Abstract
Background: The imputability of neutralizing antibodies (NABs) in secondary non-response (SnR) to botulinum toxin (BoNT) injections for limb spasticity is still debated. Objective: This systematic literature review aimed to determine the prevalence of NABs after BoNT injections for limb spasticity and analyze their determinants and their causal role in SnR. Methods: We searched MEDLINE via PubMed, Cochrane and Embase databases for articles published during 1990-2018. Two independent reviewers extracted the data and assessed the quality of studies with a specific scale (according to PRISMA and STROBE guidelines). Because the techniques used to detect NABs did not influence the results, we calculated the global (all studies) sensitivity and specificity of NAB positivity to reveal SnR. Results: We included 14 articles published from 2002 to 2018 (including an epublication) describing 5 randomized controlled trials and 5 interventional and 4 observational studies. The quality was satisfactory (mean score 18/28 arbitrary units). NAB detection was the primary criterion in 5 studies and a secondary criterion in 9. In total, 1234 serum samples for 1234 participants (91% with stroke) were tested after injection. NAB prevalence was about 1%, with no significant difference among formulations. NAB positivity seemed favoured by long-duration therapy with high doses and a short interval between injections. The identification of non-response by NAB positivity had poor global sensitivity (56%) but very high specificity (99.6%). No consensual criteria were used to diagnose non-response to BoNT injection. Conclusions: NAB prevalence is much lower after BoNT treatment for limb spasticity than cervical dystonia. Consensual criteria must be defined to diagnose non-response to BoNT injection. Because immunogenicity is not the most common cause of non-response to BoNT injection, NABs should be sought in individuals with SnR with no other cause explaining the treatment inefficacy. A test with 100% specificity is recommended. In cases for which immunogenicity is the most likely cause of non-response to BoNT injections, some biological arguments suggest trying another BoNT, but no clinical evidence supports this strategy.