Customize your JAMA Network experience by selecting one or more topics from the list below.
Carpenter PS, Shepherd HM, McCrary H, et al. Association of Celecoxib Use With Decreased Opioid Requirements After Head and Neck Cancer Surgery With Free Tissue Reconstruction. JAMA Otolaryngol Head Neck Surg. 2018;144(11):988–994. doi:10.1001/jamaoto.2018.0284
Does the addition of celecoxib to the standard analgesia regimen decrease the amount of opioid use in patients after head and neck cancer (HNC) surgery and reconstruction with free tissue?
In this matched-cohort study, 51 patients who received celecoxib after HNC surgery with free tissue reconstruction had significantly decreased perioperative opioid requirements compared with 51 matched patients who did not receive celecoxib. There were no differences in complication rates between the 2 groups.
Celecoxib was associated with a decrease in opioid requirements after HNC surgery with free tissue reconstruction and should remain a consideration for perioperative analgesia in select patients.
Head and neck cancer (HNC) surgery with free tissue reconstruction is associated with considerable postoperative pain. Opioids are typically used but can have adverse effects, including respiratory depression and high rates of dependence and addiction. Safe alternative analgesics that minimize opioid requirements are beneficial in HNC surgery.
To investigate the association of celecoxib use with opioid requirements in the postoperative setting after HNC surgery with free tissue reconstruction.
Design, Setting, and Participants
A retrospective, matched-cohort study of 147 patients who had undergone HNC surgery with free tissue reconstruction between June 2015 and Sept 2017 in an academic cancer hospital. Patients were separated into groups based on whether celecoxib had been used perioperatively or not. These groups were then matched by stage and site resulting in 102 included participants (51 celecoxib, 51 control).
Main Outcomes and Measures
Oral, intravenous (IV), and total morphine equivalents used in the postoperative setting per patient per day.
There were 51 patients in the celecoxib cohort (19 women and 32 men) and 51 patients in the control cohort (20 women and 31 men) who met inclusion criteria after clinicopathologic data were matched. The mean age of the celecoxib and control cohorts was 61.6 years and 66.1 years, respectively. Treatment with celecoxib in the postoperative setting was associated with decreased mean use of opioids in oral (mean difference, 9.9 mg/d; 95% CI, −1.2 to 21.1), IV (mean difference, 3.9 mg/d; 95% CI, 1.0-6.8), and total (mean difference, 14 mg/d; 95% CI, 2.6-25.4) amount of morphine equivalents per day. When patients were matched to surgical procedure, the effect was more significant. Patients who underwent composite oral resection and received celecoxib had decreased opioid use in oral (mean difference, 25 mg/d; 95% CI, 12.5-25.4), IV (mean difference, 3.4 mg/d; 95% CI, 1.5-5.5), and total (mean difference, 28.4 mg/d; 95% CI, 15.7-41.5) amounts compared with those in the control group. There was no significant difference in complication rates between the 2 cohorts.
Conclusions and Relevance
Use of celecoxib after head and neck cancer surgery and reconstruction with free tissue transfer was associated with a decrease in oral, IV, and total opioid requirements without increasing surgical or flap-related complications.
Malignant tumors of the head and neck upper aerodigestive tract account for approximately 5% of all newly diagnosed cancers.1-3 Pain is experienced by more than 80% of patients with head and neck cancer (HNC),4 and more than two-thirds of patients with advanced head and neck tumors require opioid analgesia.5 This pain is multifactorial in etiology, including nociceptive pain from direct invasion of richly innervated bone and soft tissue, neuropathic pain owing to cranial nerve involvement, or pain resulting from edema, inflammation and/or superimposed infection.6-8 Patients treated surgically for HNC have been found to have considerably more pain immediately in the postoperative period, and possibly extending chronically.7 Postoperative pain has been associated with longer hospital stays, lower patient satisfaction, higher costs of care, as well as compromised microvascular free flap viability.8,9 Classically, surgical pain management for HNC has required considerable amounts of opioids.10 However, opioids have considerable risk factors including respiratory depression, nausea, urinary retention, constipation, and high rates of dependence and addiction.11 According to the Centers for Disease Control and Prevention (CDC), on average, about 115 Americans die each day from opiate overdose.12 In 2013, clinicians wrote one-quarter billion prescriptions for opioids, which, put another way, is enough for every single American adult to have their own bottle of opioids.13 Despite the importance of optimal pain management, it is imperative that physicians minimize prescribing of opioids and narcotics to counteract this epidemic.
Guidelines for postoperative pain management in HNC surgery with free-flap reconstruction recommend using a multimodal approach combining opioid and nonopioid analgesics to achieve optimal analgesia, reduce the risk of opiate-related complications, and facilitate rapid recovery.10 The major advantage of incorporating both opioid and nonopioid analgesics following surgical trauma is the additive effect seen when different classes of drugs are used in combination, particularly as opiate tolerance is developed.14 Opiates bind to central nervous system opioid receptors and cause a reduction in neurotransmitter release, which inhibits transmission of pain. Nonsteroidal anti-inflammatory drugs (NSAIDs) cause cyclooxygenase (COX) inhibition, which suppresses prostaglandin production in both the spinal cord and peripheral nervous system resulting in pain-relieving, antipyretic, and anti-inflammatory properties.15,16 Cyclooxygenase-1 is constitutive, whereas COX-2 is more selective and is only activated at sites of inflammation. Therefore, this specificity theoretically results in fewer negative systemic effects, such as gastric ulceration and bleeding.15 Celecoxib has been shown to be effective for postoperative pain relief in surgical subspecialties such as neurosurgery,17 orthopedics,18-20 and plastic surgery,21-24 and has been shown to have no deleterious effect on microvascular free flap survival or wound healing.25 A recent 2013 Cochrane Review of all randomized, double-blind, placebo-controlled trials of adults prescribed celecoxib or placebo for acute postoperative pain (10 studies; 1785 participants met inclusion criteria) demonstrated improved pain relief and reduced need for rescue medications, particularly in moderate-to-severe pain, and only 1 serious adverse event of rhabdomyolysis occurred that may have been related to celecoxib, although the patient had received a number of other medications.26 Rofecoxib, another COX-2 inhibitor, was taken off the market in 2004 owing to concerns about increased cardiovascular risk.27 However, this concern has not been shown with celecoxib after long-term multi-institutional randomized clinical trials on numerous patients (n = 8059) have demonstrated no difference in incidence of cardiovascular adverse events.28 In addition, experimental and clinical studies have shown compelling evidence that COX-2 inhibitors have anticancer effects, particularly for breast, non–small cell lung, neuroblastoma, and colorectal cancers, as well as precancerous growths, such as colonic adenomatous polyps.29-31
In September 2016 at our institution a multidisciplinary task force including the Acute Pain Service was developed to reassess our postoperative pain management regimen for patients undergoing HNC surgery with free flap reconstruction. Routine, scheduled use of celecoxib was subsequently incorporated into our protocol for these patients and this study was designed to monitor and report the results of this quality improvement project. The primary objective was to investigate the association of postoperative celecoxib administration with pain management and opiate use for HNC patients undergoing surgery with free flap reconstruction, as well as report adverse complications.
A retrospective matched-cohort study was conducted evaluating all adult patients (age >18 years) at the University of Utah and Huntsman Cancer Institute who underwent HNC surgery with free tissue reconstruction between June 2015 and Sept 2017. The University of Utah institutional review board approved the study and a waiver of written informed consent was granted from for retrospective analysis. The Figure illustrates the flowchart of the study design. A total of 145 patients met initial criteria for inclusion in the study, and 2 consecutive cohorts were constructed owing to the quality improvement project enacted in September 2016, which incorporated celecoxib into our standard postoperative protocol for all head and neck cancer free flap patients in an effort to reduce opioid use. The entire cohort was then divided into recent patients receiving celecoxib postoperatively (n = 66) and a historical control group (n = 79). Patients were excluded from both cohorts for preoperative chronic opioid use, kidney disease, or a history of gastric ulcers. There were 51 patients who met inclusion criteria for the celecoxib cohort and 51 patients included in the historical control cohort. Medical records of both cohorts were then reviewed and clinicopathologic factors were extracted including age, sex, body mass index (BMI, calculated as weight in kilograms divided by height in meters squared), tobacco use, preoperative risk factors, cancer type, cancer site, stage, type of surgery, free tissue type, length of surgery, length of hospitalization, and 30-day complications. A composite resection was defined as an intraoral resection that included a segmental mandibulectomy. Patients in both cohorts had the same postoperative order set that had been created and standardized in EPIC since 2014. All patients received aspirin daily and subcutaneous heparin with prophylactic dosing.
The patients included in the celecoxib cohort received 200 mg celecoxib scheduled twice daily through the feeding tube for a minimum of 5 days starting on postoperative day 1. Oral opioids were offered to patients both in a scheduled manner (low-dose oxycodone or hydrocodone) and as needed (PRN), which could be increased depending on the individual patient’s pain. Patients were also provided with PRN intravenous (IV) opioids for severe pain that could be up-titrated depending on pain levels as well. The standardized otolaryngology free flap postoperative order set was used for all patients in both arms of the study. The only change to the order set after the multidisciplinary task force was the addition of celecoxib to the postoperative analgesic options.
Daily total amounts (mg) of both intravenous (IV) and oral opioids administered were identified using our electronic medical record medication administration history. Total daily opioid administration was calculated by adding the total IV per day and total oral per day. Opioids included in the study were hydrocodone, oxycodone, morphine, and hydromorphone. To unify the data from the various opioids used and to make it more clinically relevant, we used the 2016 CDC guideline morphine milligram equivalent (MME) calculator (Table 1). The MME is a formal way to compare milligram amounts of different opioids based on strength in relation to morphine. An example of our most commonly used opioids converted into MME per day is as follows: oxycodone (1 mg/d = 1.5 MME), morphine (1 mg/d = 1 MME), and hydromorphone (1 mg/d = 4 MME).
Levels of significance for quantitative variables were analyzed with a paired 2-sample t test for comparison of means. Statistical significance with confidence intervals was established as P≤.05. Odds ratios were used for outcome comparisons and reported with confidence intervals using a P≤.05. Statistical analysis was conducted with StatPlus statistical software (version 6.2.50, AnalystSoft Inc).
A total of 102 patients (51 in the non-celecoxib cohort, 51 in the celecoxib cohort) met the study inclusion criteria. There were no significant differences between the 2 groups in regard to age, sex, and preoperative comorbidities (Table 2). Ablative and reconstructive characteristics were also similar between the cohorts (Table 3). There was an increased population of patients who underwent lateral temporal bone/parotidectomy in the non-celecoxib group (7 [13.7%] vs 0). Squamous cell carcinoma was the most common pathology requiring head and neck oncologic procedures in both non-celecoxib (36 [71%]) and celecoxib (37 [73%]) cohorts. Composite oral resection was the most common surgical procedure performed in both cohorts. There was no significant difference between the cohorts in terms of reconstructive free tissue types, ischemia time, or length of hospitalization.
Treatment with celecoxib after HNC surgery with free tissue reconstruction was associated with significantly decreased mean use of intravenous and total (IV plus oral) opioids in patients who received celecoxib (Table 4). Celecoxib patients used a mean of 1.5 mg of IV MME per day vs 5.3 mg of IV MME per day in the control group (difference 3.9 mg/d; 95% CI, 1.0-6.8). Similarly, celecoxib-treated patients used a mean of 30.9 mg total MME per day compared with 44.6 mg total MME per day in the control group (difference 14 mg/d; 95% CI, 2.6-25.4). Celecoxib-treated patients also had an overall decrease in mean oral MME per day (difference 9.9 mg/d; 95% CI, −1.2 to 21.1). The addition of celecoxib resulted in a 3.5-fold total decrease in IV opioid used in patients after HNC surgery with free tissue reconstruction.
To further match our cohorts we chose to compare groups that underwent the most common ablative procedure (composite oral resection). Treatment with celecoxib in patients who underwent composite oral resection was associated with a clinically meaningful decreased mean oral MME per day; 26.0 mg vs 51.1 mg (difference, 25.1 mg/d; 95% CI, 12.5-37.5), decreased IV MME; 0.9 mg vs 4.4 mg (difference, 3.5 mg/d; 95% CI, 1.5-5.5), and decreased total MME; 26.9 mg vs 55.5 mg (difference, 28.4 mg/d; 95% CI, 15.7-41.5) compared with the non-celecoxib group after composite oral resection. There was a 4.8-fold decrease in IV opioids and 2-fold decrease in oral opioids used in patients after composite oral resection who received celecoxib.
There was no statistically significant difference in the incidence of complications between patients who received celecoxib after head and neck cancer surgery and free tissue reconstruction compared with those who did not. Flap dehiscence/surgical site infection after celecoxib was 16% vs 13% in the control group (OR, 1.3; 95% CI, 0.7-1.9). Hematomas occurred in 1 (1.9%) of 51 patients in the celecoxib group vs 2 (4.0%) in the control group (OR, 0.9; 95% CI, 0.2-1.5), and free flap failure rate 4.0% vs 4.0% compared with controls. There was no difference in gastrointestinal complications in celebrex group 4.0% vs 5.8% in controls (OR, 0.9; 95% CI, 0.4-1.6). Similarly, there was no difference between cardiovascular complications in celebrex patients (2 [4.0%] vs 2 [4.0%]) in controls between the celecoxib and non-celecoxib cohorts. There was no difference in 30-day mortality between the groups (1 [1.9%] vs 1 [1.9%]).
Despite the importance of pain management in patient satisfaction and health outcomes, many patients undergoing surgical intervention for head and neck cancer experience suboptimal postoperative pain control.2,32-35 One study found that 53% of patients after HNC surgery had pain scores greater than 3 using a 10-point rating scale, exceeding the typical threshold for initiating pain therapy.36 However, it remains imperative that we minimize the use of opiates to decrease the known adverse effects and risk of dependence or addiction, which is already an epidemic that plagues medical practice.
Prior studies have explored the use of COX-2 inhibitors in HNC postsurgical pain management and have shown promising results. Abdelhalim et al37 reported lornoxicam to be comparable to tramadol for postoperative pain relief and reduced opioid requirements after ENT surgery. Issioui and colleagues38,39 found that premedication with rofecoxib or celecoxib with acetaminophen reduced opiate use and improved patient satisfaction after otolaryngology surgery.
Importantly, COX-2 inhibitors have undergone some scrutiny and some (Rofecoxib) have even been taken off the market owing to risk of increased myocardial infarction. The theorized mechanism is that selective reversible inhibition of COX-2 enzyme results in decreased amount of prostaglandin I2 and a relative increased amount of thromboxane A2 (COX-1 derivative), which promotes platelet aggregation.40 In an attempt to counter this potential risk, we elected to continue providing all of our patients with 325 mg aspirin daily after free tissue transfer. Aspirin functions as a nonselective irreversible inhibitor of COX-1 and COX-2, which has a strong effect on blocking formation of thromboxane A2 in platelets.41 We hypothesized that once-daily dosing of aspirin would provide adequate platelet aggregation inhibition but would not contribute in a clinically meanningful way to analgesia.
Our data suggests that celecoxib is an appropriate adjunct for treatment of postoperative pain in patients after HNC surgery and reconstruction without increasing the risk of complications. This is in agreement with prior otolaryngology studies evaluating celecoxib.24,25 We have also demonstrated that using COX-2 inhibitors may decrease the amount of opioids used postoperatively by patients with HNC.
Specifically, in this patient population, the addition of celecoxib was associated with a decreased mean difference in total MME per day of about 14 mg per day. To further delineate, we attempted to match our cohorts based on the most common ablative procedure, which was a composite oral resection. This procedure was also hypothesized by our group to represent the most painful procedure owing to the intraoral and bony components involved with increased neurovasculature. When compared directly, the composite resection patients receiving celecoxib had an even more significant decrease in oral, IV, and total MME per patient per day during hospitalization. This effect occurred likely because comparing the procedures directly allowed for the cohorts to be more similar. Another reason hypothesized for the less pronounced decrease in opioid use after celecoxib when comparing our larger cohort of patients with HNC was that the non-celecoxib group had 7 lateral temporal bone resections that were all reconstructed with anterolateral thigh free tissue, whereas the celecoxib group had none (Table 3). This procedure was hypothesized by our group to potentially be a less painful procedure than oral resection or glossectomy and may have actually favored the null hypothesis in the cohorts.
It is important to note that the decrease in opioid use associated with celecoxib is clinically significant. The CDC has shown that dosages of opioids at 50 MME (equivalent of 33 mg oxycodone) or more per day doubles the risk of overdose and other opioid-related complications.12 Therefore, our data showing a decrease from 55.5 MME per day (36 mg oxycodone) to 26.9 MME per day (18 mg oxycodone) in patients with HNC after composite resection who received celecoxib suggests that those patients had almost one half the risk of opioid-related complications compared with those in the control group. Our data are encouraging but larger cohorts are needed to compare opioid-related complications and measure effect.
Notably, there was 1 patient in the study population (patient x) who actually met inclusion criteria for both arms of our study populations. This patient underwent composite oral resection with a fibular free tissue reconstruction prior to celecoxib use in 2015. Unfortunately, he developed a local recurrence and subsequently underwent the same ablative and reconstructive procedure with celecoxib in 2016. After celecoxib he had decreased oral MME per day (35.6 mg vs 80.4 mg), IV MME per day (0.82 mg vs 12.12 mg), and total MME per day (36.3 mg vs 92.4 mg). Both hospital stays were equal in length (7 days) and no differences in length of surgery. Although these data are interesting, patient x may have had some axonal degeneration and hypesthesia from the first surgery and interim radiation that could have confounded the decreased MME seen after celecoxib.
Although the data are promising, there are limitations to our study. It is worthwhile to note that celecoxib was added as an adjunct to the surgical pain regimen after a multidisciplinary task force was created to look at and minimize opioid use in surgical patients. The heightened awareness of opioid use that ensued after the task force may have effectively biased the data. Even though there were no formalized changes after the task force in regard to the amount of opioids made available for our patients, there is potential observer bias to confound our data that cannot be overlooked.
Some other limitations of this study include small sample size and retrospective data collection using medical record review without blinding or placebos. The small sample size particularly limits our ability to evaluate differences in complication rates thoroughly. Also, it is important to realize that we only looked at amounts of opioids used, which cannot necessarily be correlated with qualitative or quantitative measures of pain control. A small amount of variation in timing of medication administration should also be considered, though a standardized protocol was established for all patients in the study.
Our data has limitations but represents a pilot study to evaluate potential mechanisms to decrease opioid use in patients who undergo HNC surgery. Future studies should include prospective placebo-controlled trials including patient pain scores and patient-centered outcome measurements. Trials like this will be essential in providing further insight into the qualitative pain levels while minimizing potential observer bias.
Celecoxib use in patients with HNC after surgical resection and reconstruction with free tissue transfer resulted in a significant decrease in oral, IV, and total opioid requirements without increasing surgery and microvascular flap-related complications. The decrease in opioid use in these patients with HNC was clinically significant, suggesting that celecoxib should be considered as an adjunct pain medication after head and neck cancer surgery.
Corresponding Author: Richard B. Cannon, MD, Otolaryngology–Head and Neck Surgery, University of Utah, 50 N Medical Dr, SOM 3C-120, Salt Lake City, UT 84132 (firstname.lastname@example.org).
Accepted for Publication: March 27, 2018.
Published Online: April 18, 2018. doi:10.1001/jamaoto.2018.0284
Author Contributions: Drs Carpenter and Cannon had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: McCrary, Hunt, Buchmann, Cannon.
Acquisition, analysis, or interpretation of data: Carpenter, Shepherd, McCrary, Torrecillas, Kull, Monroe, Cannon.
Drafting of the manuscript: Carpenter, Shepherd, McCrary, Torrecillas, Buchmann, Cannon.
Critical revision of the manuscript for important intellectual content: Carpenter, McCrary, Torrecillas, Kull, Hunt, Monroe, Buchmann, Cannon.
Statistical analysis: Carpenter, Shepherd, Monroe, Cannon.
Administrative, technical, or material support: McCrary, Torrecillas, Hunt, Cannon.
Study supervision: Monroe, Buchmann, Cannon.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest, and none were reported.
Meeting Presentation: This study was presented at the AHNS 2018 Annual Meeting; April 18, 2018; National Harbor, Maryland.
Create a personal account or sign in to: