Cortical inhibition in neurofibromatosis type 1 is modulated by lovastatin, as demonstrated by a randomized, triple-blind, placebo-controlled clinical trial

Abstract

Neurofibromatosis type 1 (NF1) is associated with GABAergic dysfunction which has been suggested as the underlying cause of cognitive impairments. Previous intervention trials investigated the statins' effects using cognitive outcome measures. However, available outcome measures have led to inconclusive results and there is a need to identify other options. Here, we aimed at investigating alternative outcome measures in a feasibility trial targeting cortical inhibition mechanisms known to be altered in NF1. We explored the neurochemical and physiological changes elicited by lovastatin, with magnetic resonance spectroscopy and transcranial magnetic stimulation (TMS). Fifteen NF1 adults participated in this randomized, triple-blind, placebo-controlled crossover trial (Clinicaltrials.gov NCT03826940) composed of one baseline and two reassessment visits after lovastatin/placebo intake (60 mg/day, 3-days). Motor cortex GABA+ and Glx concentrations were measured using HERMES and PRESS sequences, respectively. Cortical inhibition was investigated by paired-pulse, input-output curve, and cortical silent period (CSP) TMS protocols. CSP ratios were significantly increased by lovastatin (relative: p = 0.027; absolute: p = 0.034) but not by placebo. CSP durations showed a negative correlation with the LICI 50 ms amplitude ratio. Lovastatin was able to modulate cortical inhibition in NF1, as assessed by TMS CSP ratios. The link between this modulation of cortical inhibition and clinical improvements should be addressed by future large-scale studies.

Figure 1

Differences in relative (A) and absolute (B) cortical silent period ratios, comparing measures taken after lovastatin and placebo intake with the baseline assessment. Dots represent, for each participant, the difference between lovastatin and baseline or placebo and baseline CSP:MEP ratios. Lines represent median and 95% CI. CSP cortical silent period, MEP motor-evoked potential, ms millisecond, mV millivolt, CI confidence interval.

Figure 2

MEP peak-to-peak amplitude ratios for SICI3ms, SICI5ms, ICF10ms, ICF15ms, LICI50ms and LICI100ms, at baseline and after lovastatin and placebo administration. The horizontal line represents a null effect, wherein conditioned stimulus amplitude equals the amplitude from baseline pulses. Inhibition occurs for bars below the horizontal line, whereas excitation stands above the line. MEP motor-evoked potential, ms millisecond.

Figure 3

Spearman correlations between relative and absolute silent periods and MEP amplitude ratio in LICI mean intervals, both following lovastatin (A, C) and placebo (B, D). Shaded area represents the 95% CI for the best-fit line. CSP cortical silent period, ms millisecond, CI confidence interval, LICI long-interval intracortical inhibition.

Figure 4

Study design, including the procedures performed in each visit. MR magnetic resonance, TMS transcranial magnetic stimulation.

Figure 5

A schematic representation of the voxel placement (A). In (B), it is presented an example of Gannet output, from the HERMES sequence, used to estimate GABA+ concentration. Glx levels were quantified through the PRESS sequence acquisition, processed in LCModel, as represented by the example spectrum (C). Glx glutamate + glutamine, GABA+ gamma-aminobutyric acid, mI myo-inositol, tCho total choline, tCr total creatine, tNAA total N-acetylaspartate, ppm parts per million.