Commissural myelotomy is an ablative neurosurgical procedure that blocks transmission of pain fibers crossing the midline from the dorsal horn to the contralateral spinothalamic pathway. The classical spinal commissural myelotomy is performed three spinal levels above the source of painful stimulus, reflecting the fact that pain fibers typically ascend two or more levels within the dorsal horn before crossing at the midline. The procedure was first proposed by Greenfield in 1926 as a means to address bilateral or midline pain without the morbidity and mortality of bilateral cordotomy. Armor reported on the first successful commissural myelotomy in 1927 and the general technique he proposed continues to be employed today (Gildenberg 2006) (Fig. 31.1).

In theory, sphincter disturbances and respiratory dysfunction are less likely to occur with myelotomy than with cordotomy. Given the attractiveness of a single procedure producing bilateral pain relief without high risk of damaging essential axonal tracts, one would expect the procedure to be of great interest to the neurosurgical community. However, in spite of its potential advantages, the procedure has never gained wide acceptance as compared to other ablative neurosurgical pain interventions. Commissural myelotomy was applied on a relatively wide scale in the 1940s and early 1950s by French and German neurosurgeons Wertheimer, Lecuire, Mansuy, and others. However, the procedure soon fell out of favor, and Wertheimer’s group eventually abandoned the procedure in favor of unilateral or bilateral cordotomy (Papo and Luongo 1976). The procedure has endured periods of resurgence since this time, but only a few hundred myelotomies are reported in the literature.

A commissural myelotomy may be performed either percutaneously or open. The percutaneous method is less invasive but has only been described by a handful of surgeons. The more common open myelotomy technique is performed by means of a laminectomy. After making a midline incision in the dura, the midline septum of the dorsal spinal cord is exposed by laterally displacing, or where necessary, delicately dividing the network of overlying dorsal veins.

Success rates for commissural myelotomy of 60–70% have been reported, although the inclusion criteria, technique, and endpoints vary greatly among studies. Complications of commissural myelotomy include dysesthesias, bowel and bladder dysfunction, decreased proprioception, paralysis, and death. It is not uncommon for patients to report an “absent” sensation in their lower extremities that subsides within a few weeks of the operation (Sourek 1969; King 1977; Cook and Kawakami 1977). Pain may also recur with time; therefore, it may be advisable to select only patients with a relatively short life expectancy for this procedure. At the same time the open commissural myelotomy is a relatively invasive procedure requiring general anesthesia and, as such, may not be suitable for fragile patients in poor medical condition (Gybels and Sweet 1989).

The indications for high cervical myelotomy, performed at the C1-3 level, are more limited than for spinal commissurotomy. It has been suggested that high cervical myelotomy may be attempted to provide relief of shoulder, arm, and upper chest pain without the high rate of respiratory complications of high cervical cordotomy. Hitchcock showed in a series published in 1974 that such lesions may also relieve lower body pain. The finding that a focused central lesion could relieve upper and lower pain prompted the study of cervical lesions as an alternative to the longitudinal commissural procedures described above. However, the success rate of high cervical myelotomy reported in a small series by Papo was disappointing, as only three of ten patients had successful pain relief. Furthermore, pain returned to these patients within 4–6 weeks (Papo and Luongo 1976). Nonetheless, in cases where cordotomy is contraindicated because of respiratory impairment or a short life expectancy, a high cervical myelotomy may still be considered for patients with upper body bilateral or midline pain.

Since the advent of the classic commissural myelotomy, several modifications of this procedure have been developed. These modifications are intended to maintain similar visceral pain relief while potentially reducing the frequency of sensory perception disturbances. One such procedure, the limited midline myelotomy, was first described by Gildenberg and Hirshberg (Gildenberg and Hirshberg 1984). With this procedure, a small laminectomy is performed at approximately the thoracolumbar junction. The vertical extent of the midline dissection within the adjacent dorsal spinal cord need only span 5–7 mm with a depth of 6 mm. The procedure provides remarkably widespread visceral pain relief, and the authors reported satisfactory results in ten of 14 patients treated. No reported sensory loss, weakness, or other complications were reported in this series. Gildenberg and Hirshberg proposed that the procedure interrupts a previously unrecognized ascending midline pain pathway (Gildenberg 2006; Gildenberg and Hirshberg 1984). Although this claim was met with controversy, subsequent postmortem specimens and animal experiments have since supported the existence of a long midline dorsal column pathway involved in the perception of pelvic visceral pain (Willis and Westlund 2001).

Building on the anatomical discovery of an ascending midline dorsal column visceral pain pathway, Nauta introduced the punctuate midline myelotomy, a minimally invasive procedure that specifically lesions this newly recognized pathway while sparing the crossing spinothalamic fibers. Nauta’s punctuate midline myelotomy was produced with 5-mm deep puncture using a 16-gauge needle on either side of the median septum in the dorsal column of the spinal cord (T-8) in a patient with pelvic cancer pain (Nauta et al. 1997). Following Nauta’s initial positive report, the procedure has been reproduced successfully in a dozen patients with a variety of sources of visceral pain (Nauta et al. 2000; Hwang et al. 2004). In each case the procedure provided immediate pain reduction without neurological side-effects. Because punctuate midline myelotomy deliberately spares crossing spinothalamic fibers, the success of this operation corroborates the existence of an ascending visceral nociceptive pathway in the midline of the dorsal column.

Dorsal root entry zone (DREZ) lesioning is a neuroablative procedure of the posterolateral aspect of the spinal cord, including the medial portion of Lissauer’s tract and the superficial five layers of Rexed’s laminae, where sensory root fibers synapse with axons that join the spinothalamic tract (Nashold and Ostdahl 1979). The procedure has classically been used to treat deafferentation pain arising from a brachial plexus injury, a pain syndrome that may be poorly managed with other medical or surgical interventions (Nashold and Ostdahl 1980). DREZ-lesioning has also been used to treat other forms of chronic pain including spinal cord “end zone” pain, trigeminal neuralgia, phantom limb pain, and radiation plexopathy (Saris et al. 1985; Zeidman et al. 1993; Friedman et al. 1984) (Fig. 31.1).

The concept of DREZ-lesioning was first introduced by Sindou and colleagues in 1972. Sindou successfully used microsurgical incisions to lesion the dorsal root and a small portion of the lateral portion of the dorsal horn as a means to relieve pain in a patient suffering from a malignancy within the brachial plexus (Sindou 1972). In 1974, Nashold built on Sindou’s experience and developed a wider and deeper lesion of the DREZ using radiofrequency thermocoagulation. Nashold greatly popularized the DREZ lesion and in time introduced the complete DREZ lesion as a means of controlling pain attributable to brachial plexus avulsion (Nashold and Ostdahl 1979). Nashold’s procedure requires complete laminectomy extending three levels on either side of the affected spinal cord. After opening the midline dura, an electrode is inserted approximately 2 mm into the DREZ portion of the cord at a 25° angle to the vertical plane; multiple lesions are produced approximately 1–2 mm apart in the inferior–superior direction. Other successful modalities of lesioning the DREZ have been introduced including laser light (Levy et al. 1983) and ultrasound (Dreval 1993).

DREZ-lesioning provides long-term efficacy in the management of pain from brachial plexus avulsion. In one series of 55 patients who underwent treatment of a severe brachial plexus avulsion with Sindou’s microsurgical DREZ technique, 95% reported excellent pain relief at discharge. At 3 months, 82% reported excellent or good pain relief, and at 29 months 66% of the patients not lost to follow-up had continued excellent or good pain relief, and 71% reported an improvement in activity level (Sindou et al. 2005). In a separate study, 77% of patients with brachial plexus avulsion treated by DREZ-lesioning had good or fair pain relief at 63 months (Thomas and Kitchen 1994). The success of DREZ-lesioning has been more varied when applied to other pain syndromes.

The introduction of the operating microscope and modern electrodes has greatly reduced complications of DREZ-lesioning. Nonetheless, neurological deficits remain a relatively common complication of the procedure, including ipsilateral lower-limb weakness, sensory changes, or both. These neurological deficits are typically mild and improve with time. Other reported complications include CSF leaks and transient loss of continence or bladder spasm (Thomas and Kitchen 1994), (Nashold and Ostdahl 1980).

The cingulate gyrus is a part of the limbic system and is involved in pain perception. The history of surgery in this brain region as a treatment for pain, i.e. cingulotomy, dates back to the mid-twentieth century discovery that prefrontal lobotomy often provides significant chronic pain relief (Foltz and White 1962). Although “lobotomy” lesions were relatively nonspecific and typically destroyed a large area of white matter tracts arising in the prefrontal cortex, the associated pain-reducing effects are now believed to stem from injury to the adjacent cingulate gyrus. Despite initial enthusiasm for prefrontal lobotomy as an innovative approach to managing severe pain, and even much more commonly treating disorders of the psyche, the lobotomy soon fell out of favor plagued by reports of devastating side-effects and changes in personality. After a period of dormancy, the field of psychosurgery was partially rejuvenated by the advent of stereotactic surgical method, which more accurately targets specific brain regions while simultaneously minimizes damage to adjacent structures (Brotis et al. 2009). Modern studies have clearly shown cingulotomy to be effective in the treatment of intractable pain when medical, surgical, and pharmacological interventions have failed (Wilkinson et al. 1999).

The cingulate gyrus is a paramedian cortical structure that is embedded deep within the interhemispheric fissure and wraps around the corpus collosum. A number of studies have implicated the cingulate gyrus in the perception of pain. Injection of lidocaine or morphine in the anterior cingulum of rats produces analgesia (Vaccarino and Melzack 1989), (LaGraize et al. 2006), and increased neuronal activity in areas of the cingulate gyrus of rabbits has been reported in response to noxious stimuli (Sikes and Vogt 1992). In particular, fibers of the cingulum bundle, which passes through the gyrus and connects the anterior and posterior cingulate cortex, appear to play a major role in pain perception (Vaccarino and Melzack 1989). Position emission tomography (PET) analysis of regional blood flow has confirmed these animal findings in humans. PET scans have identified the contralateral cingulate cortex as one of the centers in the brain that responds to a noxious, painful stimuli compared to a non-painful stimuli (Casey et al. 1994; Jones et al. 1991; Coghill et al. 1994).

Although it is clear that the cingulate cortex plays an important role in pain perception, its exact role remains poorly understood. The diffuse distribution of cortical and thalamic activation after a painful stimulus highlights the complex nature of pain and the difficulty in treating it (Coghill et al. 1994). The anterior cingulate cortex is highly interconnected with other limbic structures and cortical regions, which allows it to integrate affective, motor, and memory stimuli to modify the perception of a painful stimulus without actually changing the sensation itself. The region of the cingulate cortex responsible for this emotional component of pain has numerous connections with the amygdala, frontal cortex, and periaqueductal gray. Input from the hippocampus may allow the cingulate cortex to integrate nociceptive input with memory to evaluate the danger of a situation. The affective area of the anterior cingulate is implicated in conditioned emotional learning, vocalizations of emotion, motivation, and determining emotional relevance of stimuli. Another region of the cingulate cortex is the cognitive division, which is involved in information processing and determining appropriate motor responses to painful stimuli (Devinsky et al. 1995).

In 1962, Foltz and White reported the results from a study assessing the outcome of frontal cingulotomy in 16 patients with chronic, intractable pain. To test their hypothesis that chronic pain has a strong emotional component that is controlled by the cingulate, the authors chose a study population with intractable pain in whom emotional factors were thought to exacerbate their pain. The authors found that bilateral frontal cingulotomy in 12 of the 16 patients had a good or excellent outcome. Although the patients continued to sense pain, they did not react to the pain as strongly, saying that the pain “is not particularly bothersome” or “doesn’t worry me anymore” (Foltz and White 1962). Hurt and Ballantine in 1974 found that 66–67% of a total of 68 patients with chronic pain experienced some degree of pain relief after stereotactic anterior cingulate lesions (Hurt and Ballantine 1974). In subsequent studies the percent of patients who experience pain relief after an anterior cingulotomy ranges from 45 to 67% (Faillace et al. 1971; Hurt and Ballantine 1974; Yen et al. 2005; Wilkinson et al. 1999; Wong et al. 1997). A recent report published in 1999 found that the majority of 18 patients who received a bilateral anterior cingulotomy for chronic noncancer pain experienced improvement in their pain, were no longer taking narcotics, noted improvements in their family life and social interactions, and thought that cingulotomy was beneficial (Wilkinson et al. 1999).

Hassenbusch was the first to propose using magnetic resonance-guided stereotaxis for cingulotomies. Prior to this, stereotactic localization had been achieved using air ventriculograms, which requires lumbar or cisternal puncture to inject filtered air into the ventricles and was associated with a higher risk of meningitis and hemorrhage. MRI-guided stereotaxis also has the benefit of not requiring general anesthesia, is more accurate because it images the cingulate gyrus directly, and is useful for postoperative follow-up of the lesion (Hassenbusch et al. 1990; Pillay and Hassenbusch 1992). Since the 1990s, MRI-guided stereotaxis is routinely used when performing cingulotomies (Brotis et al. 2009).

Despite recent improvements in technology that have made cingulotomies more accurate and less destructive, cingulotomy remains an invasive procedure that can have serious side-effects including seizures or lasting behavioral deficits. Neurocognitive function is generally preserved after cingulotomy, although some studies have reported deficits in focused and sustained attention (Yen et al. 2009; Cohen, Kaplan, Moser et al. 1999; Cohen, Kaplan, Zuffante et al. 1999), learning and organizing verbal material (Faillace et al. 1971), and self-initiated behaviour (Cohen, Kaplan, Zuffante et al. 1999). Another study reported altered emotional experiences of patients who underwent cingulotomy, especially in regards to emotional agitation and tension. Although family members often noted personality changes after the surgery and described the patients as being more “relaxed” (or even more apathetic in a few cases), the patients were not functionally crippled by these emotional changes. The patients felt less tension and anger, but did not report a significant reduction in self-perceived energy level (Cohen et al. 2001). Despite the severity of pre-existing pain and the subtle nature of possible changes, cingulotomy remains a controversial procedure; any risk of altering a patient’s affect and personality appears to preclude widespread acceptance of this approach.

Several methods of pituitary ablation, a procedure called hypophysectomy, have been described for the treatment of chronic pain. These include transcranial hypophysectomy and microsurgical hypophysectomy. The mechanisms underlying the analgesic effect of hypophysectomy are not completely understood. The pituitary gland secretes a number of essential endocrine hormones, but the analgesic effect does not appear to directly relate to the interruption of these hormones, leading some to postulate that a direct neuronal mechanism is involved. Another theory is that the analgesic effect of hypophysectomy is mediated by a change in hypothalamic-pituitary axis (HPA) peptides in the cerebrospinal fluid (Takeda et al. 1986).

Hypophysectomy has classically been used to relieve pain from severe and diffuse cancer such as seen with widely metastatic breast and prostate adenocarcinoma. Surgical and chemical hypophysectomy with injection of alcohol are fundamentally similar procedures that have been shown to produce comparable levels of pain relief (Ramirez and Levin 1984). Chemical hypophysectomy is carried out by introducing a needle through the nostril to puncture the sphenoid sinus. Once the sinus has been punctured and rinsed, a 19-gauge or 20-gauge needle is advanced using biplanar fluoroscopy guidance to the floor of the sella turcica. Once the needle is in place, 1–2 ml of absolute alcohol is slowly injected and the patient’s pupillary reactions and body movements are closely monitored. Afterwards, the hole in the sella is sealed and the nasal cavity is packed (Ramirez and Levin 1984), (Lloyd et al. 1981).

Despite the unknown mechanism of action of hypophysectomy, it is highly effective, providing satisfactory or excellent relief of pain in 70–83% of patients with metastatic cancer (Ramirez and Levin 1984; Tindall et al. 1976; Lloyd et al. 1981). Interruption of normal pituitary function is a common complication of the procedure, resulting in the insufficiency of one or multiple pituitary hormones (Evans et al. 1982). In one published series, central diabetes insipidus occurred in 17% of patients, resulting from interruption of ADH secretion from the posterior pituitary gland. Prolonged visual disturbance was another frequent complication in this study (Lloyd et al. 1981).