Optimizing Pain Management in Spine Patients
Andre M. Samuel, MD
Avani Vaishnav, BA
Catherine Gang, MS
Sheeraz Qureshi, MD, MBA
Prescription drug abuse is now considered an epidemic in the United States with opioid analgesics resulting in more deaths than suicide and motor vehicle accidents combined.1 In total joint arthroplasty, opioid-related adverse events occur in 8% of patients accounting for over half of all measured adverse events.2 The estimated economic burden of opioid abuse in the United States exceeds $50 billion per year, most attributable to criminal justice costs and lost productivity.3 As a result, in 2012, the Joint Commission issued a Sentinel Event Alter regarding the epidemic stating that “an individual, multimodal treatment plan should be used to manage pain” and that “the best approach may be to start with a nonnarcotic.4
Nevertheless, opioid prescriptions remain high, and narcotic prescriptions increased every year from 2006 to 2011 resulting in some states with the equivalent of one opioid prescription per capita.5 Over 5 million people used nonprescription pain relievers in 2010, second only to marijuana use in terms of illicit drug use. And one in six of these patients received this medication as a prescription from a physician. In fact, the number one prescribed prescription drug in the United States in 2011 was the narcotic hydrocodone. Orthopedic surgeons, including spine surgeons, have been shown to prescribe 8% of all narcotic analgesics, although they represent only 2.5% of all U.S. physicians.6
Adequate analgesia remains a necessity and a challenge after spine surgery. In a review of 179 surgical procedures, lumbar spine fusion and complex spinal reconstruction were associated with the highest pain scores on the first day after surgery.7 This is likely due to dissection of the paraspinous muscles such as the deep paramedian transversospinalis and the more superficial and lateral erector spinae resulting in significant pain and spasm. The abundantly innervated periarticular facet tissue, which is also damaged on approach, contributes to postoperative pain as well.
Balancing the need for adequate postoperative analgesia after spine surgery with the narcotic epidemic has resulted in significant interest in the use of multimodal analgesia (MMA) regimens, with synergistic and additive effects that can reduce the required doses of individual agents and potentially lower the prevalence of associated adverse events.
Opioids remain the hallmark of postoperative pain control after spine surgery with evidence of efficacy with scheduled dose regimens, which must be tapered down.8 Grant et al. conducted a retrospective review of 72 pediatric patients undergoing spinal fusion surgery for adolescent idiopathic scoliosis (AIS).9 Postoperative opioid usage was monitored, and the mean number of pills taken in the middle-third of patients (average-use group) was 49 pills. By postoperative week 4, the mean number of pills used by the highest use group was only 2.9. Pain scores also reached preoperative levels by postoperative week 4. After minimally invasive transforaminal lumbar interbody fusion (MIS TLIF), patients on average take 7-8 pills (5 mg) of oxycodone during an average inpatient stay of 2 days.10,11 Similarly, after anterior cervical discectomy and fusion (ACDF), patients take on average 6 pills of oxycodone per day over an average admission of 1 day.12
Long-acting oxycodone, which some have suggested as the precursor to the current opioid epidemic, has also shown effectiveness in postoperative pain after spine surgery. Blumenthal et al. conducted a prospective randomized controlled trial (RCT) of 40 patients undergoing elective lumbar spine discectomy at one or two levels comparing postoperative controlled-release oxycodone or placebo.13 Postoperative patient-controlled intravenous morphine consumption was significantly lower in patients taking controlled-release oxycodone. In addition, pain scores were lower during the first 48 hours postoperatively, and nausea was decreased during the first 24 hours.
However, after lumbar spine fusion, 50% of patients continue taking prescription opioids 3 months postoperatively, 30% continue use 12 months postoperatively, and 17% continue use 24 months postoperatively.14 In opioid-naive patients undergoing elective spine surgery, 26% had continued prescription opioid use at 12 months postoperatively.15
Opioid analgesia also has multiple well-known associated complications, including sedation, respiratory compromise, ileus, and nausea. In both ACDF and MIS TLIF cohorts, postoperative opioid pain regimens were associated with longer inpatient length of stay and increased postoperative nausea and vomiting compared to MMA regimens.10,12 Long-acting oxycodone has similarly been associated with longer inpatient length of stay after MIS TLIF.16
Preexisting opioid use is common in spine surgery patients and has been associated with poor outcomes as well.17 In patients undergoing elective orthopedic surgery, preoperative opioid misuse has been shown to be associated with increased inpatient mortality, respiratory failure, surgical site infection, pneumonia, myocardial infarction, postoperative ileus, and prolonged length of stay.18 In a Workman’s Compensation cohort of patients undergoing cervical fusion for radiculopathy, patients who had prolonged (>3 months) preoperative opioid use had lower rates of return to work.19
Wick et al. determined that elective spine surgery patients with a preoperative opioid dose exceeding 29 mg/d had worsening surgical outcomes (based on Neck Disability Index or Oswestry Disability Index [ODI]) with increasing opioid doses.20 In this patient population, preoperative opioid weaning may be indicated. In patients with identified preexisting opioid use, a 6- to 7-week opioid reduction program has been shown to be effective in reducing opioid use from the start of the program to day of surgery.21 However, it has also been shown that 28% of spine surgery patients in one cohort inaccurately reported preoperative opioid consumption, making identification of preexisting users difficult.22
Therefore, identification of risk factors for preoperative and postoperative opioid misuse is essential. In patients undergoing anterior or lateral lumbar interbody fusion, preoperative opioid intake and Workman’s Compensation status were associated with continued opioid dependence at the first and second postoperative visits.23 Interestingly, while they are more likely to have postoperative opioid use, Workman’s Compensation patients undergoing MIS TLIF were shown to have equivalent rates of inpatient opioid use as patients who were non-Workman’s Compensation, while also being younger and having fewer medical comorbidities.11
NONSTEROIDAL ANTI-INFLAMMATORY MEDICATIONS
Nonsteroidal anti-inflammatory drugs (NSAIDs) work by inhibiting cyclooxygenase (COX) and therefore reducing production of prostaglandins, which mediate inflammatory reactions. COX-2 enzymes are specific to inflammatory tissues compared to COX-1 enzymes, which are found throughout the body. Therefore, COX-2 inhibitors have more specific anti-inflammatory effects without additional side effect on gastric mucosa or wound healing.
Traditionally, NSAIDs have been thought to interfere with spine fusion possibly resulting in pseudarthrosis. For example, Martin et al. demonstrated a 50% lower rate of spinal fusion in rabbits treated with ketorolac compared to controls (75% vs 35%).24 Glassman et al. conducted a retrospective review of 288 patients undergoing lumbar fusion and determined that nonunion rates were much higher in patients who received intramuscular ketorolac compared to controls (17% vs 4%).25 Li et al. conducted a meta-analysis of five retrospective cohort studies and 1403 patients undergoing spinal fusion surgery, studying the use of high-dose ketorolac.26 Patients receiving high-dose ketorolac (>120 mg/d) had higher risk of nonunion, while no normal detrimental effects were noted at normal, therapeutic doses.
Interestingly, at lower than normal doses of NSAIDs, patients have been shown to have similar analgesic effects as at high doses. Duttchen et al. conducted a prospective RCT of 50 patients undergoing lumbar decompression surgery.27 The authors compared 30 and 15 mg of intravenous intraoperative ketorolac and found that 30 mg dosing was not significantly superior.
Postoperative NSAIDs have also been shown to have good efficacy in reducing postoperative opioid consumption. Jirarattanaphochai et al. conducted an RCT of 120 patients undergoing lumbar discectomy, decompression, or spinal fusion, receiving either parecoxib or morphine preoperative and every 12 hours postoperatively for a total of 48 hours.28 Patients receiving parecoxib had a reduction in morphine use, greater satisfaction, and reduced pain at rest. Siribumrungwong et al. conducted an RCT of 96 patients undergoing posterior lumbar spinal fusion surgery, comparing preoperative parecoxib vs ketorolac vs placebo.29 Patients in both parecoxib and ketorolac groups had improved reduction of postoperative pain compared to the control group.
Finally, Jirarattanaphochai and Jung conducted a meta-analysis of 17 different RCTs and 789 patients studying the efficacy of NSAIDs compared to opioid analgesics for lumbar spine surgery.30 Patients receiving NSAIDs had lower pain scores and lower opioid use compared to those receiving opioids alone. As a result, we now have strong evidence supporting postoperative use of NSAIDs at moderate doses after spine surgery.
Acetaminophen inhibits COX both centrally and peripherally creating anti-inflammatory and antipyretic effects. While it remains a common adjunct to opioids with studies showing good analgesic effects, studies in spine surgery populations have not demonstrated an effect in reducing opioid intake or opioid-related complications.
Mörwald et al. reviewed a database of 117 269 patients undergoing lumbar spine surgery.31 The authors determined that the use of acetaminophen postoperatively was not associated with any clinically significant decrease in opioid prescriptions or reduced adverse events. Shimia et al. conducted an RCT of 52 patients undergoing lumbar discectomy receiving either intraoperative intravenous acetaminophen or placebo.32 Patients receiving acetaminophen had significantly lower pain scores but not significant decrease in opioid consumption. In an RCT of children and adolescents undergoing surgery for AIS or spondylolisthesis, intravenous postoperative acetaminophen was also similarly associated with decreased pain scores but no decrease in oxycodone consumption.33
Certain studies have demonstrated that acetaminophen has lower analgesic effects compared to other agents. Tunali et al. studied patients undergoing lumbar disc surgery and found that intravenous acetaminophen was not associated with any significant difference in pain scores compared to placebo over 24 hours postoperatively.34 In another prospective study of lumbar spine surgery patients, acetaminophen had lower analgesic effects and led to a greater need for additional analgesia when compared with tramadol patient-controlled analgesia.35 Therefore, acetaminophen may be effective as an adjuvant, but no evidence exists for its use in isolation for postoperative pain control.
Neuromodulator agents, such as pregabalin or gabapentin, work by inhibiting calcium-gated channels on presynaptic axons, reducing neuronal excitability. These drugs have long been used as anticonvulsants and for control of neuropathic pain. Interestingly, a number of RCTs have been conducted recently to study their effect on controlling postoperative pain after spine surgery and reducing opioid intake.
Khurana et al. conducted an RCT of 90 patients undergoing lumbar spine surgery and taking pregabalin, gabapentin, or placebo postoperatively.36 Patients taking gabapentin or pregabalin had lower ODI, pain scores, and opioid use compared to placebo. Kim et al. conducted an RCT of 84 patients undergoing elective lumbar spine surgery and taking either pregabalin (high or low dose) or placebo postoperatively.37 Patients receiving high-dose pregabalin had less intravenous patient-controlled analgesia use than those receiving placebo. However, there was no difference in pain scores or other effects.