Effect of Lenalidomide On TNF-Alpha Elevation and Behavior in Autism

Background: Autism currently affects 1:100 births. Research has shown markers of neuroglial inflammation with elevated cytokines. Elevation of tumor necrosis factor-alpha (TNF-α) in the CSF vs. serum in patients with regressive autism has been reported. Lenalidomide (Revlimid®) is an oral immunomodulatory agent predicted to modulate elevated TNF-α levels.

Objectives: This study was to determine if lenalidomide safely reduces CSF and serum tumor necrosis factor-alpha (TNF-α) and correlates to improved behavior and speech function in children with autism spectrum disorder who have elevated CSF and serum TNF-α levels.

Methods: Six males (6 to 12 years) with elevated TNF-α were included in this open label trial and received Revlimid® 2.5 mgs per day for 12 weeks. CSF and serum TNF-α were measured before baseline and at 12-weeks post treatment. The Expressive and Receptive one-word tests, and the Childhood Autism Rating Scale (CARS) were obtained at baseline, 6-weeks, and at 12-weeks in all subjects. Repeated measures t-tests were used to compare pre and post test differences.

---

Results: The CSF-TNF-α declined an average of 57% from 80.5 ±41.03 to 38.0 ± 31.27 (P = 0.065) in four children who completed the study at 12 weeks. Serum TNF-α showed decline of 57% (92.50 ± 68.92 to 40.25 ± 44.53 (P = 0.048). At the 6-week follow-up, language in the 4/6 patients with initial expressive language as indicated by an ADOS module of 2 or 3 was analyzed. Mean receptive language in the 4 children increased from 50.00± 19.5 to 57.75 ± 19.03 (P = 0.049) and showed decreased symptoms of autism based on the CARS scores (baseline 37.38, SD±4.11 vs. 6 weeks 34.50±4.92, P = 0.090). Patients also had Clinical Global improvement (CGI) ratings by the primary investigator at each visit and also at 6 and 12 week points. The CGI data for socialization and expressive and receptive speech improved in all 4 patients above at 6 weeks, and 3 patients at 12 weeks.
Side effects included rash causing two patients to drop out of study. One patient was dropped out of study after eight weeks for transient drop of absolute neutrophil count of 1200 (safety cutoff 1500).

Conclusions: This study successfully looked at a novel mechanism for autism with regression using elevated TNF-α markers in CSF and serum and treatment with oral TNF-α inhibitor lenalidomide. This study shows clear clinical lowering of TNF-α in both serum and CSF in 4 patients completing 12 week study. Despite the small number of patients, 4/6 patients with measurable language improved by one-word vocabulary testing with receptive language showing statistically significant and clinically meaningful changes by 6 weeks of treatment. CARS improved on average by greater than 2 points. Although some measures did not reach significance statistically, this was probably due to low number of patients. Larger and placebo controlled studies looking at TNF-α treatment in autism patients with elevation in CSF and/or serum is warranted.