COX-2

There are eight high quality (Schroer 2011, Boonrioing 2010, Chen 2015, Xu 2018, Kahlenberg 2017, Mardani-Kivi 2013, Zhu 2018, Gong 2013) and four moderate quality studies (Ittichaikulthol 2010, Jianda 2016, Zhu 2014, Zhu 2016) that met inclusion criteria for this recommendation.

In a randomized double-blind placebo-controlled trial in patients undergoing TKA, Schroer et al (2011) studied celecoxib given twice daily for 6 weeks after discharge vs. placebo. Patients in the study group took significantly fewer opioids over the course of 12 months. Pain scores at rest and with activity were significantly lower in the celecoxib group out to 3 weeks after surgery, and pain scores at night were significantly lower out to 6 weeks after surgery. Patients in the celecoxib group had a significant increase in function across multiple patient reported outcome measures/functional domains (Knee Society Score – Function, Oxford, SF-12 – physical) at 6 weeks.  No differences were noted in adverse events.

Boonriong et al (2010) evaluated preoperative celecoxib and placebo as one-time dose in patients undergoing anterior cruciate ligament reconstruction (ACLR). No significant differences were found between pain, narcotic use, or function between celecoxib and placebo. There were no other differences found to include adverse events between groups.

Chen et al (2015) evaluated patients after total hip arthroplasty receiving oral celecoxib preoperatively and postoperatively through day 5 vs placebo. Patients in the celecoxib group had improved VAS pain scores through 72 hours postop, but no difference in Harris Hip Scores. Patients in the celecoxib group were noted to ambulate more than 1 day earlier than the placebo group however the mean times to ambulation were 4.5 +1.2 days vs. 5.83 +2.04 days in the celecoxib vs. control group which calls into question the applicability of this outcome in most modern rapid recovery protcols.  There were no differences in adverse events between groups.

Xu et al (2018) evaluated markers of inflammation in patients taking celecoxib with tramadol vs. tramadol alone after TKA. Indicators of aseptic inflammation (skin temperature, WBC, ESR and CRP) were all significantly reduced in the celecoxib group and Knee Society Scores were improved. Pain was not assessed.

Kahlenberg (2017) performed a randomized double-blind placebo-controlled trial evaluating preoperative celecoxib vs. placebo when given as 1 preoperative dose prior to hip arthroscopy. There were significant improvements in postoperative pain scores up to 2 hours after surgery. There were no differences noted in opioid consumption and small but not significant decreases in PACU time were found in the celecoxib group.

Mardani-Kivi (2013) performed a randomized triple blind placebo-controlled trial evaluating one dose of preoperative celecoxib vs. placebo for patients undergoing isolated ACLR or partial meniscectomy. Pain scores were significantly lower in the celecoxib group for both ACLR and partial meniscectomy at 6 and 24 hours after surgery. Opioid usage was significantly less in the celecoxib group for partial meniscectomy at 6 and 24 hours and for the ACLR group at 6 hours.  Function was not evaluated.  There were no significant differences in adverse events.

Ittichaikulthol et al (2010) performed a randomized double-blind placebo-controlled trial in patients undergoing TKA/THA comparing 1 dose of preoperative celecoxib vs parecoxib vs placebo. Decreased opioid consumption was found in the first 24 hours in both celecoxib and parecoxib groups vs placebo. There were improved pain scores in the parecoxib group through 12 hours compared to both celecoxib and placebo but no difference in pain scores between celecoxib and placebo. Function was not assessed. There was less sedation in the parecoxib group but otherwise no difference in adverse events.

Jianda (2016) performed a randomized placebo-controlled trial looking at one dose of celecoxib preoperatively vs placebo in patients undergoing TKA. All patients received multimodal analgesia throughout hospital stay and the same postoperative discharge medications (to include celecoxib). Two-minute walking test at 3 days after surgery was better in the celecoxib group. Significant improvements in pain scores and decreased opioid consumption were found in the celecoxib group.

Zhu et al (2018) performed a randomized double-blind placebo-controlled trial evaluating the use of celecoxib in patients undergoing TKA. The primary outcome evaluated postoperative cognitive deficiencies, with pain scores evaluated as a secondary outcome. Postoperatively they noted statistically lower pain scores out to 7 days after surgery and less cognitive dysfunction in the celecoxib group.

Benefits/Harms of Implementation

Use of Cox-2 selective agents carries the risks associated with the known side effect profile of these medications, with the possible benefit of decreased gastrointestinal irritation and effect known to occur with non-cox-2 selective agents. Special caution should be exercised in patients with renal insufficiency and a known history of cardiovascular disease, as these medications may be contraindicated in this patient population.

Cost Effectiveness/Resource Utilization

Cox-2 agents may cost more than their non-selective counterparts, but as generic formulations come to market the differences in cost may gradually decrease.

Acceptability

The use of cox-2 selective agents is well accepted under conditions in which NSAIDs are typically used. Patients and clinicians should be aware of the potential risks of these medications and understand the presentation and treatment of potential adverse events.

Feasibility

As cox-2 selective agents become available in generic formulations, their availability to clinicians should become more common and encouraged.

Future Research

Cox-2 selective agents may play an important role in pain alleviation strategies after orthopaedic surgery. Further research into the ideal combination of medications to minimize opioid requirements and consumption after surgery are warranted. Future research should focus on non-arthroplasty surgery and evaluate outcomes and adverse events for longer periods to determine the safety of these medications in orthopaedic surgery patients’ long term.