Where to aim TMS for smoking cessation, from two independent lines of evidence

Design

within-participant study of multiple TMS sites with lesion network mapping and normative connectome analysis from 1,000 healthy individuals

Population

72 participants with tobacco use disorder, abstinent at least 12 hours, receiving TMS at four left-hemisphere regions across 243 stimulation sites

Primary outcome

association between TMS site connectivity to a lesion-derived addiction remission circuit and change in Shiffman-Jarvik withdrawal score

Effect

site connectivity significantly predicted percentage withdrawal improvement, robust to adjustment for sex and baseline dependence; data-driven and lesion-based circuits matched at Pearson r = -0.74

Transcranial magnetic stimulation is cleared for smoking cessation, but the target has been chosen largely by convention rather than by evidence about which circuit matters. This study tested a circuit derived from brain lesions that happen to abolish nicotine addiction — a causal source of evidence, because a lesion precedes the remission — against actual TMS outcomes.

Seventy-two people with tobacco use disorder abstained for at least 12 hours and received TMS at four left-hemisphere regions across separate sessions, generating 243 stimulation sites. Withdrawal was measured before and after each session on the Shiffman-Jarvik scale, and each site's functional connectivity to the lesion-derived remission circuit was estimated from a normative connectome of 1,000 healthy people. Connectivity to that circuit significantly predicted withdrawal improvement, and remained significant after adjusting for sex and baseline dependence. A circuit derived independently from the TMS data closely matched the lesion-based one (Pearson r = -0.74), and combining both pointed at the frontopolar cortex, posterior parietal cortex, lateral temporal lobe and superior frontal gyrus.

What changes is how a TMS service should choose its target: connectivity-based targeting has a rationale here that convention does not, and the convergence of two independent methods is more persuasive than either alone. What has not changed is the evidence on outcomes. The endpoint was acute withdrawal symptoms after a single session, not abstinence, not at six months, and not in an unselected clinic population. Anyone offering TMS for smoking should present this as the reason a target was chosen, not as evidence of how well the treatment works.

• Use connectivity to the identified circuit, rather than convention, when selecting a TMS target for tobacco use disorder
• Do not extrapolate to cessation rates — the outcome measured was acute withdrawal after single sessions
• Note the convergence of lesion-derived and TMS-derived circuits; that agreement is the study's real strength
• Normative connectomes come from healthy volunteers, so individual targeting still requires validation
• Seventy-two participants, 243 sites, no clinical abstinence endpoint — a targeting hypothesis ready for trial

The statistics, in plain English

A Pearson correlation of -0.74 between the two independently derived circuit maps indicates strong spatial agreement; the negative sign reflects the direction in which the maps were scaled rather than an inverse relationship. Because each participant contributed multiple stimulation sites, the 243 sites are not 243 independent observations, and analyses that treat them as such overstate precision. The outcome — change on a withdrawal scale after a single session in people abstinent for 12 hours — is a short-term proxy that has historically correlated poorly with whether people actually stop smoking.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.