| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the Department of Anesthesiology, Duke University Medical Center, Durham, North Carolina.
Address correspondence and reprint requests to Tong J. Gan, MD, Department of Anesthesiology, Duke University Medical Center, Durham, NC 27710. Address e-mail to GAN00001{at}mc.duke.edu.
Abstract
Adenosine, a ubiquitous metabolic intermediate in the body, is involved in nearly every aspect of cell function, including neuromodulation and neurotransmission. Adenosine A1 and A2 receptors are widely distributed in the brain and spinal cord, and are a novel, non-opiate target for pain management. The potential of adenosine as a non-narcotic analgesic in anesthetized patients has been explored in clinical trials, including double-blind studies versus placebo and remifentanil infusion. These studies suggest that, compared to placebo or remifentanil, an intraoperative adenosine infusion stabilizes core hemodynamics and reduces the requirement for anesthesia during surgery. Further, adenosine improves postoperative recovery, as indicated by lower pain scores and less opioid consumption. The safety profile of adenosine has been well characterized based on use of currently approved adenosine products. The most common adverse events associated with its use include flushing, chest discomfort, dyspnea, headache, gastrointestinal discomfort, and lightheadedness. These effects are generally well tolerated and transient. Further studies are warranted to investigate the full potential of adenosine as a non-opioid analgesic in the perioperative setting.
Acute pain is an expected outcome after surgery. Results of one survey indicate that approximately 80% of patients who undergo surgery experience acute pain during the postoperative period (1). Of these, most patients rate their pain as moderate, severe, or extreme. Pain after surgery can impede recovery, increase duration of hospital stay, increase health care costs, and adversely affect patients’ general activity, walking ability, and sleep (2,3). Despite increased focus on pain management and development of pain management guidelines, the under-treatment of postoperative pain remains an important issue in the perioperative setting (1).
Opioids remain the mainstay for postoperative analgesia, especially after major surgery. However, pain is a multifactorial phenomenon that may not be adequately controlled with opioid monotherapy alone (4). Furthermore, opioid use may be associated with dose-related adverse effects such as respiratory depression, nausea, vomiting, urinary retention, itching, and sedation. Opioids can also reduce gastrointestinal motility, contributing to postoperative ileus. When given a choice, most patients prefer non-opioid to opioid analgesics (1).
To improve pain relief and reduce the incidence and severity of side effects, a multimodal approach to postoperative analgesia should be used (5). This involves the use of different classes of analgesics, the incorporation of adjunct analgesics, as well as use of different sites of administration of the analgesics (3,6). Nonsteroidal antiinflammatory drugs (NSAIDs), including selective cyclooxygenase (COX)-2 inhibitors, improve analgesia, and reduce opioid-related side effects when used during the perioperative period (7,8). However, NSAIDs may cause bleeding, renal dysfunction, and cardiovascular adverse events. Although COX-2 specific inhibitors do not increase perioperative bleeding, their cardiorenal side effects may be similar to that of traditional NSAIDs. Hence, there is a need for other adjunctive analgesics in the perioperative period.
Adenosine, one of the most ubiquitous metabolic intermediates in the body, is involved in nearly every aspect of cell function. It is an important neuromodulator and regulates neuronal and non-neuronal cellular function by actions on adenosine receptors (9). Adenosine is also an important signaling molecule in immunity and inflammation, acting as an endogenous antiinflammatory agent (10,11). It also acts as a neurotransmitter and regulates pain transmission in the spinal cord and in the periphery. Adenosine A1 and A2 receptors are widely distributed in the brain and spinal cord, and these receptors are a novel, non-opiate target for pain management.
Two formulations of adenosine are currently on the market for cardiovascular indications (Adenocard® and Adenoscan®). The second formulation (Adenoscan) is produced in 60- and 90-mL vials, which contain sufficient volume for use in the perioperative setting. Adenocard is approved for conversion to sinus rhythm of paroxysmal supraventricular tachycardia, including that associated with accessory bypass tracts (Wolff-Parkinson-White Syndrome) (12), and Adenoscan is indicated as an adjunct to thallium-201 myocardial perfusion scintigraphy in patients who are unable to exercise adequately for diagnosis of coronary artery disease (13). We reviewed the clinical pharmacology of adenosine and its efficacy and safety when used as an adjunct in postoperative pain management.
PRECLINICAL PHARMACOLOGY
Early rodent studies suggest that adenosine and adenosine analogues have antinociceptive properties after subcutaneous or intrathecal (IT) administration (14,15). Studies using adenosine analogues with different efficacy for the A1 and A2 adenosine receptors have indicated that spinal adenosine A1 receptors are involved in inducing the antinociceptive effects (16,17). The role of the A1 adenosine receptor in inhibiting spinal sensory transmission has been confirmed by the inhibitory effect of A1 analogues on the C-fiber-evoked responses of wind-up and postdischarge of dorsal horn neurons (18). Recordings of evoked potentials from the spinal cord in rats further support these findings (19). Adenosine agonists dose-dependently inhibit the slow ventral root potential, which is the C-fiber-evoked excitatory response associated with nociceptive information. The rank order of agonist potency indicates that adenosine agonists inhibit spinal sensory transmission by acting on A1 receptors (19).
Animal models have been developed to induce nerve injuries that mimic clinical findings observed in patients with neuropathic pain, with hypersensitive reactions to low threshold mechanical stimulation and heat. Bennett and Xie (20) introduced a technique involving four loose ligatures around the sciatic nerve, called the chronic constriction injury model (20). In this model, an adenosine agonist reduced the hypersensitive reaction to heat stimuli at doses that did not affect the normal paw latencies in rats (21). Further, the adenosine agonist R-phenyl-isopropyl adenosine (R-PIA) reduced the scratch behavior induced by chronic constriction injury in the rat, after both IV administration (30 nmol) and IT injection (3 nmol) (22). In rats subjected to chronic constriction injury, a dose-dependent reduction of hypersensitivity to tactile stimulation has been observed after IT R-PIA (1–10 nmol) (23). This effect was abolished by IT injection of the A1 receptor antagonist cyclopentylxanthine. Enhancement of the effect on spinal cord stimulation was observed after concomitant IT injection of a submaximal dose of R-PIA (3 nmol).
IT adenosine induces a dose-dependent reduction of tactile hypersensitivity in rats after spinal nerve ligation (24). Interestingly, a single spinal injection of adenosine reduced hypersensitivity for >24 h. A delay in onset of drug effect was observed, with a maximal effect after 1–2 h.
In a rat model of acute thermal nociception and postoperative hypersensitivity, IT adenosine had no antinociceptive properties to acute thermal stimulation, but enhanced the antinociceptive effects of clonidine. Furthermore, IT adenosine was effective against postoperative hypersensitivity by an adrenergic mechanism (25).
The above indications that spinal administration of adenosine or adenosine agonists with selectivity for the A1 receptor may be of potential analgesic use in humans have prompted neurotoxicological studies. Studies have used laser Doppler flowmetry (26) and [14C]-iodantipyrine autoradiography (27) to test the adenosine analog R-PIA, selective for the A1 receptor, for effects on spinal cord blood flow in rats. In both studies, R-PIA induced a slight, but statistically significant, increase in spinal cord blood flow. A neurotoxicological evaluation using morphologic and morphometric measurements showed that chronic IT administration of R-PIA once daily for 14 days (5–25 nM in two groups) induced no neurotoxic changes in the rat spinal cord (28). In a similar study, IT adenosine (100 µg) administered to rats twice daily for 14 days showed no neurotoxic effects (29).
In a volunteer study involving 12 subjects, IT adenosine (500–2000 µg) reduced, in a non-dose-dependent manner, the areas of secondary allodynia after skin inflammation and reduced forearm tourniquet ischemic pain ratings. Adenosine also prevented the reduction in tactile pain thresholds by mustard oil inflammation. The ice-water-induced cold pain rating was not influenced by adenosine. Apart from transient lumbar pain in one volunteer, IT adenosine injection lacked side effects in healthy volunteers (30).
In another volunteer study, two doses of IT adenosine (0.5 and 2 mg), reduced areas of allodynia and hyperalgesia from capsaicin, with no differences between doses. Headache and back/groin pain occurred more commonly with the 2 mg dose of adenosine. Aminophylline failed to reverse adenosine’s effects (31). Adenosine also interacts with other analgesics. For instance, IV adenosine had an additive analgesic effect with morphine and ketamine in experimentally induced ischemic pain in healthy volunteers (32).
EFFICACY IN THE PERIOPERATIVE SETTING
Published reports of randomized, controlled trials that investigated perioperative adenosine administration for postoperative pain management were searched in Medline (1966–2006), The Cochrane Central Register of Controlled Trials (2006), Scopus, and CINAHL. Free text and MeSH terms "adenosine," "pain," "analgesia," "analgesic," "postoperative," and "surgery" were used for searching. Search was performed without language restriction, but limited to randomized trials in humans. The last electronic search was in November 2006. Additional studies from the bibliographies of reviews or reports were also identified.
Three randomized, double-blind, placebo-controlled parallel-group trials have been conducted to investigate the effects of IV adenosine on anesthetic requirements and postoperative recovery in patients undergoing major surgical procedures with general anesthesia (33–35) (Table 1). Collectively, these studies involved 145 evaluable patients who underwent shoulder surgery, breast surgery, or hysterectomy. These surgical populations represent different pain modalities, including deep somatic/joint-associated pain, cutaneous or subcutaneous pain, and visceral pain. In each study, patients received an IV infusion of 80 µg · kg–1 · min–1 adenosine during surgery. In the study involving hysterectomy, patients also received a reduced dose of IV adenosine (40 µg · kg–1 · min–1) for 3 h postsurgery. The anesthetic technique used in the three studies was similar. The adenosine or placebo infusion was started 10 min after induction of anesthesia. Isoflurane with an end-tidal concentration of 0.5%–2% was used to maintain anesthesia. If an end-tidal concentration of isoflurane of 2% was required for >5 min, IV fentanyl or alfentanil was administered. The mean duration of surgery in patients receiving adenosine ranged between 55 and 80 min in these three studies. Perioperative infusion of adenosine resulted in a statistically significant reduction in the requirement for isoflurane during surgery while maintaining stable hemodynamics. The anesthetic-sparing effect of adenosine was particularly marked in the hysterectomy study, where isoflurane requirement was reduced by 36% (P < 0.002) in patients who received adenosine versus placebo (34). Two of the three studies reported that postoperative opioid use was statistically significantly less in patients who received adenosine during surgery (versus placebo). In one of these studies (the hysterectomy study in which adenosine was administered postoperatively), the patients’ opioid requirement during the first 24 h after surgery was 18% less in patients who received adenosine versus placebo (P < 0.05). In the other study in females undergoing breast surgery, the postoperative 24-h opioid requirement was reduced by 27% in the adenosine group compared with that in the placebo (P < 0.03). The postoperative administration of adenosine in the hysterectomy study did not, therefore, result in an enhanced opioid-sparing effect compared with intraoperative administration only. This opioid-sparing effect of adenosine is therefore similar or less than that achieved by COX-2 inhibitors and NSAIDS, where a 20%–40% opioid-sparing effect has been reported in different studies (41).
|
|
These three studies were performed by the same group and have some limitations. In the study with breast surgery patients, no well defined variables were used to titrate the depth of anesthesia. Furthermore, the dose of intraoperative opioid was not reported, nor was the method of delivery of postoperative analgesia. Also, the authors did not report the triggers to administer postoperative analgesia. These same limitations also apply to the study in patients undergoing shoulder surgery, with the exception that in this study the authors standardized the method of postoperative analgesia (IV or IM morphine). Individual titration of postoperative analgesics is best achieved with a patient-controlled analgesia system, rather than the PRN method, which was used in these three studies. In this latter shoulder study, the authors reported that the administration of adenosine reduced the requirements for isoflurane during surgery. However, the end-tidal concentration of isoflurane was only recorded at five fixed time intervals rather than continuously. Furthermore, if the patient received an intraoperative opioid, the authors assumed that the end-tidal concentration of isoflurane was 2% for the remainder of surgery. The more meaningful outcomes in these three studies were also not documented. For instance, all patients stayed in postanesthesia care unit (PACU) for 3 h. Thus, we do not know if the beneficial effects of adenosine could have resulted in an earlier discharge from PACU. Furthermore, the opioid consumption is reported only for the whole 24 h postoperative period. Therefore, it is difficult to determine whether any opioid-sparing effect lasted for the entire 24 h or was mainly confined to the early postoperative period. Furthermore, these studies did not show a reduction in opioid-related adverse effects including postoperative nausea, vomiting, and sedation. The authors also did not report a difference in the speed of recovery secondary to reduced requirement for inhaled anesthetics. In these three studies, there was no difference in pain scores between the adenosine and placebo groups.
Other trials indicate that IV adenosine is appreciably better than IV remifentanil in providing pain relief and reducing the need for opioid analgesic after surgery under general anesthesia (38,40). These studies involved 71 patients who underwent surgery for a hysterectomy or major orthopedic procedure. The anesthetic technique was similar in both studies and involved induction with propofol and fentanyl and maintenance of anesthesia with desflurane 2%–3% with nitrous oxide 65%. The infusion of the study drug (adenosine or remifentanil) was adjusted to maintain hemodynamic stability according to clearly predefined variables. In one study, patients received a perioperative infusion of adenosine (range: 50–500 µg · kg–1 · min–1) or remifentanil (0.05–0.5 µg · kg–1 · min–1), adjusting the dose as needed to maintain acceptable hemodynamic stability. Infusion rates were also variable in the second study, ranging from 72 to 290 µg · kg–1 · min–1 for adenosine and 0.02 to 0.38 µg · kg–1 · min–1 for remifentanil. In both studies, excellent hemodynamic stability was maintained intraoperatively by both drugs. In one study, the time to complete orientation was significantly faster with adenosine compared with remifentanil (mean ± sd: 6 ± 2 min vs 31 ± 12 min, P < 0.05). In this study, postoperative sedation and nausea were also significantly less in the adenosine group (38). Postoperative pain scores were also significantly lower during the postoperative period in patients who received adenosine (reduced by 60% in PACU in one study) (38), with pronounced and sustained relief lasting up to 48 h. Adenosine-treated patients also required appreciably less opioid analgesics with a reduction of up to 71% in PACU, and up to 45% at 48 h.
These two studies were well conducted, and provided clear descriptions of the protocol used for titrating the study drug and managing postoperative analgesia. The effect of adenosine on postoperative pain in these studies outlasted the duration of the infusion. Similarly, in patients suffering from chronic neuropathic pain, the effect of adenosine lasted significantly longer than the duration of exposure (43). It is therefore speculated that adenosine may affect neuronal mechanisms involved in central hyperexcitability (44), and that such an effect would persist longer than the period of administration of the drug.
The method of drug administration appears to influence the analgesic properties of adenosine in patients who undergo surgical procedures with general anesthesia. In a double-blind, randomized, placebo-controlled study conducted in 40 women who underwent surgery for hysterectomy, IT administration of adenosine (500 µg in 1 mL) before anesthesia did not reduce the requirement for anesthetic during surgery. Further, adenosine had no significant effect on 48-h pain scores or the amount of analgesic consumed during the postoperative period (39).
The anesthetic technique used during surgery may also affect the efficacy of adenosine. In two double-blind, prospective, randomized, placebo-controlled studies conducted in 110 patients who underwent surgery of the upper extremity with brachial plexus block, addition of IV adenosine to local anesthetics had no significant effect on 24-h pain scores, time to first rescue analgesic, or analgesic consumption during the postoperative period. Adenosine significantly prolonged time to first pain sensation in one study, but not in the other (36,37).
SAFETY OF ADENOSINE
Safety of Adenosine as an Analgesic in the Perioperative Setting
In the three placebo-controlled trials that investigated the effects of IV adenosine (80 µg · kg–1 · min–1) on anesthetic requirements and postoperative recovery in 145 patients undergoing major surgical procedures with general anesthesia, no unusual safety events were reported (33–35) (Table 1). In the study involving females who underwent breast surgery, 3 of 36 evaluable adenosine-treated patients reported one or more of the following (33): hematoma; fever; reoperation because of postoperative bleeding caused by ligature insufficiency; and postoperative headache. Two of 36 evaluable patients in the placebo group experienced prolonged drain exudation or allergic reaction to an adhesive bandage complicated with wound infection. In the study involving females undergoing abdominal hysterectomy, 2 of 23 adenosine-treated patients experienced an adverse event consisting of transient atrioventricular (AV) block II or accidental overdose, resulting in a transient decrease in systolic blood pressure (34). One of 20 placebo-treated patients experienced dizziness; another placebo patient experienced respiratory arrest after administration of ketobemidone for severe pain. In the study of patients who underwent shoulder surgery, no adverse events were recorded (35).
In the two trials conducted in 110 patients who underwent surgery of the upper extremity with brachial plexus block and received IV adenosine in addition to local anesthetics, 4 patients (3 adenosine, 1 placebo) experienced an adverse event, consisting of headache (adenosine), faintness (adenosine), palpitation/tightness in chest (adenosine), and itching (placebo) (36,37).
In the study that compared IV adenosine to IV remifentanil in patients who underwent a hysterectomy, 1 of 15 patients in the adenosine group experienced severe bronchospasm shortly after initiation of the infusion. Because of its temporal relationship to the administration of study drug, adenosine was thought to possibly contribute to this event, although "light" anesthesia was also implicated (40).
Safety of Adenosine in Cardiac Indications
In United States regulatory trials of Adenoscan, events most commonly associated with the drug were flushing (44%), chest discomfort (40%), dyspnea or urge to breathe deeply (28%), headache (18%), throat, neck, or jaw discomfort (15%), gastrointestinal discomfort (13%), and lightheadedness/dizziness (12%) (13). Adenoscan is contraindicated for a similar population as reported for Adenocard, including patients with known or suspected bronchoconstrictive or bronchospastic lung disease (e.g., asthma). The package insert for Adenoscan includes warnings for fatal cardiac arrest; life threatening ventricular arrhythmias; myocardial infarction; sinoatrial and AV block; hypo- and hypertension; and bronchoconstriction (13).
In a multicenter study that assessed the safety of Adenoscan (140 µg · kg–1 · min–1) in conjunction with radionuclide imaging, information was collected from 9256 patients during and immediately after adenosine infusion (45). Side effects were generally mild and well tolerated by most patients (81%). The most notable event was severe bronchospasm, reported in seven patients. There were no deaths, and only one case each of myocardial infarction and pulmonary edema. Transient AV block occurred in 8% of the patients and generally resolved without change in the adenosine infusion. There were no sustained episodes of AV block. The safety profile of Adenocard is very similar to that described for Adenoscan.
Several studies have been conducted to investigate the effects of IV adenosine in acute myocardial infarction and reperfusion. In a controlled study that investigated the safety and feasibility of adenosine adjunct to primary percutaneous transluminal coronary angioplasty in patients with acute myocardial infarction, administration of intracoronary adenosine was well tolerated (46). Adverse cardiac events occurred in 5 of 27 (19%) patients in the adenosine group and 13 of 27 (48%) patients in the saline group (P = 0.03). No cases of AV block or worsening chest pain were reported. In the Acute Myocardial Infarction Study of Adenosine trial (AMISTAD I), patients especially those with nonanterior myocardial infarction who received adenosine, showed a trend towards more cardiac adverse events (47). However, in the Attenuation by Adenosine of Cardiac Complications study, transient cardiac effects were observed with both adenosine and placebo at similar frequencies (e.g., hypotension, bradycardia, AV block, recurrent chest pain, and congestive heart failure) (48).
The AMISTAD II trial involved a large cohort of patients (n = 2118) with evolving anterior ST-segment elevation myocardial infarction, who received thrombolysis or primary angioplasty (49). In this trial, patients received a 3-h infusion of adenosine (50 or 70 µg · kg–1 · min–1) or placebo during the procedure. No differences were observed between adenosine and placebo with regard to the primary end-point of the study (occurrence of new congestive heart failure beginning >24 h after randomization, or first rehospitalization for congestive heart failure, or death from any cause within 6 mo). In a subset of patients (n = 243), infarct size was reduced with the high dose of adenosine (70 µg · kg–1 · min–1), a finding that correlated with fewer clinical events. Adenosine was generally well tolerated in this study. The incidence of premature discontinuation due to an adverse event ranged from 4% to 6% in the placebo and 50 µg · kg–1 · min–1 adenosine groups, respectively. Hypotension, the most common adverse event reported, occurred in approximately 18.5% of adenosine-treated patients versus 14% of patients in the placebo group. The incidence of AV block was very low in each treatment group.
In a study conducted in patients undergoing coronary artery bypass surgery, the infusion had to be stopped early in 4 of 15 patients who received a 7-min infusion of adenosine (total dose, 650 µg/kg) due to profound hypotension (systolic arterial blood pressure <70 mm Hg) (50). However, in a small study of patients with severe primary pulmonary hypertension, adenosine (50–100 µg · kg–1 · min–1 titrated up to 400 ng · kg–1 · min–1 as tolerated) had no effect on pulmonary or systemic arterial blood pressures (51). Results of another study showed that the systemic hypotensive response to adenosine infusion (50 µg · kg–1 · min–1 titrated up to a maximally tolerated dose) in patients with primary pulmonary hypertension was mild and not clinically significant when combined with calcium channel blockers (52).
Given the extremely short half-life of adenosine in whole blood (<10 s) (12,13), concerns over managing cardiovascular side effects are minimized. However, it is important to emphasize that the safety of adenosine in the nonsurgical practice may not reflect its safety in the perioperative period. Unlike nonsurgical use, there may be a number of drug-to-drug interactions between adenosine and inhaled anesthetics and opioids that may worsen the hemodynamic effects of adenosine. In addition, with increased bronchotracheal stimulation during the perioperative period, the potential for bronchoconstriction may be higher than in a nonsurgical population. These potential side effects need to be investigated further in the perioperative setting.
We conclude that adenosine appears to demonstrate opioid-sparing, anesthetic-sparing, and analgesic properties and may be used as an adjunct in the intraoperative and postoperative periods. Although studies suggest that adenosine might be a useful adjunct analgesic in the perioperative period, more studies are required to confirm these findings. Furthermore, dose-finding clinical studies are warranted to establish the optimal dose for achieving a balance between efficacy and side effects profile for adenosine use in the perioperative setting.
ACKNOWLEDGMENTS
The authors appreciate the editorial assistance of Joan C. Rivers.
Footnotes
Accepted for publication April 2, 2007.
Editorial assistance supported in part by Xsira Pharmaceuticals, Inc. The authors have received research grants from Xsira Pharmaceuticals, Inc.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
