Article Text

Download PDFPDF

Research paper
Open-label trial of anterior limb of internal capsule–nucleus accumbens deep brain stimulation for obsessive-compulsive disorder: insights gained
  1. Daniel Huys1,
  2. Sina Kohl1,
  3. Juan Carlos Baldermann1,
  4. Lars Timmermann2,
  5. Volker Sturm3,
  6. Veerle Visser-Vandewalle1,4,
  7. Jens Kuhn1,5
  1. 1 Department of Psychiatry and Psychotherapy, University Hospital of Cologne, Cologne, Germany
  2. 2 Department of Neurology, University Hospital Giessen and Marburg, Campus Marburg, Marburg, Germany
  3. 3 Department of Stereotaxy and Functional Neurosurgery, University Hospital of Cologne, Cologne, Germany
  4. 4 Department of Stereotaxy and Functional Neurosurgery, University Hospital of Cologne, Cologne, Germany
  5. 5 Johanniter Hospital Oberhausen, Department of Psychiatry, Psychotherapy and Psychosomatics, Oberhausen, Germany
  1. Correspondence to Dr. med. Daniel Huys, Klinik und Poliklinik fur Psychiatrie und Psychotherapie, Uniklinik Köln, 50937 Köln, Germany; daniel.huys{at}uk-koeln.de

Abstract

Background For more than 15 years, deep brain stimulation (DBS) has served as a last-resort treatment for severe treatment-resistant obsessive-compulsive disorder (OCD).

Methods From 2010 to 2016, 20 patients with OCD (10 men/10 women) were included in a single-centre trial with a naturalistic open-label design over 1 year to evaluate the effects of DBS in the anterior limb of the internal capsule and nucleus accumbens region (ALIC-NAcc) on OCD symptoms, executive functions, and personality traits.

Results ALIC-NAcc-DBS significantly decreased OCD symptoms (mean Yale-Brown Obsessive Compulsive Scale reduction 33%, 40% full responders) and improves global functioning without loss of efficacy over 1 year. No significant changes were found in depressive or anxiety symptoms. Our study did not show any effect of ALIC-NAcc-DBS on personality traits or executive functions, and no potential outcome predictors were identified in a post hoc analysis. Other than several individual minor adverse events, ALIC-NAcc-DBS has been shown to be safe, but 35% of patients reported a sudden increase in anxiety and anhedonia after acute cessation of stimulation.

Conclusions We conclude that ALIC-NAcc-DBS is a well-tolerated and promising last-resort treatment option for OCD. The cause of variability in the outcome remains unclear, and the aspect of reversibility must be examined critically. The present data from one of the largest samples of patients with OCD treated with DBS thus far support the results of previous studies with smaller samples.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

With a lifetime prevalence of 2%–3%, obsessive-compulsive disorder (OCD) is among the five most common mental disorders. OCD is characterised by recurrent and intrusive thoughts or repetitive behaviours, or both, that are time consuming, cause remarkable distress and often perceived as inappropriate by the affected person who often unsuccessfully tries to resist. OCD bears a high level of suffering and greatly impact psychosocial functioning and quality of life.1 Guidelines agree on cognitive-behavioural therapy (CBT) as first-line treatment, combined with pharmacotherapy if applicable.2 However, guideline-conform treatments provide only 40%–60% symptom reduction in approximately 70% of patients with OCD.3

In the last decade, deep brain stimulation (DBS), a stereotactic procedure that applies chronic stimulation in subcortical regions of the brain, has gained attention as a useful therapy for patients with treatment-resistant OCD. Across all targets employed for DBS treatment of OCD, a response rate of 58.2% and a mean symptom reduction of 47.7% have been reported.4 However, no target has stood out with respect to effectiveness.5 The rationale for applying DBS is based on experiences from ablative studies and the idea that OCD is related to neurobiological alterations within the cortico-striato-thalamo-cortical (CSTC) circuits. Supporting this hypothesis, increased activity in orbitofrontal cortex–ventromedial striatum connexions via repetitive optogenetic stimulation induces persistent grooming behaviour in mice, demonstrating a causal relationship between hyperactivity within the CSTC pathway and the generation of OCD-like behaviour.6 In humans, functional imaging studies have shown hyperactivity in the CSTC network in patients with OCD compared with controls, though results have been inconsistent.7The exact mechanisms of DBS are not fully understood, but with regard to the nucleus accumbens (NAcc)/ventral striatum as the target, it can be assumed that stimulation restores the CSTC connectivity. For example, Figee et al found decreased connectivity in the CSTC loop with stimulation ON compared with OFF in patients with OCD.8

DBS of the ventral striatum has been approved by the US Food and Drug Administration under the Humanitarian Device Exemption based on studies with small sample sizes, allowing patients access to this intervention and thereby removing the requirement for further clinical trials although many aspects are still unknown and additional clinical trials necessary.9

Here, we present a naturalistic open-label trial with 20 patients with OCD. We hypothesised that DBS of the anterior limb of the internal capsule and nucleus accumbens region (ALIC-NAcc) improves OCD symptoms over 1 year. We also wanted to test whether ALIC-NAcc-DBS induces changes in executive functions and personality traits. Based on our findings and 15 years of experience at our centre, we would like to reopen the debate on the application of DBS for OCD.

Materials and methods

Patients

Patients were recruited from the interdisciplinary outpatient clinic for obsessive-compulsive spectrum disorders at University Hospital Cologne from 2010 to 2016. Demographic information and clinical characteristics at baseline are summarised in table 1.

Table 1

Demographic and clinical information

Inclusion and exclusion criteria

All patients included in this study had a primary diagnosis of OCD according to DSM-IV criteria using the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I) and the SCID-II in order to examine axis II disorders. Patients were between 21 and 65 years of age and had chronic and severe OCD, as indicated by a score of 25 or higher on the Yale-Brown Obsessive Compulsive Scale (Y-BOCS),10 and a history of OCD for over 5 years. Inclusion criteria also included highly impaired functioning, as indicated by a Global Assessment of Functioning (GAF)11 score <40. Treatment resistance or an insufficient response to conventional therapies had to be well documented, and previous treatment trials had to include at least one trial of CBT with exposure and response management over more than 45 sessions, at least two treatment trials with selective serotonin reuptake inhibitors (SSRIs) over at least 10 weeks at maximum dosage, one trial with clomipramine over 10 weeks at maximum dosage, and an augmentation approach with antipsychotic medication, lithium or buspirone. If prior CBT trials had not been well documented, we referred patients to hospitals specialising in CBT for OCD in order to rule out ‘pseudo treatment-resistance’.

Exclusion criteria were current or previous diagnosis of psychosis, drug abuse or drug addiction during the last 6 months, traumatic brain injury in the past, clinically significant internal or neurological disorders, pregnancy, lactation or mental retardation.

Surgical procedure and stimulation adjustment

Under local or general anaesthesia, quadripolar leads (Model 3387 or 3389 DBS Lead; Medtronic, Minneapolis, Minnesota, USA) were stereotactically implanted bilaterally, guided by MRI and stereotactic cerebral CT. The two distal contacts (nos. 0, 1 on the left electrode and 8, 9 on the right electrode) were placed bilaterally in the NAcc, whereas the more proximal contacts (2, 3 and 10, 11) were located in the ventral part of the ALIC. Correct positioning of the electrodes was confirmed postoperatively by stereotactic X-ray procedures in the operating room or with a CT scan the day after surgery (figures 1 and 2).12 The leads were connected to an implantable pulse generator (IPG) at a second stage (Model 7426 Soletra, or Activa PC, Medtronic) one to several days after electrode implantation under general anaesthesia. Thereafter, systematic testing of each monopolar contact was performed. Patients were interviewed for changes in mood, calmness, anxiety and OCD symptoms. Over the course of 1–2 weeks, the two best contacts per side were chosen for the final stimulation setting. Frequency was mostly >120 Hz; pulse widths and frequency were adjusted to compensate for side effects. In the following months, the stimulation settings were optimised depending on the individual outcome. Amplitude was increased stepwise beneath the threshold for side effects. In cases of non-response, this procedure for monopolar testing of the individual contacts was repeated. Mean stimulation amplitude at 12 months of follow-up was 4855 (SD 1.1). Table 2 provides an overview of the final stimulation settings.

Figure 1

Combined individual preoperative MRI and postoperative CT scans through the location of active contacts. A and B show axial (A) and coronal (B) MRI scans of patient 19 with bipolar stimulation. Top row: MRI scans through active contacts of the right electrode; bottom row: MRI scans through active contacts of the left electrode. C and D show axial (C) and coronal (D) MRI scans of patient 20 with bipolar stimulation. Top row: MRI scans through active contacts of the right electrode; bottom row: MRI scans through active contacts of the left electrode.

Figure 2

Schematic display of electrode localisation in standard MNI space. Reconstruction of electrodes was performed using the LEAD-DBS toolbox12 in patients where postoperative CT imaging was available (n=17). Patients were implanted with bilateral leads targeting the nucleus accumbens, the apical/dorsal contacts of each patient were located within the ventral part of the anterior limb of the internal capsule.

Table 2

Stimulation parameters at 12-month follow-up (these settings were chosen according to the best clinical outcome)

Study design and outcome measure

After implantation of the DBS system, patients entered an open-label phase of 12 months. The week after surgery, stimulation settings were adjusted systematically as described above. Symptom severity was assessed 6 (T1) and 12 months (T2) after surgery, along with neuropsychological tests and personality scales. The primary outcome measure was the Y-BOCS, ranging from 0 to 40 points, with higher scores indicating more severe symptoms. A decrease of 35% or more was classified as a full response to DBS treatment, a decrease between 25% and 35% as a partial response, and a decrease below 25% as a non-response.13 In order to assess the different dimensions of OCD symptoms, we used the German self-report measure ‘Hamburger Zwangsinventar’ (HZI).14 To measure each patient’s social, occupational and psychological functioning, we used the GAF scale. The Symptom Checklist (SCL-90) was used to capture the subjective burden of physical and mental health.15 Comorbid depression were assessed using the Beck Depression Inventory (BDI),16 anxiety was rated with the State-Trait Anxiety Inventory (STAI)17 and the Yale Global Tic Severity Scale was used to measure severity of any present comorbid tics. We also assessed patients’ personality traits before and after surgery to investigate personality changes as potential side effects of DBS treatment. We used a German inventory for clinical personality traits (Inventar Klinischer Persönlichkeitsakzentuierungen)18 and the Dimensional Assessment of Personality Pathology–Basic Questionnaire (DAPPQ).19

Within the clinical trial, the following neuropsychological paradigms were employed: The Tower of London test was used to assess deficits in planning. The continuous performance test is a computerised method used to measure sustained and selective attention and impulsivity. The Stroop test was used to assess working memory, inhibition and the ability to control impulses of acting. The go/no-go test was used to assess the capacity for sustained attention and response control. Inhibition of return was used to measure covert visual attention.

Statistical analysis

All statistical analyses were performed with commercially available statistical software (IBM SPSS, V.23). Due to a non-normal distribution of most parameters, the clinical outcome measures were evaluated using non-parametric tests. We performed a Friedman test for correlated data with list-wise exclusion of missing data for all variables measured at T0, T1 and T2. If the Friedman test indicated a significant difference, a post hoc analysis was performed with the Wilcoxon signed-rank test. Spearman rank correlations were performed in order to investigate the predictive value of preoperative psychopathological parameters.

Results

Primary outcome parameter (Y-BOCS)

Twenty patients were included in this study (10 women/10 men) with a mean age of 43.15 (±13.75) years and a mean total Y-BOCS score of 30.9. At 12 months, the mean Y-BOCS reduction was 33.33% (±21.50), 40% of the patients were full responders (figure 3), five showed a remission of symptoms (Y-BOCS score <16) but none fully recovered (Y-BOCS score <8).13 Thirty per cent of the patients exhibited a partial response to treatment. A non-parametric Friedman test of differences among repeated measures was performed and rendered a significant χ2 value of 22.487 (p<0.001). Wilcoxon signed-rank tests indicated that median test ranks were significantly higher at baseline (Mdn=26) than at 6 months (Mdn=23, Z=−3.495, p=0.001) or 12 months (Mdn=21.5, Z=−3.725, p<0.001). The Wilcoxon signed-rank test was also significant when comparing 6-month and 12-month follow-up (Z=−2.952, p=0.003).

Figure 3

Individual outcomes as the percentage change in Yale-Brown Obsessive Compulsive Scale (Y-BOCS) scores after 12 months. The yellow line marks the threshold for a partial response (>25%), the green line a full response (>35%).

Secondary outcome parameters

Non-parametric Friedman tests of differences among repeated measures were also performed for secondary outcome parameters BDI, STAI and GAF. For the BDI, the Friedman test rendered a χ2 value of 0.25 (p=0.882). The Friedman tests for the STAI scores rendered a χ2 values of 0.192 for STAI 1 and 0.873 for STAI 2 (p=0.909 and p=0.646, respectively). In contrast, the Friedman test revealed a significant difference for the GAF scores, with a χ2 value of 26.00 (p<0.001). The Wilcoxon signed-rank test indicated that GAF scores were significantly lower at baseline (Mdn=35) than at 6 months (Mdn=47, Z=−3.789, p<0.001). Similarly, GAF scores were significantly lower at baseline (Mdn=35) than 12 months (Mdn=52, Z=−3.745, p<0.001), and significantly lower at 6 months (Mdn=47) compared with 12 months (Mdn=52, Z=−2.703, p=0.007) (figure 4). No significant results were found with the neuropsychological test data and personality traits (online supplementary tables 1 and 2). Mean battery life (excluding cases with early replacement due to reasons other than battery depletion) was 14.76 months (±5.42).

Supplemental material

Figure 4

Changes in symptom severity and psychosocial functioning. Both obsessions and compulsions significantly decreased after 6 and 12 months of continuous deep brain stimulation as indicated by reduced Yale-Brown Obsessive Compulsive Scale (Y-BOCS) scores, whereas psychosocial functioning significantly increased at both time points, as indicated by higher Global Assessment of Functioning (GAF) scores.

Adverse events

Adverse events included infection of the IPG pocket (n=1) and traction of IPG and cables (n=1), both with the need for operative replacement. Stimulation-related adverse events consisted of transient hypomanic states (n=1), disinhibition (n=3), lack of concentration (n=2), transient loss of energy (n=1), sleep disturbances (n=2) and significant weight gain (>20%) (n=2). Thirty-five per cent (n=7) of patients reported a sudden increase in anxiety and anhedonia after acute cessation of stimulation, with one patient reporting suicidal tendencies. These symptoms disappeared immediately after restarting stimulation. Cessation of stimulation was not part of the study protocol but sometimes necessary during clinical care (eg, ECG).

Outcome predictors

To assess potential predictors of treatment, a post hoc Spearman correlation analysis between percentage improvement rates after 12 months and preoperative demographic data, personality traits, OCD symptom domains and investigational subitems of the Y-BOCS was performed. No significant difference was found between the outcomes of female or male patients (p=0.854), and there was no significant correlation between outcome and age (r=−0.104; p=0.663) or preoperative Y-BOCS (r=0.093; p=0.697). We also found no significant correlation between subscores of the Dimensional Assessment of Personality Pathology - Basic Questionnaire (DAPPQ), HZI or investigational subitems of the Y-BOCS (see online supplementary tables 1–3 for detailed correlations).

Discussion

Here, we present 1-year follow-up data on ALIC-NAcc-DBS for OCD in one of the largest samples published thus far. Although comparable data from other centres and meta-analyses on this subject exist, many aspects that can be elaborated from our study may help in further evaluations of the clinical benefit of this procedure.

Our results underline that ALIC-NAcc-DBS can be effective for otherwise treatment-refractory patients. A mean improvement of OCD symptoms of 33.33% after 1 year of DBS with a full response rate of 40% was observed, and 70% of the patients exhibited at least a partial response (25% reduction of the Y-BOCS).13 A recent meta-analysis pooling data from studies using striatal targets (ALIC, ventral striatum, NAcc) reported slightly better improvements (39.0%) and full response rates (55.5%).20 Although our outcomes lie within the SD of that meta-analysis, there may also be other reasons for the weaker effects in our study. One reason could be that patient selection was rather strict, as we demanded onward psychotherapy in clinics specialised on behavioural therapy for OCD to rule out ‘pseudo treatment-resistance’. Thus, our sample may have been even ‘more’ treatment refractory than samples in other studies. Yet, 70% of the severely affected patients exhibited more than 25% symptom reduction, underlining the capability and efficacy of the procedure.

If patients responded to the therapy, the beneficial effects lasted over the observation time of 1 year. Although not statistically tested, we did not observe a loss of efficacy over time, that is, through habituation. Such phenomena are seen after CBT in roughly 20% of cases,21 making relapse prevention an essential element of CBT for OCD. Continuous DBS may provide more persistent protection for a relapse o OCD symptoms if patients responded in the first place. Still, patients with DBS should not consider this as a solitary procedure but rather as a trailblazer for another attempt in psychotherapy.22

A major concern is whether DBS may change personality traits and cognitive function. This is often based on reports on DBS of the subthalamic nucleus in Parkinson’s disease (PD), in which such changes with both altered cognitive functions and impulsive behaviours are not rare. Our study did not show any effect of ALIC-NAcc-DBS on both measurements of personality traits and executive function. Notably, the concept of personality may differ between psychiatrists and the popular point of view,23 and alterations in OCD symptoms may already imply changes in behaviour and, therefore, in the personality concept of patients. However, standardised measurements of personality traits are suitable tools for facing these concerns, indicating that, although behaviour may change through symptom improvement, trait-based personality as a concept of emotional pattern, self-concept and interaction with one’s surroundings is not altered.

Apart from symptom severity, patients in our study benefited significantly in terms of psychosocial functioning. This notion is not trivial, as one could argue that an ongoing brain implant may come in hand with constraints in one’s work or social life. Improving both mental health and psychosocial functioning in these highly impaired patients could have positive effects on the overall cost-effectiveness of the treatment, as well as reduce health and assistance costs. The positive long-term cost-effectiveness of DBS for OCD was already stated in systematic investigation.24

Despite more than 15 years of experiences with DBS for OCD, the field is still in its early stages compared with DBS for movement disorders. Overall, the field of DBS is developing and new stimulation procedures and protocols emerge.25 26 Researchers are also focusing on carving out treatment predictors. In our post hoc analysis, we did not find relevant demographic or clinical predictors. Often discussed aspects that may reduce the effectiveness of DBS for OCD, such as poor insight into the irrationality of obsessions and compulsions,27 personality traits and symptom categories as measured by the HZI, did not significantly correlate with the outcome. Thus, further research on which patient qualifies best for neuromodulation is highly important to improving this treatment strategy.

Although DBS is an invasive procedure, the operation was generally well tolerated. We observed one infection of the IPG and one case with pain due to cable traction, which necessitated hardware removal in both. Stimulation-related adverse events were largely congruent with those reported in a large meta-analysis20 including transient hypomania and disinhibition. These side effects were generally reversible by adjusting stimulation parameters. In two patients, we observed significant weight gain after 1 year. Here, cause-and-effect relation is unknown and the mechanisms of striatal DBS on body weight are still poorly understood and need further research.

Another important observation of this study was that acute cessation of stimulation resulted in a prominent relapse of OCD symptoms, anxiety and mood deterioration in 35% of the patients. One patient even reported an acute appearance of suicidal tendencies when switching the stimulator off. In our trial, many patients reported a worsening of obsessions and compulsions as reported previously,28 and also a sudden decline in mood or the appearance of excessive fear. Fortunately, this deterioration and rebound phenomena were immediately reversed by reinitiating the stimulation in the present trial. However, these phenomena are poorly understood. Comparable but usually weaker effects are observed with serotonergic agents used as first-line pharmacological treatment for OCD, known as SSRI Discontinuation Syndrome, and a common mechanism might be speculated for this kind of adverse event.29 For patients, this aspect is crucial, as they may experience a kind of dependency from the stimulation. Notably, we did not observe other indications that DBS induces actual physical dependence, such as the development of tolerance or the need for a progressive increase in current. Yet, the acute deterioration when stopping the stimulation is an important aspect in discussing the reversibility as a much-touted advantage of DBS,30 especially as these phenomena did also occur in non-responders. Moreover, the apparent need for continuous stimulation is crucial for another aspect of this treatment; amplitudes were rather high compared with, for example, DBS for PD. In our sample, the mean battery life was 14.76 months (±5.42), leading to the need for implantation of a new IPG. However, this problem can be solved by using rechargeable devices, even though these do not last for a lifetime either.

The present study has some limitations, one of which is the small sample size, though this is one of the largest samples published in the field of DBS for OCD thus far. Only a minority of patients with OCD are candidates for DBS, making it difficult to set up studies with larger sample sizes, which would help increase the external validity of findings; thus, DBS research often has limited generalisability. In the long term, meta-analyses of methodologically sound original studies may eventually overcome this obstacle. Another limitation refers to the open design of the study and the missing control condition. In this evolving field, blinded cross-over studies are of great importance in order to assess the efficacy of DBS in OCD. However, with regard to the acute psychological deterioration in the state of ‘stimulation off’, questions regarding the reasonableness towards the patient and the risk of a partial unblinding were raised. Finally, given the relatively slow progress, a 1-year postoperative observation period may be too short to reveal the full outcome. Based on our clinical experience and studies with a longer follow-up, OCD symptoms may continue to improve over a longer period of time.31 To further examine this aspect, long-term observation is required.

In conclusion, a major strength of the present investigation is its critical examination of variations in outcome and undesirable effects rather than focusing on effectiveness only. Why ALIC-NAcc-DBS appears to have exceptional results in some patients but unsatisfying effects in others is unclear. We conclude that ALIC-NAcc-DBS is a promising last-resort treatment option for OCD that should be applied carefully given the risks of a surgical procedure and undesirable effects, in particular the rebound phenomena. Although DBS is approved for OCD under Humanitarian Device Exemption and certified for Europe, this treatment must still be considered to be in a stage of development. Before being offered as a standard treatment option, DBS should be applied further and carefully examined in reviews based on more original data collected in registries and with long follow-ups. To determine potential differences in the suitability of patients with different OCD subtypes or comorbidities, continued research in larger patient samples is needed, as well as fundamental research resolving the mechanisms of stimulation in specific targets.

Acknowledgments

We thank Andreas Gierich from the department of stereotactic and functional neurosurgery, University of Cologne, for his support in the preparation of figure 1. We also thank the German Federal Ministry of Education and Research (ELSA-DBS grant) and the Deutsche Forschungsgemeinschaft (KFO-219 grant; KU2665/1-2) for their support.

References

Footnotes

  • DH, SK, VV-V and JK contributed equally.

  • Contributors JK and DH designed the study. DH, SK, JCB and JK did the clinical examinations. VS and VV-V performed the implantations of the electrodes. SK and JCB analysed the data. JCB designed the figures. DH and SK contributed equally to this paper and wrote the manuscript in consultation with JK, JCB and VV-V. JK and LT supervised the project. All authors provided critical feedback and helped shape the research, analysis and manuscript.

  • Disclaimer The funding sources supported the study financially and had no influence on study design, collection, analysis and interpretation of data, writing of report and in the decision to submit the paper for publication.

  • Competing interests The authors report no biomedical financial interests or potential conflicts of interest. LT received payments as a consultant for Medtronic Inc, Boston Scientific, SAPIENS, St. Jude Medical, GE Medical, Bayer Healthcare, UCB Schwarz Pharma, Archimedes Pharma. LT received honoraria as a speaker on symposia sponsored by Zambon Pharma, TEVA Pharma, Lundbeck Pharma, Bracco, Gianni PR, Medas Pharma, UCB Schwarz Pharma, Desitin Pharma, Boehringer Ingelheim, GlaxoSmithKline, Eumecom, Orion Pharma, Medtronic, Boston Scientific, Cephalon, Abott, GE Medical, Archimedes, Bayer, ProsStrakan Pharma. The institution of LT, not LT personally, received funding by the German Research Foundation, the German Ministry of Education and Research, Manfred und Ursula Müller Stiftung, Klüh Stiftung, Hoffnungsbaum e.V., NBIA Disorders Society USA, Köln Fortune, Medtronic, Deutsche Parkinson Vereinigung, Archimedes Pharma, Abott, Bayer, UCB, Zur Rose Pharma, TEVA. Neither LT nor any member of his family holds stocks, stock options, patents or financial interests in any of the abovementioned companies or their competitors. VV-V has received payments for travelling, lodging and financial compensation for contributions to advisory boards or workshops (mostly 2/year) by Medtronic, Abbott and St. Jude Medical. JK has received financial support for Investigator initiated trials from Medtronic GmbH and grants from the German Research Foundation and the Marga and Walter Boll Foundation.

  • Patient consent for publication Obtained.

  • Ethics approval The study was approved by the Ethics Committee of the University of Cologne and registered in the German Clinical Trials Register (www.drks.de, DRKS00005316).

  • Provenance and peer review Not commissioned; externally peer reviewed.