Chronic Pain Treatment With Opioid Analgesics

Benefits versus Harms of Long-term Therapy

Nalini Sehgal; James Colson; Howard S Smith


Expert Rev Neurother. 2013;13(11):1201-1220. 

In This Article

Opioids & Pain Relief

Opioid efficacy in acute and cancer pain is well established. In CP conditions, randomized clinical trials (RCT) show a statistically significant and clinically meaningful pain relief (30% decrease) with modest doses (<101 mg/day morphine equivalent) of opioids in approximately 30–50% of the participants. Majority of the RCT's are 1–12 weeks in duration. Table 1 shows compilation of RCT's of 12 weeks or greater duration, and shows study duration was 16 weeks in one trial,[33] 15 weeks in four trials[34–36,143] and 12 weeks in nine trials. All of the trials, except one,[37] reported a significant decrease in pain from baseline pain intensity for pain of nociceptive or neuropathic origin. Pain outcomes, derived from these and other short duration trials, form the evidence basis for opioid efficacy in CP. RCTs that evaluate long-term (16 weeks or longer) efficacy of opioids are lacking. No study has shown clinically significant (≥50% pain reduction) and sustained pain relief with opioids over several months and years.

Some studies include an uncontrolled, open-label extension phase, in which participants completing the randomized trial receive the study drug (opioid) for a longer duration. These and other uncontrolled prospective studies (Table 2) reported clinically meaningful and sustained pain relief with opioids in CNCP patients followed for 4–12 months (ranging from 12 weeks to 3 years).[31,34,38–44] Clinically meaningful pain relief is defined as at least 30% reduction in pain, however one study reported 20% average reduction in baseline pain intensity as clinically meaningful pain relief.[40] Opioids decreased pain scores (-2.94 ± 0.27), increased percent pain relief (33.7 ± 14%) and average daytime activity (37.4% increase) in patients with neuropathic pain.[45] Less than 25% (39/174) of CP patients with osteoarthritis (OA), diabetic polyneuropathy (DPN) or low back pain (LBP), who continued to use opioids, experienced sustained opioid analgesia with modest dose escalation over 3 years (mean duration 541.5 days) in an open-label uncontrolled registry study.[31] No other study has prospectively followed CP subjects for several years to assess long-term effects of opioids on pain intensity, function and QoL. Evidence on long term effects of chronic opioid therapy has emerged from population health studies. For example, the Danish health survey of CNCP patients (n = 1906) found that opioid users, as compared to non-opioid users, reported more severe pain, poorer QoL and function and higher health care utilization.[6]

Several systematic reviews (SR) of RCT and uncontrolled trials synthesized evidence on opioid efficacy in CNCP (Table 3). The duration of trials included in these reviews varied from 4 days to 16 weeks. Only one review included trials with at least 6 months duration,[46] but 25 of the 26 studies were either uncontrolled extension trials or case series. Some of the reviews reanalyzed pooled data from multiple RCTs and showed opioids have limited efficacy in reducing pain (Table 3). Two Cochrane reviews found non-significant pain reduction with opioids in chronic low back pain (LBP) patients.[32,47] Opioids decreased pain in osteoarthritis (OA) of hip or knee joint, but the overall benefit was small, corresponding to a difference in pain scores of 0.9 cm on a 10 cm visual analog scale (VAS) or 15% pain reduction.[48] The estimated difference in response rates was 4%, in other words, four more people responded to treatment with opioids than with placebo.[48] The NNT to cause one additional treatment response on OA pain was 8 (95% CI: 7–11).[48] In rheumatoid arthritis, there was no difference in pain outcomes between opioid and placebo groups (risk ratio [RR]: 0.82; 95% CI: 0.34–2.01).[49] A meta-analysis of four RCTs on neuropathic pain estimated opioids decreased mean pain intensity scores by 13 points on a 0–101 VAS, suggesting a 'likely clinically important effect'.[50] Furlan found a medium effect size (ES = 0.58) for pain relief with opioids for both nociceptive and neuropathic pain.[51] A similar effect size (ES = 0.557) for pain was seen with opioids in the elderly (>60 years) with CNCP.[52]

Studies that compared analgesic effect of opioids with other drugs found opioids are not superior to NSAID's, tricyclic or anticonvulsant drugs with respect to pain and disability.[51,53] White et al., do not recommend opioids for chronic LBP treatment due to lack of convincing superiority and a significant rate of side effects.[53] A meta-analysis that pooled data from trials with weak and strong opioids also found no difference in pain outcomes for opioids and NSAID's, but when data for strong opioid was analyzed separately, strong opioid was more effective than 1000 mg/day of naproxen.[51] The same review also reported that opioids (either weak or strong) were not more effective than tricyclic antidepressants or anticonvulsants for pain or function.[51]

Many patients, who initially respond to opioids, abandon treatment over time for various reasons. Dropout rates in clinical trials ranges from 11.9 to 64% (Table 1, Table 2 & Table 3).[34,35,39–41,45,54–60] Furlan estimated an average dropout rate of 35% in the group on opioid drugs where 15% discontinued because of inadequate pain relief, 21% withdrew due to side effects and some dropped for both reasons.[51] Another reason for abandoning treatment is the loss of analgesic effect with prolonged use. Open-label studies show only 44% patients continue with opioids for 7–24 months.[61] The waning of opioid analgesic effect with continued use is possibly related to appearance of analgesic tolerance or opioid-induced hyperalgesia. Physiologic tolerance to opioid analgesics is common, which develops within a few weeks to months of initiating opioid therapy and requires dose escalations to sustain pain relief.[31,39,42] A 3 year registry study showed most dose escalations occurred in the first year: 44% at 3 months, 23% at 4–6 months and 17% at 10–12 months.[31] Pain relief was sustained or minimally worse in 56% patients and significantly worse in 44% patients anytime from 6–36 months. In 8% of the patients multiple dose escalations led to no pain improvement or increasing pain intensity.[31] The evidence suggests that opioid analgesic efficacy is not always sustained during continuous and long-term opioid therapy, even in patients with stable disease and despite dose escalation.[32,46]

Based on this evidence it is inferred that: all patients with CNCP do not respond to opioid analgesics, only 30–50% of carefully screened subjects report decrease in pain with opioids; the results of RCT's cannot be generalized to the CNCP population because clinical trials do not include CNCP patients with multiple pain complaints, anxiety, depression or other psychiatric comorbidities – the type of patient most likely to be on long-term and high dose opioid therapy; opioids do not eliminate pain, initial pain relief is around 30% in most studies; the evidence that clinically significant pain relief is sustained in those who are able to continue with long-term opioid treatment, is weak,[46] there is no controlled study that has evaluated long term efficacy; uncontrolled studies show a high attrition rate over time, either due to declining analgesia or drug related adverse effects[62–64] and opioid analgesics are not superior to NSAIDs, tricyclic antidepressants or anticonvulsants in reducing pain.