Abstract
Introduction: Post-transplant lymphoproliferative disease (PTLD) is a well-known complication of solid organ transplantation, and similar lymphoid neoplasms are seen in patients immunosuppressed for the treatment of autoimmune conditions or with untreated HIV/AIDS. Following the introduction of renal transplantation in the late 1960's, the central nervous system (CNS) was the most commonly involved site, comprising approximately one half of cases (Lancet 297:983-6). However, the diagnosis of primary CNS PTLD became increasingly uncommon over time, and until recently, only rare cases were seen in our practice. In this study we systematically evaluate the incidence of primary CNS lymphoproliferative disease and the relationship of the anatomic site in which it develops to the immunosuppressive regimen.
Methods: All cases of immunosuppression-related lymphoproliferative disease diagnosed at our institution between 10/86 and 5/14 were identified, including in-house and consultation cases. Patient age, sex, type and reason for transplant, immunosuppressive regimen, timing of diagnosis from transplant, and survival were recorded. In addition, a United Network for Organ Sharing (UNOS) Standard Transplant Analysis and Research (STAR) file of kidney transplant recipients using Organ Transplant Procurement Network data from 10/25/99 (introduction of PTLD-specific data fields) to 4/14/14 was analyzed. Odds ratios were calculated using Fisher's exact test.
Results: We identified 111 cases of immunosuppression-related lymphoproliferative disease at our institution, including 29 diagnosed in the CNS. The fraction of CNS cases compared to those diagnosed in other sites significantly increased over the period of study; while no CNS cases were diagnosed between 1986 and 1994, 40% of cases between 2011 and 2014 were diagnosed in the CNS (Fig 1). A similar trend was evident when consult cases were excluded.
Both CNS and non-CNS PTLDs were predominantly monomorphic (76%, 86%) and mostly large B-cell lymphomas (91%, 60%), though those outside the CNS were more morphologically diverse. CNS PTLDs were more strongly associated with EBV (96% vs. 58%) and tended to have an earlier onset (median 2.9 vs. 5.3 years post transplant) and higher mortality (48% vs. 39%). CNS PTLDs were more likely to arise in patients taking mycophenolate mofetil (MMF; 89% vs. 40%), particularly those on MMF in the absence of calcineurin inhibitors (CNI) with an odds ratio of 21.4 (p<0.005, Table 1). The data suggested a protective effect of CNI if taken with MMF (0.4 decrease in odds of CNS), but did not achieve statistical significance. Analysis of a larger, UNOS STAR datafile that included a total of 2958 PTLD cases, 59 of whom were recorded as primary CNS PTLD, however, confirmed the protective effect of CNI. Patients taking MMF in the absence of CNI were 7.6-fold more likely to develop PTLD in the CNS compared to those patients taking both MMF and CNI (p<0.001) and 20.3-fold more likely than those taking CNI without MMF. Both tacrolimus and cyclosporine when evaluated separately conferred a decreased CNS PTLD risk and were not statistically different.
Conclusions: Our data demonstrate a rise in the incidence of CNS PTLD and a relationship between the involved anatomic site and the immunosuppressive drug regimen. Specifically, these findings suggest that the decline in use of CNI following the introduction of MMF and other new immunosupressants may be responsible for the recent rise in CNS immunosuppression-related lymphoproliferative disease. A potential role of MMF in promoting CNS involvement cannot be excluded; however, review of historic data, including a several thousand fold increased risk of primary brain lymphoma in untreated HIV/AIDS (J Nat Can Inst 88:675-9), suggests the CNS may have an inherent susceptibility to lymphoproliferative disease in the context of immunosuppression. While mechanisms underlying the protective effect of CNI remain speculative, these findings may have therapeutic implications, particularly as new transplant regimens lacking CNI are introduced and use of MMF in the treatment of autoimmune disease becomes increasingly common.
Figure 1 Figure 1.
Table 1 JHH UNOS CNS Non-CNS CNS Non-CNS Cases (with drug info) 29(18) 82(66) 59(28) 2899(1677) MMF/ no calcineurin 4 3 9 68 MMF/any calcineurin 12 23 14 810 No MMF/tacrolimus 1 21 1 384 No MMF/cyclosporine 1 17 4 400
Disclosures
No relevant conflicts of interest to declare.
Review of eating disorders and oxytocin receptor polymorphisms