Overview of opioid analgesics

There are significant difficulties translating evidence from clinical pain trials into pain management in practice. This is not only because of issues around the number, quality, bias, duration and construction of studies,^90,96,286 but also because pain is a subjective experience influenced by a complex range of factors.

Additionally, labels such as CNCP do not just describe one condition, but a variety of conditions with diverse aetiologies, for which the evidence for therapeutic impact varies. Patients with the same condition will have unique pain experiences and respond differently to therapeutic interventions.

No analgesic drug works well in all patients. Most analgesics work well in a small proportion of patients. There is often a strong placebo (contextual) effect. Pain relief from therapeutic interventions is not normally distributed but is usually bimodal, being either very good (above 50%) or poor (below 15%).^287–289 Hence, using averages is unhelpful and misleading because few (if any) patients experience ‘average’ pain relief and it tells us nothing about how many patients will experience clinically useful analgesia.²²¹

Clinical trials designed for regulatory purposes consider single interventions and fixed dose regimens, which may exacerbate adverse events and withdrawals, resulting in higher failure rates.²²¹ For example:

• failure rates for NSAIDs are ≥70% in osteoarthritis, ≥80% in chronic low back pain and 58–72% in ankylosing spondylitis²²¹
• for neuropathic conditions, antidepressants and anticonvulsants have failure rates of ≥70% in painful diabetic neuropathy and post-herpetic neuralgia, and ≥87% in fibromyalgia.²²¹

This does not reflect the clinical reality of choosing options and individualising doses. For example, about half of osteoarthritis patients with moderate or severe pain on treatment had a significant (30%) reduction in pain intensity when switched to another NSAID.²⁹⁰

On review of pain trial studies, it was noted that patient response distributions are U-shaped, not Gaussian, making average values inappropriate. In fact, the ‘average pain score’ data may mislead. Since 2011, the editors of the Cochrane Pain, Palliative and Supportive Care Systematic Review Group have established new criteria for examining evidence in pain.²⁹¹ It is now standard to measure ‘responder’ analyses; reporting the proportion of patients achieving outcomes that patients consider worthwhile²²¹ (that is, the proportion of patients who experience at least 30–50% pain reduction).

A minority of patients achieve very large reductions in pain (responders) whereas the majority achieve little relief (non-responders).²⁹² Individual patient analyses for chronic pain interventions have shown that people who respond also experience improvements in fatigue, depression, and sleep interference.^53,293

It is standard for patients to delegate ‘at least 50% pain reduction’ as a successful outcome. It is the minimum outcome that patients want,²⁹³ and may be associated with restoration of function, work, and quality of life lost with chronic pain.²⁹ Clinically, this has major implications for practice. Use of responder analysis changes judgement of benefit and risk and suggests that classical trials in pain using ‘averages’ may underestimate efficacy.²²¹

Responder analysis also supports clinical practice focus on individual responses to therapy. It enables trialling numerous treatment options to achieve pain relief for the individual.²²¹ Similarly, non-responders should stop treatment that does not work.

The evidence of failure for paracetamol²⁹⁵ and NSAIDs^296,297 in musculoskeletal pain and the poor efficacy for opioids, anticonvulsants (gabapentin, pregabalin) and antidepressants for neuropathic pains¹⁰⁰ need to be reconsidered in the light of responders and non-responders. Evidence about a single intervention needs to be considered with individual patient circumstances, tempered with wisdom and experience to be used sensibly in clinical setting.²⁸⁶

It is with these caveats in mind that evidence of effect of pain medications in chronic pain should be considered.

The efficacy of opioid therapy in acute pain is supported by strong evidence from RCTs^5,7 and by systematic reviews in cancer pain,^298,299 palliative care³⁰⁰ and opioid dependence.³⁰¹

CNCP is very different. It is not a diagnosis, but a group of entities with various aetiologies. The pathophysiologic descriptors of these aetiologies are still changing. Studies examining chronic pain often have methodological weaknesses that make interpretation difficult, and transfer into clinical practice requires care.

The evidence on long-term opioid therapy for chronic pain outside of end-of-life care remains limited; however, this does not mean there is no evidence to guide care.

Evidence in different chronic pain conditions

Musculoskeletal pain

Musculoskeletal conditions account for a large proportion of general practice opioid prescriptions.³⁰² In trials with at least 12 weeks’ duration of opioids for managing osteoarthritis, there is:

• fair evidence for tramadol³⁰³
• limited evidence for transdermal buprenorphine³⁰³ – it has been shown to be effective and well tolerated (with analgesic effects similar to tramadol)³⁰⁴
• limited evidence for tapentadol for arthritic pain.^303,305

Reviews of opioid therapy in chronic low back pain provide some support for short-term use, but evidence beyond three months is lacking.^306,307 In trials with at least 12 weeks’ duration of opioids for managing chronic low back pain, there is some evidence for transdermal buprenorphine,³⁰⁸ tapentadol³⁰⁵ and tramadol/paracetamol combinations.^309–311

Caution should be used in extrapolating evidence from short-term trials into longer-term care. Analysis of openlabel extension trials provides some support for sustained opioid effect, but in only 25% of patients originally enrolled.³¹² The harms (abuse of prescribed opioids, mortality) of long-term opioid therapy in clinical practice are underestimated by long-term extension studies, probably because patients with major medical diseases and mental disorders were excluded.³¹²

Neuropathic pain

Several guidelines support opioid use in neuropathic pain, but not as a first-line treatment. When first-line medications fail or provide inadequate pain relief, tramadol or a conventional opioid analgesic may be useful as a second-line or third-line treatment.^94,96 For management of neuropathic pain:

• tramadol has weak GRADE recommendations for its use; generally it is considered a second-line treatment  because of safety and tolerability.⁹⁶ The tramadol NNT for 50% pain reduction is approximately five⁹⁶
• strong opioids, particularly oxycodone and morphine, have weak GRADE recommendations for use and are recommended as third-line treatments mainly because of safety concerns.⁹⁶ The oxycodone NNT for 50% pain reduction is approximately four.⁹⁶ Other reviews are less favourable for oxycodone and report a moderate benefit (at least 30% pain relief) NNT at 5.7.³¹³

The NNT for benefit in opioids appears similar to other drugs (eg antidepressants, anticonvulsants) used in painful neuropathies such as diabetic neuropathy, post-herpetic neuralgia, peripheral nerve injury, HIV neuropathy, central pain, trigeminal neuralgia and mixed neuropathic pain. The NNT for these non-opioid medications range from around four to eight. Continual critical appraisal of all classes of medication used in long-term pain management is warranted.

Combination therapies are common though few are studied. One meta-analysis demonstrated modest superiority of gabapentin plus opioid versus gabapentin alone, although the combination produced significantly more dropouts due to accentuated side effects related to combination treatments.³¹⁴

Summary

There is limited evidence for opioids for management of CNCP and insufficient evidence to determine long-term benefits. For well-selected patients with no history of SUDs, proper management with opioids can contribute to long-term pain relief.¹¹However, long-term opioid treatment (≥26 weeks) benefits only about 25%

Presented in alphabetical order

Buprenorphine

Buprenorphine is a partial agonist at mu opioid receptors and an antagonist at delta and kappa receptors. It is typically used for analgesia (in low-dose patch formulation) and in ORT, where sublingual formulations are usually used.

Musculoskeletal pain

There is limited evidence regarding buprenorphine for CNCP due to a lack of high-quality RCTs.³⁰³ However, transdermal buprenorphine for osteoarthritis has been shown to be effective and well tolerated, with analgesic effects similar to tramadol.³⁰⁴

Neuropathic pain

Case reports suggest that buprenorphine is effective in peripheral^315,316 and central neuropathic pain in the clinical setting.³¹⁷ However, large trials are lacking and currently there is not enough evidence to support or dispute efficacy of buprenorphine in any neuropathic pain condition.³¹⁸

Addiction medicine

Buprenorphine is listed for use in ORT (as Section100 [S100]).

In practice

Buprenorphine is PBS listed for chronic severe pain and ORT.

Transdermal patches (used for pain, not ORT) generally provide a week of analgesia. Occasionally, patients complain that there is release of the drug from the transdermal patch for only six, or rarely five, days. In these instances, the patches may need to be changed more frequently than weekly.

Buprenorphine can be safely used in patients with renal impairment and has less immunosuppressive effect than pure mu-opioid agonists.³¹⁹

As long as sedative medication is not given concurrently, the risk of respiratory depression with buprenorphine is low compared to morphine, methadone, hydromorphone and fentanyl.³²⁰ There is a ceiling effect for respiratory depression but not for analgesia.³²¹ If buprenorphine-induced respiratory depression occurs it may be completely reversed with naloxone,³¹⁹ although higher than usual doses and a longer duration infusion of naloxone are required.³²²

Withdrawal symptoms may occur if buprenorphine is ceased after long-term treatment; however, these symptoms are milder and more delayed in onset (≥72 hours) compared with other opioids.³²⁰

Buprenorphine binds strongly to the mu receptor site, but does not fully activate it.³²³ Therefore, if buprenorphine is combined with pure mu agonists (eg morphine, fentanyl), interactions may occur. For example, if a pure mu agonist is given to a person on maintenance buprenorphine it may be less effective. Conversely, buprenorphine could theoretically cause a withdrawal reaction if given to a patient taking longer-term opioid (mu) therapy.³²³

Antagonism of response to pure mu agonists (precipitated withdrawal) can occur with buprenorphine but it has only been demonstrated at buprenorphine doses exceeding the ranges used for analgesia (eg at dosages for ORT). In practice, these drug interactions are unlikely.

Codeine

Codeine is a weak mu receptor agonist (200-fold weaker affinity than morphine) and its analgesic action depends on the metabolism of about 10% of the dose to morphine, via CYP2D6.^324,325 Ultrarapid metabolisers have significantly higher levels of morphine and morphine metabolites after the same dose of codeine.³²⁶ Poor metabolisers do not produce any morphine or gain any analgesic effect.

Codeine is subject to misuse and dependence, and is the commonest prescription opioid associated with fatal overdoses in Victoria.³²⁷ Rates of misuse average between 21% and 29%, and dependence average between 8% and 12%.³²⁷

Musculoskeletal pain

Codeine is commonly used in combination with other minor analgesics (eg paracetamol, ibuprofen). There is highquality evidence that combination codeine medicines provide clinically important pain relief in the immediate term, but this is mostly in acute pain.¹⁶

In practice

Codeine is classified as a weak opioid. It is listed by the PBS for mild to moderate pain. There is no role for codeine in chronic pain.

A single 60 mg dose provides good analgesia to few adults: 12 patients need to be treated for one to achieve a 50% reduction in postoperative pain.³²⁸ OTC preparations containing low doses of 8–15 mg codeine phosphate are considered sub-therapeutic.

Combining codeine with non-opioid analgesics provides limited additional analgesic benefit: seven patients need to be treated with ibuprofen 400 mg/codeine 25.6–60 mg for one to obtain at least a 50% reduction in postoperative pain when compared to treatment with ibuprofen 400 mg alone.^328,329

Given the variability in response and risk of harm, use of codeine should be closely monitored.

Dextropropoxyphene

In November 2011, the TGA decided to remove the registration of dextropropoxyphene in Australia.³³⁰ It was withdrawn from the Food and Drug Administration (FDA) in the US due to risks of QT-interval prolongation and possibility of Torsades de Pointes (TdP) and cardiogenic death.

Oral dextropropoxyphene alone is a poorly effective analgesic.³³¹ In combination with paracetamol, it also provides little benefit above paracetamol alone.³³²

In practice

Dextropropoxyphene has now been limited to authorised users for previous users only. To prescribe this medication, GPs need to:

• be aware that the medicine is only approved for use in patients not able to be adequately treated with other mild pain killers
• have considered the contraindications for the medicine outlined in the product information and have explained them to the patient at the time of prescribing
• have considered any recent changes to the patient’s clinical presentation or biochemical status
• have warned the patient at the time of prescribing about appropriate use of the medicine
• be satisfied at the time of prescribing that the patient’s history does not indicate that the patient is at risk of accidental or intentional self-harm.

The conditions also require that a signed Prescriber Confirmation form is presented to the pharmacist dispensing these medicines before supplying them to the patient every time a patient presents for a prescription.

Fentanyl

Fentanyl is a highly potent opioid, which is active at the mu receptor. It is metabolised almost exclusively in the liver to minimally active metabolites. This makes it particularly useful in renal failure: <10% of unmetabolised fentanyl is renally excreted.³³³

It is available as transdermal patches, oral transmucosal lozenges or lollipops and injectable preparations. The transdermal system offers an excellent option for long-term treatment of cancer pain, but the RACGP believes it is not suitable for CNCP. A 25 ug/hour fentanyl patch is equivalent to approximately 90 mg of oral morphine per day. Oral transmucosal fentanyl rapidly achieves high plasma concentrations and is indicated to treat breakthrough pain in cancer patients who are not opioid naïve.³³³

Fentanyl-related mortality is currently relatively low in Australia compared to the US and parts of Europe. However, fentanyl misuse is on the rise in Australia with a large proportion of these deaths occurring among at-risk groups who inject drugs.³³⁴ Because of the misuse potential, this drug should be used only as indicated. It has known diversional potential, extremely high street value and risk of misuse.

In practice

Fentanyl is PBS listed for severe disabling pain and is usually used in cancer care or in acute hospital settings.

In the opioid-naïve patient, there is a significant risk of toxicity and overdose. Fentanyl patches are not suitable to be used as the initial agent in the management of pain for opioid-naïve patients due to high morphine-equivalent doses. Fentanyl should only be used in the case of cancer pain when all other options have been exhausted.

Be aware that local heat (eg hydrotherapy pool) may increase absorption from the patch.

Hydromorphone

Hydromorphone is an effective strong opioid acting as a mu receptor agonist. It is approximately five times as potent as morphine and provides slightly better clinical analgesia than morphine, but has similar adverse effects.^335,336 The main metabolite of hydromorphone is hydromorphone-3-glucuronide (H3G), which is dependent on the kidneys for excretion, has no analgesic action and can lead to dose-dependent neurotoxic effects.³³⁷

It is available as solution for injection, oral liquid and tablets. It also has extremely high potential for misuse and high street value for those who divert this drug.

In practice

Hydromorphone is PBS listed for severe disabling pain, but in practice is usually restricted to malignant pain, or patients undergoing dialysis. It is not suitable to be used as the initial agent in the management of pain for opioidnaïve patients.

Methadone

Methadone is a synthetic opioid acting as an agonist at the mu receptor with additional ketamine-like antagonism at the N-methyl-D-aspartate receptor. It is commonly used for the maintenance treatment of patients with an addiction to opioids and in patients with chronic pain.

It has good oral bioavailability (70–80%), high potency, a long duration of action and no active metabolites.³³⁸ But it also has a long and unpredictable half-life (mean of 22 hours; range 4–190 hours), which increases the risk of accumulation.³³⁹

Concurrent administration of other drugs that are metabolised by the P450 enzyme system may have significant effects. P450 inducers (eg carbamazepine, rifampicin, phenytoin, St John’s wort (Hypericum perforatum), some antiretroviral agents) may increase methadone metabolism, which lowers methadone blood levels and leads to potential reduced efficacy or even withdrawal.³⁴⁰ Use of P450 inhibitors (eg other antiretroviral agents, some SSRIs, grapefruit juice, antifungal agents) may lead to raised methadone levels, which increases risk of adverse effects or overdose.³⁴⁰ Checking for drug interactions with methadone can be done online.

In practice

Methadone is PBS listed for severe disabling pain and for ORT (as S100). Two formulations are available in Australia. Methadone liquid is used once daily for maintenance in opioid dependent patients. Methadone tablets may be used two to four times daily to manage persistent pain.³⁴¹

Methadone use is usually confined to specialist pain medicine areas³⁴² as it has complicated and unpredictable pharmacokinetics. Extreme caution must be taken when inducting a person onto an appropriate dose of methadone, with a slow titration regimen and close monitoring required. It may take up to two weeks to reach steady state levels, and drug accumulation may cause excessive sedation and high risk of overdose and death if the dose is increased rapidly.³⁴¹

Morphine

Morphine has been the most widely used opioid in acute, persistent and cancer pain, and remains the standard against which other opioids are compared.

The main metabolites of morphine (primarily formed by hepatic glucuronidation) are morphine-6-glucuronide (M6G) and morphine-3-glucuronide (M3G). M6G is a mu opioid receptor agonist and is the main mediator of analgesia.³⁴³ M3G has very low affinity for opioid receptors and no analgesic activity, but may be responsible for the neurotoxic symptoms such as hyperalgesia, allodynia and myoclonus, sometimes associated with high doses of morphine.³²⁵ Both metabolites are renally eliminated.

Higher doses, older age, impaired renal function and the oral administration (due to first-pass metabolism) are associated with higher M3G and M6G concentrations and therefore with the potential risk of severe long-lasting sedation and respiratory depression.^344,345

While the clinical significance is uncertain, morphine is the most immunosuppressive of the currently available opioids.^346,347

There has been a decrease in morphine prescribing in Australia.³³⁴ Prescriptions are most prevalent among older Australians.

Musculoskeletal pain

The evidence for morphine in managing CNCP, including low back pain, is poor.³⁰³

Neuropathic pain

Strong opioids including morphine have weak GRADE recommendations for use and are recommended as third line mainly because of safety concerns.⁹⁶

In practice

Morphine formulations are indicated by the PBS for severe disabling pain (cancer, palliative care) and chronic severe pain. Commencement doses vary according to patient selection and age.

Oxycodone

Oxycodone action appears to be mediated primarily by mu receptor agonism. Oxycodone contributes the majority of drug effect, as its metabolites, noroxycodone and oxymorphone (via CYP3A4), are only weakly active. However, oxycodone concentration may be dependent on CYP2D6 activity, resulting in ultrarapid metabolisers experiencing better analgesic effects than poor metabolisers, but also higher toxicity.^348,349

Paradoxically, in acute postoperative pain, the CYP2D6 genotype does not appear to influence oxycodone requirements.³⁵⁰ There is an increasing use of oxycodone in the acute, hospital and perioperative settings as it has a faster onset of action than morphine, better oral bioavailability, longer duration of action, fewer concerns about metabolites and lower rate of adverse effects based on these pharmacological properties.^350–352

Oxycodone-related deaths are currently relatively low in Australia; they are not comparable to numbers reported in the US.³³⁴

Musculoskeletal pain

The evidence for oxycodone in the management of CNCP is poor.³⁰³

Neuropathic pain

Strong opioids including oxycodone have weak GRADE recommendations for use and are recommended as third line mainly because of safety concerns.⁹⁶

In practice

Oxycodone is PBS listed for severe disabling pain and chronic severe pain. It is particularly popular in hospital and acute pain settings. Care should be used in rehabilitation settings to minimise chronic use.

Care should also be taken by GPs continuing to prescribe oxycodone in the community post discharge from the hospital setting. All patients should have plans to be weaned off their opioid analgesics post discharge.

The use of oxycodone is increasing rapidly and addiction specialists report that it is often a drug of choice for misuse. A combination of oxycodone with naloxone has recently been released in Australia. This combination substantially reduces the chance of constipation,³⁵³ but the risks of misuse and diversion still exist.

Note that St John’s wort (H. perforatum) induces metabolism of oxycodone, significantly reducing its plasma concentrations and efficacy.³⁵⁴

Pethidine

Pethidine is a synthetic opioid active at the mu receptor. IM pethidine has been widely used in Australia for a range of pain problems. Its use is decreasing because of multiple disadvantages compared to other opioids. Repeated dosing or renal failure leads to accumulation of its active metabolite (norpethidine), which is associated with neuroexcitatory effects that range from nervousness to tremors, twitches, multifocal myoclonus and seizures.³⁵⁵

When used parenterally, pethidine does not provide better analgesia than morphine, but does induce more nausea and vomiting than morphine.³⁵⁶

In practice

Use of pethidine is discouraged in favour of other opioids.^357,358

It has high addiction potential and is not recommended for the treatment of persistent pain.

Pethidine is no longer indicated for the treatment of migraines.

Tapentadol

Tapentadol is a combined weak mu agonist and noradrenaline reuptake inhibitor (acting on descending pain inhibition pathways) with no active metabolites.^359–361 In a number of chronic pain conditions, tapentadol shows efficacy that is comparable or better than conventional opioids but with reduced rates of gastrointestinal adverse effects (eg nausea, vomiting, constipation), which results in less treatment discontinuation.³⁶²

At doses up to the maximum recommended 500 mg/day, tapentadol has no effect on heart rate or blood pressure due to noradrenaline reuptake inhibition, even in patients with hypertension and/or on antihypertensives.³⁶³ However, as it is metabolised by the liver, impaired hepatic function may require dose adjustment.³⁶⁴

Despite widespread use over several years in the US and Europe, there are only two reported cases of an overdose death.³⁶⁵ Although it is a controlled medicine in all countries, tapentadol shows a lower rate of misuse and diversion than oxycodone and hydrocodone and a rate comparable to tramadol.^366,367 There are limited data to support a role for tapentadol in cancer pain.³⁶⁸

Musculoskeletal pain

Currently, relatively few RCTs have studied tapentadol. There is evidence of benefit in osteoarthritis, low back pain and postoperative pain.^{305,369–371} Three randomised trials studying tapentadol for managing chronic pain of osteoarthritis and low back found that 32% of patients received greater than 50% pain relief.³⁰³

Neuropathic pain

Due to effect of noradrenaline uptake inhibition on descending pathways of pain, tapentadol modulates increased conditioned pain seen with neuropathic pain.³⁷² This effect has been confirmed in diabetic neuropathy.³⁶⁰

In practice

Tapentadol is PBS listed for chronic severe pain.

Start at low dose 50 mg and titrate the dose according to response increase: every three days, increase the dose by 50 mg for each twice-daily dose until adequate analgesia or the 50 mg OME dose of 125 mg/day is reached.

Tramadol

Tramadol acts as both a weak opioid agonist and as a serotonin and noradrenaline reuptake inhibitor. Due to the combined effects, it is commonly referred to as an atypical centrally acting analgesic.^361,373

Tramadol is metabolised by CYP2D6 to an active metabolite, O-desmethyltramadol (M1), which is a more potent mu opioid receptor agonist than the parent drug.³⁷⁴ Hence, patients who are poor metabolisers receive less analgesic effect from tramadol.³⁷⁵

The adverse-effect profile of tramadol is different from other opioids. The most common side effects are nausea and vomiting, which occur at rates similar to morphine.^376,377 However, tramadol has less effect on gastrointestinal motor function than morphine.^377,378 It causes less respiratory depression than other opioids at equianalgesic doses.^379,380 Tramadol does not increase the incidence of seizures compared with other analgesic agents,^381,382 although there is a risk of inducing serotonin toxicity when tramadol is combined with other serotonergic medicines, in particular SSRIs.³⁸³

Tramadol has a lower potential for misuse than conventional opioids.³⁸⁴

Musculoskeletal pain

There is fair evidence for tramadol in managing osteoarthritis.³⁰³

Neuropathic pain

Tramadol has a weak GRADE recommendation for use in neuropathic pain,⁹⁶ and is regarded as generally second line because due to tolerability and safety.^96,385

In practice

Tramadol is listed on the PBS for acute or chronic pain not responding to aspirin and/or paracetamol; short-term treatment of acute pain.

Side effects often limit use, but tramadol can be useful if tolerated.

Formulations

The practical usefulness of opioids is related to the available formulations (Table 10).

Approximate equivalence doses

Oral morphine is the standard that other opioids are measured against. Full opioid agonists given in equianalgesic doses produce the same analgesic effect.³⁸⁷ However, accurate determination of equianalgesic doses is difficult due to individual variability in pharmacokinetics and dynamics.¹⁴⁷

There are several published tables providing approximate equianalgesic doses. These are typically based on singledose studies in opioid-naïve subjects and may not be as relevant when conversions are made after repeated doses of an opioid.³³⁷ They also do not take into account incomplete cross-tolerance and patient-specific factors.³⁴² 

Converting to methadone requires special caution. Regardless of how much other opioid the patient is being prescribed, commence methadone at low doses in accordance with the National guidelines for medicationassisted treatment for opioid dependence  or in consultation with pain or addiction specialists.

Commencing and increasing dosage

Starting doses are a guide only and may vary according to the clinical situation, condition of the patient and previous analgesic requirements. For example, older patients generally require lower opioid doses.³⁸⁸

At each review, assess pain intensity, cardiorespiratory status, level of sedation and other adverse effects. Titrate dose according to response, sedation score (an early indicator of respiratory depression) and respiratory rate. Use small dose increments as the dose required may vary more than 10-fold between patients of similar age, irrespective of weight.³⁸⁸

Adjust the dose of controlled-release (CR) opioids, not the frequency of administration. However, if increasing the dose fails, it may occasionally be appropriate to administer doses more frequently for patients with pain that regularly occurs shortly before the next dose is due.³⁸⁸

Opioid ceiling doses

Use caution when prescribing opioids at any dosage. Many harms are dose related, so aim for the lowest effective dose then carefully reassess for evidence of individual benefits and risks, especially when increasing dosage to 50 mg OME or more per day. GPs must be able to justify a decision to titrate dosage to 100 mg or more OME per day and should avoid increasing dosage to 100 mg or more OME per day without specialist involvement.⁸ Higher opioid doses may be acceptable in cancer-related pain.

Tolerance is a predictable state of adaption in which exposure to a drug induces changes that result in diminution of one or more of the drug’s effects over time.³⁸⁹ The patient become ‘desensitised’ to the drug and increased doses are then needed to get the same effect.

The decrease in the effectiveness of opioid analgesia has traditionally been attributed to opioid tolerance (a desensitisation of anti-nociceptive pathways to opioids). However, it is now known that administration of opioids can also result in opioid-induced hyperalgesia (OIH), which is a sensitisation of pro-nociceptive pathways leading to pain hypersensitivity. Both tolerance and OIH can significantly reduce the analgesic effect of opioids.^390,391

The predictable and physiological decrease in the effect of a drug over time may be referred to as ‘pharmacological tolerance’. ‘Apparent tolerance’ occurs when both tolerance and OIH contribute to a decrease in the effectiveness of opioids.^392,393

There is some evidence that administration of ‘commonly used’ dosages of oral opioids does not result in abnormal pain sensitivity.³⁹⁴

In an individual patient displaying decreased effectiveness of opioid therapy, it can be impossible to determine whether tolerance or OIH is causing a reduction in pain control, creating a management dilemma: inadequate pain relief due to tolerance may improve with opioid dose escalation, while improvements in analgesia in the presence of OIH may follow a reduction in opioid dose.³⁹² The only reasonable action in these circumstances is to reduce opioid doses.

Tolerance also occurs to some of the adverse effects of opioids. Rapid tolerance may develop to sedation, cognitive effects, nausea and respiratory depression. However, there is little, if any, change in miosis or constipation.³⁹²

‘Dependence’ has historically been defined in pharmacological terms: a time-limited state that develops during chronic drug treatment in which cessation elicits an abstinence reaction (withdrawal) and is reversed by renewed administration of the drug.¹⁵⁷

Opioid withdrawal syndrome is characterised by signs and symptoms of sympathetic stimulation due to decreased sympathetic antagonism by opioids (Table 12).¹⁵⁷ Symptoms start two to three half-lives after the last dose of opioid. For example, oxycodone has a half-life of 3–4 hours: symptoms would start after 6–12 hours, peak at approximately 48–72 hours, and resolve within 7–14 days.¹⁵⁷ Timelines and symptoms vary depending on the duration of action,¹⁹ specific dose, speed of taper, and duration of use.¹⁵⁷

Withdrawal can be minimised by gradual reduction of opioid use. Where it does occur, unless a patient has significant comorbidity or is otherwise medically unstable, withdrawal is not life threatening, although it may be very distressing.^19,157 Acute withdrawal (when opioids are stopped suddenly, or an antagonist such as naloxone or naltrexone is administered) should be treated by reintroducing opioids or by IV fluids, glucose, and adrenergic-blocking drugs. Clonidine is useful in this situation.¹⁵⁷ Reassurance and comfort measures may also be required.¹⁵⁷

Common opioid-related adverse effects are sedation, pruritus, nausea, vomiting, slowing of gastrointestinal function and urinary retention.^395–397 Uncommonly, opioids (methadone, oxycodone) are associated with prolonged QT-interval with a risk of TdP and cardiac arrest.^398,399 These effects are dose related.