Three main techniques delineate a possible role for intracranial ablative procedures in patients with chronic pain. Recent studies demonstrate a continued need for clinical investigation into central mechanisms of neuroablation to best define its role in the care of patients with otherwise intractable and severe pain syndromes. Cingulotomy can result in long-term pain relief. Although it can be associated with subtle impairments of attention, there is little risk to other cognitive domains.
Key points
- •
Stereotactic ablative procedures have been in use for several years. Technique, efficacy, and side effects for each of these procedures have not changed significantly in the past several years. Central ablation can be considered for some patients with intractable pain, although these procedures are irreversible, unlike neuromodulatory procedures, which are generally both reversible and adjustable.
- •
Lesions of the centromedian and centrolateral nuclei of the thalamus are still used for the treatment of pain.
- •
Mesencephalic tractotomy of extralemniscal pathways lateral to the spinothalamic tract (STT) and medial leminiscus can result in relief of intractable pain without loss of sensation or dysesthesia.
- •
Cingulotomy can result in long-term pain relief, although it can be associated with impairments of attention and risk to other cognitive domains.
Introduction: nature of the problem
Pain is a multimodal subjective experience. The experience of pain is described as distributed in 3 dimensions: cognitive, affective, and sensory. Frontal and limbic areas are believed to subserve the cognitive-evaluative component of pain. Limbic cortex, cingulum, hypothalamus, thalamus, and various midbrain regions are thought to contribute to the motivational-affective component of pain. Primary somatosensory cortex, thalamus, STT, and local nerve endings are involved in the sensory component of pain. When peripheral treatments fail, and modulatory procedures are not possible, central ablative procedures can be used for the treatment of pain, even though the use of these procedures has declined in recent years with the advent of neuromodulatory techniques.
Introduction: nature of the problem
Pain is a multimodal subjective experience. The experience of pain is described as distributed in 3 dimensions: cognitive, affective, and sensory. Frontal and limbic areas are believed to subserve the cognitive-evaluative component of pain. Limbic cortex, cingulum, hypothalamus, thalamus, and various midbrain regions are thought to contribute to the motivational-affective component of pain. Primary somatosensory cortex, thalamus, STT, and local nerve endings are involved in the sensory component of pain. When peripheral treatments fail, and modulatory procedures are not possible, central ablative procedures can be used for the treatment of pain, even though the use of these procedures has declined in recent years with the advent of neuromodulatory techniques.
Diagnosis
The characterization and diagnosis of pain is one of the most common tasks in medicine. The experience of pain has somatosensory, affective, and cognitive components. From early attempts, including the McGill inventory, to current taxonomies, the accurate diagnosis and classification of chronic pain remains a challenging problem. The 3-component model of pain (sensory, affective, and cognitive), however, still forms the basis of most pain treatment regimens and is a good framework to study the role of central ablative surgical procedures for pain.
Clinical outcomes and trends in the literature
An excellent review of destructive procedures for nonmalignant pain was conducted recently and includes several of the most recent articles regarding ablative procedures for pain. Of the 146 articles, 131 constituted class III evidence. Most of the class I and class II evidence studies were for radiofrequency rhizotomy and did not include central procedures for pain. This demonstrates a need for prospective trials to better describe the efficacy of these procedures and to compare them with more recent modulatory approaches ( Fig. 1 ).
Thalamotomy
Thalamotomy has enjoyed the most attention recently, largely due to significant contributions by many investigators to identify the pathophysiological biomarkers of the manifestation of chronic pain in the thalamus and of their modulation by various ablative treatments. Patients with greater than 1 year of medically refractory neuropathic pain are reported the best candidates for contralateral thalamotomy.
The STT and spinoreticulothalamic tract terminate in the medial and lateral thalamus (ventral posterolateral nucleus [VPL]/ventral posteromedial nucleus [VPM]), before projecting to the lateral prefrontal cortex. Destruction of the VPL/VPM interrupts these specific pain pathways, which project somatotopically to primary sensory cortex and are thought to subserve the discriminatory aspect of pain.
The medial, nonspecific thalamic nuclei include the centralis lateralis (CL) and centromedian (CM)/parafascicular (PF). The efferents from these thalamic nuclei to associative and paralimbic areas provides an additional target to modulate the affective component of pain. The lateral nuclei include the ventrocaudal nucleus, the medial posterior nucleus, and the posterior centrolateral nucleus (CLp). Bilateral receptive fields from spinal Rexed laminae V–VII travel via the STT to the CLp. CLp has diffuse efferent projections to the primary somatosensory cortex, insula, and anterior cingulate cortex as well shorter projections to the thalamic reticular nucleus.
Theories of imbalance between the medial (nonspecific) and lateral (specific) groups form the basis for some thalamic procedures for pain. In one theory, the medial thalamus overinhibits the lateral thalamus. A second theory states that uninhibited output of heat and burning signals from the medial thalamus result from the loss of cold inputs from the lateral thalamus. Finally, diffuse low-threshold calcium bursts from around the CLp implicate the reticular thalamic nuclei as well. Taken together, it is possible to postulate that the STT has excitatory inputs to the medial and lateral thalamus. The thalamic reticular nucleus has inhibitory inputs to the medial and lateral thalamus. These are in turn driven by excitatory outputs from the medial and lateral thalamus in a feedback loop. In the central pain condition, often characterized by anesthesia dolorosa, it is proposed that the lateral thalamus no longer receives STT excitatory input and is further inhibited by the reticular nucleus, resulting in loss of information to the thalamus and loss of modulation of thalamocortical networks by ascending pathways. Support for these theories can be found in electrophysiological and imaging studies. Intraoperative recordings in patients with central pain demonstrate decreased spontaneous activity in the lateral versus medial thalamus. Imaging studies of patients with neuropathic pain show decreased regional blood flow in the posterior-dorsal part of the thalamus corresponding to the dorsal CL posterior nucleus.
Centromedial Lesions
Medial thalamic lesions have focused on the destruction of the CM/PF. The target is located approximately 8 to 10 mm lateral from midline, approximately 4 mm superior to the anterior commissure (AC)–posterior commissure (PC) plane, and approximately 5 mm anterior to the PC but exact stereotactic location can vary depending on patient anatomy and prior lesions, as for other thalamic targets. Intraoperative localization of the CM/PF can be aided by microelectrode recordings, which may identify characteristic bursting patterns elicited by intraoperative sensory evoked potentials. Intraoperative test macrostimulation of these nuclei may evoke a contralateral light tingling or painful burning sensation.
In addition, 2 populations of nociceptive neurons with wide bilateral fields exist in the CM/PF. One group is represented diffusely in the nucleus and is characterized by neurons that respond to noxious stimulus after a long latency and continue to respond after the stimulation ends. The other group is restricted to the medial basal CM/PF and is characterized by short latency response as well as short latency cessation after stimulus termination. Furthermore, Jeanmonod and colleagues reported on spontaneously bursting cells unrelated to noxious stimulus and without receptive fields in the CM of the chronic pain patients.
Lenz has proposed that localization of the CM/PF complex can be aided by initially targeting the ventral caudal parvocellular pars internus, where a taste response can be elicited. This landmark can help define the inferior border and lateral third of the CM. Stimulating other medial thalamic nuclei (medial dorsal and periventricular nuclei) causes a generalized uncomfortable sensation.
A recent clinical results study targeting the mesial thalamus shows that greater than 50% pain relief was achieved in 67% of patients, with 20% of patients achieving complete relief. The best results were in patients with deafferentation pain, where 71% experienced more than 50% pain relief. Complications of treatment did occur, including mixed somatosensory deficits in 9%, transient oculomotor deficits in 5%, and significant verbal deficits in 1 patient of 69 total patients treated. Dysesthesia was associated with inadvertent lesions of the lateral thalamus.
Central Lateral Lesions
Lateral lesions focus on the posterior CLp. The approximate stereotactic coordinates for this nucleus are 2 mm posterior to the PC, 6 mm lateral to the lateral to the thalamoventricular border, at the AC-PC plane. As for other targets, these can vary—sometimes substantially—depending on patient anatomy and lesions. One of the proposed advantages for targeting the CLp is its distance from the somatosensory nuclei, which may reduce the risk for sensory deficits with lesioning. Additionally, there is strong evidence that the CLp has direct STT input.
Lesions are placed 1 to 2 mm above the intercommisural plane to avoid inadvertent injury to the pretectum if the lesion is too posterior. The mediodorsal nucleus and putamen are directly posterior to the CLp. Inadvertent lesions to the putamen alleviate pain but only for a short time. Inadvertent mediodorsal nucleus lesions may cause no obvious clinical effect. Technical difficulties may arise when targeting the nucleus due to mechanical deflection at the ependyma, encountered in typical transventricular approaches to thermoablation of this area. The mechanical deflection can result in errors of 1 mm in the anteroposterior and medial-lateral direction and up to 1.9 mm in the dorsoventral.
Intraoperative targeting of the CLp can also be refined by microelectrode physiology, which may show pathologic low-threshold spikes and low-frequency (4 Hz) oscillations. The low-threshold spike activity has been found in a variety of diseases with so-called positive signs, including tinnitus, movement disorders, multifocal epilepsy, and central and peripheral neuropathic pain. Normal cortical low-frequency oscillations occur during certain stages of sleep and cognitive activation. Widespread, continuous, state-independent overproduction of low-frequency oscillations, however, is pathologic and associated with a pathologic state termed, thalamocortical dysrhythmia (TCD).
In the largest trial of CL thalamotomy, in 96 patients, more than 50% of patients had greater than 50% pain relief whereas 20% had complete pain relief. Mean duration of pain was decreased by 65% to 90%. Intermittent-type pain syndromes were improved in 54% of patients. Constant-type pain was decreased in 30% of patients and allodynia pain was alleviated in 60% of patients. Overall, 30% of patients decreased their drug intake postoperatively. Complications were also reported and included partial and partially reversible pretectal visual deficits. Serious complications, such as intraventricular hemorrhage, thalamic edema, and thalamic hemorrhage, were reported in 5% of patients.
Recent Advances
To avoid radiation, infection, and mechanical brain shift, Jeanmonod’s group has been investigating the use of MR imaging–guided focused ultrasound CLp thalamotomy for chronic neuropathic pain. Multiple sources of ultrasound are focused at a single point in the CLp, creating a thermal lesion comparable to those made with radiofrequency. MR imaging thermography allows for the accurate visualization of the lesion and estimation of temperature to ensure adequate lesion. In a study of 12 patients treated for chronic refractory neuropathic pain, the first patient had an inadequate lesion at maximum temperature of 42°C and was excluded from the study. Of the remaining 11, the next 2 patients received lesions that were too small on T2-weighted and diffusion tensor imaging (DTI) to have clinical effect; 9 patients had adequate lesioning with maximum temps of 51°C to 64°C and only data for these patients were included in analysis. The investigators reported mean pain relief of 49% (9 patients followed up) at 3 months and 57% at 1 year (8 patients followed up at 1 year) ( Fig. 2 ).
