British Journal of Neurosurgery, 2014; Early Online: 1–3 © 2014 The Neurosurgical Foundation ISSN: 0268-8697 print / ISSN 1360-046X online DOI: 10.3109/02688697.2014.971708
SHORT REPORT
Long-term motor cortex stimulation for phantom limb pain
Br J Neurosurg Downloaded from informahealthcare.com by SUNY State University of New York at Stony Brook on 02/02/15 For personal use only.
Erlick A.C. Pereira, Tom Moore, Liz Moir & Tipu Z. Aziz Nuffield Department of Surgical Sciences and Department of Neurological Surgery, Oxford Functional Neurosurgery, Oxford University Hospitals, Oxford, UK
Abstract We present the long-term course of motor cortex stimulation to relieve a case of severe burning phantom arm pain after brachial plexus injury and amputation. During 16-year follow-up the device continued to provide efficacious analgesia. However, several adjustments of stimulation parameters were required, as were multiple pulse generator changes, antibiotics for infection and one electrode revision due to lead migration. Steady increases in stimulation parameters over time were required. One of the longest follow-ups of motor cortex stimulation is described; the case illustrates challenges and pitfalls in neuromodulation for chronic pain, demonstrating strategies for maintaining analgesia and overcoming tolerance.
secured in position by suturing to the dura prior to dural closure, bone flap replacement and skin closure. A pain diary with twice-daily 10-point visual analogue scores (VAS) was recorded while the stimulator was trialled either on or off randomly each half day with the patient blinded to its settings. Significant reductions in pain were reported during eight out of ten on trials versus none off, and implantable pulse generator (IPG) insertion was therefore decided upon. In a second operation performed by awake craniotomy under local anaesthesia, electrode position was moved slightly laterally to obtain better phantom arm analgesic coverage (Fig. 1). General anaesthesia was then induced and extension leads tunnelled to a left subclavicular IPG (Medtronic Itrel 2). Initial stimulator settings were amplitude of 3.0 V, pulse width of 450 ms, frequency of 14 Hz, and cycling for 15 min on and 4 h off. Analgesia covered the whole phantom arm except for thumb and little finger with VAS reported in pain diary reduced by 78% from 9 to 2 at 6-month follow-up. At the last clinic follow-up 16 years after MCS insertion, VAS was reported to be reduced by 75% from 8 with stimulation off to 2 with stimulation on, with concomitant improvements also seen in McGill Pain Questionnaire and activities of daily living compared with those before surgery. The patient’s neurosurgical course after MCS insertion included several IPG changes and electrode revision surgeries as summarised in Table II. Of note, the initial IPG required replacement with a more advanced magnetically shielded model (Medtronic Itrel 3) due to episodes of interference with magnetic fields including airport scanners. Due to loss of efficacy a year after initial surgery, it required electrode replacement which was complicated by superficial wound infection managed successfully by antibiotics. Side-effects of vocal arrest and seizure with high voltage stimulation two years after initial implantation were observed. Due to loss of efficacy a few months later, it required an electrode revision which was complicated by high impedance requiring electrode replacement in the same surgery. Lack of efficacy despite optimal electrode placement necessitated extension lead and IPG replacement later that year. This was followed by eight years of successful stimulation before IPG replacement
Keywords: long-term outcome; motor cortex stimulation; neuromodulation; phantom limb pain
Introduction A 38-year-old woman presented in 1979 for an excisional biopsy of a left brachial plexus neuroma of the upper trunk. The operation resulted in severe C5 and C6 dermatomal pain, refractory to both conventional analgesics and neuropathic pain medications including carbamazepine and gabapentin. Over the next four years she underwent several unsuccessful operations attempting to reduce the pain (Table I). In 1986, she opted for arm amputation at the shoulder. Immediately after surgery, she experienced severe phantom limb pain in the entire amputated arm which she described as burning in character. Over a decade later, in 1997 the patient was offered motor cortex stimulation (MCS) at 56 years of age. Electrode insertion (Medtronic Resume) was performed under local anaesthesia after right parietal craniotomy under general anaesthesia with lead externalisation for trialling of stimulation. The extradural electrode position was optimised by titration of the patient’s shoulder motor response and perceived phantom arm motor response to intra-operative stimulation. A good ‘motor’ response was reported in the left phantom hand, site of most severe pain, and the electrode
Correspondence: Mr. Erlick Pereira, Nuffield Department of Surgery and Department of Neurological Surgery, The West Wing, Oxford University Hospitals, Oxford, OX3 9DU, UK. Tel: 44 1865 741166. Fax: 44 1865 231885. E-mail [email protected] Received for publication 10 November 2013; accepted 28 September 2014
1
2 E. A.C. Pereira et al.
Br J Neurosurg Downloaded from informahealthcare.com by SUNY State University of New York at Stony Brook on 02/02/15 For personal use only.
Table I. Timeline of surgical procedures. 1980 Sympathetic blockade 1981 Right thalamotomy 1981 Cervical dorsal rhizotomy 1982 Revision right thalamotomy 1983 Left thalamotomy 1986 Left arm amputation 1997 Right motor cortex stimulator implantation
in 2007 (Medtronic Synergy) with the patient requiring IPG changes every three years (Medtronic Kinetra in 2010 and 2013) with relatively high-energy bipolar stimulation being delivered (6.4 V, 450 ms, 80 Hz).
Discussion Phantom limb pain is thought to affect two-thirds of amputees, occurring more commonly with proximal amputations preceded by chronic pain.1 Relieving the pain is challenging; initial treatment options include various psychological and physical therapies such as mirror box use and pharmacological therapies, with drugs of various classes including antidepressants, opioids, anticonvulsants and g-aminobutyric acid receptor agonists showing some efficacy. Non-responders are increasingly considered for neuromodulation procedures. Common targets for neuromodulation include the dorsal spinal cord, sensory thalamus, periventricular grey and motor cortex.2 Dorsal root entry zone lesions are also used for deafferentation pain with good effect, albeit with small risks of weakness. Deep brain stimulation targeting the contralateral ventroposterolateral thalamus or periaqueductal grey matter has been utilised to relieve deafferentation pain for over half a century. We favour sensory thalamic stimulation for phantom limb pain after amputation or brachial plexus avulsion.3 Prior thalamotomy elsewhere, however, precluded its use in the patient described here and thus MCS was preferred. This
case illustrates that MCS can provide long-lasting analgesia and moreover that typical complications of lead migration, infection, seizures, vocal arrest and tolerance can be overcome by judicious neurosurgery and programming. MCS is a recent surgical treatment first described a couple of decades ago. A recent review of 10 case series comprising 157 patients treated with MCS determined the most common adverse effects to be seizures (occurring in 12% of patients), infections (5.7%) and hardware problems (5.1%).4 Follow-up was limited to a decade with a mean of 2.5 years in the series reviewed in contrast to the 16-year follow-up described here. Importantly, there were no published reports suggesting that MCS induces chronic epilepsy. Overall, our results are consistent with a literature reporting MCS to be a relatively safe procedure. Rationales for targeting and mechanism of action in MCS have been recently reviewed.5 Electrode positioning aims for anatomical coverage of the painful somatotopic representation. While the electrode placement refined by awake patient responses appears quite lateral to facial motor cortex in Fig. 1, subsequent functional magnetic resonance imaging (fMRI) evidence during facial and imagined arm movements suggests that the area corresponding to an amputated arm is generally smaller and laterally shifted into the facial area compared with its normal cortical representation.6,7 MRI also implicates aberrant sensorimotor connectivity in phantom limb pain.8 Repetitive transcranial magnetic stimulation over the contiguous rather than precise motor area corresponding to a phantom limb has also conferred temporary analgesia,9 a finding consistent with our eventual MCS placement in facial motor cortex contiguous and adjacent to the patient’s arm cortical area. Optimal stimulation parameters for MCS described in the literature vary, and ours were initially consistent with Nguyen who suggested optimal ranges of amplitude, 1–6 V;
Fig. 1. Coronal and sagittal skull radiographs showing right parietal craniotomy and motor cortex stimulator in situ.
Motor cortex stimulation for amputation pain 3 Table II. Timeline of motor cortex stimulation-related interventions. Date Intervention 1997 1998 1999
Br J Neurosurg Downloaded from informahealthcare.com by SUNY State University of New York at Stony Brook on 02/02/15 For personal use only.
1999
1999
2007 2010 2011 2013
Medtronic Itrel 2 IPG replaced with magnetically shielded Itrel 3 following episodes of interference around magnetic fields including in airports. Wound infection following reopening of right parietal craniotomy and new electrode implantation. Treated with intravenous antibiotics. Patient complained of increasing pain. Initial re-titration of stimulator by altering pulse width, rate and amplitude was unable to elicit motor response. Later adjustment induced vocal arrest and seizure (at 6.5 V), which stopped when stimulator was turned off. Revision of stimulator following six months of burning pain in phantom arm, forearm and hand. At operation impedance was very high, suggesting the electrode had slipped, so it was replaced. No significant improvement in pain relief. Pain still not controlled. Itrel 3 IPG replaced, leads externalised and demonstrated to be in correct position. Cable between IPG and electrodes replaced. Achieved 70% pain relief compared to no stimulation. Itrel 3 IPG replaced with Synergy IPG. Settings: 0 to 1, 4.0 V, 450 ms, 75 Hz, cycling 1 s on, 1 s off Synergy IPG replaced with Kinetra IPG Settings changed: 0 to 1, 4.5 V, 450 ms, 80 Hz IPG replaced. Settings 0 to 1, 6.4 V, 450 ms, 80 Hz
pulse width, 40–450 ms and frequency, 20–65 Voltage and frequency have, however, both increased beyond those recommendations at the last follow-up and it may be that increased energy delivery or frequency changes are desirable to overcome potential tolerance during such long-term use. As with other forms of neuromodulation, MCS efficacy is known to attenuate with time in longitudinal cohort studies. In one meta-analysis 57% of 152 patients responded well to MCS initially; only 79% of that subgroup reported significant analgesia at 1–10 years after surgery.4 The patient described experienced tolerance on several occasions over the 16-year follow-up period, some episodes attributable to loss of IPG charge and one due to lead displacement. It is noteworthy that she was unsuitable for a rechargeable IPG since she had only one functioning arm as two hands would be required to recharge it. Multiple programming episodes were undertaken with a gradual trend towards increasing all stimulation parameters and hence energy delivery to deliver efficacious analgesia. This phenomenon is not established in MCS Hz.5
literature, and indeed refuted in some recent reviews.5 However, duration of MCS beyond a decade is sparsely reported. Further characterisation of its efficacy and stimulation parameters at long-term follow-up, alongside investigations to further delineate the mechanisms of action of MCS and an understanding of putative tolerance phenomena over time, are worthy endeavours for future neuromodulation research.
Acknowledgments The authors alone managed the patient clinically and wrote the manuscript. They have no conflicts of interest. Declaration of interest: The authors report no declarations of interest. The authors alone are responsible for the content and writing of the paper.
References 1. Hall GC, Carroll D, Parry D, McQuay HJ. Epidemiology and treatment of neuropathic pain: the UK primary care perspective. Pain 2006;122:156–62. 2. Cruccu G, Aziz TZ, Garcia-Larrea L, et al. EFNS guidelines on neurostimulation therapy for neuropathic pain. Eur J Neurol 2007;14:952–70. 3. Pereira EA, Boccard SG, Linhares P, et al. Thalamic deep brain stimulation relieves neuropathic pain after amputation or brachial plexus avulsion. Neurosurg Focus 2013; 35:1–11. 4. Fontaine D, Hamani C, Lozano A. Efficacy and safety of motor cortex stimulation for chronic neuropathic pain: critical review of the literature. J Neurosurg 2009;110:251–6. 5. Nguyen JP, Nizard J, Keravel Y, Lefaucheur JP. Invasive brain stimulation for the treatment of neuropathic pain. Nat Rev 2011;7:699–709. 6. Lotze M, Flor H, Grodd W, Larbig W, Birbaumer N. Phantom movements and pain. An fMRI study in upper limb amputees. Brain 2001;124:2268–77. 7. MacIver K, Lloyd DM, Kelly S, Roberts N, Nurmikko T. Phantom limb pain, cortical reorganization and the therapeutic effect of mental imagery. Brain 2008;131:2181–91. 8. Makin TR, Scholz J, Filippini N, et al. Phantom pain is associated with preserved structure and function in the former hand area. Nat Commun 2013;4:1570. 9. Lefaucheur JP, Hatem S, Nineb A, et al. Somatotopic organization of the analgesic effects of motor cortex rTMS in neuropathic pain. Neurology 2006;67:1998–2004.
SHORT REPORT
Long-term motor cortex stimulation for phantom limb pain
Br J Neurosurg Downloaded from informahealthcare.com by SUNY State University of New York at Stony Brook on 02/02/15 For personal use only.
Erlick A.C. Pereira, Tom Moore, Liz Moir & Tipu Z. Aziz Nuffield Department of Surgical Sciences and Department of Neurological Surgery, Oxford Functional Neurosurgery, Oxford University Hospitals, Oxford, UK
Abstract We present the long-term course of motor cortex stimulation to relieve a case of severe burning phantom arm pain after brachial plexus injury and amputation. During 16-year follow-up the device continued to provide efficacious analgesia. However, several adjustments of stimulation parameters were required, as were multiple pulse generator changes, antibiotics for infection and one electrode revision due to lead migration. Steady increases in stimulation parameters over time were required. One of the longest follow-ups of motor cortex stimulation is described; the case illustrates challenges and pitfalls in neuromodulation for chronic pain, demonstrating strategies for maintaining analgesia and overcoming tolerance.
secured in position by suturing to the dura prior to dural closure, bone flap replacement and skin closure. A pain diary with twice-daily 10-point visual analogue scores (VAS) was recorded while the stimulator was trialled either on or off randomly each half day with the patient blinded to its settings. Significant reductions in pain were reported during eight out of ten on trials versus none off, and implantable pulse generator (IPG) insertion was therefore decided upon. In a second operation performed by awake craniotomy under local anaesthesia, electrode position was moved slightly laterally to obtain better phantom arm analgesic coverage (Fig. 1). General anaesthesia was then induced and extension leads tunnelled to a left subclavicular IPG (Medtronic Itrel 2). Initial stimulator settings were amplitude of 3.0 V, pulse width of 450 ms, frequency of 14 Hz, and cycling for 15 min on and 4 h off. Analgesia covered the whole phantom arm except for thumb and little finger with VAS reported in pain diary reduced by 78% from 9 to 2 at 6-month follow-up. At the last clinic follow-up 16 years after MCS insertion, VAS was reported to be reduced by 75% from 8 with stimulation off to 2 with stimulation on, with concomitant improvements also seen in McGill Pain Questionnaire and activities of daily living compared with those before surgery. The patient’s neurosurgical course after MCS insertion included several IPG changes and electrode revision surgeries as summarised in Table II. Of note, the initial IPG required replacement with a more advanced magnetically shielded model (Medtronic Itrel 3) due to episodes of interference with magnetic fields including airport scanners. Due to loss of efficacy a year after initial surgery, it required electrode replacement which was complicated by superficial wound infection managed successfully by antibiotics. Side-effects of vocal arrest and seizure with high voltage stimulation two years after initial implantation were observed. Due to loss of efficacy a few months later, it required an electrode revision which was complicated by high impedance requiring electrode replacement in the same surgery. Lack of efficacy despite optimal electrode placement necessitated extension lead and IPG replacement later that year. This was followed by eight years of successful stimulation before IPG replacement
Keywords: long-term outcome; motor cortex stimulation; neuromodulation; phantom limb pain
Introduction A 38-year-old woman presented in 1979 for an excisional biopsy of a left brachial plexus neuroma of the upper trunk. The operation resulted in severe C5 and C6 dermatomal pain, refractory to both conventional analgesics and neuropathic pain medications including carbamazepine and gabapentin. Over the next four years she underwent several unsuccessful operations attempting to reduce the pain (Table I). In 1986, she opted for arm amputation at the shoulder. Immediately after surgery, she experienced severe phantom limb pain in the entire amputated arm which she described as burning in character. Over a decade later, in 1997 the patient was offered motor cortex stimulation (MCS) at 56 years of age. Electrode insertion (Medtronic Resume) was performed under local anaesthesia after right parietal craniotomy under general anaesthesia with lead externalisation for trialling of stimulation. The extradural electrode position was optimised by titration of the patient’s shoulder motor response and perceived phantom arm motor response to intra-operative stimulation. A good ‘motor’ response was reported in the left phantom hand, site of most severe pain, and the electrode
Correspondence: Mr. Erlick Pereira, Nuffield Department of Surgery and Department of Neurological Surgery, The West Wing, Oxford University Hospitals, Oxford, OX3 9DU, UK. Tel: 44 1865 741166. Fax: 44 1865 231885. E-mail [email protected] Received for publication 10 November 2013; accepted 28 September 2014
1
2 E. A.C. Pereira et al.
Br J Neurosurg Downloaded from informahealthcare.com by SUNY State University of New York at Stony Brook on 02/02/15 For personal use only.
Table I. Timeline of surgical procedures. 1980 Sympathetic blockade 1981 Right thalamotomy 1981 Cervical dorsal rhizotomy 1982 Revision right thalamotomy 1983 Left thalamotomy 1986 Left arm amputation 1997 Right motor cortex stimulator implantation
in 2007 (Medtronic Synergy) with the patient requiring IPG changes every three years (Medtronic Kinetra in 2010 and 2013) with relatively high-energy bipolar stimulation being delivered (6.4 V, 450 ms, 80 Hz).
Discussion Phantom limb pain is thought to affect two-thirds of amputees, occurring more commonly with proximal amputations preceded by chronic pain.1 Relieving the pain is challenging; initial treatment options include various psychological and physical therapies such as mirror box use and pharmacological therapies, with drugs of various classes including antidepressants, opioids, anticonvulsants and g-aminobutyric acid receptor agonists showing some efficacy. Non-responders are increasingly considered for neuromodulation procedures. Common targets for neuromodulation include the dorsal spinal cord, sensory thalamus, periventricular grey and motor cortex.2 Dorsal root entry zone lesions are also used for deafferentation pain with good effect, albeit with small risks of weakness. Deep brain stimulation targeting the contralateral ventroposterolateral thalamus or periaqueductal grey matter has been utilised to relieve deafferentation pain for over half a century. We favour sensory thalamic stimulation for phantom limb pain after amputation or brachial plexus avulsion.3 Prior thalamotomy elsewhere, however, precluded its use in the patient described here and thus MCS was preferred. This
case illustrates that MCS can provide long-lasting analgesia and moreover that typical complications of lead migration, infection, seizures, vocal arrest and tolerance can be overcome by judicious neurosurgery and programming. MCS is a recent surgical treatment first described a couple of decades ago. A recent review of 10 case series comprising 157 patients treated with MCS determined the most common adverse effects to be seizures (occurring in 12% of patients), infections (5.7%) and hardware problems (5.1%).4 Follow-up was limited to a decade with a mean of 2.5 years in the series reviewed in contrast to the 16-year follow-up described here. Importantly, there were no published reports suggesting that MCS induces chronic epilepsy. Overall, our results are consistent with a literature reporting MCS to be a relatively safe procedure. Rationales for targeting and mechanism of action in MCS have been recently reviewed.5 Electrode positioning aims for anatomical coverage of the painful somatotopic representation. While the electrode placement refined by awake patient responses appears quite lateral to facial motor cortex in Fig. 1, subsequent functional magnetic resonance imaging (fMRI) evidence during facial and imagined arm movements suggests that the area corresponding to an amputated arm is generally smaller and laterally shifted into the facial area compared with its normal cortical representation.6,7 MRI also implicates aberrant sensorimotor connectivity in phantom limb pain.8 Repetitive transcranial magnetic stimulation over the contiguous rather than precise motor area corresponding to a phantom limb has also conferred temporary analgesia,9 a finding consistent with our eventual MCS placement in facial motor cortex contiguous and adjacent to the patient’s arm cortical area. Optimal stimulation parameters for MCS described in the literature vary, and ours were initially consistent with Nguyen who suggested optimal ranges of amplitude, 1–6 V;
Fig. 1. Coronal and sagittal skull radiographs showing right parietal craniotomy and motor cortex stimulator in situ.
Motor cortex stimulation for amputation pain 3 Table II. Timeline of motor cortex stimulation-related interventions. Date Intervention 1997 1998 1999
Br J Neurosurg Downloaded from informahealthcare.com by SUNY State University of New York at Stony Brook on 02/02/15 For personal use only.
1999
1999
2007 2010 2011 2013
Medtronic Itrel 2 IPG replaced with magnetically shielded Itrel 3 following episodes of interference around magnetic fields including in airports. Wound infection following reopening of right parietal craniotomy and new electrode implantation. Treated with intravenous antibiotics. Patient complained of increasing pain. Initial re-titration of stimulator by altering pulse width, rate and amplitude was unable to elicit motor response. Later adjustment induced vocal arrest and seizure (at 6.5 V), which stopped when stimulator was turned off. Revision of stimulator following six months of burning pain in phantom arm, forearm and hand. At operation impedance was very high, suggesting the electrode had slipped, so it was replaced. No significant improvement in pain relief. Pain still not controlled. Itrel 3 IPG replaced, leads externalised and demonstrated to be in correct position. Cable between IPG and electrodes replaced. Achieved 70% pain relief compared to no stimulation. Itrel 3 IPG replaced with Synergy IPG. Settings: 0 to 1, 4.0 V, 450 ms, 75 Hz, cycling 1 s on, 1 s off Synergy IPG replaced with Kinetra IPG Settings changed: 0 to 1, 4.5 V, 450 ms, 80 Hz IPG replaced. Settings 0 to 1, 6.4 V, 450 ms, 80 Hz
pulse width, 40–450 ms and frequency, 20–65 Voltage and frequency have, however, both increased beyond those recommendations at the last follow-up and it may be that increased energy delivery or frequency changes are desirable to overcome potential tolerance during such long-term use. As with other forms of neuromodulation, MCS efficacy is known to attenuate with time in longitudinal cohort studies. In one meta-analysis 57% of 152 patients responded well to MCS initially; only 79% of that subgroup reported significant analgesia at 1–10 years after surgery.4 The patient described experienced tolerance on several occasions over the 16-year follow-up period, some episodes attributable to loss of IPG charge and one due to lead displacement. It is noteworthy that she was unsuitable for a rechargeable IPG since she had only one functioning arm as two hands would be required to recharge it. Multiple programming episodes were undertaken with a gradual trend towards increasing all stimulation parameters and hence energy delivery to deliver efficacious analgesia. This phenomenon is not established in MCS Hz.5
literature, and indeed refuted in some recent reviews.5 However, duration of MCS beyond a decade is sparsely reported. Further characterisation of its efficacy and stimulation parameters at long-term follow-up, alongside investigations to further delineate the mechanisms of action of MCS and an understanding of putative tolerance phenomena over time, are worthy endeavours for future neuromodulation research.
Acknowledgments The authors alone managed the patient clinically and wrote the manuscript. They have no conflicts of interest. Declaration of interest: The authors report no declarations of interest. The authors alone are responsible for the content and writing of the paper.
References 1. Hall GC, Carroll D, Parry D, McQuay HJ. Epidemiology and treatment of neuropathic pain: the UK primary care perspective. Pain 2006;122:156–62. 2. Cruccu G, Aziz TZ, Garcia-Larrea L, et al. EFNS guidelines on neurostimulation therapy for neuropathic pain. Eur J Neurol 2007;14:952–70. 3. Pereira EA, Boccard SG, Linhares P, et al. Thalamic deep brain stimulation relieves neuropathic pain after amputation or brachial plexus avulsion. Neurosurg Focus 2013; 35:1–11. 4. Fontaine D, Hamani C, Lozano A. Efficacy and safety of motor cortex stimulation for chronic neuropathic pain: critical review of the literature. J Neurosurg 2009;110:251–6. 5. Nguyen JP, Nizard J, Keravel Y, Lefaucheur JP. Invasive brain stimulation for the treatment of neuropathic pain. Nat Rev 2011;7:699–709. 6. Lotze M, Flor H, Grodd W, Larbig W, Birbaumer N. Phantom movements and pain. An fMRI study in upper limb amputees. Brain 2001;124:2268–77. 7. MacIver K, Lloyd DM, Kelly S, Roberts N, Nurmikko T. Phantom limb pain, cortical reorganization and the therapeutic effect of mental imagery. Brain 2008;131:2181–91. 8. Makin TR, Scholz J, Filippini N, et al. Phantom pain is associated with preserved structure and function in the former hand area. Nat Commun 2013;4:1570. 9. Lefaucheur JP, Hatem S, Nineb A, et al. Somatotopic organization of the analgesic effects of motor cortex rTMS in neuropathic pain. Neurology 2006;67:1998–2004.
