(Melbourne, 2010)

Click here to download the lecture in document format .

Over forty year ago, my mentor, Geoff Toakley, introduced me to surgery for pain. This was for both neuropathic and cancer pain. Since then, I have come to know many of our neurosurgical forefathers who have raised the question, "Does surgery work for pain?"

 
These neurosurgeons have included William Sweet, Bill Nordenbos, Sean Mullan, Don Long, Peter Jannetta, Sira Bunyaratavej, Kim Burchiel, Mark Sindou, John Loeser, Richard North and Hugh Rosomoff. However, in recent years, there has been a significant loss of interest in the neurosurgical input into chronic pain due to the influence of John Bonica and Sam Lipton.

Bill Sweet and James White produced an anecdotal account of their experiences in the surgery of chronic pain over a forty year period and published this in 1969. Kim Burchiel provided a modern summary of the surgical management of pain in his textbook published in 2002.  

John Loeser drew our attention to the biopsychosocial nature of persistent pain with his classic rings, noting that the tissue injury was compounded by the thoughts and beliefs of the patient, the loss and suffering of the patient and, lastly, behavioural changes.

The surgical procedures for pain can be classified into anatomical, ablative and augmentative.

Hip replacement is an anatomical example, the cordotomy in an ablative approach and augmentative procedures include spinal cord stimulation. 

Today, I reserve the approach to surgery for pain to neuropathic pain.

This is the pain caused by a lesion or a disease in the somato-sensory area of the nervous system - lesions in the peripheral nerves, the spinal cord and the brain. It is associated with complex molecular changes in the electrical conduction in the nerve, as well as in the synapse and dorsal horn of the cord.  

When we come to measurement and comparison results, there is greater credibility in our understanding of medications using the "numbers needed to treat" convention. The Oxford League of Tables shows the expectation for each pain relieving drug, with "1" being the ideal level to reduce pain by 30-50%. This is also being used for some interventions, such as lateral facet joint injections.

Over the past century, the neurosurgeon has carried out the afferentation procedures at all levels of the nervous system as identified here. However, in 2010, involvement in these surgical procedures has contracted.

The anterolateral cordotomy is a fine example of a valuable ablative procedure. It was first reported by Spiller in 1912, based on earlier records by Brown Sequard in 1850. If the surgeon makes a 4.5 centimetre incision in the anterolateral aspect of the spinal cord, there is a relief of pain from six segments down on the contralateral side. The procedure is done at the high cervical or high thoracic level.  But bilateral procedures are likely to be complicated by Ondine's Syndrome (respiratory failure), as well as urinary and bowel incontinence. 

The innovative Sean Mullan, in Chicago, developed a percutaneous approach which he reported in 1963. 

We started using this at the Mater Hospital in Brisbane in 1966. It provided the advantage that electrical testing prior to making the lesion ensured that the correct area of the body was appropriately covered. The procedure could be done under local anaesthetic or with a Neurolept anaesthetic. Still, it was stressful for the patient and the surgeon and, like Peter Teddy, I preferred the open approach.  

The results of anterolateral cordotomy are difficult to interpret. There were large series by Cowrie, Lorenz and Hugh Rosomoff, but these included unilateral and bilateral procedures for malignant and non-malignant pain.  

Cordotomy remains an excellent operation, either open or percutaneously. It seems to be a lost art in Australia. It is best reserved for patients with unilateral cancer pain with more than six months to live. For the most part, bilateral cordotomy in cancer pain has been replaced by intrathecal opioids.

This picture of Norman Dott at his 70th birthday party reminds us that his open cordotomy for non-malignant pain, carried out by Sir Hugh Cairns, did not work and that continues today with our approach to non-cancer pain and cordotomies. The initial good result falls away quickly, leaving the patient with dysaesthesia and unhappiness.

In this picture, the superior cerebellar artery compresses the entry zone of the 5th cranial nerve into the brainstem, and in 91% of cases this is thought to be the cause of trigeminal neuralgia. In 8% it is caused by base of skull tumours, and 1% is caused by multiple sclerosis. 

In most of the procedures for managing trigeminal neuralgia, there is an initial 90-98% success rate, but the variables relate to the recurrence rate with these procedures.

The dorsal root entry zone procedure was probably first recognised by Sindou. It meant a deafferentation procedure at this place in the cervical spinal cord using coagulation or an incision.

Mark Sindou published an account of this in 1972.

While the subsequent results were impressive, there were some unspoken problems with litigation which has resulted in a loss of interest in the procedure.

These are examples of spinal cord stimulators, first given to me by Bill Sweet in 1969, and the modern version in 2010. 

 
The procedure has been accepted enthusiastically as an augmentative procedure which is reversible. There is Level 2 evidence that it works in the failed back syndrome, Level 4 that it works in chronic regional pain syndrome, but little support for it in brachial plexus avulsion or in post herpetic neuralgia, spinal cord injury or phantom limb pain. 

In 1977, Tony Yaksh drew our attention to the value of intrathecal opioids. By the 1990's, the Medtronic computerised pump allowed longterm management of chronic pain with intrathecal opioids. It came at some expense. Then again, it was overused.  

 
The results indicated that it was an excellent procedure for midline cancer pain of a somatic or visceral origin, provided the patients had a successful four-day trial and provided they had more than six months to live. Increasingly we are recognising that it was inappropriate for non-malignant pain and many of these pumps are now being removed.

In all these cases, the correct patient selection and appropriate exclusions are imperative. It is best avoided in patients with:

• Cognitive impairment;
• Unresolved psychological disorders;
• Active or untreated abuse of medications;
• Unresolved issues of secondary gain;
• Short life expectation (cancer).

In conclusion, it can be said that deep brain stimulation for pain is an experimental procedure at present.  Microvascular decompression is well accepted with remarkable results. Sectioning of the 5th and 9th cranial nerves for pain in malignancy has significant benefits. The cordotomy for unilateral cancer pain is a worthwhile procedure. Spinal cord stimulation again is particularly helpful in 64% of patients with the failed back syndrome, provided they are appropriately selected by a multidisciplinary pain clinic. The morphine pump has a place in the management of cancer pain, but little place in non-malignant pain.