Immune mediated demyelinating disease (IMDD) after allogeneic hemopoietic stem cell transplant (HSCT) is rare and its etiology unclear. In this retrospective study, we identified patients who underwent HSCT between January 1992 and December 2010 and had IMDD post transplant. A total of 1,484 patients received HSCT and 7 (0.5 %) suffered from IMDD; five were men, and the median age was 54 years (range, 29-64 years). HSCT treated acute myeloid leukemia (n = 5), myelodysplastic syndrome (n = 1), and Waldenström macroglobulinemia (n = 1). All received an HLA matched donor graft, related (6), unrelated (1); from the bone marrow (1), peripheral blood stem cell (6); and T-cell depleted, ex vivo (6) or in vivo (1). The median time from transplant to neurologic symptoms was 120 days (range, 60-390 days). Three had acute demyelinating encephalomyelitis (ADEM), three acute inflammatory demyelinating polyradiculopathy (AIDP) and one autonomic neuropathy. Four of six patients tested had hemopoietic mixed chimerism prior to neurologic symptoms and low CD4(+) T-cell counts, median 76 (15-500 cells/μL). Two patients had simultaneous systemic graft versus host disease (GVHD). Two patients with ADEM had a spinal cord or brain biopsy which revealed demyelination. No patients had a viral etiology identified in the cerebrospinal fluid. Patients were treated with IV immunoglobulin, high dose steroids and/or rituximab. Five patients had a significant recovery. Response to immune modulators suggests an immune-based etiology. The incidence of de novo autoimmune disease after HSCT for hematological diseases is rare and may be difficult to differentiate from GVHD.

No abstract

2028 Background: Salvage treatment with bevacizumab (BEV) has been increasingly used in grade III AG, but limited data are available, with conflicting results reported. Because of differences in molecular characteristics, angiogenesis mechanisms, growth rate and chemosensitivity, results observed in glioblastoma (GBM) treated with BEV may not apply to AG, and may differ between anaplastic astrocytomas (AA) and oligodendrogliomas (AO). Methods: IRB approved retrospective review of all pts with recurrent WHO grade III AG treated with BEV at MSKCC. Response and progression were determined by RANO criteria. Results: BEV was given to 39 pts with recurrent grade III AG (AA: 26; AO: 10; anaplastic oligoastrocytoma [AOA]: 3); median (med) age: 49 (range 20-75); med KPS: 80 (60-100). MGMT promoter methylation was determined in 17 pts (methylated 8, unmethylated 9). Amongst AO/AOA, 1p/19q co-deletion was present in 6 and absent in 5. IDH1/ IDH2 mutations were present in 3/5 tested tumors. Med time from diagnosis of AG to BEV treatment was 11.5m (3 – 112.5). The med number of previous recurrences was 1 (range 1-4). BEV was given as single agent to 16 pts and combined with a cytotoxic agent in 23. Objective response rate (ORR) was 41% (complete response: 5; partial: 11). The med progression-free survival (PFS) was 4.8m (6-m PFS: 35% [95% CI 20-50]). The med overall survival (OS) was 11.5 m (1y-OS: 45% [95% CI 29-61]). In pts achieving ORR, med OS was 13 m vs 4 m in pts with stable/progressive disease (P=0.02). Pts at first recurrence fared better than pts with more than one recurrence (med PFS: 6 vs 3.4 months, P=0.04). AO/AOA tended to fare worse than AA (med PFS 3 vs 5.5 m, P=0.07). There were no differences in PFS (P=0.8) or OS (P=0.4) between single agent vs BEV + cytotoxic agent. Toxicity included one grade 4 brain hemorrhage. Conclusions: In this relatively large series, BEV was associated with higher ORR and PFS as compared to historical controls, although improvements in those endpoints were of a lower magnitude than in GBM. While ORR predicted OS, improvements in OS were not apparent; results in AO were particularly disappointing. The use of BEV in this population requires reappraisal in a randomized study with adequate stratification for histology and number of previous recurrences.

No abstract