Opioids for chronic pain management in patients with dialysis-dependent kidney failure

Abstract

Chronic pain is highly prevalent among adults treated with maintenance haemodialysis (HD) and has profound negative effects. Over four decades, research has demonstrated that 50–80% of adult patients treated with HD report having pain. Half of patients with HD-dependent kidney failure (HDKF) have chronic moderate-to-severe pain, which is similar to the burden of pain in patients with cancer. However, pain management in patients with HDKF is often ineffective as most patients report that their pain is inadequately treated. Opioid analgesics are prescribed more frequently for patients receiving HD than for individuals in the general population with chronic pain, and are associated with increased morbidity, mortality and health-care resource use. Furthermore, current opioid prescribing patterns are frequently inconsistent with guideline-recommended care. Evidence for the effectiveness of opioids in pain management in general, and in patients with HDKF specifically, is lacking. Nonetheless, long-term opioid therapy has a role in the treatment of some patients when used selectively, carefully and combined with an ongoing assessment of risks and benefits. Here, we provide a comprehensive overview of the use of opioid therapy in patients with HDKF and chronic pain, including a discussion of buprenorphine, which has potential as an analgesic option for patients receiving HD owing to its unique pharmacological properties.

Alleviating pain and suffering is a core ethical obligation of physicians. Unfortunately, the prevalence of chronic pain is high (50–80%) among adults treated with maintenance haemodialysis (HD)^1–6. At least 50% of these patients have chronic moderate-to-severe pain, making the burden of pain among patients receiving HD similar to that among patients with cancer^1–4,7–10. Nephrologists must safely and effectively manage pain in patients with HD-dependent kidney failure (HDKF), which might require the use of opioids.

Pain is often under-recognized and under-treated in patients with kidney failure^6,10–12. Pain management in this population can be quite challenging owing to the presence of multifactorial and often coexisting aetiologies and types of pain, altered perception of pain due to comorbid psychosocial symptoms, as well as lack of provider training, incentives, quality metrics and care coordination. Moreover, treatment options for pain in these patients can be limited. Pharmacological pain management options might have altered efficacy and safety due to differential pharmacokinetics and dialysability of medications in patients with kidney failure treated with HD. For example, the use of NSAIDs can increase the risk of gastrointestinal bleeding and of loss of residual kidney function in these patients, thus warranting caution^13,14; the risk of residual kidney function loss is less relevant to patients with HDKF and anuria, but the potential gastrointestinal complications remain important. Nevertheless, such concerns might lead to an underuse of NSAIDs and over-reliance on medications such as opioids in patients with HDKF. Moreover, evidence regarding the effectiveness of non-pharmacological pain management options is sparse and such treatments might not be widely available or accessible.

Opioids, either derived from the poppy plant, or partially or fully synthesized to mimic the natural compound, are widely prescribed for both acute and chronic pain in short-acting formulations (for example, oxycodone and hydrocodone) or, particularly for chronic pain, in long-ac ting formulations (for example, morphine sustained-action). These medications exert their analgesic effect primarily through agonism of G protein-coupled receptors — mainly the μ-opioid receptor but other opioid receptors (κ-opioid and δ-opioid receptors, and the nociceptin receptor (ORL1)) have important roles in drug pharmacodynamics^15,16.

In this Review, we discuss how and when to use opioid therapy in patients with HDKF. We explore the scope and relative paucity of data supporting the use of opioid therapy and offer practical advice for implementing opioid therapy as safely as possible. Pain management with buprenorphine in patients with HDKF is also highlighted as an emerging, and possibly safer, alternative to traditional opioid therapy. Although we focus on the evidence-based use of opioid therapy in patients with HDKF experiencing pain, pain management should be carefully customized to treat the underlying aetiology of the pain syndrome of the patient, whenever possible. Depending on the source, severity, functional impact and presumed mechanism of pain experienced by the individual (that is, nociceptive pain, neuropathic pain, or mixed), a non-pharmacological or non-opioid therapeutic option may be preferred.

Opioid therapy in patients with HDKF

Understanding the pain that many patients with HDKF experience is important before considering opioid therapy for analgesia. Chronic pain is associated with substantial morbidity and mortality in this population and is often inadequately controlled. Moreover, opioid therapy is frequently used but is associated with a negative impact on health outcomes and resource utilization. Below, we examine these factors in more detail and consider the potential role of opioids in patients with HDKF.

Pain experience and management in HDKF

Pain has profound negative effects in adult patients receiving HD and is associated with up to a 1.5-fold increase in mortality compared with patients with HDKF without chronic pain^17,18. Pain that limits daily activities is associated with premature mortality, functional impairment and poor self-rated health, and has a negative effect on lifestyle and physical activity^19,20. Patients receiving HD with severe pain are more likely to use emergency department services and be hospitalized than those without severe pain^18,21. Pain frequency and intensity also correlate with poor health outcomes^1,17. One potential explanation is that patients with pain are more likely to shorten, skip or not adhere to HD treatments^18,21. Pain also contributes to morbidity (for example, irritability, sleep disturbances, anxiety and depression), which in turn lowers quality of life and satisfaction^17,18,22,22. Pain, depression and anxiety create a vicious cycle — depression or anxiety can intensify the perception of pain, and pain can worsen the symptoms of depression and anxiety^22,23.

Pain among patients receiving HD can be caused by the HD procedure itself (for example, pain due to needle insertion, fluid shifts, cramps or headache), but manifests most often as chronic pain related to the disease aetiology (for example, polycystic kidney disease), complications (for example, bone disease, or uraemic or diabetic neuropathy) or comorbidities (for example, osteoarthritis, vascular disease or diabetes). Among patients treated with HD with chronic pain, musculoskeletal pain is the most prevalent (59%), but multifactorial sources of pain are also highly common and include neuropathy (40%), the HD procedure (35%) and vascular disease (20%)^1,6,8,24. Of note, musculoskeletal pain is a broad term that encapsulates a poorly defined heterogeneous pain phenotype in these patients. To advance the study of pain in patients receiving HD, data about pain aetiology, location, quality, frequency, severity and other associated factors should also be collected and analysed.

Historically, studies have found differences in pain experiences among patients of different sexes and races^26–29. Biological factors seem to drive differences in pain experiences between men and women²⁶, whereas differences in pain experiences among racial groups are often attributed to variation in psychological responses that are associated with environmental stressors (such as anxiety and hypervigilance) and coping strategies^27,29. These hypotheses have not yet been explored in the dialysis population². Nonetheless, sex and race drive disparities in the assessment of pain by treating clinicians. For example, women are more likely to be perceived as exaggerating their pain symptoms and are rated as having less pain than men³⁰. Similarly, Black individuals might be thought to experience less pain than white individuals owing to the false assumption that biological traits, such as skin thickness or sensitivity of nerve endings, differ between these populations³¹. Explicit and implicit clinician biases might explain why both women and individuals from minority ethnic groups are less likely to receive opioids and other pain medications when clinically indicated^32–34. In addition to increasing awareness of individual implicit biases, clinicians should consider engaging in pain treatment simulations that focus on empathy and perspective-taking to reduce disparities in pain assessments^35–37.

Pain management in patients treated with maintenance HD is often ineffective¹⁸ — approximately 75% of patients report that their pain is inadequately treated^6,10. The reasons for this inadequacy vary but include factors related to health-c are providers, such as failure to recognize symptoms, insufficient pain management education or reluctance to prescribe analgesics; nephrology care team providers frequently under-recognize pain symptoms and underestimate pain severity in patients receiving HD^11,12. Moreover, up to 75% of patients who report intense pain, and 35–75% of those who report moderate-to-severe pain, are not prescribed any analgesics^8,10,38. Other important factors that affect pain management relate to patients (for example, the belief that pain is ineviTABLE or fear of medication-related adverse events) and systems (for example, lack of tools for pain assessment and management)^39–42. Notably, 17–50% of patients who are treated with opioids still report moderate-to-severe pain^8,10,38 — pain of some aetiologies might not respond effectively to opioids (for example, neuropathic pain syndromes) and opioid therapy might therefore not always be the appropriate choice.

Prevalence of opioid use and patient outcomes

Considering the burden of pain among patients treated with HD, opioid use is unsurprisingly common in this population. A 2011 study found that 64% of 140,899 patients with HDKF in the USA received opioids⁴³. Although opioid prescriptions over the past several years have plateaued overall, the use of these drugs in patients treated with HD remains more than threefold higher than in the general US population^44–46. An analysis from the Dialysis Outcomes and Practice Patterns Study found that 50% of patients treated with HD received opioids for >12 months³⁸. Moreover, a study based on the US Renal Data System found that ~20% of patients receiving HD had a >90-d ay prescription of opioids each year⁴⁶. The subgroups of patients who were most likely to receive opioids included those of non-Hispanic white ethnicity, female sex, or those with a psychiatric disease; younger age, lower income or older dialysis vintage (that is, a longer time on dialysis) were also associated with higher opioid use^38,45,46.

Several reports have raised concerns about opioid use among patients treated with maintenance HD. Opioid prescriptions are associated with increased morbidity, mortality, and health-care resource utilization. Both short-term and long-term (>90 days) opioid use among these patients is associated with multifold increases in dialysis discontinuation, hospitalization and mortality⁴⁶. Additionally, patients treated with opioids have a greater prevalence of falls, fractures, altered mental status, poor sleep and sexual dysfunction than those who do not receive opioids^42,43. These findings are dose-dependent but are observed even at doses recommended by current practice guidelines⁴³.

Among patients with HDKF, the adverse effects of opioids might be a consequence of decreased drug clearance in the kidney, polypharmacy and drug–drug interactions, as well as comorbid conditions^25,42,47. Additionally, opioid prescribing patterns are frequently inappropriate. More than 25% of patients treated with HD who were prescribed opioids received doses exceeding recommendations from the Centers for Disease Control and Prevention (CDC) (that is, >50 morphine milligram equivalents (MMEs) per day); beyond this dose, the risk of sedation and respiratory depression increases sharply^41,46. Moreover, the choice of opioids for pain management in patients receiving HD is often poorly substantiated²⁵. In a Canadian study of 205 patients with HDKF and pain, almost 50% of those prescribed opioids had pain aetiologies (such as neuropathy) for which the evidence supporting the benefits of opioid use is particularly weak¹⁰.

Role of opioids in patients with HDKF

Leading organizations acknowledge that opioids were excessively promoted and prescribed in the late 1990s and early 2000s^48,49. In addition to laying the foundation for the opioid overdose crisis, opioid-centric pain management does not reliably improve function compared with non-opioid therapies ^50,51. These observations led to widespread calls by pain experts to return to holistic pain management, where improved function and wellbeing are target outcomes, and non-pharmacological modalities, including manual therapies such as chiropractic care, behavioural therapies such as cognitive behavioural therapy, and exercise or movement therapies such as yoga, are the preferred treatments. Focal pain generators (for example, severely arthritic knees) should also be treated, whenever possible.

Non-opioid pharmacotherapies, including topical agents such as lidocaine, acetaminophen (also known as paracetamol), anticonvulsants such as gabapentin, and certain antidepressants such as serotonin–norepinephrine reuptake inhibitors and tricyclic antidepressants, should also be considered and dose-optimized before considering opioid use⁵². Moreover, although NSAIDs have been associated with kidney failure, hypertension, volume overload and hyperkalaemia, they can still be used in patients with HDKF — at the lowest effective doses and for short periods — if patients and providers are willing to accept these risks⁵³; the magnitude of these risks should be carefully assessed in each patient. Of note, topical NSAIDs have minimal systemic exposure (2–3% of oral dose) and can be safely used in patients with kidney failure as long as the maximum approved total body daily dose (for example, 32 g of diclofenac sodium 1% gel) is not exceeded⁵⁴.

Nonetheless, most guidelines suggest that opioid therapy might be considered for persistent disabling pain if non-pharmacological and non-opioid pharmacological options fail to meet the personalized analgesic goals for the patient despite being optimized, and if the potential benefits outweigh the risks^55–57. Guidelines also emphasize that opioids should be administered in combination with other therapies to synergize active pain management strategies, such as exercise, meditation or cognitive behavioural therapy, and passive strategies, such as the use of medications.

Of note, where the pain experienced at the end-of-life merits the use of opioids, the consensus is that dosing should be proportionate to the level of pain and should be neither withheld nor under-prescribed. Patients with HDKF, especially older individuals, might have mortality rates similar to those of patients with cancer but prognosis is often hard to estimate⁵⁸. Health-care providers might be concerned that, by denying opioids, they are unwittingly denying palliation for dying patients.

Starting opioid therapy

Enthusiasm for opioid therapy is tempered by its potential harms and the lack of high-quality evidence supporting its benefits. Nonetheless, opioids are neither a panacea to be embraced nor a poison to be universally avoided. Instead, opioids are potentially helpful but dangerous tools that must be used with great care. Below, we examine the risks and benefits of opioids to facilitate their rational and appropriate use, and we introduce prescribing principles to minimize the risk of harm.

The risk–benefit framework

Short-term opioid therapy for acute severe disabling pain following trauma or surgery is consistent with the standard of care, regardless of whether the patient has HDKF. However, because 3–5% of patients given an opioid prescription go on to use opioids long-term^59,60, the supply of these drugs should be limited to what is medically necessary — more than a 7-day supply is rarely needed⁵⁶. Some evidence supports opioid use for up to 3 months in patients with persistent pain syndromes but, although statistically significant, the benefits were small^50,61.

Long-term opioid therapy (LTOT) for chronic pain management is more controversial than its short-term use⁵¹. Given the many potential risks of LTOT, the decision to prescribe opioids is nuanced and complex. To approach this challenge ethically and responsibly, the prescriber must have a detailed understanding of the potential benefits and harms of LTOT, and should use this risk–benefit framework in the context of shared decision-making to guide clinical care⁶².

Evidence supporting LTOT.

The lack of randomized controlled trials (RCTs) of extended duration (>3 months) that compared opioids with placebo or assessed functional improvement following opioid therapy limits our understanding of the benefits of LTOT⁵¹. A systematic review and meta-analysis of open-label extension studies of RCTs showed benefit of opioid therapy over the course of months to years in the 31% of patients with chronic low back pain, painful diabetic polyneuropathy or osteoarthritis who completed the open label period⁶³; however, this conclusion is markedly limited by the ‘very low’ evidence grade for every outcome of the component trials, as well as the fact that 13 of the 15 trials were industry-sponsored, making generalization difficult. Moreover, another RCT found no statistically significant difference in the efficacy of opioid versus non-opioid therapy in improving function limited by chronic back or hip pain over 12 months⁶⁴. However, that study excluded patients who were already on LTOT, some of whom undoubtedly benefited. Despite the relative absence of evidence, the expectation that LTOT provides a clinically meaningful improvement in pain and function in some patients seems reasonable. Unfortunately, adequately powered studies with larger sample sizes are needed to enable the prediction of which patients might benefit.

Risks of LTOT.

In contrast to the uncertain benefits of LTOT, the risks are well-defined and include misuse (that is, using opioids differently than directed), addiction, overdose and a myriad of potential adverse effects, even when used only as directed. The incidence of opioid misuse, regardless of the presence or absence of harm, is estimated to range from 21% to 29%⁶⁵. Opioid misuse can jeopardize safety, for example, by increasing the risk of adverse events or overdose, so must be considered seriously even in the absence of addiction. Screening for a history of mental illness, and addiction to alcohol or illicit drugs is vital because it predicts future problems with opioids⁶⁶. Unfortunately, clinicians are notoriously poor at predicting who will misuse opioids⁶⁷, so clinical impressions alone are not sufficient to guide care. Several validated tools are available to help clinicians assess the risk of developing aberrant drug-related behaviours from prescribed opioid therapy, such as the Screener and Opioid Assessment for Patients with Pain (SOAPP), which has a sensitivity of up to 90% when combined with the clinical interview⁶⁸. However, these tools can be cumbersome to use in practice, so nephrologists might prefer a more accessible screening tool such as the revised unweighted Opioid Risk Tool (ORT-R)⁶⁹ (BOX 1). All patients are likely to have some underlying risk, thus clear communication and close follow-up are required in all patients initiating LTOT.

Box 1 |. Revised opioid risk tool.

The revised opioid risk tool (ORT-R) is a simple and brief (<1 min) validated instrument used to rapidly screen for the risk of developing opioid use disorder in patients with chronic non-malignant pain when long-term opioid therapy is being considered⁶⁹. The ORT-R should be performed prior to beginning or continuing opioid therapy for pain management; it can be self-administered and reported by the patient, or completed by the health-care provider. A score of ≤2 indicates low risk of future opioid use disorder, whereas a score of ≥3 indicates a high risk.

Mark each box that applies	YES	No
Family history of substance abuse
Alcohol	1	0
Illegal drugs	1	0
Prescription drugs	1	0
Personal history of substance abuse
Alcohol	1	0
Illegal drugs	1	0
Prescription drugs	1	0
Age between 16 and 45 years	1	0
Psychological disease
ADD, OCD, bipolar, schizophrenia	1	0
Depression	1	0
Scoring totals

ADD, attention deficit disorder; OCD, obsessive compulsive disorder.

Table reprinted with permission from ref.⁶⁹, Elsevier.

The risk of overdose in patients receiving LTOT has been widely reported. In 2009, unintentional drug overdoses, which occurred most commonly owing to opioid use, surpassed motor vehicle accidents as the leading cause of accidental death in the USA⁷⁰. Coma and apnoea occur in the final phases of an overdose and clinical monitoring of even mild sedation and respiratory depression is therefore crucial. Other common opioid adverse effects include nausea and vomiting, constipation, pruritus and urinary retention, but these can typically be managed by switching opioids, reducing the dose or using adjunctive medications (for example, laxatives or antihistamines)⁷¹. LTOT also has potential to induce immunosuppression and organ toxicity, and is associated with increased risk of invasive pneumococcal disease and community-acquired pneumonia^72,73, fractures⁷⁴ and opioid-induced endocrinopathies such as hypogonadism⁷⁵. Paradoxically, LTOT can also be associated with an increase in pain, which is related to a poorly understood process of increased pain sensitization called opioid-induced hyperalgesia⁷⁶. Collectively, these findings confirm that opioid use can be associated with considerable morbidity and mortality.

The risk–benefit ratio.

The potential risks and benefits of opioid use in an individual patient are dynamic and can be difficult to assess. Prescribers must take a deliberate and measured approach to starting opioid therapy and frequently reassess appropriateness and safety. Initial therapeutic opioid use should be framed as a trial and not a commitment to ongoing opioid prescribing⁵⁵. Considering the precarious balance of potential harms and benefits, patients should be counselled about these issues in advance for optimal shared decision-making and informed consent. However, if the physician believes that the risks of opioid prescribing clearly outweigh any potential benefit, their goal should be to communicate their clinical judgement to the patient in a patient-centred way and to focus on alternative therapies⁶².

Choosing an opioid

The μ-opioid receptor (encoded by OPMR1) is an essential mediator of the pain relief actions of opioids and OPMR1 single nucleotide polymorphisms can therefore affect opioid metabolism and activity⁷⁷. Consequently, no opioid is universally more effective than another for the treatment of chronic pain. The choice of which opioid to use in patients receiving HD should therefore be based on general potency, expected pharmacokinetics, dialysability, accessibility, response to prior use and other individual considerations (TABLEs 1,2 and 3). The response of individual patients might be difficult to predict, so some trial and error might be required before identifying the best drug.

Table 1 |.

Pharmacological considerations for opioid use in patients with HDKF

Medication (MW in g/mol)	Typical pharmacokinetic properties			Pharmacokinetics in patients with HDKF	Effects of dialysis	Use and dosing information in hDKF
	PB (%)	VD (L/kg)	t1/2 (hr)
Codeine (299.4)	7–25	3–6	3	t1/2 of metabolites substantially greater in patients with HDKF than in healthy controls166,167	Metabolites substantially removed by HD166	Not recommended Risk of death increased in patients with CYP2D6 polymorphism Drug accumulation might result in serious adverse drug events, including severe hypotension, narcolepsy and respiratory depression
Fentanyl (336.5)	85	4	7	Inverse relationship between degree of azotaemia and fentanyl clearance168,169	No significant removal by HD170, including high-efficiency or high-flux HD membranes169	May be used but consider a 25% dose reduction53 No supplemental dose needed with HD
Hydrocodone (299.4)	36	5	IR: 4 ER: 9	t1/2 increased in patients with HDKF compared with healthy controls171	Minimal clearance by HD171	Initiate treatment at a low dose (generally 50% reduction)53 Carefully monitor patients during titration to an effective dose
Hydromorphone (285.3)	8–19	4	IR: 2–3 ER: 8–15	Primary metabolite H3G accumulates in patients with kidney failure172,173	Parent drug and primary metabolite H3G efficiently removed by HD172,174	May be used but consider a 75% dose reduction53 No supplemental dose needed with HD
Meperidine (247.3)	60–80	3–4	3–5	Meperidine and its active metabolite normeperidine accumulate in patients with kidney failure175	HD efficiently removes meperidine and its metabolite normeperidine175,176	Not recommended Meperidine and normeperidine increase the risk of seizures and are highly neurotoxic HD has been used to treat normeperidine neurotoxicity176
Methadone (309.4)	85–90	1–8	7–59	t1/2 of methadone is variable and may be increased in patients with HDKF172 177–179	Methadone and primary metabolite EDDp not significantly removed by HD or PD172 177 178	May be used but consider a 25–50% dose reduction53,180 No supplemental dose needed with HD
Morphine (285.3)	35	5.3	2–3	Morphine and its active metabolites M6G and M3G accumulate substantially in patients with HDKF181	Minimal clearance of morphine by PD181 Average 37% extraction of morphine during HD session170	Not recommended Accumulation of M3G and M6G is associated with serious adverse events including myoclonus, seizures, sedation and respiratory depression Serious adverse drug events related to use of morphine in patients with kidney failure are well documented104,182,183 Use with extreme caution; reduce dose by 75%53,180
Oxycodone (315.4)	45	2.6	IR: 3.2 ER: 4.5	t1/2 of oxycodone increased in patients with HDKF184–186	Oxycodone removed by HDF (54%) and HD (22%); primary metabolite noroxycodone removed by HDF (27%) and HD (17%)185	May be used but consider a 75% dose reduction53 No supplemental dose needed with HD Uncertainty regarding variable individual response and prolonged effects with long-acting formulations Drug concentration might be significantly altered in patients with CYP2D6 polymorphism and patients might be at increased risk of drug interactions with CYP2D6 inducers or inhibitors104,183,187
Tramadol (263.4)	20	2.9	6	t1/2 of tramadol may be increased in patients with HDKF188	Substantial removal by HD	Dose every 12 h53 Reduce dose by 50%188

CYP, cytochrome P450; EDDP, 2- ethylidene-1,5- dimethyl1–3,3- diphenylpyrrolidine; ER, extended release; H3G, hydromorphone-3- glucuronide; HD, haemodialysis; HDF, haemodiafiltration; HDKF, HD- dependent kidney failure; IR, immediate release; M3G, morphine-3- glucuronide; M6G, morphine-6- glucuronide; MW, molecular weight; PB, protein binding; PD, peritoneal dialysis; t_1/2, drug half-life; V_D, volume of distribution.

Table 2 |.

Clinical considerations for the use of specific opioids in patients with kidney failure

Medication	Advantages	Disadvantages	Other considerations
Fentanyl	Generally favourable pharmacokinetic properties in patients with kidney failure53,105,180 Available in multiple formulations, including transdermal	50–100 times more potent than morphine189, thus increases the risk of overdose Variable transdermal absorption; wide range of morphine equivalents listed for each patch strength Multiple potential drug interactions	Transdermal formulation might have metal foil backing that is not compatible with MRI Because of the potentially dangerous disadvantages, should only be used by appropriately trained prescribers
Hydrocodone	Limited data do not suggest substantial accumulation or toxicity	Metabolized in the liver via CYP to hydromorphone Drug concentration might be significantly altered in patients with CYP2D6 polymorphism and patients may be at increased risk of drug interactions with CYP2D6 inducers or inhibitors104,183 187	Semisynthetic opioid derived from codeine Should be used cautiously until more data on drug safety in kidney failure are available53 105 180
Hydromorphone	Potent semisynthetic opioid; well-studied in patients with kidney failure Fewer side effects than morphine Parent drug does not accumulate in patients with kidney failure174	Primary metabolite H3G accumulates in patients with kidney failure and can lead to neuroexcitation (for example, tremor or myoclonus) and cognitive impairment190 Easily dialysed, which might result in opioid withdrawal symptoms, although supplemental dosing is generally not recommended	Short half-life and the need for frequent dosing complicate its use for effective management of chronic pain
Methadone	Generally favourable pharmacokinetic properties in patients with kidney failure53,105,180 No substantial parent drug or metabolite accumulation owing to elimination via faecal route NMDA receptor antagonist and i-opioid receptor agonist; might improve efficacy for neuropathic pain and reduce analgesic tolerance191	Long, variable, unpredictable half-life (serum half-life 15–60 h; up to 120 h in some studies) with an analgesic effect of only 6–8 h; risk of drug accumulation and delayed respiratory depression Risk of dangerous QTc prolongation192 and numerous drug interactions	Inexpensive; no active metabolites; available in small dosage units; tablets may be broken without altering formulation Because of the potentially dangerous disadvantages, should only be used by appropriately trained prescribers
Tramadol	Might enhance central suppression of ascending pain signalling in addition to analgesia from opioid agonism; helpful for both nociceptive and neuropathic pain193 Weak (μ-opioid receptor agonist and SNRI, theoretical benefit for comorbid mood disorder	Tramadol and its metabolite accumulate in kidney failure and increase risk of respiratory depression and seizures139 Increased risk of serotonin syndromea, especially when used concurrently with other serotonergic agents, or CYP2D6 and CYP3A4 inhibitors104,187 Concentrations might be significantly altered in patients with CYP2D6 polymorphism	Limited data in patients with kidney failure but metabolites are predominantly excreted in the kidney and caution is therefore advised53,105,180

CYP, cytochrome P450; H3G, hydromorphone-3- glucuronide, NMDA, N- methyl- d- aspartate; QTc, corrected QT interval; SNRI, serotonin–norepinephrine reuptake inhibitor.

^aSerotonin syndrome consists of a combination of mental status changes, neuromuscular hyperactivity and autonomic hyperactivity caused by overactivation of the serotonin receptor.

Table 3 |.

Buprenorphine formulations used to treat chronic pain

Formulation	FDA-approved indication	Available strengths	Dosing	Conversions	Other considerations
Buprenorphine transdermal patch (Butrans)194	Moderate-to- severe chronic pain	5, 7.5, 10, 15, 20 μg/h	Available as a 7-day patch Opioid-naive: 5 μg/h If switching from full agonist195: 30–80 mg morphine equivalent (10 μg/h) For >80 mg morphine equivalent consider buprenorphine buccal film or sublingual buprenorphine Titration interval: minimum 72 h	Maximum dose 20 μg/h patch, which is equivalent to ~1 mg/day sublingual	Local dermatitis is common at the patch site and might respond to application of a thin film of topical steroid prior to patch application
Buprenorphine buccal film (Belbuca)196	Moderate-to- severe chronic pain	75, 150, 300, 450, 600, 750, 900 μg	Opioid-naive: 75 μg daily or twice daily <30 mg morphine equivalent: 75 μg twice daily 30–89 mg morphine equivalent: 150 μg twice daily 90–160 mg morphine equivalent: 300 μg twice daily >160 mg morphine equivalent: use sublingual buprenorphine Titration interval: minimum 4 days	Buccal-to-sublingual dose ratio: 2:1	A buccal film combination buprenorphinenaloxone product (Bunavail) is FDA-approved for treatment of opioid use disorder
Buprenorphinenaloxone sublingual tablets or films (Suboxone and others)	Opioid use disordera	Suboxone: 2 mg/0.5 mg, 8 mg/2 mg	2–32 mg/day, typically divided doses when used off-label for pain management Individuals with comorbid opioid use disorder typically require higher daily doses (≥12–16 mg) to prevent opioid cravings	No dose adjustments needed when changing between combination and mono product If converting from sublingual tablets to sublingual films, dose reduction might be required as absorption might be increased	Although the DEA allows off-label use for pain, pharmacies and health insurers might have policies restricting its use for pain indication197
Buprenorphine sublingual tablets or films (Subutex and others)	Opioid use disordera	Subutex: 2 mg, 8 mg

DEA, US Drug Enforcement Administration.

^aDEA allows off- label use for pain management¹⁹⁸.

Prescribing principles to reduce risk

The prediction of opioid risk and misuse is inexact, and prescribers must adopt a set of so-called ‘universal precautions’ for all patients on LTOT⁷⁸. Some precautions have already been described above (for example, assessing the risk of misuse and addiction, obtaining informed consent, using opioids on a trial basis and frequent reassessment) but other risk-reduction measures are also available.

First, in accordance with patient-centred clinical practice guidelines, a controlled substance agreement (CSA) should be developed to transparently share safe opioid use and monitoring expectations, and as a tool to document informed consent and shared decision-making^55,79–81. CSAs are commonly and incorrectly referred to as ‘pain contracts’ but this terminology should be avoided because it can be highly stigmatizing and CSAs are not legally binding agreements. CSAs that are written coercively can erode physician–patient trust and imply that failure to agree to its terms will result in loss of access to pain medication^82,83. By contrast, implementing CSAs using non-stigmatizing and understand-able phrasing that is framed in terms of safety might improve the completeness and clarity of documentation in the medical record and help to mitigate risks⁶².

Reducing the risk of opioid-related overdose death is crucial. Importantly, this risk correlates directly with the total daily dose of opioids^84–87 and current guidelines therefore recommend using the lowest effective dose, exercising caution when escalating doses and avoiding high-dose opioids (for example, >90 MMEs) whenever possible⁵⁶. This consideration is particularly important in patients with HDKF who might not be able to metabolize and eliminate the drug effectively (discussed in more detail below). Moreover, when prescribing opioids to an opioid-naive patient without tolerance, extended-release or long-acting formulations should be avoided because their use is associated with an increased risk of overdose⁸⁸ and some formulations of extended-release opioids transform into immediate-release drugs when adulterated⁸⁹; abuse-deterrent opioid formulations should also be used where available. The risk of overdose can increase up to tenfold when opioids are prescribed concurrently with benzodiazepines⁹⁰ and this combination should therefore generally be avoided. Similar to opioids, the potential risks and benefits of benzodiazepines should be scrutinized carefully and the FDA has enhanced the boxed warning associated with this medication class⁹¹. Emerging research has also demonstrated that combining opioids with GABAergic medications (that is, medications that affect the action of γ-aminobutyric acid), such as gabapentin and pregabalin^71,92,93 also increases the risk of overdose, especially in patients with HDKF^94,95. A prudent approach might be to lower the opioid dose before initiating a GABAergic drug or to avoid the combination altogether.

Prescribers should also attempt to minimize the risk of diversion — the unlawful redirection of prescription drugs to the illicit marketplace or the giving, trading or selling of those drugs to an individual for whom they were not prescribed⁹⁶. Prescribers should begin by advising patients that diversion is illegal and dangerous, and must not occur. Prescribers should also prevent prescription forgery by safeguarding prescription pads and e-prescribing credentials, and writing prescriptions in a way that minimizes tampering (for example, by writing the number of pills in words instead of using numbers, which are easier to alter)⁹⁷. We also recommend issuing 28-day supplies rather than 30; a 28-day supply will predictably be refilled on the same day of the week, thus avoiding the need to refill a prescription over the weekend⁹⁸. Furthermore, prescribers should avoid trade-name-only prescriptions. Generic and trade-name opioids typically have equal efficacy but generics usually have a lower street value since they are less recognizable as real opioids⁹⁹. Additionally, prescribers should consider using prescription drug monitoring programs (PDMPs) before initiating opioid therapy and periodically thereafter. Reviewing PDMP data might detect instances in which a patient obtains controlled substances from multiple health-care practitioners without disclosing this information to the prescriber (so-called ‘doctor shopping’); such findings might reflect undisclosed medical procedures that required the use of pain medication, but could also be driven by inadequate pain management, an underlying opioid use disorder (OUD) or an attempt to acquire and subsequently divert opioids for monetary gain. Availability of PDMPs is variable throughout the world, but in the USA nearly every state has its own PDMP and many legally require its use. Prescribers should familiarize themselves with all local, state or provincial regulatory requirements before writing the first opioid prescription. Consistent with FDA recommendations¹⁰⁰, prescribers should counsel patients on how to safely store and dispose of unused opioids before the first prescription is filled.

Pharmacological concerns in patients with HDKF

The effects of kidney failure and dialysis on the disposition (that is, the pharmacokinetics) and response (that is, the pharmacodynamics) of medications must be considered when prescribing opioids to patients with kidney failure¹⁰¹ (TABLE 1; Supplementary TABLE 1). Relevant factors include the effects of kidney failure on hepatic enzymes from the cytochrome P450 (CYP) superfamily, particularly CYP1A2, CYP2C9, CYP2C19, CYP2D6 and CYP3A4^102–104, on drug transporters¹⁰³, and on the renal clearance and dialysability of parent drugs and/or metabolites (FIG. 1). Pharmacokinetic alterations affect systemic drug exposure and the subsequent drug response (for example, the magnitude of the analgesic effect). Accumulation of opioids and opioid metabolites that are normally excreted by the kidneys increases the risk of adverse events, including respiratory inhibition, central nervous system (CNS) depression, seizures, hypotension and, potentially, death¹⁰⁵. Rigorous and consistent data that inform the use of opioids in the setting of kidney failure, especially in patients treated with dialysis (TABLE 2), are lacking¹⁰⁵. Most studies include small sample sizes without long-term follow up and therefore have numerous methodological limitations. Moreover, drug manufacturers offer insufficient evidence-based guidance on dosing in patients with HDKF and recommendations do not account for inter-individual metabolic variability^101,106. Additional research addressing the effects of kidney failure and HD on opioid metabolism and tolerability is needed to ensure that clinicians are neither excluding potentially beneficial pharmacological treatments nor putting patients with HDKF at greater risk of adverse drug events.

Fig. 1 |. Factors contributing to altered pharmacokinetics in patients with hDKF.

Kidney failure and dialysis treatment can impact pharmacokinetics, which might change drug responses. For parent drugs and/or metabolites that are typically eliminated in the kidneys, reduced renal clearance in patients with haemodialysis-dependent kidney failure (HDKF) can potentially lead to drug and/or metabolite accumulation and increased systemic exposure. Hepatic drug metabolism and drug transporter activity are similarly decreased in patients with HDKF, which can additionally impair non-renal clearance^101–103. Although the net effects of HDKF and altered non-renal clearance on opioid disposition are complex, accumulation of drug substrates and increased systemic exposure is likely in these patients. Concurrent use of medications that inhibit or induce the activity of drug-metabolizing enzymes further complicate opioid management in patients with HDKF. CYP, cytochrome P450; OATPs, organic anion-transporting polypeptides; P-gp, P-glycopr otein.

Buprenorphine

Buprenorphine is a fully synthetic opioid that was first developed in the 1960s and is most commonly used to treat OUD but, given its unique pharmacological properties (discussed below), it is now emerging as a viable option to treat pain, especially in patients with medical comorbidities that place them at an above-a verage risk of opioid overdose, including patients with HDKF. We discuss this treatment option in more detail below because although its use is increasing it might be unfamiliar to many clinicians and requires a nuanced approach for effective prescription. Multiple buprenorphine formulations are FDA-approved for either pain or OUD; pain formulations are generally of lower strength (TABLE 3). The use of buprenorphine to treat OUD is beyond the scope of this Review, but nephrologists should communicate and coordinate care with an addiction medicine specialist where appropriate.

General considerations

Here we discuss the general properties and considerations for the use of buprenorphine among all adult patients, irrespective of kidney function, followed by unique considerations for patients with kidney failure. Of note, clinicians should be mindful about racial disparities in buprenorphine prescribing in the USA¹⁰⁷. Individuals from minority ethnic groups are more likely to receive methadone than buprenorphine, possibly owing to racial neighbourhood segregation and local availability of this medication¹⁰⁸.

Utility of partial μ-receptor agonist for the treatment of OUD or pain.

Buprenorphine is a non-selective mixed agonist–antagonist opioid receptor modulator — it functions as a partial opioid agonist on the μ-opioid receptor, and as an opioid antagonist on the κ-opioid and δ-opioid receptors¹⁵. This drug is a first-line treatment for OUD in the general population and has beneficial effects in individuals with acute and chronic pain^109,110. In an observational study in patients with chronic pain who switched from full opioid agonists to buprenorphine, the change to buprenorphine resulted in opioid dose reduction and improvement in pain intensity¹¹¹. Similarly, in two studies, patients who transitioned from high-dos e full opioid agonists to buprenorphine had decreased pain intensity and improved functional status as well as decreased opioid-related adverse effects^111,112. Buprenorphine is often viewed as a safer alternative to full agonist opioids (discussed below)^113,114.

Although partial μ-agonist effects are the most widely studied aspect of the favourable safety profile of buprenorphine, its antagonistic effects on the δ-opioid and κ-opioid receptors might also decrease the risk of constipation and might have anti-depressant and anti-anxiety effects¹⁶. Moreover, the weak agonism of buprenorphine on the ORL1 receptor has been shown to exert analgesic effects on the spinal cord¹⁶.

Reduced risk of overdose and respiratory depression.

Similar to other opioids, buprenorphine use can cause respiratory depression and overdose but, owing to its partial agonism, the risk of respiratory depression is substantially lower than that of full agonist opioids^113,114. Nevertheless, concomitant use of benzodiazepines, alcohol or other CNS depressants accentuates this risk. Current guidelines recommend screening all patients for use of these substances before initiating buprenorphine therapy, educating patients about these risks and tapering prescribed benzodiazepines or other CNS depressants if possible¹¹⁵; abstinence from alcohol is suggested in patients prescribed buprenorphine, in accordance with US National Institute on Alcohol Abuse and Alcoholism guidelines.

QTc prolongation potential.

In vitro studies of many opioids, including buprenorphine, have shown that these drugs block the inward potassium rectifier activity of potassium voltage-gated channel subfamily H member 2 (also known as hERG1), which potentially leads to corrected QT interval (QTc) prolongation and ventricular dysrhythmias, including torsades de pointes. Two randomized, placebo-controlled, dose-escalating clinical studies evaluated QT interval prolongation by transdermal buprenorphine in healthy adult participants and found that, at therapeutic doses, transdermal buprenorphine did not have a significant effect on QTc¹¹⁶. In a large retrospective pharmaco-epidemiological study that examined adverse drug events reported to the FDA from 1969 to 2011 from nearly 200 countries, including the USA¹¹⁷, only 132 cases of ventricular arrhythmia or cardiac arrest and 19 cases of QTc prolongation and/or torsades de pointes were associated with buprenorphine; these events represented 1.8% and 0.3%, respectively, of all adverse events reported for buprenorphine. By contrast, the number of cases and frequency of all adverse event cases associated with methadone, which is a full μ-opioid receptor agonist that is also commonly used for the treatment of OUD and, less commonly, pain, was about tenfold higher.

Precipitated withdrawal risk.

A high risk of precipitated withdrawal exists when an opioid-tolerant individual with recent use of full agonist opioids — either an illicit substance or a prescribed medication — switches to a therapeutic dose of buprenorphine. Precipitated withdrawal is experienced as the abrupt onset of potentially severe opioid withdrawal symptoms (for example, nausea, vomiting, abdominal pain, restlessness, anxiety, sweating, tachycardia, insomnia or tremor) and occurs because buprenorphine has a higher affinity for the μ-receptor than nearly all full opioid agonists and therefore rapidly displaces any full agonists from the μ-receptors¹¹⁸. However, precipitated withdrawal can be avoided in two ways. The traditional method involves discontinuing opioids, awaiting withdrawal symptoms, and then initiating therapeutic doses of buprenorphine once the full agonist has cleared the μ-receptors. Alternatively, the prescriber can start very low-dose buprenorphine while continuing full agonist opioids, then increase the buprenorphine to a therapeutic dose over 4–5 days while gradually discontinuing full opioid agonists; this newer approach is termed the ‘micro-induction’ or ‘overlap’ method and, although it is less well studied than the traditional method, it seems to be very well-tolerated by patients^119,120.

Unique considerations or concerns in HDKF

The pharmacological properties of buprenorphine make it appealing for use in patients treated with HD, although this approach has not been evaluated in any large studies. Buprenorphine and two-thirds of its metabolites are excreted via the gastrointestinal tract, which minimizes its accumulation and toxicity in patients with impaired kidney excretion compared with most other opioids^41,42,121. Only ~10–30% of buprenorphine is excreted in urine¹²². Buprenorphine is a drug with a large volume of distribution and high protein-binding capacity (>90%)¹²², and might therefore not be removed substantially during HD. A small study in ten patients with HDKF and chronic pain receiving buprenorphine supports its wider use, as investigators found that blood concentrations of the parent drug and its metabolites were not significantly affected by dialysis, and pain relief was maintained¹²¹. These findings also suggest that the dose of buprenorphine might not require substantial adjustment in patients receiving HD¹²². However, a case of breakthrough pain associated with decreased free buprenorphine serum concentrations after dialysis in a patient with hypoalbuminaemia who was receiving buprenorphine suggests that some patients with hypoalbuminaemia might require supplemental buprenorphine doses to maintain pain relief¹²³.

When and how to prescribe buprenorphine.

The use of buprenorphine to treat chronic pain in the absence of OUD is rapidly evolving. Among opioid-naive individuals with chronic pain, the CDC guideline recommends considering pain-approved buprenorphine formulations as a third line treatment for chronic pain when non-opioid and non-pharmacological pain treatments are inadequate⁵⁶. For the prescription of buprenorphine for chronic pain, we advise following the guidance from the latest Treatment Improvement Protocol from the Substance Abuse and Mental Health Services Administration¹²⁴; although this resource focuses on OUD, the general prescribing principles are similar.

Of note, sublingual buprenorphine is being increasingly used to help patients without OUD taper down from high-dose long-term full opioid agonists prescribed for chronic pain^112,125. At present, this approach is an off-label use of buprenorphine for pain, but emerging research might lead to its approval by the FDA as a new indication (physiological opioid dependence or inability to taper) for sublingual buprenorphine.

Initiation and stabilization.

Treatment of chronic pain with buprenorphine should begin with a discussion of the risks, benefits and alternatives. An informed consent document signed by the patient is not required but is consistent with patient-centred clinical practice. A urine toxicology screen to confirm clinical history findings and detect the use of other undisclosed substances is also consistent with the general standard of care. If a patient is already using full agonist opioid therapy, precipitated withdrawal can be avoided as previously discussed. For many years, guidelines recommended that buprenorphine initiation should take place in the office to assess opioid withdrawal symptoms but newer guidance recognizes that home initiation is well-tolerated and generally more convenient for patients. A dose adjustment period of 2–4 weeks is typically necessary to maximize the balance of efficacy and safety.

Opioid monitoring and risk reduction

The assessment and mitigation of risks associated with opioid use is not limited to the first prescription, regardless of which opioid is chosen. Rather, all opioid analgesics must be monitored for safety and each refill represents an opportunity to reassess the appropriateness and safety of the care plan. The treating clinician should routinely check toxicology tests to confirm that the opioid is being taken and that non-prescribed or other illicit substances are not being used. The biopsychosocial functioning of the patient should also be assessed and plans for the management of other concurrent problems that might potentiate pain perception, such as mental health conditions, should be outlined and confirmed. The prescriber must remain alert to the emergence of new diagnoses, drug interactions and aberrant drug use behaviours (for example, early refill requests, unsanctioned dose escalations or prescription forgery) that might undermine safety and contraindicate the ongoing use of LTOT, as previously discussed. Below, we consider additional monitoring and risk reduction opportunities for patients already on LTOT.

Drug monitoring and toxicology testing

Urine drug monitoring and toxicology testing provides a means to monitor opioid use objectively and is superior to self-reported drug use history¹²⁶. In addition to confirming adherence, testing might reveal illicit drug use or potential diversion when the prescribed opioid is not detected¹²⁷. Clinicians should partner with patients to implement an opioid monitoring plan transparently before initiating therapy as part of the patient–provider CSA^56,62; early planning helps set expectations and mitigates any perceived stigmatization that might result from scheduled or random drug testing^62,128.

Currently, no established guidelines detail preferred sample collection matrices or timing of drug screening in patients with HDKF. The 2016 CDC Guideline for Prescribing Opioids for Chronic Pain recommends drug screening before starting opioid drug therapy and at least annually thereafter, although increased frequency based on assessed risk for misuse or diversion is also sometimes recommended^56,127. When interpreting toxicology results, clinicians must consider the type of assay used. For example, rapid screening immunoassays typically provide only qualitative or semiquantitative data and are prone to both false-positive results (due to interfering substances) and false-negative results (due to the drug being present in quantities below the reporting threshold). By contrast, gas chromatography coupled with mass spectrometry is a more expensive and slower method, but it can quantitatively identify specific substances when they are present in the body. In addition, physiological factors that affect absorption, metabolism and elimination of opioids and other drugs in patients with HDKF can also confound toxicology testing¹²⁹.

Drug monitoring in patients with anuria.

Drug monitoring presents unique challenges in patients with HDKF. Generally, urine immunoassay testing is the preferred initial test because sample collection is non-invasive, inexpensive, can be performed at the point of care and has a longer window of drug detection than blood or serum testing^126,130. However, urine collection might not be feasible in patients with kidney failure and anuria — sampling of blood, serum or plasma is a viable alternative and avoids the risk of sample adulteration associated with urine screening¹³¹. Serum testing is relatively simple in patients treated with HD and can also be used to investigate individual variations in drug absorption or metabolism in patients with repeated unexpected results^126,130. Serum assays use gas chromatography–tandem mass spectrometry or liquid chromatography–tandem mass spectrometry. Disadvantages include the need for a laboratory with expertise in mass spectrometry specimen analysis, as well as increased cost of specimen processing and a narrow window of detection for drugs with a short half-life compared with urine testing^126,130.

Saliva sampling is also increasingly used for drug monitoring owing to its minimally invasive collection procedure but its use in patients with HDKF has not been tested yet. Saliva testing has a lower risk of sample adulteration than urine but has substantial limitations due to the lack of standardization of collection devices, and samples might be difficult to collect in patients with kidney failure who commonly have xerostomia^132,133.

Polypharmacy in patients with HDKF

Polypharmacy, which is defined as the concomitant use of four or more medications, is common among patients treated with HD owing to multiple comorbidities^134–136. As the number of older patients treated with dialysis has increased, the prevalence and potential for adverse events associated with polypharmacy has also risen¹³⁷. On average, patients with HDKF take 10–12 prescribed and over-the-counter medications and an average of 19 pills per day^134,138. Frequent and thorough medication reconciliation is therefore essential to reduce the risk of polypharmacy-r elated adverse events, particularly when opioids are used given their tenuous risk–benefit profile. Prescribers must also monitor for QTc prolongation when a QT-prolonging medication such as methadone or buprenorphine is used with another QT-prolonging drug. Tramadol might increase the risk of seizures¹³⁹ and prescribers should consider this risk when using other medications that lower the seizure threshold (for example, bupropion). Other drugs might affect the metabolism of various opioids via CYP induction or inhibition. Great care must be taken to screen for potential drug–drug and drug–disease interactions, and consultation with a clinical pharmacist might be beneficial in complex cases. Additional research is needed in patients with HDKF to understand better the prevalence of polypharmacy, to determine its effect on patient outcomes and health-care utilization, and to evaluate the efficacy of mitigation strategies.

Naloxone co- prescription

Clinicians should prescribe the opioid reversal agent naloxone to patients on high-dose opioid therapy (>50 MMEs per day) or who might have other reasons for being at increased risk of overdose (for example, patients with comorbid obstructive sleep apnoea, concomitant use of other sedating medications or prior overdose)⁵⁶. Naloxone is not a controlled substance, requires no special certification to prescribe and can quickly reverse the effects of a life-threatening opioid overdose. Dose adjustments for naloxone are not typically required in patients with HDKF but cases of prolonged opioid antagonism with late relapse of overdose symptoms in patients with HDKF have been reported¹⁴⁰. Patients should be monitored closely whenever naloxone is administered for overdose. To encourage naloxone co-prescribing, many US states have legislation protecting against civil and criminal liability for administering naloxone, and/or require naloxone co-prescribing; in some states pharmacists are also allowed to directly dispense naloxone without a prescription¹⁴¹. However, in 2018 only one naloxone prescription was dispensed for every 69 prescriptions of high-dose opioids in the USA¹⁴² so this risk reduction tool remains woefully underutilized.

Screening and responding to OUD

Unfortunately, 8–12% of patients taking opioids for chronic pain develop OUD⁶⁵, which is defined as a problematic pattern of opioid use that leads to serious impairment or distress¹⁴³. The diagnostic criteria and an assessment of severity can be determined using the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) Diagnostic Criteria for Opioid Use Disorder¹⁴³. Distinguishing between pain relief-seeking and drug-seeking behaviours can be difficult and patients might not be forthcoming about opioid misuse; therefore, health-care providers must proactively screen for opioid misuse and evolving addiction. Self-administered screening tools such as the Prescribed Opioids Difficulties Scale (PODS)¹⁴⁴ can be valuable. Of note, proof that a patient has an addiction is not a requirement for the discontinuation of opioid treatment. Rather, the prescriber must determine whether the benefits of ongoing use no longer outweigh risks for that patient at that time.

Discontinuing opioids might require a difficult conversation between the patient and the prescriber and should be approached with equanimity. Keeping in mind that a patient with an addiction might have a stronger relationship with the drug than with the prescriber can be helpful when speaking with patients who are angry or adopt manipulative behaviours in an attempt to preserve their opioid supply (BOX 2)¹⁴⁵.

Box 2 |. Discontinuing opioids: ‘Do’s and Do nots’.

A patient with chronic pain, opioid use disorder or a combination of the two might have a stronger relationship with their opioid medication than they do with their prescriber. When the decision to discontinue opioids is shared, some patients might become confrontational or manipulative in an attempt to maintain their access to opioids. The following ‘Do’s and Do nots’ offer suggestions to promote patient-centred and non-confrontational communication.

Do

• Judge the medical appropriateness of the treatment.

• Communicate with the patient when opioids are being stopped due to safety or efficacy concerns.

• Explain your decisions in a reasoned and transparent manner.

• Commit to the well-being of the patient and collaborate on the development of a new treatment plan.

• Respond to emotional distress with empathy.

Do not

• Judge the character of the patient or whether they ‘deserve’ opioid therapy.

• Use labels, or accusatory or blaming language (for example, ‘dirty urine’ or ‘drug addict’).

• Debate your decision with the patient or allow a ‘no’ to become a ‘yes’ under pressure.

• Abandon the patient (for example, by discharging the patient from the medical practice) or refuse to treat the patient’s pain with an alternative approach.

• Allow empathy to change your decision on discontinuation.

• Tolerate abusive behaviour.

Copyright © 2016 The Cleveland Clinic Foundation. All rights reserved (rEf.¹⁴⁵).

Management of OUD in patients with HDKF is beyond the scope of this Review but guidelines and resources about the referral and treatment of OUD are available^{56,129,146–148}. Nephrologists should become familiar with the addiction medicine resources in their community and, once a diagnosis of OUD is made, they should leverage these resources and refer patients for addiction care.

Future directions

Evidence for the effectiveness of opioid treatment in pain management in general, and in HDKF patients specifically, is lacking¹⁰⁴. In addition to further study of the safety and efficacy of opioids and other analgesics in this patient population, the effect of kidney failure and HD on opioid metabolism and drug–drug interactions, and alternatives to urine testing for opioid safety monitoring in patients with anuria need to be investigated. Some studies have suggested that long-term opioid use might contribute to further decline in glomerular filtration rate but these findings are controversial^104,149,150 and need to be evaluated further in clinical trials. RCTs that assess effective pain treatment strategies for specific pain syndromes (for example, osteoarthritis or calciphylaxis) in the context of HDKF are also needed, as well as studies that assess the needs of the growing population of older individuals receiving dialysis. Further study is needed to understand the prevalence and impact of OUD among patients with HDKF and to enhance the capacity for referral and care coordination with addiction medicine specialists. Unfortunately, patients with kidney disease are notoriously under-represented in clinical trials^151,152, which has probably contributed to the paucity of data related to opioid use in patients with HDKF. In recognition of the urgency of enhancing diversity of clinical trial participants, including special patient populations and under-represented minority populations (who comprise a large proportion of patients with kidney disease in the USA), the FDA issued a new guidance document in November 2020 that specifically addressed clinical trial eligibility criteria, enrolment strategies and study designs¹⁵³. The FDA also issued an important draft guidance document in September 2020 that provides recommendations for assessing pharmacokinetics and dosing of drugs in patients receiving HD¹⁵⁴. Collectively, the recommendations set forth in these two guidance documents will probably translate to greater inclusion of patients with HDKF in future drug clinical trials, including opioid trials, along with improved standardization of pharmacokinetic study design and reporting of results.

Owing to the potential dangers of opioid use, particularly in the context of an ongoing opioid epidemic, clinicians should have in-depth knowledge of opioid pharmacology and effective treatment strategies. The American Medical Association offers a free and comprehensive opioid training course, and several FDA-approved Risk Evaluation and Mitigation Strategies courses are also available.

Given the variability in pain perception, responses to treatment and adverse events among individuals with HDKF who experience pain, ‘omics’ studies (for example, genomics, microbiomics or metabolomics) are also warranted. As gene sequencing becomes increasingly economically feasible, barriers to its use in clinical research and practice are reduced. Several studies have demonstrated the utility of genomics in understanding pain^155–158; for example, gene expression patterns associated with pain have been identified and linked to neuropathic pain and depression^159,160. Moreover, gut microbes can moderate non-visceral pain through a brain–gut–microbiome axis^161–163 and pharmacogenomic studies can predict drug metabolism^164,165.

Finally, RCTs evaluating combinations of pharmaceutical, psychosocial and behavioural interventions on health-related quality of life are needed to inform evidence-based clinical guidelines. In April 2018, the NIH launched the Helping to End Addiction Long-term (HEAL) initiative — an ambitious effort to accelerate the development of scientific solutions that can address the opioid public health crisis. In addition, in early 2019, the National Institute of Diabetes and Digestive and Kidney Diseases developed the Hemodialysis Opioid Prescribing Effort (HOPE) Consortium to address opioid use within the US End-Stage Renal Disease (ESRD) HD programme. The HOPE Consortium will focus on the prevalence of pain in US patients receiving HD and explore multimodal strategies to reduce long-term opioid prescriptions in this population using a combination of behavioural and medical interventions.

Conclusion

Chronic pain in patients with HDKF is highly prevalent and is associated with substantial morbidity. Unfortunately, management of opioid use, particularly LTOT, in this patient population has not been well studied. Potential risks and benefits for individual patients can be difficult to determine and can change over time. The complexity of opioid metabolism in patients with HDKF challenges nephrologists who are trying to relieve the pain of their patients without doing harm. However, several practical approaches to pain management and safer opioid use, including various risk-reduction tools and emerging therapies such as buprenorphine, can empower health-care providers and can meaningfully and positively affect patient care. With further study, the science of pain management in patients with HDKF might advance to enable personalized, safe and effective analgesia for all.

Key points.

• Patients with kidney failure treated with haemodialysis have a pain burden that is comparable to that of patients with cancer and correlates with poor health outcomes. However, pain management is often ineffective in this population.

• Opioid therapy can be an important pain management tool for patients treated with maintenance haemodialysis but opioids are frequently overprescribed and are associated with increased dialysis discontinuation, hospitalization and mortality.

• Opioid use is associated with many potential risks and their risk–benefit ratio is not always clear. The use of opioids must be personalized, based on shared decision-making and implemented cautiously.

• Kidney failure and haemodialysis can affect the pharmacokinetics and metabolism of certain opioids profoundly, which complicates their use in clinical practice; special prescribing considerations are necessary.

• Buprenorphine has several unique pharmacological properties that make it an appealing analgesic option for patients treated with maintenance haemodialysis but effective prescribing requires a nuanced understanding of its special characteristics.

• Monitoring for harm and the emergence of opioid use disorder in patients who are prescribed opioids is crucial. Screening for and managing opioid use disorder is an important aspect of safe opioid prescribing.

Glossary

Nociceptive pain

A type of pain that occurs due to the activation of nociceptors in the peripheral nervous system by noxious stimuli, including tissue injury, inflammation or disease.

Neuropathic pain

A form of pain caused by a lesion or disease of the somatosensory nervous system.

Morphine milligram equivalents

(MMEs). A dose that is calculated to be equivalent to a morphine dose based on the equianalgesic potency of the opioid relative to morphine.

Abuse-deterrent opioid formulations

Opioid formulations designed to resist or discourage physical or chemical adulteration of the drug (for example, by crushing, smoking, extracting or injecting the drug) while remaining safe and effective when used as intended.

Prescription drug monitoring programs

(PDMPs). An electronic database that tracks and records controlled substance prescribing and dispensing data for patients.

Partial opioid agonist

A drug that binds to and activates opioid receptors, but causes less receptor conformational change and activation than a full opioid agonist.

Opioid antagonist

A drug that competitively binds to an opioid receptor without activating it, thus preventing receptor conformational change, and/or displacing and reversing the effects of a drug that previously activated the receptor.

Corrected QT interval

(QTc). The time from the Q wave to the T wave (Q-T interval) on an electrocardiogram, divided by the square root of the time between successive r waves (r-r interval).

Torsades de pointes

A form of polymorphic ventricular tachycardia that is characterized by a gradual change in the amplitude and twisting of the Qrs complexes around the isoelectric line in an electrocardiogram.

Opioid-tolerant individual

A patient with previous opioid exposure who can safely receive an opioid dose that would be otherwise dangerous without earlier repeated opioid exposure.

Volume of distribution

The volume necessary to contain the total amount of an administered drug at the same concentration measured in plasma.