Anakinra in Sanfilippo syndrome: a phase 1/2 trial

Abstract

Sanfilippo syndrome is a fatal childhood neurodegenerative disorder involving neuroinflammation among multiple pathologies. We hypothesized that anakinra, a recombinant interleukin-1 receptor antagonist, could improve neurobehavioral and functional symptoms owing to its capacity to treat neuroinflammation. This phase 1/2 trial aimed to test the safety, tolerability and effects of anakinra on neurobehavioral, functional and quality-of-life outcomes in patients and their caregivers. The primary outcome was the percent of participants requiring a dose increase at week 8 or week 16. Secondary efficacy outcomes included a multi-domain responder index (MDRI). Twenty-three participants (6-26 years of age) were enrolled. Twenty continued treatment to week 8, and 15 (75%) required an increased dose at week 8 or week 16. There was an improvement in at least one domain in the MDRI in 18 of 21 (86%) at week 8 and in 15 of 16 (94%) at week 36. Seven participants withdrew (intolerability of daily injections and lost to follow-up) before week 36. Adverse events occurred in 22 of 23 (96%) participants, most commonly mild injection site reactions. No serious adverse events were related to anakinra. In conclusion, anakinra was safe and associated with improved neurobehavioral and functional outcomes, supporting continued investigation of anakinra in Sanfilippo syndrome and other mucopolysaccharidoses. ClinicalTrials.gov identifier: NCT04018755 .

Fig. 1. CONSORT flowchart.

Anakinra treatment for the 8-week open label from day 1 to week 8 with 28-week open-label extension. Treatment occurred from day 1 to week 36. Observation occurred for 8 weeks before day 1 and for another 8 weeks starting at week 36 when treatment was stopped. The primary enrollment target was 20 on treatment at week 8. n = 3 patients stopped treatment but continued the study by completing assessments through week 44: one before week 8 (stopped treatment owing to a grade 3 AE of increased agitation) and two between week 8 and week 36 (stopped treatment owing to persistent ANC < 1,500 cells per microliter).

Fig. 2. MDRI: comparison of change during 8 weeks (a), 16 weeks (b) and 36 weeks (c) of treatment to MCID.

A change during treatment with anakinra relative to the MCID was used to define ‘improved’, ‘no change’ and ‘worsened’. MCID definitions can be found in Methods, ‘Secondary outcomes’.

Fig. 3. Change in the SBRS individual clusters and domains before, during and after treatment with anakinra.

The treatment period was from week 0 to week 36. The black line graph with circle symbols shows the observed least-squares means with 95% CIs for patients with available data (observed) over time. The imputed least-squares means for the pre-specified sensitivity analysis are plotted as gray lines with square symbols. P values are for change in observed least-squares means from day 1. A return of symptoms after 8 weeks off treatment, seen by the upward slope to week 44, suggests that the reduction of symptoms from week 0 to week 36 was not part of the natural disease course. There was no statistically significant difference in SBRS scores between week 44 and day 1. All statistical tests were two-sided, including the Wald test with 23 degrees of freedom from the mixed model. No adjustment was made for multiple comparisons. LSMean, least-squares mean.

Fig. 4. Immunophenotype changes compared to the MDRI.

a, Flow cytometry plot of whole blood stained with specific antibodies to identify monocytes. Whole blood cells were gated on live cells and then sub-gated on CD19⁻ cells (to exclude B cells), CD3⁻ cells (to exclude T cells) and CD56 cells (to exclude NK cells). Cells that were not bright CD16 (neutrophils) in the side scatter plot (SSC) plot were identified as monocytes. These cells expressed CD14 and CD16. b, B lymphocytes were identified as CD19⁺ cells and CD3⁻ cells. T lymphocytes CD4 and CD8 were sub-gated from CD3⁺ cells. c, Trends for changes in CD4⁺ T cells (for example, T helper and T regulatory cells), CD8⁺ T cells (for example, cytotoxic T cells), monocytes and B cells over 16 weeks of treatment with anakinra, compared to the MDRI. The number of MDRI surveys with a treatment response that was either improved (black triangles/solid line) or worsened (open circles/dashed line), as defined by the MCID, is plotted with immunophenotype data as percent of the cell population. The y axis is absolute change in percent of the cell population. Positive numbers are an increase in cell population, and negative numbers are a decrease in cell population. An increase in CD4⁺ T cells with a decrease in CD8⁺ T cells and a decrease in monocytes is suggestive of a less pro-inflammatory environment. This less pro-inflammatory environment is the desired change with therapy. These changes coincide with a higher number of improvements on the MDRI outcomes. Anakinra is not expected to have a significant effect on B cells. The lack of B cell change and lack of relationship with MDRI outcomes may increase confidence in the findings. * Spearman’s rank-based correlation coefficient P < 0.05. SSC-A, side scatter area. Source data

Extended Data Fig. 1. Change in the Individual Clinical Response (ICR) before, during and after treatment with anakinra.

Treatment period indicated by grey box (Week Zero to Week 36). The black line graph with circle symbols shows the observed LSMeans with 95% confidence intervals for patients with available data (observed) over time. The imputed LSMeans for the prespecified sensitivity analysis are plotted as grey lines with square symbols. A return of symptoms after eight weeks off treatment, seen by the upward slope to Week 44, suggests the reduction of symptoms from Weeks Zero to 36 were not part of the natural disease course. There was no statistically significant difference in ICR between Week 44 and Day One. All statistical tests were two-sided, including the Wald test with 23 degrees of freedom from the mixed model. No adjustment was made for the multiple comparison.