Burkitt post-transplantation lymphoma in adult solid organ transplant recipients

Sequential immunochemotherapy with rituximab (R) followed by cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) or R-CHOP is safe and effective in an analysis of 8 patients

Authors


  • The German Study Group on Post-Transplantation Lymphoproliferative Disorder is a member of the German Competence Network Malignant Lymphomas.

  • R.U.T. is the principal investigator, coordinated the research and takes primary responsibility for the article. R.U.T., P.R., R.N., H.L., S.O., J.A., M.P., and H.R. recruited the patients. D.L. performed fluorescence in situ hybridization analyses, I.A. served as reference pathologist. B.G. served as reference virologist. H.Z. and R.U.T. collected the data, performed the analyses, and wrote the article.

Abstract

BACKGROUND:

Burkitt lymphoma post-transplantation lymphoproliferative disorder (Burkitt-PTLD) is a rare form of monomorphic B-cell PTLD for which no standard treatment has been established. Currently, the treatment of Burkitt lymphoma outside the post-transplantation setting involves high doses of alkylating agents, frequent dosing, and intrathecal and/or systemic central nervous system prophylaxis. In PTLD, however, such protocols are associated with considerable toxicity and mortality.

METHODS:

The authors present a retrospective series of 8 adult patients with Burkitt-PTLD. Six patients were reported to the prospective German PTLD registry or were enrolled in the PTLD-1 trial, and 2 patients had received treatment before 2000, thus allowing for comparison with the pre-rituximab era.

RESULTS:

Seven of the 8 patients were men. The median age at presentation was 38 years, and the median time since transplantation was 5.7 years. Five of 8 patients had histologically established, Epstein-Barr virus-associated disease, and 7 of 7 patients were positive for a MYC translocation. Five of 8 patients received sequential immunochemotherapy (4 courses of rituximab [R] followed by 4 cycles of cyclophosphamide, doxorubicin, vincristine, and prednisolone [CHOP] or R plus CHOP [R-CHOP]). In this group, 5 of 5 patients reached complete remission (CR), and their overall survival (OS) was significantly longer (P = .008) compared with the OS for 2 of 8 patients who received first-line CHOP and did not respond. One of 8 patients (who had stage IV disease with meningiosis) received combination therapy (cyclophosphamide pretreatment, rituximab, intrathecal chemotherapy, whole-brain irradiation, and radioimmunotherapy) and reached CR. Overall, 6 of 8 patients reached CR; and, after a median follow-up of 4.7 years (range, 1.7-4.8 years), the median OS was 36.7 months. There was no treatment-related mortality under first-line therapy.

CONCLUSIONS:

In the largest adult case series in Burkitt-PTLD to date, sequential immunochemotherapy with rituximab followed by standard CHOP or R-CHOP was a both safe and effective treatment. Cancer 2012. © 2012 American Cancer Society.

INTRODUCTION

Burkitt lymphoma (BL)1 is characterized clinically by extranodal manifestations of exceedingly rapid growth and histologically by diffuse growth of fairly uniform, medium-sized lymphoid cells with frequent mitotic figures; expression of germinal center B-cell–associated antigens, including cluster of differentiation 20 (CD20 [acute lymphoblastic leukemia antigen]), CD10 (acute lymphoblastic leukemia antigen), and B-cell lymphoma 6 (BCL6); a very high proliferative index; and (usually) the absence of BCL2 expression. In most (but not all) cases of BL, a v-myc myelocytomatosis viral oncogene homolog (avian) (MYC) translocation, most commonly to the immunoglobulin heavy-chain (IGH) gene, can be demonstrated. Epstein-Barr virus (EBV) association is very common in the endemic subtype of BL; in addition, the World Health Organization (WHO) classification recognizes sporadic BL and immunodeficiency-associated BL clinical variants.2

Post-transplantation lymphoproliferative disorder (PTLD), conversely, is a spectrum of lymphoproliferations associated with the use of potent immunosuppressive drugs after transplantation3 and ranges from polyclonal, early lesions associated with primary EBV infection to monomorphic lymphomas, as reflected in the WHO classification.4 The majority of monomorphic PTLD lesions resemble diffuse large B-cell lymphoma (DLBCL).5, 6 The prospective PTLD-1 trial recently demonstrated a 90% overall response rate and a median overall survival of 6.6 years with sequential immunochemotherapy in adult patients with CD20-positive PTLD.7 However, PTLD subtypes like BL (Burkitt-PTLD) are rare; and, currently, their treatment is not based on clinical trials.

An association with EBV, a ubiquitous lymphotropic and epitheliotropic γ1-herpesvirus first discovered in BL cell cultures,8 is detectable not only in BL but also in other lymphoma entities, including monomorphic PTLD in approximately 50% of patients.5-7 However, viral gene expression varies considerably: Although it is limited in BL (latency type I), the (highly antigenic) expression of EBV proteins, such as latent membrane protein 1 (LMP1) and Epstein-Barr nuclear antigen 2 (EBNA2), is frequent, particularly in early PTLD (latency type III).9

The oncogene MYC is a key player in lymphomagenesis and is involved in cell cycle control on multiple levels.10 MYC translocations are characteristic for BL but can be identified in other lymphoma entities, particularly DLBCL, in which they have been associated with a poor prognosis.11 In the latter, MYC translocations usually are associated with a more complex karyotype and additional translocations (for example, of BCL2 and/or BCL6).

The current treatment of (sporadic) BL in adults is based on chemotherapy, including high doses of alkylating agents, frequent dosing and intrathecal and/or systemic central nervous system (CNS) prophylaxis (for a review, see Perkins and Friedberg12). This includes human immunodeficiency virus (HIV)-associated BL in the era of highly active antiretroviral therapy13, 14 as well as patients aged >40 years.15 Addition of the monoclonal anti-CD20 antibody rituximab to BL protocols, such as combined cyclophosphamide, vincristine, doxorubicin, and methotrexate/etoposide, ifosfamide, and cytarabine (CODOX-M/IVAC) and hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone (Hyper-CVAD), has led to significantly lower recurrence rates.16, 17 In PTLD, even first-line cyclophosphamide, doxorubicin, vincristine, and prednisolone (CHOP) every 3 weeks (CHOP-21) is associated with considerable toxicity and a treatment-related mortality (TRM) rate of up to 31% in patients with DLBCL-PTLD.18 Treatment of post-transplantation BL and Burkitt-like acute lymphoblastic leukemia within an adult BL protocol led to TRM in 3 of 5 patients (60%).19 Conversely, sequential immunochemotherapy with rituximab and CHOP in the PTLD-1 trial had a TRM rate of only 10.6%.7

Other previous case series in adult (n = 5)20 and pediatric (n = 12)21 Burkitt-PTLD have been limited by their retrospective formats and limited access to clinical and pathologic data. Here, we present a retrospective series of 8 patients from the German prospective PTLD registry, the PTLD-1 trial, and the Department of Hematology, Charite Virchow Klinikum, Berlin, including clinicopathologic features and detailed information on treatment and outcome.

MATERIALS AND METHODS

Patients

We initiated the prospective German PTLD registry with the goal of assessing clinical features, treatment options, and outcomes in adult patients with rare PTLD subtypes after solid organ transplantation (SOT) in 2006. Because Burkitt-PTLD is CD20-positive, patients also were eligible for enrollment in the European prospective phase 2 trial investigating sequential treatment with rituximab and CHOP-21 in patients with CD20-positive B-cell PTLD who were unresponsive to immunosuppression reduction (IR) (PTLD-1). Here, we report the patients with Burkitt-PTLD that were enrolled on the registry or on the PTLD-1 trial (from 12/2012). To allow for comparison with the prerituximab era, we retrospectively retrieved the records of patients who presented to the Hematology Department at the Charite Virchow Clinic in Berlin from 1995 to 2003 with a diagnosis of Burkitt-PTLD. We excluded patients who were negative for MYC breaks and those who were diagnosed with B-cell PTLD, unclassifiable, with features intermediate between DLBCL and BL. Follow-up data were reviewed for all patients up to July 2011. All patients provided written informed consent.

Histology, Epstein-Barr Virus Association, and Fluorescence in Situ Hybridization

The diagnosis of PTLD was based on the examination of formalin-fixed, paraffin-embedded tissue specimens that were obtained either by open biopsy or by core-needle biopsy. All diagnostic tissue samples, including conventional histology samples (hematoxylin and eosin stain, Giemsa stain) as well as immunohistochemistry samples (a minimal panel composed of CD20, CD10, BCL6, BCL2, CD3, CD21, Ki67, and terminal deoxynucleotidyl transferase) were reviewed by an expert pathologist (I.A.) and were classified according to 2008 WHO classification criteria.4 An association with EBV was confirmed by in situ hybridization for Epstein Barr-encoded small RNAs (EBERs). Interphase fluorescence in situ hybridization (FISH) was performed on formalin-fixed, paraffin-embedded tissues using break-apart assays for MYC, BCL2, and BCL6 (all from Dako, Glostrup, Denmark) and the dual-color, dual-fusion translocation probes LSI MYC/IGH and LSI BCL2/IGH (all from Abbott/Vysis, Downers Grove, Ill). For signal detection, an Axio Imager Z1 (Zeiss, Oberkochen, Germany) and Isis software (version 5.3.1; MetaSystems, Altlussheim, Germany) was used.

Treatment

For patients who were treated as part of or on the same regimen as the PTLD-1 trial up to 2007, the protocol was sequential treatment with rituximab followed by 4 cycles of CHOP plus granulocyte-colony–stimulating factor (G-CSF) starting 4 weeks after the last dose of rituximab (Fig. 1A).7 After an interim analysis in 2007 demonstrated that the response to treatment with rituximab (complete remission [CR]/partial remission [PR] vs stable disease [SD]/progressive disease [PD]) was a significant predictor of overall survival, risk stratification was introduced according to the patients' response to rituximab treatment.22 Patients who achieved CR after 4 courses of rituximab carried on with 4 additional courses of rituximab at 3-week intervals, whereas all others received R-CHOP-21 plus G-CSF instead of CHOP-21 plus G-CSF (risk-stratified sequential treatment) (Fig. 1B). The remaining patients were treated on individual regimens.

Figure 1.

Treatment algorithms are illustrated for chemotherapy protocols with rituximab (R) and combined cyclophosphamide, doxorubicin, vincristine, and prednisolone (CHOP) in the European Post-Transplantation Lymphoproliferative Disorder 1 (PTLD-1) trial. (A) Sequential treatment is illustrated. CHOP-21 indicates CHOP every 3 weeks; G-CSF, granulocyte–colony-stimulating factor. (B) Risk-stratified sequential treatment is illustrated. D indicates day; PR, partial remission; SD, stable disease; PD, progressive disease; CR, complete remission.

Assessment of Treatment Response and Outcome

The extent of existing disease was determined through a complete patient history, physical examination, laboratory investigations (including a full blood count, lactate dehydrogenase level [upper limit, 248 U/L], renal and liver function tests, as well as determination of the EBV DNA load in peripheral blood), bone marrow biopsy, and computed tomography scans of the chest, abdomen, and pelvis. Tumor response to treatment was defined according to WHO criteria. Disease-free survival was calculated from the first evidence of CR to the date of either PD or death from any cause; response duration was calculated from the date of best response (CR or PR) to the date of either PD or death from PTLD; and overall survival (OS) was calculated from the date of PTLD diagnosis to the date of death from any cause. In addition, CNS relapse and CNS PD were defined as secondary CNS events. Clinical data on all patients were collected before, during, and at least 4 weeks, 6 months, 12 months, and 24 months after treatment. Additional data on patient characteristics were retrieved from databases at the different transplantation centers.

Data Analysis

Exploratory univariate analyses were performed by applying log-rank analyses to Kaplan-Meier statistics for time-to-event outcomes and chi-square tests to categorical variables. The level of significance in all analyses was set at P < .05. Statistical tests were performed using IBM SPSS software (version 19.0.0.1; SPSS, Inc., Chicago, Ill).

RESULTS

Baseline Findings

By mid-2011, 195 patients had been reported to the German PTLD registry or enrolled on the PTLD-1 trial. Of this cohort, 6 patients (3%) had been diagnosed with Burkitt-PTLD (3 of 6 had been treated on the PTLD-1 trial, 2 of 6 patients had received the PTLD-1 trial regimen [Patients 1-5], and the other patient had received an individual regimen [Patient 8]). One patient had a diagnosis of B-cell PTLD, unclassifiable, with features intermediate between DLBCL and BL and was not included in this series. Two other patients were identified from Charite records (Patients 6 and 7). An additional patient from the Charite records was excluded (this patient was diagnosed originally with Burkitt-PTLD but was negative for an MYC break).

The clinical baseline characteristics of the 8 patients with Burkitt-PTLD are summarized in Table 1. The median age at diagnosis was 38 years (range, 18-54 years), and 7 of the 8 patients (88%) were men. The transplanted solid organs were the kidney in 3 of 8 patients, the kidney and pancreas in 1 of 8 patients, the liver in 3 of 8 patients, and the heart in 1 of 8 patients. Underlying diseases included end-stage renal disease cirrhosis, primary sclerosing cholangits, and dilated cardiomyopathy. The median time from transplantation to diagnosis of Burkitt-PTLD was 5.7 years (range, 85 days to 12.5 years). Seven of 8 cases were late Burkitt-PTLD, and the only case of early Burkitt-PTLD (ie, within 1 year of transplantation) occurred in a man aged 54 years who was seronegative for EBV immunoglobulin G at the time of liver transplantation for cirrhosis secondary to chronic hepatitis C virus infection. Immunosuppression at the time of Burkitt-PTLD diagnosis ranged from single-agent cyclosporine A (CyA) or tacrolimus (FK) to double therapy. FK-based immunosuppression was associated with earlier disease onset after transplantation compared with CyA-based immunosuppression (median time to onset of PTLD, 2.5 years vs 8.5 years; P = .022), whereas the number of immunosuppressants (monotherapy vs combination treatment) had no significant impact on the time of onset.

Table 1. Clinical Baseline Characteristics
Patient No.Year of DxSexAge at Dx of PTLD, yTransplantationDisease Resulting in Solid Organ TransplantationTime From Transplantation to PTLD, dImmunosuppression at Dx of PTLD
CNI/mTORAntimetaboliteSteroid
  1. Abbreviations: AZA, azathioprine; CNI, calcineurin inhibitor; CyA, cyclosporin A; DCM, dilated cardiomyopathy; DM, diabetes mellitus; Dx, diagnosis; ESRD, end-stage renal disease; FK, tacrolimus; HCV, hepatitis C virus; MMF, mycophenolate mofetil; mTOR, mammalian target of rapamycin; PCKD, polycystic kidney disease; PSC, primary sclerosing cholangitis; PTLD, post-transplantation lymphoproliferative disease.

12002Woman40LiverCirrhosis (cryptogenic)4555CyA 150 mg/d
22006Man36Kidney, pancreasESRD (DM type I)2364CyA 190 mg/dMMF 1000 mg/d
32006Man38KidneyESRD3151CyA 220 mg/dMMF 720 mg/d
42007Man49KidneyESRD966CyA 250 mg/dPrednisolone 5 mg/d
52009Man18KidneyESRD (PCKD)1763FK 1.5 mg/dPrednisolone 5 mg/d
61995Man54LiverCirrhosis (chronic HCV infection)85FK 8 mg/dPrednisolone 15 mg/d
71998Man37HeartDCM2987CyA 300 mg/dAZA 150 mg/d
82008Man37LiverPSC906FK 2 mg/d

Immunohistochemistry revealed that tumor cells were positive for CD20 in all 8 patients. In all tumors examined, tumor cells were positive for CD10 (7 of 7 tumors), positive for BCL6 (5 of 5 tumors), and negative for BCL2 (7 of 7 tumors). The Ki67 level was >95% throughout. With regard to EBV status, the neoplastic cells were latently infected by EBV in 5 of 8 tumors, as demonstrated by EBER in situ hybridization (63%). Wherever examined, tumor cells did not express EBNA2 (0 of 3 tumors) or LMP1 (0 of 5 tumors), confirming EBV latency type I in this lymphoma entity. EBV DNA load in peripheral blood was positive in 1 of 3 patients examined. FISH analysis revealed evidence of an MYC translocation in all 7 of 7 tumors examined. In 5 tumors with sufficient tissue samples, the presence of additional translocations involving the BCL2 and BCL6 genes could be excluded (Table 2). In 1 tumor, an insufficient amount of tissue sampled did not allow FISH analysis, and the diagnosis, which was based on morphology, proliferative rate, and immunophenotype, was confirmed independently by 2 expert hematopathologists.

Table 2. Immunohistochemistry, Fluorescent in Situ Hybridization, and Epstein-Barr Virus Association
Patient No.DiagnosisCD20CD10BCL6BCL2MYC BreakBCL2−/BCL6- Breakt (8;14) MYC/IGHEBV Association of PTLD (EBER-ISH)
  • Abbreviations BCL2, B-cell lymphoma 2; BCL6, B-cell lymphoma 6; CD10 cluster of differentiation 10 (acute lymphoblastic leukemia antigen); CD20, cluster of differentiation 20 (B-lymphocyte antigen); EBER-ISH, in situ hybridization for Epstein-Barr–encoded small RNAs; EBV, Epstein-Barr virus; IGH, immunoglobulin heavy-chain; MYC, v-myc myelocytomatosis viral oncogene homolog (avian); ND, not done; PTLD, post-transplantation lymphoproliferative disease.

  • a

    There was an insufficient amount of tumor material.

  • b

    Dual-color dual-fusion translocation assays could not be evaluated because of insufficient tissue quality.

1Burkitt-PTLDPositivePositiveNDNegativeNDaNDa/NDaNDaNegative
2Burkitt-PTLDPositivePositivePositiveNegativePositiveNegative/negativePositivePositive
3Burkitt-PTLDPositivePositivePositiveNegativePositiveNegative/negativeNegativeNegative
4Burkitt-PTLDPositivePositivePositiveNegativePositiveNegative/negativePositivePositive
5Burkitt-PTLDPositivePositivePositiveNegativePositiveND/negativeNDPositive
6Burkitt-PTLDPositiveNDNDNegativePositiveNegative/negativeNDbNegative
7Burkitt-PTLDPositivePositiveNDNDPositiveNegative/negativeNDbPositive
8Burkitt-PTLDPositivePositivePositiveNegativePositiveND/NDNDPositive

Clinical Presentation

Superficial, space-occupying lesions and secondary effects of mass lesions ranging from VIth nerve palsy to flank pain were the most common presenting complaints. At presentation, B symptoms were present in 4 of 8 patients, and Ann Arbor stage IV was the most frequent disease stage (5 of 8 patients). Lymph node disease and extranodal disease manifestations were present in 7 of 8 patients each; and, in 6 of 8 patients, both were present. The most common extranodal site was bone marrow infiltration in 4 of 8 patients. Involvement of stomach, liver, and kidneys occurred in 2 patients each. CNS disease (meningiosis and sinus cavernosus infiltration) was present at diagnosis in 1 of 8 patients (there were no CNS mass lesions). Disease manifestations >7.5 cm (bulky) were present in 3 of 8 patients at diagnosis. Regarding baseline laboratory investigations, serum lactate dehydrogenase was elevated in 5 of 7 patients (range, 290-1846 U/L), β2 microglobulin was elevated in 4 of 4 patients (range, 2.5-10.5 mg/L), and serum albumin was low in 1 of 4 patients (Table 3).

Table 3. Clinical Presentation at Diagnosis of Burkitt-PTLD
Patient No.Presenting ComplaintOrgan InvolvementLymph Node InvolvementBulky Disease (>7.5 cm)Bone Marrow InvolvementCNS InvolvementAnn Arbor StageLDH, U/Laβ2-M, mg/mLbAlbumin, g/dLc
  • Abbreviations: B2-M, β2 microglobulin; CNS, central nervous system; LDH, lactate dehydrogenase; ND, not done.

  • a

    The upper limit of normal for LDH is <248 U/L.

  • b

    The normal range for β2-M is 0.8-2.4 mg/mL.

  • c

    The normal range for albumin is 3.6-5.0 g/dL.

1IncidentalNoneInfrahepatic and peripancreaticYesNoNoIIA2413.5ND
2IncidentalStomachMesenteric, para-aortic, retroperitonealNoNoNoIIAE29042.1
3Neck swellingStomachRight submandibular, right cervical, right supraclavicular, mediastinalNoNoNoIIIAE1902.53.8
4Neck swellingNoneSubmandibular, cervical, supraclavicularNoYesNoIVB1157NDND
5Neck swellingRight submandibular gland, bone, liverRight cervicalYesNoNoIVAENDNDND
6Left flank painLeft perirenal bulk (20 cm)NoneYesYesNoIVB172010.5Within normal range
7Left flank painLiver, both kidneys, heartAxillaryNoYesNoIVBE493ND4.14
8Right-sided VIth nerve palsySinus cavernosus, meningiosisMediastinal, paraaortic, mesenteric, liver hilusNoYesYesIVBE1846NDND

Treatment and Treatment Outcome

Three of the patients presented here were enrolled in the PTLD-1 trial, and 2 were treated similarly. Therefore, 5 of 8 patients (Patients 1-5) received immunochemotherapy with rituximab (Roche Pharma AG, Basel, Switzerland) and CHOP as first-line treatment, either as sequential treatment (3 patients) or in sequence and in combination (risk-stratified sequential treatment; 2 patients). Five of 5 patients responded (4 CRs and 1 PR that was converted to CR with 2 cycles of carboplatin plus etoposide [CE]) (see Table 4). After a median follow-up of 4.7 years (range, 1.7-4.8 years), only 1 in 5 patients who received rituximab and CHOP had relapsed. Having received successful second-line salvage therapy (R-CE), that patient remained alive without evidence of disease at follow-up 8.8 months after second CR. There was only 1 death after sequential therapy (not PTLD-related). In contrast, 2 patients (Patients 6 and 7) who were diagnosed in 1995 and 1998 and received CHOP alone developed PD on first-line therapy. Both patients rapidly developed secondary CNS events and died shortly thereafter. One patient (Patient 8) who had stage IV disease with meningiosis received cyclophosphamide pretreatment (200 mg/m2 on days 1-5), rituximab combined with intrathecal chemotherapy (methotrexate, cytarabine), followed by whole-brain irradiation and consolidating radioimmunotherapy (yttrium-90 ibritumomab tiuxetan; Bayer Pharma AG, Berlin, Germany). CNS relapse occurred shortly thereafter, and the patient died 8.4 months after the first diagnosis of Burkitt-PTLD. Therefore, in this series, the OS of patients who received sequential rituximab and CHOP was significantly longer compared with the OS of patients who received CHOP alone (P = .008) or any other treatment (P = .004) (Fig. 2).

Figure 2.

Overall survival is illustrated for 8 adult patients with Burkitt lymphoma post-transplantation lymphoproliferative disorder who received sequential immunochemotherapy consisting of rituximab plus combined cyclophosphamide, doxorubicin, vincristine, and prednisolone (CHOP or R-CHOP) (solid line; n = 5 patients) versus any other therapy (dashed line, n = 3 patients; P = .004).

Table 4. First-Line Treatmenta
Patient No.IRSubsequent TreatmentGrade 3/4 Hematologic and Infectious ToxicityResponseResponse Duration, d
  • Abbreviations: Ara-C, cytarabine; AZA, azathioprine; CE, carboplatin and etoposide; CHOP, cyclophosphamide, doxorubicin, vincristine, and prednisolone; CR, complete remission; CyA, cyclosporin A; FK, tacrolimus; G-CSF: granulocyte-colony–stimulating factor; IR: immunosuppression reduction; IT, intrathecal; MBq, megabecquerels; MMF, mycophenolate mofetil; MTX, methotrexate; ND, not documented; PD, progressive disease; PR, partial remission; R-CHOP, rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisolone.

  • a

    The following chemotherapy protocols were followed: sequential treatment (PTLD-1 trial) consisted of 4 cycles of rituximab (375 mg/m2 every 8 days) and 4 cycles of CHOP-21 plus G-CSF; risk-stratified, sequential treatment (PTLD-1 third amendment) consisted of 4 cycles of rituximab followed by 4 cycles of R-CHOP-21 plus G-CSF (unless the patient achieved CR after 4 cycles of rituximab; this was not applicable for the patients discussed here); CE consisted of carboplatin at dose based on the area under the curve (AUC) of 4 on day 1 and etoposide 120 mg/m2 on days 1 through 3 every 22 days; CHOP-21 consisted of cyclophosphamide 750 mg/m2 on day 1, doxorubicin 50 mg/m2 on day 1, vincristine 1.4 mg/m2 (maximum, 2 mg) on day 1, and prednisolone 100 mg on days 1 through 5 every 22 days; cyclophosphamide pretreatment consisted of cyclophosphamide 200 mg/m2 on days 1 through 5; R-CHOP-21 consisted of rituximab 375 mg/m2 on day 1, cyclophosphamide 750 mg/m2 on day 1, doxorubicin 50 mg/m2 on day 1; vincristine 1.4 mg/m2 (maximum, 2 mg) on day 1, and prednisolone 100 mg on days 1-5 every 22 days; rituximab therapy consisted of rituximab 375 mg/m2 every 8 days; and Ara-C/MTX IT consisted of Ara-C 40mg IT and MTX 15mg IT.

  • b

    The following chemotherapy dose reduction (% of target dose administered) was used: The first cycle of CHOP was given at an average dose of 50% with a subsequent, stepwise dose increase up to 90% in Cycle 4 and no vincristine in Cycles 5 and 6 because of peripheral neuropathy.

  • c

    The following chemotherapy dose reduction (% of target dose administered) was used: 80% carboplatin in Cycle 2.

  • d

    The following chemotherapy dose reduction was used: vincristine 2 mg on day 1, cyclophosphamide 750 mg on day 3, doxorubicin 50 mg on day 3, and prednisolone 200 mg on days 1 through 5.

  • e

    The following chemotherapy dose reduction (% of target dose administered) was used: 73% cyclophosphamide, 80% doxorubicin, and 0% vincristine.

1No changeSequential treatment with 4 cycles of rituximab and 6 cycles of CHOP plus G-CSFbGrade 3 leukopeniaPR>930
  2 Cycles of CE plus G-CSFcGrade 4 thrombocytopenia, grade 3 febrile neutropeniaCR 
  Involved-field irradiationNDCR 
2CyA target trough serum levels lowered from 150-200 ng/mL to 100-150 ng/mLSequential treatment with 4 cycles of rituximab and 4 cycles of CHOP plus G-CSFGrade 3 leukopeniaCR>1545
3CyA switched to sirolimus (3 mg); MMF reduced to 50%Sequential treatment with 4 cycles of rituximab and 4 cycles of CHOP plus G-CSFGrade 3 leukopeniaCR>1635
4CyA stopped; prednisolone increased to 20 mgRisk-stratified sequential treatment with 4 cycles of rituximab and 4 cycles of R-CHOP plus G-CSFNDCR>1234
5No changeCyclophosphamide pretreatment followed by risk-stratified sequential treatment with 4 cycles of rituximab and 4 cycles of R-CHOP plus G-CSFNDCR77
6No change1 Cycle of CHOPdGrade 4 leukopenia, grade 3 pneumoniaPD 
7AZA stopped3 Cycles of CHOPeNonePD 
8FK reduced by 50%Cyclophosphamide pretreatment followed by 5 cycles of rituximab monotherapy and whole-brain irradiation and 1 cycle of Ara-C/MTX ITNoneCR120
  1003 MBq yttrium-90 ibritumomab tiuxetanNDCR 

Regarding the roles of other treatment components, IR produced no effect in 4 patients who had interim staging data available to allow an evaluation of IR in combination with single-agent rituximab or CHOP. One PR was achieved with the combination of IR, cyclophosphamide pretreatment, and rituximab. Intrathecal chemoprophylaxis (triple therapy with dexamethasone, methotrexate, and cytarabine) was received by only 1 of 7 patients without CNS disease at diagnosis (Patient 5). That patient did not develop a secondary CNS event. Because there was only 1 such patient, the number was too small to draw any meaningful conclusions about the impact of this prophylaxis.

Overall, 6 of 8 patients responded to treatment and reached CR. Two of 6 patients in CR developed a disease relapse (at 77 days and 120 days, respectively). Disease relapse was systemic in 1 patient and meningeal in the other. Both patients who had PD on first-line therapy developed CNS disease (Table 5). Subsequent treatment strategies for these patients included chemotherapy with rituximab, carboplatin, and etoposide23 for systemic relapse and intrathecal methotrexate or systemic HD cytarabine for the treatment of secondary CNS events. The median disease-free survival for the 6 patients who achieved CR was 30.6 months, whereas the median OS for all patients was 36.7 months. Histologic EBV association had no impact on OS (P = .281; data not shown).

Table 5. Long-Term Outcome and Overall Survival
Patient No.Current StatusRelapse/Progressive DiseaseOS, dCause of Death
  • Abbreviations: CR, complete remission; DR, disease related; NDR, not disease related; NTR, not treatment related; OS, overall survival; TRM, treatment-related mortality.

  • a

    This patient developed sepsis after undergoing retransplantation because of cirrhosis (in CR).

  • b

    This patient had respiratory arrest after receiving intrathecal methotrexate for meningiosis that was affecting the brainstem.

1Dead (first CR) 1118Sepsis (NTR/NDR)a
2First CR >1701 
3First CR >1768 
4Second CR >1372 
5Second CRSubmandibular (after 77 d)>614 
6DeadMeningeal29Respiratory arrest (DR/TRM)b
7DeadMeningeal139Pneumonia (TRM)
8DeadMeningeal, cerebral (after 120 d)310Hemorrhage (TRM)

The development of secondary CNS events was a significant predictor of OS in univariate analysis (P = .004) and was associated with the presence of bone marrow infiltration (P = .028) and B-symptoms P = .028) at diagnosis. Crucially, the receipt of rituximab and especially its combination with CHOP significantly reduced the risk of secondary CNS events (P = .035 and P = .005, respectively). Correlations with other baseline characteristics or treatments did not reach statistical significance (data not shown).

Treatment Toxicity

There was no treatment-related mortality (TRM) on first-line therapy. Grade 3 hematologic toxicity was common despite supportive treatment with G-CSF, and grade 4 hematologic toxicity occurred without G-CSF. We observed no graft toxicity during chemotherapy. Patient 1 died from sepsis after complications of retransplantation for transplant cirrhosis 15 years after undergoing liver transplantation and 3 years after successful treatment of PTLD. Second-line therapy, however, was associated with a TRM rate of 50% (2 of 4 patients: hemorrhage secondary to pancytopenia after HD cytarabine and pneumonia after CHOP-21 and irradiation) (see Table 5).

DISCUSSION

Our case series confirms Burkitt-PTLD as a rare form of adult PTLD, constituting only 3% of the PTLD patients reported to our registry and the PTLD-1 trial. The median time from transplantation to diagnosis (5.7 years) in our series also confirmed the results from previous case series of adult Burkitt-PTLD20 (4.5 years) and pediatric Burkitt-PTLD21 (4.3 years) and is nearly identical to the 5.5 years we observed in patients with DLBCL-PTLD.7 Other rare forms of PTLD, such as plasmacytoma-like PTLD24 and plasmablastic PTLD,25 manifest later (median, 8.3 years and 12.8 years, respectively). The clinical features of Burkitt-PTLD are similar to those of BL in the nontransplantation population,2 including frequent early CNS relapse with poor response to subsequent therapy.26 We can also confirm the high rate of latent EBV infection in Burkitt-PTLD compared with sporadic BL and immunodeficiency-associated BL (63% in our series; 82% in the pediatric Burkitt-PTLD series21).

The excellent response to and long-term survival after immunochemotherapy with rituximab and CHOP or R-CHOP in Burkitt-PTLD, however, may seem surprising. Because of the success of intensive treatment protocols,12 published data for direct comparison with R-CHOP in BL are limited. Historically, a 53% CR rate (5-year OS, 23%) for CHOP with systemic and intrathecal methotrexate27 and a 50% CR rate (projected relapse-free survival, 62.5%) for a cyclophosphamide-increased, CHOP-like protocol plus systemic and intrathecal methotrexate28 have been reported in 2 small studies that included 13 patients and 8 patients, respectively. In contrast, an abstract that was presented at the 2006 American Society of Hematology meeting29 reported a CR/unconfirmed CR rate of 100% and an OS rate of 100% after a median potential follow-up of 28 months with the DA-EPOCH protocol (combined etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin)30 plus systemic rituximab and intrathecal methotrexate in 17 patients with BL. A potential explanation for the excellent outcome with sequential immunochemotherapy in PTLD may be the additional effect of IR.31

CRs have been reported in adult Burkitt-PTLD with treatment ranging from IR plus rituximab monotherapy to IR plus high-dose protocols, such as ProMace/CytaBom (combined cyclophosphamide, doxorubicin, etoposide, cytozar, bleomycin, vincristine, methotrexate, and prednisone) or the Cancer and Leukemia Group B 9251 protocol.20 However, the latter approach is associated with significant mortality in adults (60%),19 and the complications observed in pediatric Burkitt-PTLD by Picarsic et al also support our reservations regarding intensive therapy, because 2 of 4 patients who had received cytarabine died of therapy complication and graft (liver) failure, respectively.21

In contrast, our data support the idea that sequential therapy is sufficient to treat Burkitt-PTLD, thus eliminating the risks associated with more aggressive chemotherapy regimens. The sequential combination of rituximab and CHOP chemotherapy has achieved a response rate of 100% without TRM in 5 prospectively analyzed patients. Only 1 of those 5 patients suffered a relapse. Furthermore, the addition of rituximab to CHOP significantly lowered the risk of secondary CNS disease compared with historic controls who received CHOP alone. The value of additional intrathecal prophylaxis in patients with Burkitt-PTLD remains to be evaluated, particularly in patients who have bone marrow involvement or B symptoms.

FUNDING SOURCES

The German PTLD registry is supported by grants from Amgen, CLS Behring, Mundipharma GmbH, and Roche Pharma AG.

CONFLICT OF INTEREST DISCLOSURES

S.O. is currently an employee and shareholder of Hoffmann-LaRoche Ltd. He treated patients presented here and designed the PTLD-1 trial while employed at the Department of Hematology, Charite Medical University of Berlin, Campus Virchow-Klinikum, Berlin, Germany, up to 2004. R.U.T. received payment for lectures and consultancy from CSL Behring, Mundipharma, and/or Roche; received grant support from Amgen, CSL Behring, Mundipharma, and Roche; and received travel support from Amgen, CLS Behring, and Roche. BG received payment for lectures from Roche and GlaxoSmithKline.

Ancillary