Characteristics, management, and outcome of pediatric patients with post‐transplant lymphoproliferative disease—A 20 years' experience from Austria

Abstract Background Management of pediatric post‐transplantation lymphoproliferative disorder (PTLD) after hematopoietic stem cell (HSCT) and solid organ transplantation (SOT) is challenging. Aim This study of 34 PTLD patients up to 19‐years old diagnosed in Austria from 2000 to 2018 aimed at assessing initial characteristics, therapy, response, and outcome as well as prognostic markers of this rare pediatric disease. Methods and results A retrospective data analysis was performed. Types of allografts were kidney (n = 12), liver (n = 7), heart (n = 5), hematopoietic stem cells (n = 4), lungs (n = 2), multi‐visceral (n = 2), small intestine (n = 1), and vessels (n = 1). Eighteen/34 were classified as monomorphic PTLD, with DLBCL accounting for 15 cases. Polymorphic disease occurred in nine, and non‐destructive lesions in six cases. One patient had a non‐classifiable PTLD. Thirteen/34 patients are surviving event‐free in first remission (non‐destructive, n = 4/6; polymorphic, n = 4/9; monomorphic, n = 6/18). Fourteen/34 patients lacked complete response to first‐line therapy, of whom seven died. Four/34 patients relapsed, of whom two died. In 3/34 patients, death occurred as a first event. The 5‐year overall and event‐free survival rates were 64% ± 9% and 35% ± 9% for the whole cohort. Among all parameters analyzed, only malignant disease as the indication for transplantation had a significantly poor influence on survival. Conclusions This study shows PTLD still to be a major cause of mortality following SOT or HSCT in children. A continued understanding of the molecular biology of the disease shall allow to decrease treatment intensity for lower risk patients and to identify patients who may benefit from newer therapy approaches to improve outcome and decrease morbidity.

The 5-year overall and event-free survival rates were 64% ± 9% and 35% ± 9% for the whole cohort. Among all parameters analyzed, only malignant disease as the indication for transplantation had a significantly poor influence on survival.
Conclusions: This study shows PTLD still to be a major cause of mortality following SOT or HSCT in children. A continued understanding of the molecular biology of the disease shall allow to decrease treatment intensity for lower risk patients and to identify patients who may benefit from newer therapy approaches to improve outcome and decrease morbidity. which is a heterogeneous disorder ranging from benign hyperplasia to malignant lymphomas. [2][3][4] In pediatric patients, PTLD represents the most frequent type of malignant diseases secondary to solid organ transplantation (SOT) or hematopoietic stem cell transplantation (HSCT), with the overall risk of developing malignancies being 45-fold higher than in healthy individuals. [5][6][7] Epstein-Barr virus (EBV) infection is detected in most cases of pediatric PTLD. [8][9][10][11][12][13][14][15] Whereas >90% of the world's adult population already harbor EBV, the infestation rate is comparatively low in children. 16,17 Transplantation of grafts from EBV-seropositive adults results in primary EBV infection of children, thereby explaining a 2-to 4-fold higher risk of PTLD in children compared to adults. 5,15,[18][19][20] Reflecting the age-dependent seroprevalence rates, EBV-positivity occurs in only 50% of adult as compared to 85%-90% of pediatric PTLD. 15,16,19,[21][22][23] Notably, pediatric EBV-negative PTLD typically develops late, that is, >1 year after transplantation. 4,8,11,16,24 The World Health Organization (WHO) classification system differentiates between (i) non-destructive, benign forms, showing polyclonal cell populations only, (ii) polymorphic subtypes, which can present with either polyclonal or monoclonal proliferation patterns, and (iii) monomorphic, monoclonal subtypes, which are genuine malignant lymphomas, mostly of B-and, less frequently, of T-or NK-cell origin. 8,9 The incidence of lymphoproliferative diseases (LPD) following transplantations varies between 1% and 30%, depending on the presence of risk factors, 25,26 the most important of which are a mismatched EBV serostatus between recipient and donor, 8,10,11,27,28 a high-intensity immunosuppressive regimen, 3,11,12,19,29 and the type of allograft. The highest risk for PTLD after SOT results from having an EBV-positive organ donor and an EBV-naïve recipient. For PTLD following HSCT, a converse serological EBV constellation of donor and recipient is an established risk factor. Transplantation of intestine or lungs poses the highest risk for PTLD (20%), 12,26,[30][31][32] followed by cardiac or liver (2%-10%), 29,33,34 and kidney transplantations (1%-4%). 5,19,29,35,36 HSCT shows an incidence of 1%-8%, likewise depending on a range of risk factors, such as transplantation of T-cell depleted bone marrow (BM) or HLA-mismatched grafts from unrelated donors, as well as the use of anti-thymocyte globulin. [37][38][39][40] Concerning the latency period after transplantation, LPDs can be grouped into early-and late-onset diseases, referring to cases within and later than the first year after transplantation. 25 Herein, we describe 34 SOT and HSCT patients with PTLD diagnosed in Austria within a period of nearly 20 years.

| Patients
Patients had to fulfill three inclusion criteria to be enrolled into this retrospective study: (i) diagnosis of PTLD as established by a reference pathologist according to the WHO classification valid at the time of diagnosis, (ii) diagnosis and treatment of PTLD between 2000 and 2018, and (iii) treatment of PTLD carried out at one of the four participating Austrian centers offering performance and aftercare of SOT and/or HSCT. Thirty-four patients aged <19 years meeting the inclusion criteria were identified and, hence, included in our study.

| METHODS
Classification of PTLD was based on histopathology according to the contemporary WHO classification system. 41,42 EBV-association was defined by a positive in-situ hybridization test of EBV-encoded RNA (EBER) in the tissue(s) analyzed. Patient data included demographics and information on disease, treatment, response, and outcome. Serum analysis was performed to detect EBV in peripheral blood. In dependence of the laboratory, to which the samples were submitted, the threshold for EBV positivity was either ≥5 or ≥ 100 copies/ml. Tumor stage was retrospectively defined according to the All patients were treated after informed consent from the patient, patient's parents or legal guardians had been obtained.

| Statistical analysis
Event-free survival (EFS) was calculated from the date of diagnosis to the date of first event. Events considered were lacking complete response at the end of a therapy line, relapse, or death, whichever occurred first. Overall survival (OS) was defined as the time from diagnosis to death from any cause or the date of last follow-up. EFS and OS were estimated according to the Kaplan-Meier method; differences between groups were evaluated with the log-rank test. P-values ≤0.05 were considered statistically significant.
We aimed at staging the 34 patients according to the IPNHLSS.

| Treatment and response
A detailed description of the therapy, response and outcome of the 34 patients is given in Figure 1 and Table 3. One/34 patients died from disease before any therapy. Reduction of immunosuppression (RIS) was the first therapeutic step taken in 32 of the remaining Abbreviations: No., number; Tx, transplantation; LDH; lactate dehydrogenase; PTLD, post-transplant lymphoproliferative disease. a There was one patient included who developed the PTLD after liver transplantation, but had also undergone a previous kidney transplantation. 33

| Outcome according to histology
Of all six patients with non-destructive disease, four received rituximab±steroids, and two lacked complete response and received second-line therapy ( Table 3). All patients finally achieved a CCR.
Among the nine patients with polymorphic disease, eight received RIS upfront, and one patient died before having received any therapy.
Two/eight patients received RIS only, by which one achieved a CCR.
The other patient's lesions regressed from polymorphic to nondestructive PTLD (classified as PR). Six/8 patients received rituximab±steroids, with four achieving a CR and two who did not (both died). Two of the four patients relapsed, of whom one died from the PTLD, and the other one achieved a second CCR. Of all eight patients with polymorphic disease having received treatment, three achieved a first CCR, and one a second CCR, while one showed a PR and three patients died (PTLD-related, n = 2; PTLD-unrelated, n = 1) ( Table 3).

| Events
Thirteen/34 patients experienced no events and are alive in first CCR.
Fourteen/34 patients lacked complete response to first-line therapy.
One patient died amidst relapse therapy due to an unknown cause.
One case of candida septicemia was considered a PTLD-treatmentrelated death. The group of 4/12 PTLD-unrelated deaths consisted of a septicemia, an acute respiratory distress syndrome combined with secondary graft failure following HSCT, a diffuse alveolar hemorrhage syndrome, and a cardiac arrest, respectively. and EFS rates were 64% ± 9% and 35% ± 9% for the whole study cohort, respectively (Figure 2). When evaluating the 5-year OS according to the PTLD subtypes, it was 100% for non-destructive, 53% ± 17% for polymorphic, and 65% ± 12% for monomorphic PTLD.
A variety of factors were evaluated concerning their impact on OS and EFS (Supplemental Table 2 among the EBV-positive patients. Accordingly, all four EBER-negative patients had a CD20-negative tumor, but no reduced survival. The incidence of LPD following transplantations shows a biphasic distribution, with a first peak occurring within 12 months, and another either at 3 to 5 or at 7 to 10 years after transplantation. 11,25,[49][50][51][52] Our study showed that half of the patients presented with early-onset dis- assessed differently by the respective organ specialist. 11,35,57,58 Remarkably, none of the 32 patients included in our study who received RIS suffered from graft-failure or -loss due to PTLD-therapy. Non-destructive PTLD and low-risk patients show high response rates to RIS alone, but it is not recommended as a sole therapy in cases of monomorphic PTLD, aggressive disease or patients with a high tumor burden. 11,35 Among our cohort, three underwent RIS only (non-destructive, n = 1; polymorphic, n = 2) and thereby achieved a CCR (n = 2) or PR (n = 1), and another 2 patients achieved CCR by RIS and surgery only (non-destructive, n = 1; monomorphic, n = 1).
Another substantial part of PTLD-therapy is B-cell depletion due to the use of rituximab. It is recommended as first-line treatment for all CD20-positive subtypes sequential or in parallel to RIS. Altogether, treatment with rituximab has led to better overall survival in children and adolescents with PTLD. 22 to the relatively small number of patients included in our study.
Children diagnosed with PTLD tend to have a better prognosis compared to adults, which seems to be due to more favorable PTLDsubtypes (EBV-positive) and less treatment-related complications. 15,20 The prospective PTLD-1 trial implemented sequential treatment with rituximab followed by a CHOP-regimen in adult CD20-positive patients with mostly monomorphic PTLD, resulting in a median OS of 6.6 years. Considering that response to rituximab predicted favorable OS, another prospective study subsequently treated the patients with CR to 4 weeks of rituximab with further rituximab consolidation, whereas non-responders switched to R-CHOP. Thereby, a 3-year OS of 78% (compared to 69% in PTLD-1) was achieved. 59,61 The prospective Ped-PTLD 2005 trial treated pediatric patients with CD20-positive PTLD following SOT with 3 weeks of rituximab followed by either rituximab or chemotherapy (mCOMP) depending on initial response, achieving a 2-year OS of 86%, and 67% surviving event-free. 64 According to the literature available, 5-year survival rates for pediatric PTLD are around 53-80%. 25,61,65,74,75 The 3-and 5-year OS rates in our patient cohort were 69% ± 8% and 64% ± 9%, respectively, fitting well to the hitherto reports.
The primary aim of this study was to investigate the characteristics and outcome of PTLD in Austrian patients over a period of 20 years. Special attention was paid to factors influencing the patient's prognosis. In childhood and adolescence, PTLD is the largest group of secondary malignant diseases, making it essential to define coherent diagnostic and therapeutic guidelines addressing the heterogeneity of this disease. Considering the absolute number of affected people, it is still defined as an Orphan Disease, and both, a transnational cooperation and a compilation of comprehensive scientific studies, including series like the present one, are essential to optimize the management of PTLD, as it is still a major cause of mortality following transplantation.