In this retrospective study, we examined rTMS therapeutic efficacy in 65 pharmaco-resistant OCD outpatients. An overall significant reduction in OCD symptoms and anxiety / depressive states were observed after 20 sessions of rTMS. 46.2% of the patients responded to rTMS treatment (based on the criterion of 30% reduction of Y-BOCS baseline scores), and a significant reduction in OCD symptoms was observed in the whole patient group, including non-responders (less than 30% symptom reduction). No significant difference in treatment efficacy between the intervention protocols (i.e., bilateral DLPFC vs. SMA rTMS) was found. Regarding the predictors of rTMS response, no demographic predictor (i.e., age, gender, marital status) was identified. Both obsession and compulsion -related items/factors were related to response rate. “Obsessions severity”, “disturbance”, and “resistance” were the clinical factors that significantly predicted response to rTMS. In this line, items 2 and 9 of the Y-BOCS (i.e., “interference due to obsessive thought” and “resistance against compulsions”) were the most relevant clinical predictors of response to rTMS based on individual items regression analysis.
The response rate to rTMS in our OCD sample is in line with previous studies. The first meta-analysis in the field included 10 randomized controlled rTMS studies (with ≥25–40% reduction in Y-BOCS scores) and reported a 35% response rate in 120 OCD patients that received rTMS[19]. Other recent studies reported a response rate of 40%-55% based on the 30% reduction versus Y-BOCS baseline score criterion [30-33]. In all of these studies, rTMS was applied over the DLPFC, SMA or pre-SMA except for [32], that targeted the medial PFC. Another recent meta-analysis showed that the therapeutic outcome of DLPFC vs SMA rTMS protocols is not significantly different, which was confirmed by our study results [35]. A recent rTMS study in OCD patients, which targeted the DMPFC, reported, however, a success rate of 50% with ≥50% reduction in post-treatment Y-BOCS scores specifically in those OCD patients with hyperconnectivity of fronto-striatal circuits [9].
This pattern of results suggests that response to rTMS in OCD patients depends on the pathophysiology of target region/s and the appropriate modulation of the involved regions. OCD is a heterogeneous disorder not only at the symptom, but also the pathophysiological level [4, 5, 60]. It can be speculated that nonresponders to rTMS in our study resemble OCD subtypes with specific pathophysiological features, including involved cortico-subcortical regions in deeper brain areas (e.g. OFC) that were not adequately modulated by SMA or DLPFC rTMS. The rationale for targeting SMA and DLPFC with these protocols is related to the activation pattern of these regions in OCD. The SMA and right DLPFC that have extensive connections with regions implicated in cognitive processes and motor control and response inhibition [61, 62], show hyperactivation in OCD patients. The 1 Hz stimulation applied over the SMA and right DLPFC has an inhibitory effect and is thus expected to reduce activation in these regions [42]. On the other hand, the left DLPFC is involved in cognitive control [63] and increasing its activation with NIBS has been associated with improved control over intrusive thoughts [35]. Increasing the left DLPFC with excitatory and right DLPFC with inhibitory stimulation is in line with the putative regions that are affected in OCD and involved in cognitive control and response inhibition / affect respectively [10]. Nevertheless, whether and how these protocols are effective in OCD patients depends on the individualized stimulation protocol taking into account the underlying pathophysiology, relevant symptoms, and comorbid diagnosis.
The length and number of rTMS sessions can also affect the response, with a higher number of sessions providing more symptom reduction [30, 64]. Our findings support this assumption since we observed a significant reduction of Y-BOCS scores even in non-responders with a relatively intensive intervention protocol. With respect to the efficacy of the protocols, lack of sham condition does not allow us to rule out a potential placebo effect despite the baseline-control condition we have. Secondly, similar efficacy of DLPFC vs SMA rTMS protocols should be considered in the absence of demographic and questionnaire-based group differences. Although allocation to DLPFC protocol was based on one higher report of affective states in patients, no significant difference was found between the patients in the baseline depressive scores based on the BDI-II. Future studies should use a more strict cut-off point, using objective measures, for determining depression states and group allocation accordingly.
Our finding of the absence of relevant demographic predictors of rTMS response is in line with a recent report of a missing correlation between baseline psychometric factors, including age, and rTMS treatment outcome[9]. Age, however, seems to be a predictor of rTMS response in depression [65-68]. Regarding clinical predictors, our model showed that significant factors and items that predicted response to rTMS were related to both obsession and compulsion. The obsession severity, resistance and “disturbance” were the factors with the highest predictive ability in rTMS response and the last two factors (resistance, disturbance factors) include items related to both obsession and compulsion. Our model based on Y-BOCS items showed a similar pattern of predictors. Specifically, we found that higher scores in “interference due to obsessions” (item 2), which is related to disturbance factor, and “resistance against compulsion” (item 9), related to the resistance factor, determined response failure to rTMS treatment. The common factor underlying both obsession and compulsion related items is the disturbance factor and this implicates that those OCD symptoms (including both obsession and compulsion-related ones) that result in more disturbance/interference are important in predicting response to rTMS.
The predictive value of these factors and relevant items could have clinical implications for treatment response. Patients with more “severe obsessions” have more intrusive thoughts and experience greater overall difficulties due to obsession interference [33, 69]. This is in line with recent findings showing that obsessions are usually not targeted by rTMS protocols while they are important in determining treatment response and thus should be primarily targeted in future interventions [70]. Furthermore, the relationship between intrusive thoughts and development and maintenance of OCD symptoms[71] might also explain why severe obsessive symptoms hinder response to rTMS treatment. This, however, should not implicate that compulsions are not important. Indeed, the “disturbance” factor, which is the strongest predictor among the factors, is related to distress and interference in both obsession and compulsion indicative of [56]. The “resistance” factor similarly reflects the severity of both obsessive and compulsive symptoms that are difficult to overcome [56, 57, 72]. Items related to the “resistance” factor, do not significantly change after pharmacological treatment [57, 73]. Taken together, these factors and the relevant items indicate that the severity of symptoms and the level of interference and disturbance caused by OCD symptoms (related to both obsession and compulsions) are negative predictors of rTMS response.
Since these aspects of OCD symptoms seem to be of utmost importance for response failure to rTMS treatment, it might on the one hand help to decide about therapeutic options in specific patients. On the other hand, it also suggests that patients with a lower probability for successful rTMS treatment with conventional protocols might require alternative interventions. In this line, one important implication of our findings is the need for adapting rTMS protocols to patients’ symptoms. According to our results, high levels of the interference, disturbance and resistance aspects of OCD symptoms have a negative impact on rTMS response. Patients with prominent symptoms related to these factors might benefit more from protocols that are optimized to have a strong impact on interference or inhibition. An alternative option might be to treat these patients with a treatment approach focused on improving cognitive control strategies over both obsessions and compulsions [74]. For example, treatment strategies that are focused on improving appraisal and control strategies in response to intrusive thoughts, which are associated with distress and interference [75-78], might be beneficial to OCD patients who fail to respond to rTMS due to being highly disturbed by intrusive thoughts and compulsions.
Our findings are preliminary, and this study has some limitations. The major limitation of our analysis is the retrospective study design. On the other hand, this study has relatively high ecological validity and provides a realistic picture of rTMS application in clinical settings. Secondly, our control condition is limited to a baseline-control and lacks a sham intervention condition. Unblinded assessment by multiple raters, which could be a source of bias, and inter-rater variability, are other limitations of our work. Although the sample size was large enough for investigating the efficacy of rTMS, it may have been relatively underpowered for obtaining robust results by the regression analysis performed on individual items of the Y-BOCS. Sample size, however, was sufficiently large for factor-based regression analysis, which resulted in similar clinical predictors of rTMS response. The use of 120% of active rather than resting motor threshold could be a limiting factor in determining response rate as rTMS at AMT intensity usually delivers underdosage stimulation compared to RMT [79]. Allocation to rTMS protocols (DLPFC vs SMA) based on the prominence of depressive states determined by clinical impression and mere self-reports should be improved in future studies by applying higher cut-off points. Lastly, although we kept the medication dosage constant 12 weeks before the experiment and throughout the intervention (4-6 weeks) to minimize potential confounding and interference, and this factor did not predict response status, it should be controlled directly in future studies, as it might be a potential source of variability of rTMS effects.