Locating the ventral intermediate thalamic nucleus for deep brain stimulation surgery: analysis of a case series comparing CT and MR targeting

Abstract

Background

Deep brain stimulation (DBS) surgery targeting the ventral intermediate thalamic nucleus (Vim) has proven efficacy in the treatment of tremor.

Aims

The primary aim is to investigate whether there is a statistically significant difference in patient outcomes when CT-guided targeting of the Vim is compared with MRI-guided targeting.

Methods

This is a retrospective study concerning patients undergoing Vim-targeted DBS at the Department of Neurosurgery, Royal Victoria Infirmary in Newcastle (9^th August 2012 to 4^th January 2019). Fahn–Tolosa–Marin Tremor Scale (FTM TS) and EQ-5D scores were collected from patient notes. Statistical analysis was performed using IBM^® SPSS^® Statistics Version 24. Independent samples t-tests were used to compare means.

Results

Independent samples t-test did not reveal a statistically significant difference between CT (n = 10; FTM TS mean = 65.40, SD = 11.40; EQ-5D mean = 39.50, SD = 17.87) and MR (n = 7; FTM TS mean = 60.57, SD = 7.50; EQ-5D mean = 32.14, SD = 9.94) groups in pre-surgery FTM TS (t(15) = 0.977, p = 0.344) and EQ-5D (t(15) = 0.982, p = 0.342) scores. No statistically significant difference between the CT (FTM TS mean = 24.12, SD = 20.47; EQ-5D mean = 75.56, SD = 15.63) and MR (FTM TS mean = 22.86, SD = 6.72; EQ-5D mean = 70.43, SD = 15.48) groups was revealed at 1 year assessment of FTM TS (t(14) = 0.155, p = 0.879) and EQ-5D (t(14) = 0.654, p = 0.524). The median difference between pre- and post-surgery FTM TS and EQ-5D scores in the CT group at 1 year was 43.00 and 35.00, respectively. The MR patient group median difference in pre- and post-surgery at 1 year was 35.00 and 35.00 respectively.

Conclusion

No statistically significant difference between CT and MR image-guided targeting patient groups was detected.

Keywords:

• Deep brain stimulation
• essential tremor
• ventral intermediate nucleus
• targeting
• imaging

Introduction

Deep brain stimulation (DBS) surgery targeting the ventral intermediate thalamic nucleus (Vim) has proven efficacy in the treatment of tremor.^1,2 ET has been described as the most prevalent adult movement disorder.³ According to a population study with a clear definition of ET, prevalence ranges between 0.4–3.9%.³ The Vim was among the first nuclei to be targeted for DBS,⁴ and it comprises part of the motor thalamic nuclei group acting as a relay between the basal ganglia, cerebellum and motor cortex.⁵ It is not possible to directly visualise the Vim using imaging modalities in current standard practice. The gold standard of Vim targeting is therefore an indirect method, comprising MR, CT and/or ventriculography to identify anatomical landmarks such as the mid-commissural point (MCP) of the anterior-posterior commissures (AC-PC) from which to extrapolate.⁶ This is used in conjunction with stereotactic atlases^7,8 or the quadrilateral of Guiot⁹ to plot co-ordinates for the target. MRI-guided target planning is by far the most commonly used method currently employed.^10–14

Aims

The primary aim is to investigate whether there is a statistically significant difference in patient outcomes when comparing CT-guided targeting of the Vim with MRI-guided targeting. The secondary aim is to use these results in conjunction with evidence in the literature to ascertain the possibility of reducing the financial and time expenditure associated with the use of MRI sequences where CT is adequate.

Materials and methods

This is a retrospective study concerning patients treated with DBS targeting the Vim at the Department of Neurosurgery within the Royal Victoria Infirmary in Newcastle. Data captured concerned surgeries taking place between 9^th August 2012 and 4^th January 2019. Fahn-Tolosa-Marin Tremor Scale (FTM TS) scores and EQ-5D quality of life scores were collected from patient notes. Patients’ pre-surgery scores were recorded, as well as post-surgery scores at 6 months and 1 year. Statistical analysis of collated data was performed using IBM® SPSS® Statistics Version 24. Independent samples t tests were performed to compare cohort means, as well as the mean difference in pre- and post-surgery FTM TS and EQ-5D scores. The FTM TS ¹⁵ (Appendix 1) is validated and widely used outcome measure, making this an ideal tool for use in this study and to enable generalisability among the wider literature. The EQ-5D quality of life healthcare evaluation has been previously validated¹⁶ (Appendix 2). The scale assesses outcomes across five domains including self-care, mobility, activities of daily living, pain/discomfort, and anxiety/depression. The implantation of the electrodes was carried out using a Leksell frame. The frame is attached to the patient once they are anaesthetised and subsequently taken to scan. One group of patients would have their peri-operative scan in CT and another group in MRI. The subsequent planning and surgical procedures were equivalent between the two groups. The target was defined using the Tailarach atlas, using the anterior commissure (AC) and the posterior commissure (PC) as our landmarks (Figure 1), we establish the imaging slice that incorporated both. A measurement was made 1/3 of the way from the PC along the AC-PC line and the target was marked 13mm lateral to this point. This was sometimes altered depending upon the size of the patients’ third ventricle. This process was equivocal for both CT and MRI-guided procedures. All measurements were made using the Siemens scanning software. Importantly, the post implant programming and clinical review was also done independently of scanning modality used during the insertion of the electrodes. The surgeon for the MRI series was the same for each case, and the surgeon for the CT series was the same for each case.

Figure 1. A. example of CT axial slice to demonstrate AC-PC landmarks (red asterisk denotes AC, green asterisk denotes PC); B. example of MR axial slice to demonstrate AC-PC landmarks (red asterisk denotes AC, green asterisk denotes PC).

Figure 2. Comparison of FTM TS CT and MRI group scores. Within each colour pair (green, blue and purple) left-hand side box and whiskers represent CT patient group, right-hand side represents MR patient group. First pair (green) denote pre-operative scores, middle pair (blue) denote 6-month post-operative scores), and final pair (purple) denote 1-year post-operative scores. Abbreviations used: FTM TS: Fahn–Tolosa–Marin tremor scale; CT: computerised tomography; MRI: magnetic resonance imaging.

Figure 3. Comparison of EQ-5D CT and MRI group scores. Within each colour pair (green, blue and purple), left hand side box and whiskers represent CT patient group, right hand side box and whiskers represents MR patient group. First pair (green) denote pre-operative scores, middle pair (blue) denote 6-month post-operative scores), and final pair (purple) denote 1-year post-operative scores. Abbreviations used: CT, computerised tomography; MRI, magnetic resonance imaging.

Technical specifics

The CT scans were performed at 120 kV using a Siemens Edge machine. A J40s medium algorithm was used with Base Orbita window to generate 1.0mm sliced images. The MRI scans were performed using a Siemens 3T Magnetom Skyra machine. 2.0mm axial slices were generated using TR 6950.0ms and TE 97ms settings and coil elements HE1-4. 2.0mm coronal slices were generated using TR 6000.0 and TE 97ms also involving coil elements HE1-4. Voxel size 0.7 × 0.7 × 2.0mm, SNR 1.00. No specific planning software was used but a template within Microsoft Word was used to type in measurements between fiducials to ensure accuracy between all four measured quadrants.

Ethical statement

This study was registered with our institution as a retrospective service evaluation quality assurance study. A study-specific consent form was not required, although all patients undergoing procedures had completed a valid consent form which included permission to collect data for service and quality assurance projects.

Results

The patient population was first divided into two groups; CT- (n = 10, 4 = female) and MRI-guided (n = 7, 3 = female) targeting of Vim for the treatment of ET using DBS. Patient demographics are described in Table 1. Median pre-surgery FTM TS and EQ-5D scores in the CT group were 63 and 35 respectively, compared with 66 and 35 in the MR group (Figures 2 and 3). Independent samples t-test for FTM TS (t(13) = 0.860, p = 0.405) and EQ-5D (t(10.138) = 1.491, p = 0.166) scores during pre-surgery assessment did not display a statistically significant difference between CT and MR groups. Similarly, no statistically significant difference between the groups was revealed at 6 month assessment of FTM TS (t(13) = 0.062, p = 0.951) and EQ-5D (t(13) = 0.255, p = 0.803) scores. Finally, FTM TS (t(9) = 0.992, p = 0.347) and EQ-5D (t(9) = −0.83, p = 0.935) scores were recorded at 1 year post-surgery, still revealing no statistically significant difference between groups. The median difference between pre- and post-surgery FTM TS and EQ-5D scores in the CT group at 1 year was 40.5 and 35, respectively. The MR patient group median difference in pre- and post-surgery at 1 year was 35 and 35, respectively. In the CT group, 1 case involved standard lead insertion (1/10, 10%). In 4 cases involving multidirectional lead insertion, ring mode was activated (4/9, 44.5%). Directionality was activated in 5 out of 9 cases where multi-directional leads were inserted (5/9, 55.6%). In 4 cases, directionality was activated for side effects (4/9, 44.5%; 2 = dysarthria (2/9, 20%, 2 = ataxia (2/10, 22.3%). Directionality was activated for symptom control in 1 case (1/9, 11.2%). In the MR group, 5 cases involved standard lead insertion (5/7, 71.4%). Multidirectional leads were inserted in 2 out of 7 cases (2/7, 28.6%), of which directionality was activated in 1 case (1/2, 50.0%; 1 = dysarthria).

Discussion

We describe the comparison of indirect targeting methods (CT and MRI) for localisation of the Vim and placement of DBS electrodes for the treatment of ET. No statistically significant difference in FTM TS outcomes between CT and MR groups was detected at 6 months (p = 0.951) or 1 year (p = 0.347). Equally, no statistically significant difference between CT and MR groups in EQ-5D scores was detected at 6 months (p = 0.803) or 1 year (p = 0.935). This highlights the resulting implication that either method is efficacious based upon the experience within our centre. This is a retrospective study conducted in a single centre. However, the surgical methods are uniform across the implanting surgeons and the programming and evaluation Is done by an independent nursing team. The results have a single variable which would be hard to recreate in larger multi-centre trials.

Use of CT alone as a departure from ventriculography assisted targeting soon became more popular^17–19 and studies demonstrated comparable accuracy between these methods.²⁰ Direct targeting is based upon visualisation of anatomical structures lying adjacent to the thalamic motor nuclei, such as the third ventricular wall or internal capsule using MR imaging modalities.²¹ Most influential in the generation of MRI distortion is the level of homogeneity in the magnetic field and associated non-linear gradients.²² Benabid et al. made early comments on spatial irregularities arising from MR imaging in the form of systematic distortion.²³ Further, it has previously been suggested that, due to distortion of structures upon MR imaging and sub-optimal appearance of the AC-PC line using CT, that in fusing both modalities together a more accurate representation can be gleaned for targeting purposes.²⁴ Studies comparing MR with CT targeting for accuracy report differences in x and y coordinates reported as 0.00-1.00mm and 1.70-3.75mm respectively.^25–27 A study examined interpersonal errors in the identification of coordinates using MR (1.5T and 3.0T) and CT modalities. Two neurosurgeons identified coordinates. The mean errors were 0.48 ± 0.22mm, 0.75 ± 0.38mm, and 2.18 ± 0.76mm in CT, 1.5T T2, and 3.0T T2 respectively.²⁸ This evidence further supports our conclusion that the efficacy of CT and MR for targeting is comparable.

Conclusion

No statistically significant difference between CT and MR image-guided targeting patient groups was detected. This supports the hypothesis that CT targeting demonstrates similar accuracy to that achieved using MR imaging. This raises the possibility that patients undergoing DBS for ET using a Vim target could endure shorter imaging times, as well as of reducing the burden on resources in neurosurgical centres.

Limitations

This study is limited by the small sample size, although these results warrant further investigation for the purposes of improving services. Reproducibility may be difficult in centres where CT-guided Vim targeting is not in routine use and may introduce proficiency bias. A further limitation is that the groups were not contemporary. Despite the use of directionality in some of the CT group cases, however, this was used for side effect control in all cases except one where it was employed to improve symptom control. These findings suggest that, although MR imaging is clearer for landmark visualisation, CT has a role in resource sensitive settings where it may be employed safely as an alternative to MR-guided techniques.

Recommendations

Further studies are recommended to expand the dataset concerning this patient population for further analysis to support consideration of CT-guided Vim as standard. Future studies may also incorporate the dentate-rubro-thalamic tract target for the treatment of ET.²⁹

Disclosure statement

No potential conflict of interest was reported by the author(s).

Table 1. A. CT patient group demographics; B. MR patient group demographics.

Patient number	Age	DBS implant date	Target	DBS system	Comorbidities	Medication
A.
CT
1	74	04-01-2019	Bilateral Vim	Boston scientific re-chargeable directional leads (left vim, mono-polar ring mode, right VIM bipolar ring mode)	Hypertension Depression Iron deficiency	Atenolol Fluoxetine
2	50	26-01-2016	Bilateral Vim	Boston scientific re-chargeable standard leads	Hypertension Depression Iron deficiency	Atenolol Fluoxetine
3	55	25-08-2017	Bilateral Vim	Boston Scientific re-chargeable multi directional leads	Nil	Nil
4	86	20-08-2016	Bilateral Vim	Boston Scientific re-chargeable multi directional leads	TIA CVA Hypertension	Bisoprolol Aspirin Nitrofurantoin Ticagrelor Candasartan
5	70	29-07-2017	Bilateral Vim	Boston Scientific Primary cell multi directional leads	Hypothyrodism	Thyroxine
6	78	01-06-2018	Bilateral Vim	Boston Scientific Primary cell multi directional leads	COPD	Lansoprazole
7	85	20-08-2018	Bilateral Vim	Boston Scientific Primary cell multi directional leads	TURP OA	Nil
8	73	05-08-2016	Bilateral Vim	Boston Scientific Primary cell multi directional leads	BPH COPD Hypertension Back pain	Spiriva Uniphyllin Carbocysteine Lansoprazole Fluticasone Alendronic acid Aspirin Atorvostatin Mometasone Ventolin Furosemide Felodipine GTN Ivabradine
9	66	21-03-2017	Bilateral Vim	Boston Scientific Primary cell multi directional leads	Osteoarthritis of hip, neck and back Rosacea OSA on NIV Elevated BMI Chronic pain Cholecsystecomy	Amitriptyline Clonazepam Furosemide Dihydromorphine Co-codamol
10	72	19-08-2017	Bilateral Vim	Boston Scientific Primary cell multi directional leads	Asthma	Atorvastatin Ibuprofen Salbutamol
Patient number	Age	DBS implant date	Target	DBS system	Comorbidities	Medication
B.
MRI
1	66	13-07-2016	Bilateral Vim	Boston Scientific Primary cell multi directional leads	Hypertension Depression Hypothyroidism	Sertaline Calcipotriol Hydrocortisone Atorvostatin Mirtazepine Lisinopril Levothyroxine
2	63	14-10-2015	Bilateral Vim	Boston Scientific re-chargeable standard leads	CVA Asthma Atrial fibrillation Depression Type II diabetes	Tramadol Duloxetine Pregabalin Furosemide Symbicort Clenil Aerochamber plus flow inhaler Mebeverine Metformin Docusate
3	76	06-01-2015	Bilateral Vim	Medtronic Primary cell standard leads	Type II diabetes Hypertension	Metformin Atorvostatin Glimepiride
4	70	26-07-2016	Bilateral Vim	Boston Scientific Primary cell multi directional leads	Crohn’s disease Hypertension	Amlodipine
5	79	29-09-2015	Bilateral Vim	Boston scientific re-chargeable standard leads	Atrial fibrillation Type II diabetes COPD Hypertension	Omeprazole Salbutamol Lercanidipine Dabigatran Atenolol Insulin Glargine Levothyroxine Metformin Warfarin
6	86	09-08-2012	Bilateral Vim	Medtronic Primary cell standard leads	CKD Type II diabetes	Bisoprolol Ticagrelor Atorvostatin Aspirin Lisinopril Latanoprost
7	55	22-01-2015	Bilateral Vim	Boston Scientific RC standard leads	Depression	Fluoxetine Sertraline Propranolol

Abbreviations used: DBS: deep brain stimulation; Vim: ventral intermediate nucleus (of thalamus); TIA: transient ischaemic attack; CVA: cerebrovascular accident; COPD: chronic obstructive pulmonary disease; TURP: transurethral resection of the prostate; OA: osteoarthritis; BPH: benign prostatic hyperplasia; OSA: obstructive sleep apnoea; NIV: non-invasive ventilation; BMI: body mass index; CKD: chronic kidney disease.

Additional information

Funding

No funding was secured for the delivery of this project.