Brain Stimulation (Mar 2022)

Comparison of clinical outcomes with left unilateral and sequential bilateral Transcranial Magnetic Stimulation (TMS) treatment of major depressive disorder in a large patient registry

  • Scott T. Aaronson,
  • Linda L. Carpenter,
  • Todd M. Hutton,
  • Stacia Kraus,
  • Miriam Mina,
  • Kenneth Pages,
  • Luoxi Shi,
  • W. Scott West,
  • Harold A. Sackeim

Journal volume & issue
Vol. 15, no. 2
pp. 326 – 336

Abstract

Read online

Background: It has been suggested that sequential bilateral (SBL) TMS, combining high frequency, left dorsolateral prefrontal cortex (DLPFC) stimulation and low frequency, right DLPFC stimulation, is more effective than unilateral TMS. Objective: To contrast treatment outcomes of left unilateral (LUL) and SBL protocols. Methods: Registry data were collected at 111 practice sites. Of 10,099 patients, 3,871 comprised a modified intent-to-treat (mITT) sample, defined as a primary MDD diagnosis, age ≥18, and PHQ-9 completion before TMS and at least one PHQ-9 assessment after baseline. The mITT sample received high frequency (10 Hz) LUL TMS exclusively (N = 3,327) or SBL TMS in at least 90% of sessions (N = 544). Completers (N = 3,049) were responders or had received ≥20 sessions and had an end of acute treatment PHQ-9 assessment. To control for site effects, a Matched sample (N = 653) included Completers at sites that used both protocols. To control for selection bias, the SBL group was also compared to a Restricted LUL group, drawn from sites where no patient switched to SBL after substantial exposure to LUL TMS. Secondary analyses were conducted on CGI-S ratings. Results: The LUL group had superior outcomes compared to the SBL group for multiple PHQ-9 and CGI-S continuous and categorical measures in the mITT, Completer and Matched samples, including in the specified primary analyses. However, outcome differences were not observed when comparing the Restricted LUL and SBL groups. Within SBL protocols, the LUL-RUL order had superior outcomes compared to the RUL-LUL order in all CGI-S, but not PHQ-9, measures. Conclusions: While limited by the naturalistic design, there was no evidence that SBL TMS was superior to LUL TMS. The sequential order of RUL TMS followed by LUL TMS may have reduced efficacy compared to LUL TMS followed by RUL TMS.