Mood Biomarkers of Response to Deep Brain Stimulation in Depression Measured with a Sensing System.

in this letter describes a technique for acquiring patient-specific data from mood fluctuations occurring throughout the day using a sensing deep brain stimulation (DBS) system. Previous attempts to uncover neurophysiologic biomarkers for depression were promising, but greatly hampered by technological challenges involved in recording local field potentials (LFP) from implanted devices [[1]Veerakumar A. Tiruvadi V. Howell B. Waters A.C. Crowell A.L. Voytek B. et al.Field potential 1/f activity in the subcallosal cingulate region as a candidate signal for monitoring deep brain stimulation for treatment-resistant depression.J Neurophysiol. 2019; 122: 1023-1035Crossref PubMed Scopus (14) Google Scholar]. Recent studies have exposed the possibilities and complexities of implementing a patient-specific biomarker and stimulation selection closed-loop approach with a system already approved for the treatment of epilepsy [[2]Scangos K.W. Khambhati A.N. Daly P.M. Makhoul G.S. Sugrue L.P. Zamanian H. et al.Closed-loop neuromodulation in an individual with treatment-resistant depression.Nat Med. 2021; 27: 1696-1700Crossref PubMed Scopus (116) Google Scholar]. An alternative strategy involves having patients undergo a period of invasive electrode monitoring with concurrent deep brain stimulation testing. We implanted three patients with bilateral Medtronic 3387 electrodes connected to a Percept implantable pulse generator (IPG). The subgenual cingulum (SCC) was targeted at the intersection of the cingulum, forceps minor, uncinate fasciculus, and frontal-striatal fibers (Fig. 1A). At 2 weeks postoperatively, patients had their systems activated. Double-monopolar stimulation from middle contacts 1 and 2 was used, thus allowing LFP recordings between contacts 0 (ventral, closer to SCC grey matter) and 3 (dorsal) (Fig. 1B). At each appointment, stimulation amplitude was adjusted based on self-reported depressive symptoms (and the absence of reported or observed hypomania). Stimulation was started at 60μsec, 130 Hz and 1V. Current was increased based on clinical response at subsequent visits (Supplementary Table 1). Experiments were conducted between postoperative weeks 3–5 and 24–26. Activity in the ON DBS condition was recorded during postoperative weeks 3 and 24. Recordings in the OFF DBS condition were conducted during postoperative weeks 4 and 25. During recording sessions, patients were asked to document mood states at home using a handheld patient-programmer. They were asked to log events when they felt neutral, happy, depressed, or anxious. Logging an event involved the patient pairing the programmer and device, followed by the selection of one of the four mood options described above. Patients were asked to sit quietly for 30 seconds while the Percept device recorded continuous activity, which was fast-Fourier transformed in the frequency domain. Using power spectral densities recorded from the Percept system, differences between positive/neutral and anxious/depressed states were evaluated. For the purpose of analysis, depressed and anxious events were combined into a single ‘negative affect’ group, while neutral and happy events were grouped together as a ‘positive affect’ group. Demographics of the patients included in our trial may be found in Supplementary Table 1. Baseline Hamilton Depression Rating Scale (HAMD-17) scores and postoperative improvement at 6mo in patients 1, 2 and 3 were 28 (61%), 29 (21%) and 24 (42%), respectively. Patient 1 logged 127 events (71 negative), patient 2 logged 336 events (180 negative) and patient 3 logged 135 events (109 negative). In Patient 1, the only responder in our trial (≥50% reduction in HAMD-17 scores compared to baseline), there was a marked spectral pattern differentiating negative and positive emotions states in the left hemisphere. Negative affect states were characterized by a significant reduction in the delta band (2–4 Hz), and an increase in the alpha band (p < 0.01) (Fig. 1C). This signal was driven primarily by 6-month recordings compared with 3-5-week activity where no major differences were noted (Supplementary Figure). No differences were found in other frequency bands or when results recorded within the first weeks or 6mo after DBS were compared. To date, a few studies have either recorded activity from DBS electrodes or electroencephalography (EEG) in DBS-treated patients. When depression was compared to other psychiatric conditions or when patients were exposed to negative emotional stimuli, a common element in most studies was an increase in alpha activity [3Huang Y. Sun B. Debarros J. Zhang C. Zhan S. Li D. et al.Increased theta/alpha synchrony in the habenula-prefrontal network with negative emotional stimuli in human patients.Elife. 2021; 10Crossref Scopus (8) Google Scholar, 4Merkl A. Neumann W.J. Huebl J. Aust S. Horn A. Krauss J.K. et al.Modulation of beta-band activity in the subgenual anterior cingulate cortex during emotional empathy in treatment-resistant depression.Cerebr Cortex. 2016; 26: 2626-2638Crossref Scopus (37) Google Scholar, 5Neumann W.J. Huebl J. Brucke C. Gabriels L. Bajbouj M. Merkl A. et al.Different patterns of local field potentials from limbic DBS targets in patients with major depressive and obsessive compulsive disorder.Mol Psychiatr. 2014; 19: 1186-1192Crossref PubMed Scopus (0) Google Scholar]. Changes in lower frequency bands were not as straightforward, with low theta cordance [[6]Broadway J.M. Holtzheimer P.E. Hilimire M.R. Parks N.A. Devylder J.E. Mayberg H.S. et al.Frontal theta cordance predicts 6-month antidepressant response to subcallosal cingulate deep brain stimulation for treatment-resistant depression: a pilot study.Neuropsychopharmacology. 2012; 37: 1764-1772Crossref PubMed Scopus (86) Google Scholar] and an increase in SCC theta activity being reported [[4]Merkl A. Neumann W.J. Huebl J. Aust S. Horn A. Krauss J.K. et al.Modulation of beta-band activity in the subgenual anterior cingulate cortex during emotional empathy in treatment-resistant depression.Cerebr Cortex. 2016; 26: 2626-2638Crossref Scopus (37) Google Scholar]. As for the beta band, most studies showed low activity, desynchronization, or coherence while patients engaged in negative emotional activity [[4]Merkl A. Neumann W.J. Huebl J. Aust S. Horn A. Krauss J.K. et al.Modulation of beta-band activity in the subgenual anterior cingulate cortex during emotional empathy in treatment-resistant depression.Cerebr Cortex. 2016; 26: 2626-2638Crossref Scopus (37) Google Scholar,[7]Lipsman N. Kaping D. Westendorff S. Sankar T. Lozano A.M. Womelsdorf T. Beta coherence within human ventromedial prefrontal cortex precedes affective value choices.Neuroimage. 2014; 85: 769-778Crossref PubMed Scopus (27) Google Scholar]. Changes associated with antidepressant DBS effects include an increase in theta oscillations and cordance [[6]Broadway J.M. Holtzheimer P.E. Hilimire M.R. Parks N.A. Devylder J.E. Mayberg H.S. et al.Frontal theta cordance predicts 6-month antidepressant response to subcallosal cingulate deep brain stimulation for treatment-resistant depression: a pilot study.Neuropsychopharmacology. 2012; 37: 1764-1772Crossref PubMed Scopus (86) Google Scholar,[8]Widge A.S. Zorowitz S. Basu I. Paulk A.C. Cash S.S. Eskandar E.N. et al.Deep brain stimulation of the internal capsule enhances human cognitive control and prefrontal cortex function.Nat Commun. 2019; 10: 1536Crossref PubMed Scopus (73) Google Scholar] and the suppression of beta and gamma oscillations [[9]Sun Y. Giacobbe P. Tang C.W. Barr M.S. Rajji T. Kennedy S.H. et al.Deep brain stimulation modulates gamma oscillations and theta-gamma coupling in treatment resistant depression.Brain Stimul. 2015; 8: 1033-1042Abstract Full Text Full Text PDF PubMed Scopus (39) Google Scholar], the latter signal being successfully used as a closed-loop signal [[2]Scangos K.W. Khambhati A.N. Daly P.M. Makhoul G.S. Sugrue L.P. Zamanian H. et al.Closed-loop neuromodulation in an individual with treatment-resistant depression.Nat Med. 2021; 27: 1696-1700Crossref PubMed Scopus (116) Google Scholar]. In line with the results described above, we show for the first time in a naturalistic setting that negative affect states are associated with increased alpha and reduced delta activity in the SCC. Ultimately, electrophysiologic markers, such as the one demonstrated in this report, are important steps towards ‘closing the loop’ with neurostimulation for depression [[10]Widge A.S. Closing the loop in psychiatric deep brain stimulation: physiology, psychometrics, and plasticity.Neuropsychopharmacology. 2023; Crossref Scopus (0) Google Scholar]. In patient 1, the LFP biomarker of negative states could be used to trigger closed-loop stimulation, in theory allowing for more targeted neuromodulation, less development of ‘tolerance’ to stimulation, and increased battery longevity. Deriving such biomarkers from an at-home naturalistic method possibly presents a less resource-intensive method than using a period of inpatient recording through externalized electrodes or stereoelectroencephalography [[2]Scangos K.W. Khambhati A.N. Daly P.M. Makhoul G.S. Sugrue L.P. Zamanian H. et al.Closed-loop neuromodulation in an individual with treatment-resistant depression.Nat Med. 2021; 27: 1696-1700Crossref PubMed Scopus (116) Google Scholar]. Due to its preliminary nature, our study has a few caveats. It only included three patients. As the system only records data <250Hz, high frequency oscillations could not be captured. The number of anxiety episodes presented by the patients was relatively small. As no major differences were found between depression and anxiety events, these two categories were grouped for analysis. Whether mood fluctuations throughout the day are clinically important to be grounds for stimulation initiation is up for debate, but this approach has shown promise elsewhere [[2]Scangos K.W. Khambhati A.N. Daly P.M. Makhoul G.S. Sugrue L.P. Zamanian H. et al.Closed-loop neuromodulation in an individual with treatment-resistant depression.Nat Med. 2021; 27: 1696-1700Crossref PubMed Scopus (116) Google Scholar]. Because signals were far more evident in our responder than in non-responders, this raises the possibility that these electrophysiological signals could be potential biomarkers to trigger stimulation in a closed-loop paradigm. This, however, could not be tested in our study due to limitations of the DBS system used. The device currently in use for dual sensing and recording is capable of closed-loop stimulation, but this function requires approval to be unlocked by the company. Benjamin Davidson: Conception, data acquisition, data analysis, writing the manuscript. Maximilian Scherer: Data acquisition, data analysis. Sean Nestor: Critical revision of the manuscript. Peter Giacobbe: Critical revision of the manuscript. Agessandro Abrahao: Critical revision of the manuscript. Jennifer S. Rabin: Critical revision of the manuscript. Liane Phung: Data acquisition. Fa-Hsuan Lin: Data acquisition. Nir Lipsman: Critical revision of the manuscript, funding acquisition. Luka Milosevic: Data analysis, critical revision of the manuscript, funding acquisition. Clement Hamani: Conception, data acquisition, data analysis, writing the manuscript, funding acquisition. None of the authors have conflicts of interest related to this work. This work was conducted with funds from the Harquail Centre for Neuromodulation and a grant from the New Frontiers in Research Fund NFRFE-2021-00261 (L M.; C.H.; N.L.). The following are the Supplementary data to this article: Download .docx (.06 MB) Help with docx files Multimedia component 1 Download .docx (.01 MB) Help with docx files Multimedia component 2 Download .docx (.02 MB) Help with docx files Multimedia component 3