Opioid-induced respiratory depression

Learning objectives.

By reading this article you should be able to:

• •Explain the cardiorespiratory risks of opioids and the physiological mechanisms by which these drugs disrupt ventilatory control.
• •Define who is at risk for opioid-induced respiratory depression.
• •Discuss opioid-induced respiratory depression with naloxone and the factors that may complicate reversal in some circumstances.
• •List the alternatives to naloxone, so called agnostic respiratory stimulants.

Key points.

• •Opioids remain important in the treatment of moderate-to-severe pain, despite their adverse effects.
• •Risk factors for postoperative opioid-induced respiratory depression include older age, male sex, opioid naivety, sleep-disordered breathing and heart failure.
• •Rapid rescue treatment including cardiorespiratory resuscitation and naloxone is needed in cases of serious opioid-induced respiratory depression.
• •The efficacy of naloxone is complicated by a short duration of action, the occurrence of withdrawal symptoms and inability to reverse non-opioid respiratory depressants.
• •Alternatives to naloxone, such as agnostic respiratory stimulants, may be available in future.

In spite of their many adverse effects, opioids remain important for the management of acute (perioperative) and chronic cancer or non-cancer pain. However, opioids disrupt the generation of normal respiratory rhythm and negatively affect the central and peripheral chemoreflex loops, potentially causing severe and life-threatening respiratory depression. These respiratory effects are predominantly related to the binding of opioids to μ-opioid receptors (MORs) within the respiratory network located in the pons, medulla and possibly within the carotid bodies that house the peripheral chemoreceptors.^1,2 After an opioid overdose, breathing becomes initially irregular, followed by periodic or cyclic breathing, gasping and eventually the full cessation of breathing activity (apnoea). Sometimes, abrupt cessation of breathing may occur without any warning signs. When no prompt rescue is initiated, opioid-induced apnoea will result in asphyxia, a combination of low arterial oxygen and high arterial carbon dioxide tensions caused by the loss of gas exchange in the lungs, which potentially may cause fatal cardiac arrest.

There is a large variability in the development of opioid-induced respiratory depression (OIRD). It is therefore challenging to predict the effect that opioids have on ventilatory control in a specific individual. In this review, we describe the physiology and pharmacology of OIRD, rescue options using naloxone and emerging alternatives to naloxone to mitigate OIRD.

Control of breathing and the influence of opioids

The ventilatory control system has two major components: chemical control, in which the chemical composition of arterial blood dictates the respiratory state, and behavioural control, which adapts our breathing to multiple exogenous and endogenous behavioural components such as eating, drinking, talking or exercising. Whereas chemical control is dependent on the input from central and peripheral chemoreceptors, behavioural control receives input from cortical areas, and peripheral tissues such as peripheral muscles. Chemical control is the dominant control system in patients spontaneously breathing under general anaesthesia and during non-rapid eye movement (REM) sleep. The central chemoreceptors are spread out within the medulla and are responsible for about 70–80% of the ventilatory response to hypercapnia. The peripheral chemoreceptors are located in the carotid bodies, which are located in the neck, in the fork of the carotid arteries, and sense oxygen and are responsible for 20–30% of the hypercapnic ventilatory response (HCVR).³ The hypoxic ventilatory response originates exclusively within the carotid bodies. The HCVR has a predominant but not exclusive central origin.³ During hyperoxia, the carotid bodies are silenced, and the HCVR then originates exclusively at the central chemoreceptors.

The respiratory networks that integrate all the information from chemical and behavioural control and produce the respiratory rhythm are located in the medulla and pons.² Their integrated activity activates the motor pool that controls the diaphragm and intercostal muscles.⁴ Respiratory centres with particular sensitivity to opioids are the pre-Bötzinger complex, located in the medulla, and the Kölliker-Fuse nucleus and the parabrachial nucleus, collectively called the parabrachial complex, located in the pons.^5,6 The pre-Bötzinger complex is essential for respiratory rhythm generation.^5,6 The loss of pre-Bötzinger activity with systemic opioids depends in part on the loss of drive from the pons and chemoreceptive areas.^4,5 Interestingly, at high opioid doses, phrenic motor output is depressed even when the rhythm generator has recovered, suggesting a direct opioid effect in phrenic premotor and motoneurons.⁷ The parabrachial complex contributes excitatory input that sustains upper airway patency.⁸ It further integrates sensory input from central and peripheral chemoreceptors, adjusts ventilation in response to hypercapnia, hypoxia, or both in an effort to maintain normal Pao₂ and Paco₂ in the body, and contains MORs.^7,9,10

Measuring the effects of opioids on ventilatory control

Most studies that quantify the effects of opioids on ventilatory control do so by measuring the HCVR (i.e. the ventilatory response to inhaled carbon dioxide). Compared with baseline ventilation and baseline Pco₂, the HCVR is particularly sensitive to the effects of opioids.¹¹ The HCVR has a so-called hockey-stick shape. With increasing concentrations of inhaled CO₂ an initial flat part of the ventilation-CO₂ curve is followed at the ventilatory recruitment threshold by a linear increase in ventilation. The flat part indicates the area of the curve in which ventilation is insensitive to CO₂, the linear increasing part is the actual HCVR and is described by a slope and an apnoeic threshold: ventilation=slope×(Pco₂–apnoeic threshold) (Fig. 1). Opioids have three effects on the HCVR: (i) a rightward-shift of the curve, (ii) a reduction of the slope of the HCVR, (iii) a slow decrease in baseline ventilation, the horizontal part of the HCVR.¹¹

Fig 1.

Effect of an opioid on the hypercapnic ventilatory response (HCVR). These curves are obtained from a single subject, and show the effects of increasing levels of end-tidal Pco₂ at steady-state before and after a dose of the opioid tapentadol. The black line illustrates the HCVR curve before giving the opioid (t=0). The horizontal part of the curve illustrates resting ventilation without added CO₂. Beyond the ventilatory recruitment threshold the response to added inspired CO₂ shows a linear increase. This part of the curve is commonly analysed by the function Ventilation=Slope×(end-tidal Pco₂–B), where B is the so-called apnoeic threshold and here shown on the x-axis by the red and black dots. The opioid has an effect on the slope of the HCVR and causes a rightward shift of the curve. The HCVR is affected more rapidly in time than baseline ventilation data. Baseline ventilation slowly drops (blue horizontal line). From Hellinga and colleagues with permission from the authors; Open access article published under the terms of the Creative Commons Attribution 4.0 International License. (http://creativecommons.org/licenses/by/4.0/).¹¹

Utility functions

Opioids have therapeutic and toxic effects. The therapeutic effects (e.g. analgesia) are seldom studied in the context of the toxic effects, such as respiratory depression. One possible method to do so is by creation of utility functions. Such functions combine benefit and harm from opioid treatment into one function. In fact, the creation of utility functions is based on the estimation of the probability of analgesia minus the probability of harm, such as respiratory depression.¹² An example of such a function is given in Figure 2 and shows the difference in utilities as a function of the estimated effect-site opioid concentration between morphine and oliceridine. Agonism of surface MORs triggers activation of intracellular G-protein and β-arrestin signalling. Although both are MOR agonists, oliceridine, in contrast to morphine, is biased towards intracellular G-protein signalling with reduced β-arrestin signalling.¹² The G-protein system is predominantly (but not exclusively) associated with analgesia, the β-arrestin system is associated with opioid-related adverse events. The utility function of oliceridine is consistently positive indicating a higher probability of analgesia than respiratory depression. The reverse is true for morphine (and also for drugs like fentanyl). The only opioid that has a similar utility function compared with oliceridine is the experimental opioid R-dihydroetophine.

Fig 2.

Utility function of two opioids: oliceridine and morphine. Data are plus or minus 95% confidence interval and depict the probability of analgesia minus the probability of respiratory depression. From Dahan and colleagues, with permission.¹²

Who is at risk for OIRD?

Various studies have examined the effect of multiple predictors of OIRD. The PRODIGY (Prediction of Opioid-induced respiratory Depression in patients monitored by capnoGraphY) trial studied the occurrence of OIRD in postoperative patients on potent opioids.¹³ The trial studied >1300 patients in three continents between April 2017 and April 2018 and measured continuous capnography, oxygen saturation, ventilatory frequency and heart rate in patients in the postoperative surgical ward.¹³ The incidence of at least one OIRD event was 46%. A risk prediction tool was developed that showed that five independent patient-related variables were associated with a high likelihood of OIRD: age ≥60 yrs, male sex, opioid naivety, sleep disorders and the presence of chronic heart failure.¹³ Patients in the high-risk group had a significantly higher risk of respiratory depression than patients in the low-risk group with odds ratio of 6 (Table 1). The PRODIGY risk tool is important as it may be used to guide monitoring and treatment of postoperative patients at higher likelihood of OIRD.

Table 1.

PRODIGY risk score for the risk of postoperative respiratory depression. From Khanna and colleagues with permission.¹³ The STOP-BANG acronym stands for snoring history, tired during the day, observed stop of breathing while sleeping, high blood pressure, body mass index >35 kg m⁻², age >50 yrs, neck circumference >40 cm and male sex. CHF, chronic heart failure; SDB, sleep-disordered breathing.

Risk factor	Scoring criteria	Points
Age (yrs)	<60	0
	60–69	8
	70–79	12
	≥80	16
Sex	Male	8
	Female	0
Previous opioid use	Opioid naive	3
	Previous opioid use	0
Sleep-disordered breathing	Known SDB or high STOP-BANG score	5
Chronic heart failure	No SDB or normal STOP-BANG score	0
	Coexisting CHF	7
	No known CHF	0
		Total score = ….
PRODIGY risk score	Low risk: <8 points
	Moderate risk; 8–14 points
	High risk: ≥15

We suggest that the predictors derived from the PRODIGY trial serve as indicators of potential risk, though all patients are at risk for OIRD, including women. Apart from patient-related factors, OIRD may be associated with the doses of opioid given. In addition, we need to be aware that a significant proportion of (surgical) patients take opioids chronically, so the question arises as to how chronic opioid use affects OIRD.

Does opioid tolerance affect OIRD?

Chronic opioid use leads to analgesic tolerance, where higher doses of opioid are required to achieve a similar response. In addition, opioid tolerance affects OIRD, where chronic opioid users have a four-fold decrease in opioid sensitivity to respiratory depression compared with opioid-naive individuals.¹⁴ This may suggest that tolerance to OIRD protects the opioid user. However, this is not true for various reasons: (i) animal studies suggest that OIRD tolerance develops more slowly to analgesic tolerance; (ii) irrespective of its development, individuals with opioid tolerance will use higher opioid doses, thus increasing the likelihood of an overdose; (iii) opioid tolerance is a dynamic phenomenon and tolerance may decrease during a period of abstinence.^15,16

How does opioid affinity for the mu-opioid receptor affect OIRD?

The effect that opioids have on the ventilatory control system increases in complexity when opioids are used that have a high affinity for the opioid receptor. Some authors have suggested that we can separate opioids into three groups according to their affinity for the MOR.¹⁷ Affinity is defined by the affinity constant Ki. The greater the value of Ki, the lower the affinity, and vice versa. In general, opioids with a higher affinity display greater respiratory depression and have a greater resistance to displacement from the MOR by opioid receptor antagonists, such as naloxone. Opioids with a low affinity (i.e. with a high affinity constant), include codeine and oxycodone (Ki>25), opioids with an intermediate affinity include methadone, fentanyl and hydromorphone (Ki 1–5) and opioids with a high MOR affinity include buprenorphine, sufentanil and carfentanil (Ki<1).¹⁸ Although the dose is clearly the factor that most determines the effect on ventilation, the affinity of opioid for the receptor determines the ease at which the opioids may be displaced from the MOR in case of a serious respiratory effect.¹⁸ The higher the affinity, the more difficult it is to reverse OIRD with naloxone, the most important treatment of OIRD (see below). However, note that when the opioid dose is low, as is the case in postoperative care, naloxone will be able to reverse the opioid effect, even in case of high-affinity opioids (see next paragraph). In addition, the concept of affinity is useful when designing treatment for patients with an opioid use disorder (OUD). For example, buprenorphine, an opioid with a high MOR affinity, but modest respiratory depressant effect because it is a partial agonistic at the MOR, is a viable therapeutic option in patients with an OUD. As buprenorphine occupies the MOR, opioids with a lesser affinity, such as fentanyl, will have a much lesser effect when given to patients receiving long-term treatment with buprenorphine.¹⁹ This relates to all of the effects of fentanyl including analgesia and respiratory depression. Figure 3 shows that a high concentration of buprenorphine at the effect-site (brainstem) causes the majority of MORs to be bound by buprenorphine. As its affinity for the MOR is high, much higher than that of fentanyl, fentanyl is unable to displace buprenorphine from the MOR. As a consequence, there is little effect of fentanyl on respiration (compare panel E without buprenorphine, with panel F, with buprenorphine).

Fig 3.

Effect of high-affinity opioid buprenorphine on ventilatory depression induced by multiple doses of fentanyl. (A) and (B). Plasma concentrations of fentanyl and buprenorphine. (C) and (D). Fentanyl and buprenorphine mu-opioid receptor occupancy. (E) and (F). Isohypercapnic ventilation. These are simulated data, showing plasma concentrations of fentanyl and buprenorphine resulting from four consecutive doses of fentanyl (0.25, 0.35, 0.50 and 0.75 mg per 70 kg). From Olofsen and colleagues with permission of the authors; open access article published under the terms of the Creative Commons Attribution 4.0 International License. http://creativecommons.org/licenses/by/4.0/).¹⁹

Rescue from OIRD

Currently, naloxone is the first-line treatment in the pharmacological reversal of opioid toxicity. Naloxone is a non-selective opioid antagonist and rapidly crosses the blood–brain barrier. It produces rapid reversal of OIRD by displacing the opioid from the MOR. Naloxone works quite well in perioperative care as the concentration of opioid molecules at the receptor in perioperative patients is commonly just above the threshold for respiratory depression and relatively low naloxone doses (incremental doses up to 400 μg) are adequate to restore respiratory activity, often without compromising analgesia.¹⁸ The situation differs in cases of (massive) overdose with potent opioids in the community setting. In such cases one has to consider the short duration of action of naloxone (it has an elimination half-life of 32 min) and respiratory depression may reoccur (renarcotisation).¹⁸ Modelling data indicate that the incidence of failure of naloxone rescue varies with the naloxone dose, but may be as high as 20–30% after 1.6 mg fentanyl i.v. and 2 mg naloxone i.m.²⁰ At even higher doses of fentanyl, the failure rate with naloxone increases to 30–50%. Other limitations of naloxone include provocation of withdrawal symptoms; inability to reverse high affinity opioids; and respiratory depression in cases of co-intoxication.

Withdrawal

At high doses naloxone can trigger acute withdrawal in individuals taking opioids regularly or those with an OUD.¹⁸ The symptoms typically observed are tachycardia, hypertension, abdominal pain, malaise, yawning, agitation or anxiety. In extremely rare cases, naloxone may cause pulmonary oedema, seizures or cardiac arrest. These effects stem from the rapid systemic release of catecholamines induced by naloxone.

High-affinity opioids

Another limitation lies in the fact that naloxone faces challenges in effectively reversing the effects of highly potent opioids, such as carfentanil, or fentanyl and sufentanil at high doses. Because of their high affinity at the MOR, naloxone has a limited ability to displace these opioids from the receptor.¹⁸ High doses of naloxone, repeated doses or more concentrated formulations may be needed in these settings, all aimed at increasing the naloxone concentration at the MOR.

Co-intoxication

The efficacy of naloxone is even further compromised when opioids are given with non-opioid respiratory depressants such as benzodiazepines (e.g. etizolam), antidepressants, gabapentinoids, alcohol or α₂-adrenergic drugs (e.g. xylazine).^18,21, 22, 23 Although naloxone may reverse (part of) the opioid component of the respiratory depression, respiratory depression may be maintained because of the other often potent respiratory depressants.

Prevention of cardiac arrest and brain damage

Severe respiratory depression may lead to hypoxia, asphyxia, cardiac dysrhythmias and eventually cardiac arrest, particularly when no rescue is initiated. As discussed above, rescue includes cardiopulmonary resuscitation and rapid reversal with naloxone via the intranasal, intramuscular or intravenous route. The primary goal of restoration of respiration after an opioid overdose is preventing cardiac arrest. Recent simulation studies showed that the success of an intranasal naloxone intervention is dependent on receptor kinetics, opioid and naloxone dose and naloxone concentration.²⁰ Particularly overdoses with opioids with a high MOR affinity, such as carfentanil, are difficult to reverse with (intranasal) naloxone and may lead to a high likelihood of cardiac arrest (>50%).²⁰ Finally, also the timing of the naloxone dose is important, with a diminishing success rate when treatment with naloxone is delayed.

New developments

Longer-acting naloxone-like drugs

There are currently several longer-acting drugs that antagonise the opioid receptor such as nalmefene and methocinnamox. These drugs are being tested for treatment of OUD and respiratory depression in the context of OUD. However, these drugs have limited value in the perioperative setting.

Agnostic respiratory stimulants

Given the limitations of naloxone and the increasing use of non-opioid respiratory depressants, there is the need for respiratory stimulants that restore respiratory activity independent of the underlying cause of respiratory depression. There are several new drugs under development with different modes of action, all aimed at full restoration of breathing activity despite the presence of potent respiratory depressants.²⁴

Drugs that do show clinically promising results to date and exert their effect through actions at the carotid bodies include the background potassium channel blockers doxapram and ENA-001.²⁴ Although doxapram is currently available for clinical use, its role in reversal of OIRD seems limited; it is more often used to treat apnoea in preterm infants. ENA-001 mimics the effects of hypoxia at the carotid bodies and consequently stimulates breathing and enhances the ventilatory response to hypoxia. So far, ENA-001 showed promising results in reversal of alfentanil-induced respiratory depression and propofol-induced blunting of the hypoxic ventilatory response. Other promising agnostic respiratory stimulants include the N-methyl-d-aspartate (NMDA) receptor antagonist ketamine and ampakines (e.g. CX717) acting at the α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptor (AMPAR).²⁴ In contrast to ENA-001, these drugs act within the respiratory network of the brainstem and cause excitation of the respiratory rhythm generator via their respective receptor systems. All of these agnostic stimulants are still being investigated and are not currently available for clinical practice.

Conclusions

Opioids play a crucial role in the management of acute and chronic pain, and, unfortunately, are widely abused, often to avoid withdrawal symptoms. Opioids are associated with numerous adverse effects, among which respiratory depression can be life-threatening. Predicting the impact of opioids on ventilatory control is challenging, and various experimental and clinical data have been collected over the years to enhance our understanding of their effects on respiration.

It is essential to differentiate between the various scenarios of OIRD. Perioperatively, overdose incidents are typically limited, and OIRD occurrence is influenced by several factors including age, sex, opioid naivety, sleep disorders or chronic heart failure.¹³ Continuous monitoring of patients in clinical practice can help prevent severe OIRD from escalating into cardiac arrest and death. In case of an OIRD, low doses of naloxone (up to 400 μg) i.v. are usually sufficient to reverse the effects.¹⁸

However, a different scenario arises when dealing with an overdose in patients receiving potent opioids (e.g. fentanyl) at relatively high dose (>50 mg morphine equivalents per day) for chronic pain, or massive opioid overdoses in the community.¹⁸ In the latter setting, various rather complex and still poorly understood factors determine the degree of respiratory depression and the efficacy of rescue. These include the dose and respiratory potency of the opioid, intake of other substances of abuse and the individual's tolerance. Effective rescue in cases of serious OIRD depends on the timing of intervention, the initiation of cardiopulmonary resuscitation, the naloxone dose and concentration, the route that naloxone is given and opioid-related factors such as dose and affinity for the MOR.²⁰ The complexity of these interacting factors contributes to the challenges faced in achieving successful rescue. Although designing opioids with fewer adverse effects is beneficial in clinical settings, addressing the opioid crisis, especially concerning potent opioids such as fentanyl, requires socioeconomic interventions rather than solely relying on medical solutions.

Declaration of interests

AD has received research funding from the Dutch research Council and the US Food and Drug Administration; and consultancy fees from Trevena Inc. (USA), Takeda (Germany) and Enalare Therapeutics Inc. (USA). SCJ declares no conflicts of interest.

MCQs

The associated MCQs (to support CME/CPD activity) will be accessible at www.bjaed.org/cme/home by subscribers to BJA Education.

Biographies

Simone Jansen MD is a PHD student in the Department of Anesthesiology, Leiden University Medical Centre. She works on studies related to the control of breathing.

Professor Albert Dahan MD PhD is an anaesthetist and head of the Anesthesia and Pain Research Unit at Leiden University Medical Centre. His research is focussed on the control of breathing, opioid-induced respiratory depression and reversal of opioid-induced respiratory depression.

Matrix codes: 1A02, 2A06, 3E00