We conducted a network meta-analysis to compare the efficacy of acetyl-levo-carnitine (ALC) with other commonly used oral medications in patients with diabetic peripheral neuropathy (DPN).

We searched the Chinese and English electronic databases up to February 2020 to identify randomized controlled trial (RCT) that evaluated one or more of the following oral monotherapy interventions in the treatment of DPN in adults: ALC, α-lipoic acid (ALA), epalrestat, methylcobalamin (MC), and pancreatic kininogenase. The primary outcome was nerve conduction velocity (NCV). The secondary outcomes were amplitude, neurological symptoms, neurological signs, and pain. A random-effect network meta-analysis within a Bayesian framework was performed. We estimated the ranking probabilities for each intervention by use of surface under the cumulative ranking curve (SUCRA).

Fifty-five RCT with 6 473 participants were included in this network meta-analysis. All trials were two-arm studies, involving four treatments (ALC, ALA, eparlrestat, MC). Three trials related to ALC (two studies compared ALC versus placebo, and the other one versus MC) were included. Compared with MC, ALC had more improvement in NCV of median sensory nerve (MD=4.0 m/s, 95%CI 0.1—7.9 m/s). According to the ranking results, ALC might be better in improving median sensation nerve (SUCRA, 91.6%), ulnar motor nerve (SUCRA, 69.3%) and ulnar sensation nerve (SUCRA, 77.8%) compared with other commonly used oral drugs. However, the results of relative-effect network meta-analysis showed no significant difference between ALC and other drugs in NCV improvement of other nerves.

The findings of this network meta-analysis showed ALC might have beneficial effects on improving the sensory nerve function and pain intensity, and have similar performance with other commonly used oral medications on other outcomes.

Diabetes mellitus; Diabetic peripheral neuropathy; Acetyl-levo- carnitine; Nerve conduction velocity; Network meta-analysis