total burden of PTLD in the United States has not been reported to date. The United Network of Organ Sharing (UNOS) collects multiple data on all solid organ transplants performed in the United States. In this study, we have analyzed UNOS data on all reported cases of PTLD from 1988 to 1999 to report the total burden of PTLD in the USA. We have compared the PTLD frequency in the UNOS data with the reported incidence in smaller organ-specific transplant registries, such as the North American Pediatric Renal Transplant Cooperative Study (NAPRTCS), Studies of Pediatric Liver Transplantation (SPLIT), the International Society of Heart and Lung Transplantation (ISHLT), and the International Intestinal Transplant Registry (IITR). We have also compared transplant-related risk factors for PTLD development in UNOS and NAPRTCS data.
Materials and Methods
The United Network of Organ Sharing is a federally funded private organization that receives information on all solid organ transplants performed in the United States. The Scientific Registry of the UNOS maintains the database, which is available to investigators for research analysis. We queried the UNOS database for all cases of PTLD (listed as de novo lymphoproliferative disease) in all organ-transplant recipients. For all patients (including those that did not develop PTLD), the data requested and obtained included: organ received, recipient age at transplant, gender, race, donor source, date of transplant, report of PTLD, year of PTLD development, pretransplant EBV serostatus and pretransplant CMV serostatus. The data were then stratified into a PTLD group and a non-PTLD cohort, which were then analyzed for differences in the available demographic and transplant-related data, as described below.
In order to minimize the potential bias of under-reporting from the results and conclusions of the UNOS data, we also obtained data from smaller organ-specific transplant registries, and performed comparisons for the available data. Since its inception in January 1987, the scientific registry of the NAPRTCS has enrolled over 7000 children to date who have received kidney transplants. Besides collecting data from US-based centers, the NAPRTCS also includes centers from Latin America and Canada. For the purpose of this study, complete data were available on patients until December 2000. Data collection methods have been described previously (16). Data collected from each center are sent to the Data Coordinating Center of NAPRTCS (EMMES Corporation). Data collected were analyzed to identify the patients who had a diagnosis of PTLD post transplant. Demographic data analyzed included: age at transplant, gender, race, primary diagnosis, time to PTLD, and donor source of kidney. The data were stratified into PTLD and non-PTLD groups and analyzed for differences as described below.
The SPLIT registry was formed in 1995 and collects transplant-related data from US-based centers performing pediatric liver transplants. The registry of the ISHLT, formed in 1982, collects and publishes data on heart and lung transplants in all age groups performed at multiple international centers. Similarly, the IITR collects data on intestinal transplants in recipients of all ages from several international centers (http://www.intestinaltransplant.org).
Differences in occurrence of postulated risk factors in the UNOS and NAPRTCS databases were determined for both the PTLD and non-PTLD groups, and statistical significance was calculated using SAS® software (Cary, North Carolina, USA). Chi-square and t-tests were used to test for association between independent risk factors and the rate of PTLD over all organs and by each organ individually. Logistic regression was used to model the rate of PTLD against the independent variables that showed a significant association by chi-square testing (age < 18 years, Caucasian race, and male gender) over all organs. Trends in PTLD incidence by year of transplant were analyzed by the Cochran-Armitage trend test. A p-value < 0.05 was considered significant.
PTLD incidence by organ system
UNOS data: Data were available for the period January 1, 1988 to December 31, 1999. During this time period, a total of 205 114 solid organ transplants were performed in the United States. PTLD was reported in 2365 recipients (1.2%). The highest raw incidence (% cases of total within the 12-year time period) was reported in intestinal transplants (IN; 6.0%) and thoracic organ recipients: heart-lung (HL; 5.5%); heart (HR; 3.9%); and lung (LU; 3.7%). The raw incidence was lower in abdominal organs such as liver (LI; 0.9%) and kidney (KI; 0.6%). However, the largest absolute number of cases was in HR transplants (n = 950), followed by KI (n = 692) and LI (n = 375). The detailed breakup of the total number of transplants and PTLD cases by each organ system is depicted in Table 1.
|Organ||Number of transplants||Number of PTLD cases (%)|
|Heart||24 100||950 (3.9%)|
|Liver||39 974||375 (0.9%)|
|Kidney||124 638||692 (0.6%)|
|Total||205 114||2365 (1.2%)|
Organ-specific registry data: In order to evaluate for possible under-reporting, we compared the incidence of PTLD in specific organ populations with the incidence from registries for those specific organ systems: NAPRTCS (pediatric kidney), SPLIT (pediatric liver), ISHLT (heart) and IITR (intestine). In the NAPRTCS database of subjects aged 1–21 years (including Mexico and Canada), there were 121 PTLD cases reported from 1987 to 2001 (1.7%). Between 1988 and 1999, the UNOS registry lists 101 PTLD cases in subjects aged 1–17 years (1.27%). In heart-transplant recipients of all ages and varying nationalities who have reached 5-year follow-up, the registry of the ISHLT reported a PTLD frequency of 1.1% (17), whereas the frequency of PTLD in heart-transplant recipients of all ages in the UNOS registry from 1988 to 1999 was 3.9%. The SPLIT registry (Studies in Pediatric Liver Transplantation) has reported a 2.4% occurrence rate for PTLD in its registry data from 1995 to 2000 (18). The occurrence of PTLD in pediatric-aged liver-transplant recipients in the UNOS database is 2.3% (80 PTLD cases in 3414 recipients). The Intestinal Transplant Registry has reported 15 cases of PTLD in 193 intestinal transplants (8%) as of May 1999 (http://www.intestinaltransplant.org). In the UNOS data, from 1988 to 1999, there have been 19 cases reported (6%). Overall, these results suggest similar occurrence rates. Thus, while we cannot completely exclude under-reporting, the magnitude of under-reporting is not likely to be high. We then chose to perform a detailed comparison of the results in the UNOS data with those from NAPRTCS (see below) because of the very low degree of under-reporting to NAPRTCS. The similarities in the results from both databases testify to the accuracy of our overall conclusions.
PTLD incidence by year of transplant
UNOS data: We then analyzed the incidence of PTLD by year of transplant for all organs as a group and then individually. This analysis was restricted to the years 1988–97 (deleting the last 2 years) in order to prevent skewing of data due to shorter follow-up times in the latter 2 years. There was no significant trend in frequency by year in the combined group. However, when analyzed for each organ separately, there were striking trends in two opposite directions, as shown in Figure 1. The incidence of PTLD showed a significant rise per year for HL (p = 0.005), KI (p = 0.001), kidney-pancreas (KP) (p = 0.001), and LI (p = 0.001), but decreased significantly for HR (p = 0.001). There was a rising trend in PTLD frequency per year in lung transplants, but this did not achieve statistical significance. In other organ systems, the sample sizes were much smaller and no significant trends were detected.
NAPRTCS data: We also analyzed the NAPRTCS data for incidence of PTLD by transplant year in the same time period 1988–97. The raw incidence of PTLD by year in pediatric kidney transplant recipients did not show a rising trend by Cochran-Armitage trend test. However, in this database, we also had the ability to analyze incidence density rates (which are adjusted for years of patient follow-up and delete patients who have already developed PTLD). We have previously reported an increasing trend in incidence density from 0.254/100 patient years in the period 1987–92 to 0.395/100 patient years in the period 1992 onwards (15). In the most recent data available from NAPRTCS, we have now confirmed a statistically significant increase in incidence density from 0.320/100 patient years between 1988 and 1992 to 0.632/100 patient years between 1993 and 1997 (p < 0.0001).
Risk factors for PTLD development
UNOS data: In a logistic regression model (Table 2), the following three variables were associated with statistically higher risk (p = 0.0001 for all): recipient age less than 18 years (OR = 2.81); Caucasian race (OR = 2.22) and male gender (OR = 1.40). The presence of all three of these characteristics was found to cumulatively increase the OR for PTLD to 8.78.
|Database||Variable (n)||Percentage developing PTLD||Odds ratio (95% CI)||Relative risk||p-value|
|UNOS data||Age < 18 years (17 200)||2.63%||2.81 (2.51, 3.12)||2.80||0.0001|
|Age ≥ 18 years (187 863)||1.02%|
|Caucasian (142 793)||1.38%||2.22 (2.00, 2.48)||2.21||0.0001|
|Non-Caucasian (62 321)||0.64%|
|Male (124 797)||1.29%||1.40 (1.29, 1.53)||1.40||0.0001|
|Female (80 316)||0.94%|
|Cadaver donor (88 494)||0.52%||0.79 (0.68, 0.93)||0.79||0.004*|
|Living donor (36 144)||0.65%|
|NAPRTCS data||Caucasian (4424)||2.15%||2.20 (1.42, 3.40)||2.17||0.0001|
|(pediatric renal)||Non-Caucasian (2634)||0.99%|
|Male (4230)||1.73%||1.02 (0.70, 1.47)||1.02||NS|
|Cadaver donor (3444)||1.86%||1.18 (0.82,1.17)||1.17||NS|
|Living donor (3614)||1.58%|
By univariate chi-square testing, age < 18 years, Caucasian race and male gender were all significantly associated with PTLD development over all organs (p = 0.001 for each). When broken down by organ system, pediatric age (< 18 years) was a significant risk factor for PTLD development in HR, LI, LU, KI and IN transplants (p = 0.001 for HR, LI, LU, KI and p = 0.006 for IN). Pediatric age was not a risk factor in PA, HL and KP transplants, where the proportion of pediatric age recipients was also much smaller. Within the pediatric age group < 18 years, we further analyzed whether extremely young age (0–5 years) was a significant risk factor. In some studies, liver-transplant recipients aged less than 5 years were at higher risk for PTLD than other pediatric liver recipients. Though we found a statistically significant difference in the combined overall group (3.27% incidence in 0–5 age group vs. 2.18% in 6–17 years age group, p < 0.001), the major contribution to this significance came from HR transplants (6.34% incidence in 0–5 years age group vs. 4.09% in 6–17 years age group, p = 0.010). No statistically significant differences were found between the rates of PTLD in the 0–5 years age group and 6–17 years age group for IN, HL, LU, KI, KP, LI and PA organs. Thus, in this analysis, very young age was not a risk factor in pediatric liver transplantation, as has been previously reported in other studies (12). Interestingly, the increased risk for young age recipients of heart transplants did not apply to infants less than 1 year of age, in whom the occurrence rate of PTLD was 4.65% (p = NS in comparison to heart-transplant recipients in other age groups).
Caucasian race was analyzed separately against both African-American race and Hispanic race. Caucasian recipients had a statistically higher rate of PTLD than African-American recipients in the overall group as well as in KI, HR and LU, but not in LI, KP, PA, HL and IN transplants. Similarly, Caucasian recipients had a statistically higher rate of PTLD than Hispanic recipients in the overall group and in KI transplants, but not in other organs. The percentage of Caucasian patients with PTLD was not higher in very young (0–5 years) age groups.
We analyzed the association of gender by organ system as well. Male gender was a significant risk factor over all organs and in HR and KI transplants (p = 0.001), but not in other organ systems.
We evaluated cadaver donor source as a potential risk factor using data from KI patients only. In other organ systems, the number of living donor transplants was very low and thus not useful for analysis. Using UNOS dataset for KI adult and pediatric transplants, cadaver donor source was not a risk factor for PTLD. On the contrary, cadaver donor source was associated with a slightly lower risk (Odds Ratio = 0.79, p = 0.004).
Finally, we attempted to evaluate whether the demographic and transplant-related risk factors outlined above were secondary to pretransplant EBV or CMV seronegativity. Primary infections with either EBV or CMV have been found to be extremely important risk factors in previous studies (5,7–9). However, these data were reported in only 20 subjects within the UNOS database, and thus no meaningful statistical analysis could be performed.
NAPRTCS data: In our current analysis of the NAPRTCS data for PTLD risk factors, Caucasian race was an independent risk factor (Table 1: RR = 2.17, p = 0.0001). However, in contrast to our previous analysis, cadaver donor source was not a significant risk factor in this more recent dataset (RR = 1.17, p = NS). All registered recipients in the NAPRTCS are below the age of 21 years, so we did not look at pediatric age as a risk factor in this dataset. Neither young age (< 5 years) nor male gender has been shown to be a PTLD risk factor in either analysis of NAPRTCS data. As with the UNOS database, CMV and EBV reporting is not a requirement for the NAPRTCS and we did not have data to analyze for these variables as PTLD risk factors.
Reports of PTLD were uncommon till the 1980s, when investigators started reporting several series of cases (19,20). With the advent of more potent immunosuppression, the number of cases and the interest in this disorder have risen dramatically (4). In the 12-year time period of this analysis, 2365 cases of PTLD have been reported to the UNOS Scientific Registry, representing almost 200 cases per year.
Several trends are evident in this analysis. The raw incidence of PTLD per year has shown an increase in kidney, liver, heart-lung and kidney-pancreas transplants, but has shown a decrease in heart transplants. These raw incidence numbers do not adjust for time period of follow up, nor do they delete patients after development of PTLD. Thus, they cannot be compared directly with incidence density, as used by us (15) and other investigators previously (14). Since the UNOS database does not contain information on when PTLD developed in the recipient, we were unable to calculate incidence density rates from UNOS data in this study. In the NAPRTCS data, the incidence density for PTLD showed a statistically significant rising trend in this analysis. It is worth noting that the trends in raw incidence closely mirror the trends of increased PTLD frequency by incidence density analysis in pediatric renal and liver transplantation.
In addition, the rate of PTLD is clearly organ dependent. Previous studies have reported the incidence of PTLD to be 20–30% in intestinal transplants (21,22), 2–6% in cardiac transplants (10,23) and 0.7–2.6% in renal transplants (11,24,25). At this time in the UNOS database, the highest incidence is seen in intestinal and thoracic organ transplants. The incidence of PTLD in liver transplants in this database is much lower (0.9%) than previously reported by individual centers (3–20%) (6,13,26). This may be explained by the bias of more frequent reporting by centers with a higher incidence. However, this may also be explained by the rising incidence of PTLD on a per year basis in liver transplants, or by possible under-reporting. Under-reporting appears unlikely given the similarities in the PTLD incidence in UNOS data with the incidence results in NAPRTCS, SPLIT, ISHLT and IITR.
Previously reported risk factors for PTLD that relate to recipient and peri-transplant characteristics include recipient EBV/CMV seronegativity, age less than 5 years, Caucasian race and cadaver donor source. We have attempted to evaluate each of these in this analysis. In contrast, we have not attempted to analyze post-transplant immunosuppression risk factors, such as use of OKT3 and tacrolimus, detailed data for which are not available in the UNOS database. Caucasian race was a higher risk factor for PTLD in our previous analysis of pediatric renal transplants and was also a higher risk factor in several organ systems in this study. The mechanisms for this increased risk remain unclear. In contrast, cadaver donor source, a weak risk factor in our prior analysis of the NAPRTCS database, was not a risk factor in this analysis. It is possible that both pediatric age and Caucasian race are not independent risk factors but are secondary to a possible higher incidence of pretransplant EBV seronegativity in these demographic groups. However, to our knowledge, there have been no reports relating EBV exposure to racial background to date. We could not analyze for EBV/CMV serostatus, as these data are not routinely reported to UNOS or NAPRTCS.
The high incidence of PTLD in heart-transplant recipients aged 1–5 years may be related to the large age and size mismatching that is often necessary in this population, which may increase the chances of a primary EBV mismatch. In contrast, infants less than 1 year of age were not at significantly higher risk for PTLD. This may relate to a greater use of infant or young donors for these recipients (more likely to be EBV seronegative) or the possibility of passive maternal anti-EBV antibodies playing a protective role in neonates who receive transplants for lethal congenital heart disease.
There have been several recent reports suggesting that increased surveillance or preemptive antiviral therapy can lower the incidence of PTLD (27,28). These strategies have worked best in high-risk populations within liver transplantation, where pretransplant EBV and CMV serology are tested routinely. However, due to the heretofore low incidence of PTLD, pretransplant EBV testing has not been routine in kidney transplants. The UNOS data suggest that the incidence of PTLD is rising in several organ systems such as kidney, kidney-pancreas and heart-lung. Our data may help further identify the ideal high-risk patient for PTLD reduction strategies in different organ systems, especially where pretransplant EBV serology has not been obtained.
In summary, there has been a large number of PTLD cases in the United States in the last 12 years. There are clear trends towards increasing incidence in kidney, kidney-pancreas, heart-lung and liver transplants, though the incidence varies with the organ. Pediatric age < 18 years, Caucasian race and male gender are synergistic factors that increase the risk for PTLD development.
This study was based on UNOS Scientific Registry data from January 1988 to December 1999 and NAPRTCS data from 1987 to December 2000. The authors thank Ms. Samia Buckingham at UNOS for the data extraction and Ms. Brenda Hutson at the University of Florida for the statistical analysis of the UNOS data. This study was presented in part, in abstract form, at the Transplant 2001 conference in May 2001.
Commercial associations: none to disclose.
- 20Reversibility of lymphomas and lymphoproliferative lesions developing under cyclosporin-steroid therapy. Lancet 1984; i: 583–587., , et al.