Obsessive-compulsive disorder (OCD) occurs in the form of repetitive intrusive thoughts and/or actions, which are typically experienced as uncomfortable and often considered as nonsensical. Patients’ attempts to resist these thoughts or actions are mostly unsuccessful. For OCD, the lifetime prevalence is about 2–3% worldwide (1).

OCD causes significant distress, is time-consuming, and significantly interferes with the person’s normal daily routine, work (or school) functioning, or usual activities and relationships (2).

OCD is highly co-morbid with other mental illnesses (3). It has been reported that 90% of those who meet the criteria for OCD also meet the criteria of at least one other mental disorder over the course of their life (1). Frequent comorbidities of OCD include mood disorders and anxiety disorders (4).

The standard treatment for OCD consists of a combination of psychotherapy and pharmacotherapy, but even with these treatments, about 40–60% of patients experience only partial recovery, and about 10% of OCD patients remain treatment resistant, which leads to significant functional limitations (5–7).

One treatment option for severe, therapy-refractory obsessive-compulsive disorders represents deep brain stimulation (DBS). In DBS, electrodes are stereotactically implanted unilaterally or bilaterally in specific brain regions, which can then be stimulated via a battery located under the skin of the upper chest. This method has already proven remarkable benefits for people with a variety of neurologic conditions, and many researchers have investigated the potential benefit of DBS of selected brain regions for other disorders such as pain, depression, and obsessive compulsive disorder (8). Bilateral DBS of the anterior limb of the internal capsule (ALIC) has shown remarkable effects in patients with treatment resistant OCD (9, 10) and based on these data, the U.S. Food and Drug Administration (FDA) approved in 2009 DBS for treatment-refractory OCD as a humanitarian device exemption (HDE H050003) (11).

The exact mechanism of action of DBS in OCD is not fully understood. Since OCD might be explained through abnormal activity in corticostriatal-thalamocortical (CSTC) circuitry (12), it is hypothesized that DBS may disrupt the CSTC circuit, thereby restoring normal connectivity within and between circuits (13). Also, it has been postulated that DBS induces release of neurotransmitters, such as GABA and glutamate (14, 15). The role of the later in pathogenesis, clinical manifestations and treatment response of obsessive-compulsive disorder has been supported by evidence from animal studies, neurophysiological studies, genetic, neuroimaging studies (16). Accordingly glutamate-modulating medications have been investigated as a treatment option for OCD (17).

The effectiveness of DBS in OCD has been examined in several studies, targeting different brain regions. Mainly, the anterior limb of internal capsule (ALIC) (18–20), the bed nucleus striae terminalis (BNST) (21), ventral capsule/ventral striatum (VC/VS) (22–24), the nucleus accumbens (NA) (25–27), and the nucleus subthalamicus (STN) (28) have been targeted. The efficacy of DBS for OCD patients has been evidenced by many further studies (29, 30), summarized in various systematic reviews (31–34). Meta-analyses reported statistically significant effects of DBS in treatment-resistant OCD patients (35, 36). Alonso et al. estimated a mean reduction of the Yale–Brown Obsessive Compulsive Scale (Y-BOCS) score of 45.1% and a global percentage of responders of 60.0% (37). Moreover, long-term data demonstrated sustained improvement for the responders (38–46). Despite these impressive results, only very few potential candidates receive DBS for OCD (47). According to recent systematic reviews, only data from about 350 patients is reported in the world literature (48). Moreover, the investigated stimulation targets vary considerably across studies.

In this context, a systematic collection of clinical and demographic data of further patients would be highly desirable, to gain more knowledge about efficacy and safety, the importance of the stimulation target, successful stimulation settings, the relevance of medication as well as the significance of predictors for response and effects of DBS treatment on patients’ co-morbidities. Thus, in order to enlarge the knowledge base of DBS for OCD we report comprehensive clinical, demographic, and treatment data from 10 consecutive patients from our institution.

All patients presented have provided written informed consent to this observational study, which was approved by the ethic committee of the University of Regensburg (ethic vote: 21-2707-104). Observational means that DBS was not the issue of the study but the systematic evaluation of the change in patients’ pathology after treatment. All patients, who underwent DBS for their OCD between January 2020 and December 2022 at the multidisciplinary center of deep brain stimulation at the University of Regensburg, Germany, were included in the study.

Eleven patients have fulfilled our criteria, were identified as eligible candidates and were implanted between January 1st, 2020 and December 31st 2022. One patient has retracted his consent to publish the results of his treatment, thus his data are not included in the analysis. We are reporting the results of 10 patients (5 males, 5 females, age between 20 and 63 years “mean: 37 years”). Since the patients were implanted at different times and some of them missed visits, the number of patients varies for every visit. The demographic data are summarized in Table 1.

Since the sample size is rather small, the results are presented mainly descriptively. According to accepted standards, the treatment response was defined as a reduction in YBOCS of at least 35% compared to baseline (55). Out of our 10 patients, 6 (2 males, 4 females) have reached the response criterion indicated by a YBOCS reduction between 42 and 100 percent at last follow-up visit (after 12 months for four patients, 9 months for one patient, and 6 months for one patient). One further patient experienced a subjectively dramatic effect on her OCD symptoms after DBS, which could also be objectified with a YBOCS-reduction of 90%. Like other patients, the optimization of the stimulation parameters for this patient took place in our clinic in an inpatient setting, and the patient was dismissed after reaching satisfactory symptom improvement. Some weeks afterwards, the patient reported deterioration of depressive symptoms and also complained about vomiting (not self-induced). The medication (clomipramine) was reduced. The patient expressed the wish to stop the stimulation or extract the electrodes. Although no clear relationship between clinical worsening and the stimulation could be established, we opted to stop the stimulation temporarily. After 3 months, the patient contacted us asking for a reactivation of the DBS-therapy.

The other 3 patients showed a slight improvement of YBOCS scores (between 8.8 and 21.9% YBOCS reduction) but did not reach the response criterion (see Table 2).

The overall mean YBOCS decreased from 28.3 at baseline to 13.3 (53% reduction) at the last follow-up. For responders (6 patients) the overall mean YBOCS decreased from 25.1 at baseline to 7.6 (69.7% reduction) at last follow-up.

Since the follow-up period varied among patients, we calculated also the mean values for YBOCS of both first two visits (baseline and before stimulation) and compared them with the mean values of all follow-up visits combined (3 months, 6 months, 9 months, and 12 months). The percentage of YBOCS reduction varied slightly as compared to the comparison between baseline and last follow-up, but the response status did not change (see Table 2; Figure 1).

The improvement of the OCD symptoms was also accompanied by an improvement of depressive symptoms, as shown by a reduction in MADRS scores in six out of seven patients (MADRS scores were only available for 7 patients, see Figure 2). Overall, the mean MADRS-score improved from 28 points at baseline to 12.4 at the last follow-up.

GAF mean score increased from 34.5 at baseline to 61.5 at last follow-up, indicating an improvement of 78% (see Figure 3) and CGI-S showed an improvement of 32% (see Figure 4). CGI-I mean score was reduced from four points before stimulation to two points at last follow-up (see Figure 5). All domains of WHOQOL (physical, psychological, social, and environment) have shown improvement when comparing the mean values of baseline, 3 months follow-up, and last follow-up. The mean of total four domains’ scores increased from 44.8 at baseline to 54.7 at 3 months follow-up and to 58.5 at the last follow-up (see Table 3; Supplementary material).

Statistically, a significant effect of study visit was observed for each assessment inventory except for the domain environment of the WHOQOL-BREF. Post hoc tests revealed symptom improvement from baseline/before stimulation to post stimulation (see Figure 1).

The stimulation parameters were optimized individually for every patient according to symptom reduction and tolerance. The DBS parameter settings in our patients ranged from 4 to 6 V amplitude, 60–120 ms pulse width, 130–180 Hz frequency (see Table 3). Stimulation related side effects were mainly experienced during the titration visits, i.e. stimulation-induced and could be resolved by adjusting the stimulation parameters. There were no serious AEs. During the titration visits, two patients have experienced hypomanic symptoms like excessive talking, increased energy, euphoric mood, and partially inadequate behavior. Since those patients were inpatients in our clinic, the symptoms have been identified fast and the stimulation was reduced or stopped after some hours, which resulted in immediate disappearance of the hypomanic symptoms without further problems. Most frequent side effects during the optimization period were insomnia, restlessness and sensation of fear. One patient reported stuttering after about 6 months of the stimulation. The side effects are summarized in Table 3.

Three of the six responders reduced their medication after DBS. Also one of the non-responders discontinued his medication and reported no worsening of symptoms.

The aim of this report is the evaluation of the effectiveness and safety of DBS for therapy-resistant OCD patients in a naturalistic setting. Overall, 60% of the treated patients fulfilled the response criterion, and clinical improvement continued over the follow-up period, ranging from 6 months to 1 year. In the whole sample, there was a mean reduction of 53% in the YBOCS score. The response in the YBOCS was accompanied by improvement in GAF, CGI, and quality of life.

Our results are in keeping with other published data: Denys et al. implanted 16 patients and found a YBOCS reduction of 46%, with 9 of the patients being responders (26). Barcia et al. reported response in six out of seven patients, with a median symptomatic reduction of 50% (25). Menchon et al. reported a Y-BOCS reduction of 42% with a responder rate of 60% (19). Further studies revealed comparable results, for example a ≥ 35% YBOCS reduction in four out of 6 patients for Goodman et al. (22) and a YBOCS reduction of 48% with responder rate of 70% for Luyten et al. (21). Greenberg et al. published the results of 26 patients implanted in (VC/VS), revealing clinically significant symptom reductions and functional improvement in about two-thirds of patients (23). Tyagi et al. compared VC/VS and amSTN DBS in their study, showing that stimulation at both targets was associated with a significant improvement of YBOCS scores over baseline (24).

Remarkably, the above-mentioned results lay all in the same range, even if they come from relatively small samples, and studies using various experimental designs and brain targets.

According to our results, the symptom reduction as indicated by YBOCS was paralleled by an improvement in quality of life and the regaining of social participation. In our group, four patients started to work following DBS after many years of unemployment due to illness. This illustrates the impact of DBS on patients’ level of functioning as well as overall satisfaction. This is also consistent with previous findings, reporting a significant improvement in quality of life for DBS OCD patients (39, 56).

In our group, there was one patient who showed a significant initial response and afterwards reported a deterioration of depressive symptoms and opted to stop the stimulation. Our hypothesis for this deterioration is as follows; firstly, it is well recognized that after symptom reduction or remission, many chronically ill individuals have difficulty adjusting to their new situation (57). The relief of symptoms following DBS represents a major challenge for patients, as it goes along with major changes in identity and relationships (58). Especially the regaining of social participation is accompanied by new types of challenges, such as stress at the workplace. Secondly, this patient displayed signs of a comorbid personality problem, which manifested as instability and impulsivity. Although this was not confirmed by neuropsychological assessment, it could have contributed to the afterward deterioration.

Although three patients displayed only a minor reduction of the YBOCS score and have been considered non-responders, none of them considered stopping the stimulation. One of the non-responders was able to stop his medication without deterioration of his symptoms.

Stimulation related adverse events in our group of patients could all be resolved through adjustment of the stimulation parameters. Hypomania was the most severe adverse event.

In our group of patients, there were three patients who might be considered only moderately affected according to the baseline YBOCS score (YBOCS < 24points), yet those three patients were showing clear signs of suffering, treatment resistance and evident disturbed level of functioning. They all responded well to DBS. This raises many questions or challenges to DBS research. Should DBS be restricted to only severely affected patients? What are the criteria for this severity? And how to determine the most suitable candidates for DBS surgery, in other words, which criteria determine the likelihood of a treatment response. The good response in patients with lower YBOCS score parallels to a certain extent the situation in Parkinson’s disease where the indication for deep brain stimulation moved over the years from a treatment for severely treatment resistant patients to patients in earlier stages of the disease (59).

Interestingly, in all our patients obsessive thoughts were clearly more pronounced than compulsive acts. This relativizes the relevance of the YBOCS total score as the sole criterion for the severity of the disease, as patients, who score low in compulsive acts, have a relatively low total score, even if they are extremely impaired by obsessive thoughts. The preponderance of obsessive thoughts in our patients may reflect the fact, that compulsive acts can be better addressed by CBT than obsessive thoughts. Thus, patients suffering predominantly from obsessive thoughts may be overrepresented among treatment resistant patients.

A major clinical challenge is the individual optimization of stimulation parameters. The combination of all possible settings results in a huge parameter space, making systematic testing of all combinations impossible. A valid assessment of the efficacy and tolerability of a given stimulation setting typically requires a time period of at least several days. In clinical practice, we used a stepwise exploration of contact, voltage, frequency and stimulus width. Settings that seemed optimal in the clinic were then evaluated under real life conditions, as OCD symptoms are typically context dependent.

An ongoing matter of debate is the optimal neuroanatomical target, since several neuroanatomical regions have been targeted with comparable outcomes. All our patients were implanted in the BNST. The BNST, which is considered as a part of the “extended amygdala” (60), is a brain nucleus embedding the stria terminalis and located posterior to the nucleus accumbens (61). It is suggested that the BNST is involved in striatal circuitry that integrates descending glutamatergic input with ascending modulatory inputs (62). Through its role connecting limbic forebrain structures to hypothalamic and brainstem regions associated with autonomic and neuroendocrine functions, the BNST serves as a major component in the integration of physiological and behavioral responses (63). In addition, an interaction of neurotransmitters within the BNST has been reported, primarily via a modulation of presynaptic neurotransmitter release (64–67). The BNST was first introduced as DBS target for the treatment of OCD by Nuttin et al. (68). In their study of comparing BNST DBS with NA DBS, Islam et al. reported a better outcome for DBS in the BNST (69). Yet, according to Farrand et al., the overall effect of these two brain targets was comparable (70). Another study of 24 patients found a better result of DBS in the BNST compared with patients implanted in the anterior limb of internal capsule (21) with also reported stability of symptom reduction over time (71).

An important issue in the DBS field for psychiatric disorders, which has been recently stressed (47) is the accessibility problem. Firstly, DBS for OCD requires collaboration between psychiatric and neurosurgery departments and experienced personnel for patient selection, surgery, and therapy optimization. This is only available at a few centers.

Secondly, although DBS is an approved therapy with reported long-term cost-effectiveness (72, 73), its costs are not normally covered by health insurance companies. An application for coverage of DBS’s costs for severe OCD patients is often denied by the health insurance companies. For OCD patients, who cannot afford to pay the high costs of DBS themselves, this means depriving them of access to this therapy (74).

Thirdly, DBS seems not to be perceived by many psychiatrists as an available therapeutic option. The majority of our patients were not referred to us by their psychiatrist, but presented in our clinic on their own initiative after finding the option of deep brain stimulation on the internet.

While our data illustrate the efficacy of DBS for patients with treatment-resistant OCD, we are aware of many study limitations. First, neither randomization nor a sham control existed. Second, several patients missed some appointments, and the duration of the follow-up period varied among individuals. Thirdly, stimulation parameters were not standardized but adjusted individually for each patient. Notwithstanding these limitations, we emphasized a transparent and comprehensive presentation of the individual demographic and clinical characteristics of our patients to supply further valuable data to the DBS research field.

Our results further confirm that BNST DBS is effective for treatment-resistant OCD patients, as indicated by a reduction in symptoms and an overall improvement in functioning. Beside the need for additional research to define the patient’s selection criteria, the most appropriate anatomical target, and the most effective stimulation parameters, improved patient access for this therapy should be established.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving humans were approved by Ethic Committee of University of Regensburg, Germany. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

MA, VL-H, TH, and BL screened the possible candidates, performed the evaluation ratings for enrolled patients. MA collected the data und summed the results. SS converted the results into figures and performed the statistical analysis. JS and DD performed the neurosurgical operations. MA, BL, and JS drafted the manuscript. Other authors revised and corrected the drafts. All the authors designed the study, interpreted the data, and approved the final version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.