Portal de Periódicos da CAPES

Predicting treatment response to 1Hz rTMS using early self-rated clinical changes in major depression

2020; Elsevier BV; Volume: 13; Issue: 6 Linguagem: Inglês

10.1016/j.brs.2020.10.004

1935-861X

Marine Mondino, David Szekely, Maxime Bubrovszky, Samuel Bulteau, Jonathan Downar, Emmanuel Poulet, Jérôme Brunelin,

Neurological disorders and treatments

Repetitive transcranial magnetic stimulation (rTMS) is increasingly used as an evidence-based treatment for patients with medication-resistant major depressive episode - MDE [[1]Lefaucheur J.-P. Aleman A. Baeken C. Benninger D.H. Brunelin J. Di Lazzaro V. et al.Evidence-based guidelines on the therapeutic use of repetitive transcranial magnetic stimulation (rTMS): an update (2014-2018).Clin Neurophysiol. 2020; 131: 474-528https://doi.org/10.1016/j.clinph.2019.11.002Crossref PubMed Scopus (199) Google Scholar]. Despite a good clinical efficacy at the group level in this population, a large heterogeneity is observed at the individual level with distinct trajectories of response defining some sub populations of responders and nonresponders [[2]Fitzgerald P.B. Hoy K.E. Anderson R.J. Daskalakis Z.J. A study of the pattern of response to rTMS treatment in depression.Depress Anxiety. 2016; 33: 746-753https://doi.org/10.1002/da.22503Crossref PubMed Scopus (64) Google Scholar,[3]Kaster T.S. Downar J. Vila-Rodriguez F. Thorpe K.E. Feffer K. Noda Y. et al.Trajectories of response to dorsolateral prefrontal rTMS in major depression: a THREE-D study.Am J Psychiatr. 2019; 176: 367-375https://doi.org/10.1176/appi.ajp.2018.18091096Crossref PubMed Scopus (35) Google Scholar]. Early prediction of trajectory of response is of main interest to avoid false hope in patients and to decrease the financial and time burden associated with rTMS treatment in patients with MDE. Some clinical (e.g., retardation, smoking status [[4]Poulet E. Galvao F. Haffen E. Szekely D. Brault C. Haesebaert F. et al.Effects of smoking status and MADRS retardation factor on response to low frequency repetitive transcranial magnetic stimulation for depression.Eur Psychiatr. 2016; 38: 40-44https://doi.org/10.1016/j.eurpsy.2016.04.004Abstract Full Text Full Text PDF PubMed Scopus (3) Google Scholar]) and biological measures such as neurophysiological features (e.g., cortical excitability measures), genetic polymorphisms (e.g., BDNF or COMT) and imaging measures (with either EEG, PET, SPECT or fMRI) have been proposed as predictive markers of response to rTMS [[5]Silverstein W.K. Noda Y. Barr M.S. Vila-Rodriguez F. Rajji T.K. Fitzgerald P.B. et al.Neurobiological predictors of response to dorsolateral prefrontal cortex repetitive transcranial magnetic stimulation in depression: a systematic review.Depress Anxiety. 2015; 32: 871-891https://doi.org/10.1002/da.22424Crossref PubMed Scopus (44) Google Scholar]. However, their large-scale use in clinical settings is limited by their usefulness at an individual level as well as their costs and accessibility. Recently, Feffer and colleagues [[6]Feffer K. Lee H.H. Mansouri F. Giacobbe P. Vila-Rodriguez F. Kennedy S.H. et al.Early symptom improvement at 10 sessions as a predictor of rTMS treatment outcome in major depression.Brain Stimulation. 2018; 11: 181-189https://doi.org/10.1016/j.brs.2017.10.010Abstract Full Text Full Text PDF PubMed Scopus (14) Google Scholar] proposed to use the self-rated Beck Depression Inventory (BDI) to measure early symptoms variations as predictor of nonresponse in patients with MDE receiving rTMS. They reported that the lack of early symptom improvement at 10 sessions might predict nonresponse to high frequency rTMS (either 10 Hz or intermittent theta burst stimulation - iTBS) applied over the left dorsomedial prefrontal cortex (DmPFC) in patients with major depression whatever the stimulation protocol (10Hz or iTBS). Here, we used a similar approach in order to investigate whether early variation of BDI scores may predict clinical response in a sample of patients with medication-resistant depression who received low frequency rTMS applied over the right dorsolateral prefrontal cortex (DLPFC). We hypothesized that as observed with high frequency rTMS over the DmPFC, the lack of early improvement in BDI scores might predict nonresponse in patients who received 1Hz rTMS applied over the DLPFC given alone or in combination with an antidepressant (venlafaxine), a commonly used association. We performed a retrospective evaluation of BDI scores for 92 patients with medication-resistant unipolar MDE who received 2–6 weeks (10–30 sessions) of daily active 1Hz-rTMS combined either with active-venlafaxine (n = 46) or with placebo-venlafaxine (n = 46) in a previously published randomized-controlled study [[7]Brunelin J. Jalenques I. Trojak B. Attal J. Szekely D. Gay A. et al.The efficacy and safety of low frequency repetitive transcranial magnetic stimulation for treatment-resistant depression: the results from a large multicenter French RCT.Brain Stimul. 2014; 7: 855-863https://doi.org/10.1016/j.brs.2014.07.040Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar]. A session of stimulation consisted of 360 pulses of stimulation delivered in 6 bursts of 60 sec separated by 30 sec off. The intensity of stimulation was set a 120% of the resting motor threshold. All the patients provided a written informed consent after a detailed description of the study. The study was approved by a local ethic committee (CPP Sud Est) and registered in a database NCT00714090. The protocol and the inclusion criteria are detailed in Refs. [[7]Brunelin J. Jalenques I. Trojak B. Attal J. Szekely D. Gay A. et al.The efficacy and safety of low frequency repetitive transcranial magnetic stimulation for treatment-resistant depression: the results from a large multicenter French RCT.Brain Stimul. 2014; 7: 855-863https://doi.org/10.1016/j.brs.2014.07.040Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar]. We used the shortened 13-item self-rated BDI [[8]Beck A.T. Ward C.H. Mendelson M. Mock J. Erbaugh J. An inventory for measuring depression.Arch Gen Psychiatr. 1961; 4: 561-571https://doi.org/10.1001/archpsyc.1961.01710120031004Crossref PubMed Scopus (26699) Google Scholar] to assess depressive symptoms at baseline and weekly, after every 5 sessions of rTMS. The early-improvement assessment point was defined at week 1, after the 5 first sessions of stimulation because some patients were already in remission at week 2, after 10 sessions. The post-treatment score was defined as the last clinical assessment point in the window from week 2 to week 6 (after 10 to 30 sessions of rTMS). The prediction metric at these timepoints was defined as the percentage improvement in BDI score, calculated using the pre-treatment BDI score as the baseline. At baseline, no between-groups differences were observed for sociodemographic and clinical characteristics including severity of depression (Table 1). As described in Feffer et al.'s study [[6]Feffer K. Lee H.H. Mansouri F. Giacobbe P. Vila-Rodriguez F. Kennedy S.H. et al.Early symptom improvement at 10 sessions as a predictor of rTMS treatment outcome in major depression.Brain Stimulation. 2018; 11: 181-189https://doi.org/10.1016/j.brs.2017.10.010Abstract Full Text Full Text PDF PubMed Scopus (14) Google Scholar], we first compared the empirical cumulative distribution functions for the two treatment protocols, both at week 1 and endpoint. We used the nonparametric Kolmogorov-Smirnov two-sample test because the distributions were not normal. We found no significant differences in the distribution of clinical outcomes between rTMS-only and rTMS + venlafaxine groups either at week 1 (D = 0.15, p = 0.66) or at post-treatment (D = 0.26, p = 0.09). Hence, data from both groups were combined for all subsequent analyses.Table 1Demographic characteristics of patients receiving 1Hz-rTMS over the right DLPFC combined either with active-venlafaxine (rTMS + venlafaxine) or with placebo-venlafaxine (rTMS-only), presented as means (standard deviations). The statistical significance of between-group comparisons was assessed with Student's t tests for continuous data and Chi-square for gender.rTMS-onlyrTMS + venlafaxinepGender (N men/women)17/2915/310.66Age52.4 (12.0)54.4 (11.8)0.42Current episode duration (months)15.8 (17.0)14.7 (14.9)0.74N of previous treatment failures2.3 (1.7)2.7 (2.3)0.35N of rTMS sessions received25.1 (6.6)24.8 (6.7)0.84BDI at baseline22.3 (5.8)21.1 (5.7)0.32MADRS at baseline32.3 (6.4)33.7 (6.0)0.28 Open table in a new tab To distinguish responders from nonresponders, we used the post-treatment BDI score and applied a classical criterion of an at least 50% decrease from baseline. We performed a kernel density estimate with an Epanechnikov kernel to determine the overall distribution of response. Results showed that the distribution was trimodal with a distinct subgroup of individuals who achieved <30% improvement (30.3% exactly). We thus defined nonresponders both at the conventional <50% improvement criterion and at a more stringent <30% criterion. We first computed negative predictive values (NPVs) using the <50% improvement criterion for nonresponse at post-treatment. NPVs were computed across a range of cut-off values from 0 to 50% improvement at week 1. As depicted in the Fig. 1 panel A, NPVs of 79–85% were observed when cut-offs of <15% improvement were applied at week 1. However, NPVs dropped at ∼70% when cut-offs were relaxed beyond <15% improvement at week 1. For patients achieving less than 15% improvement after 5 sessions, it was possible to be ∼80% certain of nonresponse. The full confusion matrix for <15% improvement at week 1 showed a sensitivity of 76.3%; specificity of 66.7%, and overall accuracy of 70.7% (Fig. 1 panel C). NPVs only slightly improved when using a cutoff of <10% improvement at week 1 (Fig. 1 panel D). We next assessed the NPV using the data driven nonresponse criterion of <30% improvement at post-treatment (Fig. 1 panel B). With cutoffs of <15% at week 1, NPVs were around 70–80%. NPVs remarkably declined when cutoffs were extended beyond <15% at week 1. The confusion matrix showed a sensitivity of 75.0%, specificity of 80.0%, NPV of 71.1%, PPV of 83.0%, and overall accuracy of 77.2% (Fig. 1 panel E). NPVs were not notably improved by the use of a cutoff of <10% improvement at week 1 (Fig. 1 panel F). In the current study; we were able to predict nonresponse to rTMS with a ∼80% certainty in patients displaying less than 15% improvement after only 5 sessions. The lack of early improvement after 5 sessions might thus be a predictor of nonresponse to 1Hz rTMS applied over the right DLPFC. The observed predictive values of nonresponse are in line with those reported by Feffer et al. [[6]Feffer K. Lee H.H. Mansouri F. Giacobbe P. Vila-Rodriguez F. Kennedy S.H. et al.Early symptom improvement at 10 sessions as a predictor of rTMS treatment outcome in major depression.Brain Stimulation. 2018; 11: 181-189https://doi.org/10.1016/j.brs.2017.10.010Abstract Full Text Full Text PDF PubMed Scopus (14) Google Scholar] after 10 sessions of either iTBS or 10Hz rTMS over the DmPFC. More precisely, Feffer et al. reported that patients achieving less than 20% improvement at the BDI after 10 sessions of stimulation were correctly predicted as nonresponders at endpoint with a NPV of 88.2%. Our results are also in line with the study from Lin et al. reporting that not achieving an at least 20% improvement after 6 ECT predict non response in 75% of cases [[9]Lin C.-H. Chen M.-C. Yang W.-C. Lane H.-Y. Early improvement predicts outcome of major depressive patients treated with electroconvulsive therapy.Eur Neuropsychopharmacol. 2016; 26: 225-233https://doi.org/10.1016/j.euroneuro.2015.12.019Crossref PubMed Scopus (23) Google Scholar]. Altogether, these results suggest that the lack of early improvement might predict nonresponse to brain stimulation whatever the combined antidepressant treatment (with venlafaxine or as monotherapy) and the stimulation parameters (stimulation protocol and stimulation location). Although ∼80% certainty seems not enough to allow a strict clinical decision, we believe that identifying predictors of nonresponse might be useful in clinical settings to help rTMS practitioners taking decisions regarding treatment continuation in patients with MDE receiving 1Hz rTMS over the DLPFC. In case of early nonresponse to 1Hz rTMS after 5 sessions, practitioners could thus decide to switch or persist and quickly adapt rTMS treatment parameters and propose either HF rTMS over the DmPFC or the DLPFC or bilateral stimulation or move to another treatment such as ECT. Indeed, nonresponse to one type of rTMS protocol did not predict response rate to another [[10]Speer A.M. Benson B.E. Kimbrell T.K. Wassermann E.M. Willis M.W. Herscovitch P. et al.Opposite effects of high and low frequency rTMS on mood in depressed patients: relationship to baseline cerebral activity on PET.J Affect Disord. 2009; 115: 386-394https://doi.org/10.1016/j.jad.2008.10.006Crossref PubMed Scopus (95) Google Scholar]. The authors declare that they have no conflict of interest that may have influence this work. The authors are grateful to the French Association for Biological Psychiatry (AFPBN), STEP Group - Section for NIBS in Psychiatry. This study was supported by the French Ministry of Health, France ( DGOS - PHRC 2007 ). The placebo-venlafaxine was synthesized and delivered by the Wyeth (Pfizer) laboratory. The placebo and the active venlafaxine had identical appearance in similar capsules. The funding sources had no role in the design and conduct of the study, the data collection management, analysis and interpretation of the data, drafting or submission of the manuscript.