Opioids in chronic non-malignant pain

The use of opioids in chronic non-malignant pain is profoundly messy. A simple start is to say that if somebody has severe pain which responds to opioids and for which there is no other effective remedy then why should they not receive opioids? Two judgments are then implicit: that opioids are effective and that other remedies are not. How well do these judgments hold up? And if they do, how do we work out which opioid and formulation? A paper in this week's issue addresses, but doesn't answer, the second question (p 1154).¹

Opioids are often withheld to protect society or to protect the patient. The society argument is that the medical availability of opioid increases street addiction. There has never been any strong evidence that medical use increases street problems, and the introduction of oral morphine in Sweden in the early 1980s was shown not to increase addiction.²

Withholding the opioid to protect the individual might be done for physical or psychological reasons. Many notables across the centuries used opioids long term without deterioration in physical health: Florence Nightingale, for example, survived over 40 years after her first opium injection for back pain.³ We know too that if the opioid sensitive pain later resolves opioids can be stopped without patients becoming addicts. The grey areas here are the judgments about the patient's potential for addictive behaviour and about the opioid sensitivity of the pain. We lack good tests to help with either judgment. We fear scenarios such as patients with no identifiable cause for their back pain using escalating doses of “minor” and then “major” opioids. This fear can lead to draconian guidelines and thoughtless legislation—which restrict opioid use to the detriment of those with genuine need.

In cancer pain we claim that tolerance, the need for increasing doses to achieve the same result, is rare. Patients with stable disease stay on a constant dose for months. We need to admit that escalating doses in the absence of disease progression is a red flag in the management of chronic non-malignant pain. So the care pathway for those with severe chronic non-malignant pain has several points at which we need an injection of wisdom.⁴ Diagnosis is not always a problem (phantom pain or postherpetic neuralgia), but with back pain or abdominal pain we often struggle. If we accept that opioids are a treatment option even in the absence of a precise diagnosis, then we need to know, firstly, whether non-opioids have been tried rationally and, secondly, how to measure their success or failure.

A trial of opioid beckons only when conventional analgesics (such as non-steroidal anti-inflammatory drugs and combinations), unconventional analgesics (such as antidepressants and anticonvulsants), psychological approaches, injections, devices, or operations have failed. In neuropathic pain antidepressants and anticonvulsants can both provide at least 50% pain relief for one patient in three,⁵ but we have very little evidence of which drug is best in its class, or indeed which class is best in particular pain syndromes.

We have no reliable predictive tests for opioid efficacy other than suck it and see. We do know that opioid efficacy may be reduced in neuropathic pains⁴ and that current opioid formulations often work poorly for severe pain on movement. The trial of opioid itself needs success and failure criteria, encapsulated in the common clinical question “What is an adequate dose?” The dose needs to be increased until analgesia or adverse effects result, and sometimes this needs doses that raise eyebrows. The balance between effect and adverse effect may be fine, and active management may be needed to produce, in that coy phrase, tolerable and manageable adverse effects.

There is a complicated practical and research agenda, which requires coherent multicentre working and innovative research design. There is no evidence base on which we can rely other than common sense, our own experience, and that of others. Patients' wishes are simple but can be hard to fulfil. They want good pain relief, but not at the expense of adverse effects, particularly those affecting the central nervous system. Even when we resolve these puzzles the professional's unease will remain. Few would be uncomfortable with opioid use that allows an elderly patient with rheumatoid disease to sit without pain, but few would be comfortable prescribing strong opioids long term for a young person with a vague diagnosis of back pain.

The trial by Allan et al reported this week compares two opioids, each in a different formulation—oral or transdermal (p 1154).¹ This is a welcome trial in a difficult area. The focus is which drug (or formulation) gives the fewest problems, or is preferred by patients, at the same level of pain relief. Unfortunately the design of the trial means that we have to question the results. Rule one of drug trials that compare different formulations and use subjective outcomes such as patient preference is that the comparison should be done double blind. This may be awkward and it will be more expensive, but breaking the rule means that the conclusions may not be correct. Yet here we are with a trial which compared different formulations and used subjective outcomes and was not done double blind. The problem we are left with is whether any difference between formulations is credible, and whether any credible difference is worthwhile. Given the high prevalence of chronic pain and its major impact on quality of life it is time that we had a better grip on what works in clinical practice and when.

Primary care p 1154