Meta-analysis comparing placebo responses in clinical trials of painful HIV-associated sensory neuropathy and diabetic polyneuropathy

The placebo response has been identified as one factor responsible for the lack of therapeutic trials with positive outcomes in neuropathic pain. Reviews have suggested that certain neuropathic pain conditions, including HIV-associated sensory neuropathy (HIV-SN), exhibit a greater placebo response than other neuropathic aetiologies. If true, such a finding could substantially affect clinical trial design and therapeutic developments for these conditions. This study aimed to identify any difference in placebo response between trials of systemic pharmacological intervention in HIV-SN and a comparable neuropathic condition, diabetic polyneuropathy (DPN) and to identify factors influencing the placebo response. A systematic review search to identify randomised, double-blind studies of systemic pharmacological interventions for painful HIV-SN and DPN published between January 1966 and June 2019 was performed. A meta-analysis of the magnitude of placebo response and the proportion of placebo responders was conducted and compared between the two disease conditions. A meta-regression was used to assess for any study and participant characteristics that were associated with the placebo response. Only studies meeting a methodological quality threshold were included. Seventy-five trials were identified. There was no statistically significant difference in the proportion of placebo responders (HIV-SN = 0.35; versus DPN = 0.27, p = 0.129). The difference observed in the magnitude of the placebo response [pain reduction of 1.68 (1.47-1.88) DPN; 2.38 (1.87-2.98) in HIV-SN] was based on only 2 trials of HIV-SN and 35 of DPN. Potential factors influencing the placebo response such as psychological measures, were reported inconsistently. We found no statistically significant difference in the placebo response rate between painful HIV-SN and DPN. Too few studies were available that reported the necessary information to clarify potential differences in the magnitude of placebo response or to elucidate parameters that could be contributing such differences. The placebo response is one factor that may contribute to a lack of positive trials in neuropathic pain; some etiologies may display larger responses than others. This meta-analysis found no significant difference in placebo response between trials of HIV-associated sensory neuropathy and painful diabetic polyneuropathy, although limited data were available.