Background: There are no effective medical treatments for MPNST. Inactivation of the NF1 tumor suppressor in MPNST results in upregulation of mTOR signaling and angiogenesis contributes to disease progression. In a genetically engineered mouse model of NF1 MPNST sirolimus prolonged progression free survival and sunitinib extended these effects. Methods: We conducted a phase II study of everolimus (10 mg PO once daily, 1 cycle = 28 days) with bevacizumab (10 mg/kg IV every 2 weeks) to determine the clinical benefit rate (CBR) [complete response, partial response (PR), or stable disease (SD) ≥ 4 months]. Tumor response was assessed after every 2 cycles (WHO criteria). Patients (pts) ≥ 18 years old with chemotherapy refractory sporadic or NF1 MPNST were eligible. A two-stage design targeting a 25% CBR was used: If ≥ 1/15 pts in stage 1 responded, enrollment would be expanded by 10 pts, and if ≥ 4/25 patients had clinical benefit, the combination would be considered active. Results: Twenty-five pts (15F:10M), 17 NF1 MPNST (median age 29 years, range 19-63) and 8 sporadic MPNST (median age 62 years, range 19-72) were enrolled. Twenty-one pts had metastatic disease at enrollment. One of 15 pts in stage 1 had clinical benefit (sporadic MPNST, SD for 6 cycles). Of 10 additional pts enrolled in the second stage, 2 had clinical benefit (1 NF1 MPNST SD for 6 cycles, 1 NF1 MPNST SD 7+ cycles, still on treatment). The best response was SD except in 1 pt with a PR after cycle 2, but PD after cycle 4. The median number of completed cycles was 3 (range 1-7+). Everolimus dose was reduced/discontinued for grades 1-3 oral mucositis (n = 3), grade 2 ear pain (n = 1), and grade 3 fatigue (n = 1). Bevacizumab and everolimus were permanently discontinued in 4 patients for drug related adverse events: grade 1-2 proteinuria (n = 2) per investigator decision, grade 5 intra-abdominal intra-tumor hemorrhage (n = 1), and grade 3 ALT elevation (n = 1). Conclusions: With a CBR of 12% (3/25) the combination of everolimus and bevacizumab did not reach the study’s target response rate and is not considered active in refractory MPNST. Adverse events were similar to those known for this combination. Clinical trial information: NCT01661283