The expansion of therapeutic options for rheumatoid arthritis (RA) has improved the functional, quality of life, and radiographic outcomes for patients, although immunomodulatory and immunosuppressive effects may increase the potential for complications related to infection risk or wound healing after orthopedic surgery. Additionally, medications without significant effects on the immune system, such as nonsteroidal anti-inflammatory drugs, may have consequences in terms of bleeding or cardiovascular disease. Increased vigilance is required when using these medications, particularly in the perioperative period. This article considers the recent literature and data on the perioperative use of medications commonly used in RA.
- •
Although the rates of total joint arthroplasty (TJA) have increased markedly in the general population, fewer patients with rheumatoid arthritis (RA) undergo joint surgery.
- •
Given the fact that most patients with RA are being treated with disease-modifying antirheumatic drugs, those hospitalized for orthopedic surgery, particularly arthroplasty, are on drug regimens that can affect the outcome of surgery and therefore must be factored into the care plan.
- •
Perioperative management decisions for patients with RA must balance the potential conflict between wound healing and infection risk against ongoing disease control and the ability to participate in postoperative rehabilitation routines that are critical for TJA outcome.
- •
Patients with RA who undergo orthopedic surgery for joint destruction typically have more severe disease and are those for whom therapy has failed.
Introduction
Contemporary rheumatoid arthritis (RA) has become milder, and fewer patients develop erosive, destructive disease than described in historical cohorts. This change is theorized to be associated with changes in biology and is clearly associated with a more aggressive treatment paradigm involving the increasingly widespread early use of potent disease-modifying antirheumatic drugs (DMARDs). Population studies show a decrease in all-cause hospitalizations for RA over the past decade, including a decrease in admissions for arthroplasty.
Although the rates of total joint arthroplasty (TJA) have increased markedly in the general population, fewer patients with RA undergo joint surgery. The rate of hip surgery in patients with RA has remained stable, soft tissue procedures have decreased, and, paradoxically, knee surgery has increased, with the latter attributable to osteoarthritis and the increase in obesity in patients with RA. Given the fact that most patients with RA are being treated with DMARDs, those hospitalized for orthopedic surgery, and particularly arthroplasty, are on drug regimens that can affect the outcome of the surgery and therefore must be factored into the care plan. Perioperative management decisions regarding patients with RA must balance the potential conflict between wound healing and infection risk against ongoing disease control and the ability to participate in postoperative rehabilitation routines that are critical for TJA outcome. The significant expense in both health care dollars and individual morbidity of TJA infection further magnifies the importance of this conflict.
Patients with RA who undergo orthopedic surgery for joint destruction typically have more severe disease and are those for whom therapy has failed. From 2005 to 2007, the total knee arthroplasty rate was 220 per 100,000 patient years, a remarkable increase from 31 per 100,000 patient years between 1971 and 1976, with an estimated 3% to 4% performed in patients with RA. Predictors of severe disease logically overlap with predictors of TJA, and have been identified from large cohort studies. These include poor function on the Health Assessment Questionnaire, persistent elevations of erythrocyte sedimentation rate and C-reactive protein (CRP), and the presence of nodules, extra-articular disease, and erosions within the first year of disease. Among patients undergoing a first TJA, 25% will undergo a second TJA within the following year, and 50% will undergo a second TJA within 7 years. Mortality reports are inconsistent, however. Although an increase in mortality has been reported in one cohort of patients with RA undergoing arthroplasty, population-based studies have failed to confirm this increase.
Although the more widespread use of DMARDs has contributed to the improved outcome and a less severe course for RA, the risks and benefits of DMARD use in the perioperative period have not been well defined. DMARD use has increased in patients undergoing orthopedic surgery, as it has in the general RA population. Among patients with RA who underwent orthopedic surgery, only 30% with RA onset between 1980 and 1994 received methotrexate, compared with 64% of those with onset of RA between 1995 and 2007. None in the early cohort had received biologic response modifiers, whereas 20% in the later cohort had. Corticosteroid use was also increased in the later cohort, increasing from 51% to 81%. Given the potent immune suppressant characteristics of these medications, understandable concerns have been raised regarding perioperative use. Patients with RA intrinsically have a 2-fold increased risk of infections when not taking an immunosuppressive drug. Moreover, RA confers an infection risk 2 to 4 times that of osteoarthritis in patients undergoing arthroplasty, and RA with high disease activity has also been shown to increase the risk of infection. Concerns about the increased risk of surgical site infection and delayed wound healing have led many clinicians to withhold disease-modifying drugs preoperatively. This approach may lead to a postoperative flare, a reality that may make compliance with postoperative rehabilitation regimens more difficult. Additionally, postoperative flare has an unknown impact on long-term orthopedic outcomes. Loss of efficacy of an established medication regimen is an additional concern when medications are withheld before surgery. Recommendations for medication management in patients with RA undergoing surgery are extrapolated from the experiences of orthopedists, rheumatologists, and gastrointestinal surgeons operating on patients with inflammatory bowel disease. The unique concerns associated with infection risk when orthopedic devices are implanted into a susceptible patient, and the distinctive demands of postoperative orthopedic rehabilitation necessitate a unique approach to patients with RA undergoing orthopedic surgery. This article considers the recent literature on the topic of perioperative use of medications commonly used in RA and assesses the strengths and weaknesses of the available data.
Corticosteroids
Corticosteroids have a broad and strong impact on innate and adaptive immunity, leading to the general concept that no other immunosuppressive drug is associated with a higher risk for any type of infection. Corticosteroid use remains highly prevalent in patients with RA, and has well-recognized effects on symptom control, with more recent studies showing a disease-modifying effect. A recent study assessing corticosteroid use found that although 35% of 12,749 patients with RA were current corticosteroid users, 65% had used corticosteroids at any point in their lifetime, and use varied with disease severity. Poorer outcomes were more common in current than in prior corticosteroid users, with increases in rates of mortality (5.7% vs 2.6%), work disability (28.4% vs 17.2%), and TJA (18.5% vs 13%). Infection risk varied with cumulative and absolute doses. The risk of infection in a current user of prednisolone, 5 mg, was 30% after 3 months of therapy, 46% after 6 months of therapy, and 100% after 3 years of therapy.
Infection remains strongly associated with corticosteroid use, even at low doses, with a dose-dependant increase in risk. In addition to their use before admission, corticosteroids may be incorporated into multimodal regimens for postoperative pain and nausea control. The widespread preference for corticosteroid use over traditional and biologic DMARDs in the perioperative period to prevent flare warrants reevaluation.
Given the prevalence of corticosteroid use in RA, and the association of corticosteroid use with disease severity, many patients with RA undergoing orthopedic surgery are likely to be current corticosteroid users. Because prolonged use of corticosteroids will predictably suppress the hypothalamic-pituitary-adrenal (HPA) axis, many patients are given supraphysiologic doses (“stress doses”) of steroids to support the HPA axis during the stress of surgery and prevent catastrophic hypotension. Careful study has raised questions about this practice, however. When patients who had adrenal insufficiency based on corticotropin testing were given either their daily steroid dose plus saline plus excess corticosteroids or their usual daily dose plus saline alone, no hemodynamic difference was seen between the groups.
In a classic study, 21 of 41 patients with RA undergoing synovectomy were chronically treated with corticosteroids, and 16 of these 21 had adrenal insufficiency on testing. These patients were compared with the 20 patients with RA who had not been treated with corticosteroids who served as controls. Corticosteroid therapy was stopped 18 hours before surgery in 20 patients and 48 hours before surgery in 1 patient. Although the blood pressure in the corticosteroid-treated patients was lower than that in patients who had not received corticosteroids, the differences were not significant, except in the patient whose corticosteroid was discontinued 48 hours before surgery. Hypotension in that patient responded promptly to intravenous administration of corticosteroids.
Additional studies have confirmed the safety of administering patients’ usual daily dose of steroid only in the setting of major orthopedic procedures and allograft nephrectomy. Reviews of the available literature support the administration of the usual daily dose of steroid during major surgery, and not supraphysiologic doses. Testing the HPA axis does not provide guidance. Results in patients with presumed secondary adrenal insufficiency from prolonged corticosteroid use do not benefit from HPA testing before surgery, because the results do not predict clinical course. Patients with presumed secondary adrenal insufficiency can be given their usual daily steroid dose on the morning of surgery. Intraoperative hypotension can be easily treated with intravenous hydrocortisone should the need arise. This recommendation does not apply to patients with primary HPA insufficiency, and has not been tested in patients who have received chronic corticosteroids during childhood development, such as those with juvenile inflammatory arthritis. Perioperative surgical site infections and wound healing may benefit from a more restricted approach to perioperative corticosteroid administration.
Corticosteroids
Corticosteroids have a broad and strong impact on innate and adaptive immunity, leading to the general concept that no other immunosuppressive drug is associated with a higher risk for any type of infection. Corticosteroid use remains highly prevalent in patients with RA, and has well-recognized effects on symptom control, with more recent studies showing a disease-modifying effect. A recent study assessing corticosteroid use found that although 35% of 12,749 patients with RA were current corticosteroid users, 65% had used corticosteroids at any point in their lifetime, and use varied with disease severity. Poorer outcomes were more common in current than in prior corticosteroid users, with increases in rates of mortality (5.7% vs 2.6%), work disability (28.4% vs 17.2%), and TJA (18.5% vs 13%). Infection risk varied with cumulative and absolute doses. The risk of infection in a current user of prednisolone, 5 mg, was 30% after 3 months of therapy, 46% after 6 months of therapy, and 100% after 3 years of therapy.
Infection remains strongly associated with corticosteroid use, even at low doses, with a dose-dependant increase in risk. In addition to their use before admission, corticosteroids may be incorporated into multimodal regimens for postoperative pain and nausea control. The widespread preference for corticosteroid use over traditional and biologic DMARDs in the perioperative period to prevent flare warrants reevaluation.
Given the prevalence of corticosteroid use in RA, and the association of corticosteroid use with disease severity, many patients with RA undergoing orthopedic surgery are likely to be current corticosteroid users. Because prolonged use of corticosteroids will predictably suppress the hypothalamic-pituitary-adrenal (HPA) axis, many patients are given supraphysiologic doses (“stress doses”) of steroids to support the HPA axis during the stress of surgery and prevent catastrophic hypotension. Careful study has raised questions about this practice, however. When patients who had adrenal insufficiency based on corticotropin testing were given either their daily steroid dose plus saline plus excess corticosteroids or their usual daily dose plus saline alone, no hemodynamic difference was seen between the groups.
In a classic study, 21 of 41 patients with RA undergoing synovectomy were chronically treated with corticosteroids, and 16 of these 21 had adrenal insufficiency on testing. These patients were compared with the 20 patients with RA who had not been treated with corticosteroids who served as controls. Corticosteroid therapy was stopped 18 hours before surgery in 20 patients and 48 hours before surgery in 1 patient. Although the blood pressure in the corticosteroid-treated patients was lower than that in patients who had not received corticosteroids, the differences were not significant, except in the patient whose corticosteroid was discontinued 48 hours before surgery. Hypotension in that patient responded promptly to intravenous administration of corticosteroids.
Additional studies have confirmed the safety of administering patients’ usual daily dose of steroid only in the setting of major orthopedic procedures and allograft nephrectomy. Reviews of the available literature support the administration of the usual daily dose of steroid during major surgery, and not supraphysiologic doses. Testing the HPA axis does not provide guidance. Results in patients with presumed secondary adrenal insufficiency from prolonged corticosteroid use do not benefit from HPA testing before surgery, because the results do not predict clinical course. Patients with presumed secondary adrenal insufficiency can be given their usual daily steroid dose on the morning of surgery. Intraoperative hypotension can be easily treated with intravenous hydrocortisone should the need arise. This recommendation does not apply to patients with primary HPA insufficiency, and has not been tested in patients who have received chronic corticosteroids during childhood development, such as those with juvenile inflammatory arthritis. Perioperative surgical site infections and wound healing may benefit from a more restricted approach to perioperative corticosteroid administration.
Methotrexate
Methotrexate has been best studied in the perioperative period. In a prospective study, 388 patients undergoing orthopedic surgery were randomly assigned to either continue or discontinue methotrexate, and also compared with patients who were not taking methotrexate at the time of surgery. The incidence of surgical complications or infections was highest (15%) in the group that discontinued methotrexate, compared with 2% in the group continuing methotrexate. Flares were also high in the group discontinuing methotrexate. This finding has been confirmed in multiple other prospective and retrospective studies. More recently, 65 cases from the original 388 patients studied were followed up, and none had developed late infection, leading the authors to conclude that late infections are not increased in methotrexate users either. Although the doses used in these studies (10–15 mg/wk) may not be as high as the doses currently being used (20–25 mg/wk), continuing methotrexate during the perioperative period seems safe, without detrimental effects on wound healing or infection rate. Alternative treatment regimens using corticosteroids to replace methotrexate in the perioperative period lack evidence in the literature and may be less safe. Patients with a significant comorbidity burden, such as those with diabetes or renal functional impairment, may warrant a different approach to perioperative methotrexate dosing, although no studies support a more conservative approach.
Leflunomide
Leflunomide is a DMARD with proven efficacy comparable to methotrexate in controlling inflammation and preventing ongoing radiographic damage in patients with RA. The prevalence of infection associated with leflunomide use is not as clear, however. Two small studies addressing perioperative risk reached opposite results, one suggesting no increase in infection and wound healing complications, and the other reporting a significant increase in perioperative complications. A larger retrospective audit of 171 patients taking leflunomide suggested a small increase in infection, particularly significant in patients who had severe disease and those who were also receiving methotrexate or corticosteroids. Patients in this report also had rapidly progressive infections that responded poorly to appropriate antibiotic therapy. The prolonged elimination half-life of leflunomide also differentiates it from other traditional DMARDs. It therefore seems prudent to withhold leflunomide preoperatively, and consider a prolonged (2-week) period off-drug before surgery in patients with other significant infection-related risk factors, such as diabetes or corticosteroid use.
Biologic therapy
Tumor Necrosis Factor Inhibitors
The addition of tumor necrosis factor inhibitors (TNFis) to the RA armamentarium has completely revolutionized the expectations and goals of RA therapy. Patients frequently experience remission, an outcome that was unanticipated as recently as the past decade. Although these potent immunosuppressant medications share an unequivocal increase in the risk of serious infection, much of the initial published experience was biased, because the more severe RA cases were the early recipients of these novel drugs. These patients frequently concurrently received corticosteroids or other immunosuppressants. Later, meticulous analyses began to raise questions about the infection risk associated with TNFi use, and the importance of comorbidities and use of medications such as corticosteroids has again been noted. Although an overall risk of infection is consistently associated with TNFi and corticosteroid use, the specific risk of infection in patients undergoing arthroplasty is less clear, because most studies are either too small to detect infection or are retrospective studies or case series. Although TNFi use during wound healing may be of benefit, concerns regarding the potential for heightened infection risk have understandably taken precedence. Current guidelines developed by various national rheumatic disease societies, such as the American College of Rheumatology (ACR) and the British Society for Rheumatology, recommend stopping TNFi use 1 to 4 weeks before surgery, proportional to the drugs’ half-lives. Until studies clearly delineate the comorbidities and concurrent medications that contribute to the increase in risk, caution dictates holding these agents preoperatively. Given the disparity in national society guidelines, caution would suggest adherence to the more rigorous guidelines for patients with other infection risk factors, such as diabetes, older age, lung or cardiac disease, or concurrent therapy with corticosteroids. Short-term prophylaxis for bacterial infection after arthroplasty can also include implant fixation with antibiotic-laden cement when appropriate.
Rituximab
Rituximab is a monoclonal antibody that targets the CD20 B-cell antigen and is useful and approved for patients with RA who have not had a satisfactory response to TNFi use, and has shown efficacy in both symptom control and radiographic progression. Compared with TNFi, rituximab is associated with a lower risk for bacterial infections, which are the primary concern in perioperative management, although the presence of low immunoglobulin levels in a small proportion of patients raises the infection risk. Rituximab has been shown to be safe in patients with prior recurrent bacterial infections, which is reassuring when considering TJA, given its prolonged duration of action. The association with reactivation of Jakob-Creutzfeldt disease and resultant progressive multifocal leukoencephalopathy, and for hepatitis B reactivation, are notable but have little impact on perioperative management. Preoperative screening of immunoglobulin levels in patients at high risk for surgical site infection might be considered, but no studies support that approach and no therapy exists to modify an identified increased risk. Administration of intravenous immunoglobulin as a preoperative preventive regimen would be experimental, although implant fixation with antibiotic-laden cement, when appropriate, should be considered.
Abatacept
Abatacept, a human fusion protein of CTLA-4 and immunoglobulin, is an effective medication for the treatment of RA that acts through down-regulating T-cell activation. The risk of infection in patients treated with abatacept is not significantly increased over baseline nonbiologic-treated RA. Abatacept is administered either as a monthly infusion or a weekly subcutaneous injection, and conservative timing of surgery should be at the end of the dose cycle.
Tocilizumab
Tocilizumab is a humanized monoclonal antibody that targets the IL-6 receptor. It can be used either as monotherapy or in conjunction with other DMARDs, and is effective in treating the signs and symptoms of RA, and in slowing radiographic progression. Infection rates attributed to tocilizumab are comparable to those associated with other biologic DMARDs. A retrospective study that compared 161 orthopedic surgical patients who had received tocilizumab with those taking other DMARDs found no increase in infection, which was seen in 3 patients. Wound healing seemed delayed in 20 cases. Additionally, the usual elevation of CRP and temperature associated with surgery was not seen, a finding confirmed in another retrospective series of orthopedic cases treated with tocilizumab. Flares of RA were also seen, and in these cases the CRP was elevated. The direct effect of tocilizumab on temperature and CRP, observed in multiple settings, may lead to a delay in diagnosing infection in the postoperative period. Although clinical complaints may raise the index of suspicion for an infectious complication, these patients may be also receiving analgesics, as would be appropriate in the postoperative setting, further decreasing the clinical clues of a complication. As with other biologic DMARDs, discontinuing tocilizumab before surgery based on the drug half-life of 11 to 13 days is a safe approach to perioperative therapy. Additionally, heightened vigilance for complications is required given the absence of cues, such as fever.
Tofacitinib
This oral DMARD is likely to be approved soon for use in RA, and has shown efficacy in control of symptoms and in slowing radiographic progression (Kremer). The Janus activated kinase 3 (JAK3) has an essential role in cytokine signal transduction, and regulates lymphocyte differentiation and survival. Adverse events reported in early trials include anemia, neutropenia, and upper respiratory infections. Insufficient information is currently available to make recommendations regarding perioperative management.
Nonsteroidal Anti-Inflammatory Drugs
Nonsteroidal anti-inflammatory drugs (NSAIDs) are included in the 2012 recommendations for the management of osteoarthritis of the hip, knee, as well as therapy for RA, and although recognition of potential associated cardiac, renal, and vascular toxicity has prompted caution. NSAIDs are widely used for this indication. Osteoarthritis is the diagnosis in most patients admitted for arthroplasty, and therefore patients presenting for surgery frequently list NSAIDs among their preoperative medications. NSAIDs are widely used postoperatively in multimodal pain management regimens to decrease narcotic use and adverse narcotic effects. An additional use is to prevent heterotopic ossification, with new bone formation at the site of surgical trauma that may lead to long-term functional limitation and pain. Significant concerns exist regarding preoperative NSAID use, however, including bleeding risk, cardiovascular risk, and risk of poor bony ingrowth of noncemented prostheses and a decrease in bone fusion rates.
Excessive bleeding may theoretically result when NSAIDs are continued up to the time of surgery, because of the reversible effect of NSAIDs on platelet function. Evidence supporting a significant increase in blood loss with preoperative administration of nonselective NSAIDs compared with cyclooxygenase-2 (COX-2)–selective NSAIDs include studies showing that indomethacin was associated with 17% more blood loss than meloxicam ; diclofenac was associated with 32% greater blood loss than rofecoxib ; and ibuprofen was associated with 45% greater blood loss. Because the effect of NSAIDs on platelet function is reversible, simply discontinuing the medications using a formula based on doubling the drug half-life provides adequate safety. Lower extremity arthroplasty is typically performed under either hypotensive anesthesia or tourniquet, which diminishes blood loss and therefore lessens the potential significance of blood loss associated with NSAID use. In fact, some orthopedists at the authors’ institution continue NSAIDs through surgery.
NSAIDs may also contribute to postoperative bleeding from interactions with warfarin anticoagulants. Enzymes in the cytochrome P450 (CYP) pathway play a key role in the metabolism of both warfarin and NSAIDs, and enzyme variants have been linked to excessive anticoagulation. Among 100 patients undergoing total hip arthroplasty, 11 of 30 who were heterozygous for a CYP variant had an international normalized ratio (INR) greater than 4.9. Only those who had received NSAIDs had significant elevations of INR. In the outpatient setting, a cohort study showed a significant increase in risk with an INR greater than 6 when patients anticoagulated with warfarin had CYP variants and received NSAIDs. Additionally, the effect of inhibiting platelet aggregation and decreasing traditional vitamin K–dependant clotting factors with warfarin is additive. Given the role of warfarin as a cornerstone of thromboembolism prophylaxis after TJA, avoiding NSAIDs during the period of prophylaxis and careful monitoring of INR are strongly recommended for all patients on warfarin given the high frequency of the CYP allelic variant and multiple other potential interactions.
Cardiovascular risk associated with NSAIDs has been well described; COX-2 inhibitors were prescribed with the intent of avoiding COX-1–mediated gastrointestinal toxicity while abrogating COX-2–mediated pain and inflammation. Subsequent large studies have shown an increase in cardiovascular events and thrombotic events with NSAIDs, particularly COX-2 inhibitors. Recently, this effect was elucidated in a model in which deletion of the COX-2 gene in endothelial cells resulted in decreased production of prostacyclin (PGI 2 ), which decreases platelet aggregation, vasoconstriction, and thrombosis, through showing a decrease in the excretion of a PGI 2 metabolite. Nitrous oxide, a potent protective vasodilator, was also reduced. Although the theoretical risks are significant, when troponin levels were obtained on 1518 (14%) of 10,873 patients undergoing arthroplasty, elevations were seen in 0.8 of 9831 patients who received NSAIDs, compared with 1.8 of 1042 patients who did not, showing no increase in cardiovascular events after arthroplasty with a mean duration of therapy of 3 days. A cautious approach to NSAIDs and COX-2s would include withholding COX-2s in the perioperative period in patients at high risk for cardiovascular events, and treating for a short duration when NSAIDs or COX-2s are indicated.
Bony ingrowth inhibition by NSAIDs and COX-2s in noncemented prostheses is concerning, given the role of prostaglandins in osteoblast and osteoclast regulation. Although in vitro studies have shown a decrease in bony ingrowth in animal models, short-term administration of COX-2s do not have the deleterious effect on bone that 6-week courses of therapy have. In vivo studies assessing prosthesis migration via radiostereometric analysis has not shown a difference in patients randomized to receive either celecoxib versus placebo after TJA. Using a model in which tetracycline labeling allows measurement of bony ingrowth into plugs during staged total knee arthroplasty, no difference in bony ingrowth in patients treated with celecoxib versus placebo could be shown. The clinical concern regarding fixation of noncemented prostheses has not been substantiated, and therefore NSAIDs and COX2s can be used after arthroplasty for short-term pain relief without concern for bony in growth.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
