Arthritis & Rheumatism

Guidelines For Monitoring Drug Therapy In Rheumatoid Arthritis

American College Of Rheumatology Ad Hoc Committee On Clinical Guidelines

This report presents guidelines for monitoring the effects of medications used in the treatment of rheumatoid arthritis (RA). These guidelines are drawn from a synthesis of expert opinion, a survey of rheumatologists, published guidelines, and, whenever possible, data on toxicity. They are intended for use by primary care physicians, rheumatologists, and other health professionals involved in the care of patients with RA. It is important to emphasize the following points that were considered in putting forth these guidelines: 1) there are insufficient data to develop completely evidence-based recommendations on the extent and frequency of monitoring; and 2) it is unlikely that studies to obtain such data will be performed, because toxicities that drive the monitoring strategies occur with a frequency ranging between 0.1% and 5%. For certain medications, other reports may recommend more frequent monitoring than is recommended here. In these instances we have been unable to find supporting documentation for more frequent monitoring, and therefore, where possible we have used recommendations that will minimize cost and inconvenience to patients (1,2).

In this article, we describe the toxicity of agents used in the treatment of RA, risk factors, strategies to prevent toxicity, and our recommendations for prudent monitoring. Guidelines for the use of these drugs in the treatment of RA have recently been developed (3) and will not be considered here.

Toxicity may range from mild to serious and from reversible to irreversible. We define rare toxicities as those which occur in <1% of patients using the agent, uncommon in 1-10%, and common in >10%. Toxicities of drugs used in RA that require monitoring include gastrointestinal (GI) bleeding, hypertension, hyperglycemia, macular damage, renal damage, hepatotoxicity, and myelosuppression. Reduction in the incidence, severity, and unfavorable outcomes of these toxicities can be attempted by 1) pretreatment assessment to identify patients with risk factors for toxicity, 2) careful patient and physician education about safe dosage and the signs and symptoms of toxicity, and 3) appropriate monitoring with physician followup and periodic laboratory studies. Since multiple physicians may be following a patient with RA, an explicit plan should be made among the physicians and the patient to assign responsibility for monitoring at the beginning of treatment. This plan should also detail who will make adjustments in the antirheumatic medications.

Guidelines for monitoring drug treatment in RA are presented in Table 1. Included are listings of toxicities that require monitoring, baseline evaluation, and monitoring strategy for each drug or class of drugs. These monitoring recommendations are for patients who have uncomplicated RA with no history of or active concurrent illness and who are not receiving other medications. Situations in which there is concurrent disease or concurrent medication necessitate clinical judgments regarding dosing and monitoring that go beyond the intent of these guidelines. The discussion below is designed to supplement the information provided in Table 1.

Table 1. Recommended monitoring strategies for drug treatment of rheumatoid arthritis* Click here

Nonsteroidal antiinflammatory drugs (NSAIDs)

The toxicities of NSAIDs include dyspepsia (common), gastric or small bowel bleeding or ulceration (4-8) (uncommon), renal insufficiency (9-17) (rare), confusion, depression, rash, headache (rare), and hepatic toxicity (rare) (18). NSAIDs may also reversibly inhibit platelet function and prolong bleeding time. Patients with prior aspirin hypersensitivity are also at risk for developing bronchial spasms (rare), when taking NSAIDs. There appear to be few differences in the frequency of serious toxicities among the different NSAIDs (7,8).

Risk factors for major GI toxicity include advanced age, dosage, history of peptic ulceration or bleeding, concurrent corticosteroid use, and cardiovascular disease (19-21). Patients starting treatment with NSAIDs should be advised to take them with food in order to reduce dyspepsia and other GI side effects. Currently, only misoprostol has been shown to reduce the frequency of NSAID-induced GI complications (20,21). Misoprostol should be considered for patients who require NSAID treatment and are elderly or have a history of peptic ulcer disease, GI bleeding, or cardiovascular disease. Sucralfate, H2 blockers, and antacids are often used to treat dyspepsia, but may not prevent ulcer formation or bleeding due to NSAIDs (22-24).

All NSAIDs can cause renal complications, including reversible renal insufficiency, papillary necrosis, nephrotic syndrome, interstitial nephritis, and renal failure. High-risk groups for renal toxicity include the elderly, particularly those receiving diuretics, and patients with preexisting renal disease, congestive heart failure, cirrhosis, atherosclerotic heart disease, or any altered physiologic state in which renal blood flow is being maintained by compensatory vasodilatation (9,11,14). To prevent renal toxicity in patients who are at risk, NSAIDs should be started in modest doses and then carefully increased. Patients should be instructed to report if signs of fluid retention evidenced by weight gain or edema develop, if they become ill and dehydrated, or if they are to begin treatment with diuretics or angiotensin-converting enzyme (ACE) inhibitors.

Since renal insufficiency induced by physiologic mechanisms occurs soon after administration of NSAIDs, it is prudent to monitor serum creatinine in high-risk patients every week for several weeks after treatment is started. The immune-mediated or idiosyncratic syndromes of acute interstitial nephritis and NSAID-induced nephrotic syndrome (usually associated with interstitial nephritis) can occur immediately after starting NSAIDs or at any time up to 18 months later. The average time of drug exposure has been 6.6 months for NSAID-induced nephrotic syndrome and 15 days for allergic interstitial nephritis (9).

NSAIDs may cause elevation of liver enzyme levels, but severe hepatotoxicity is rare (25). There is no evidence that abnormal findings on liver function tests in the absence of clinical symptoms change the outcome or are associated with serious hepatotoxicity (25). The value of routine liver function test monitoring for most patients receiving NSAIDS is uncertain. Liver function should be monitored in patients who are treated with diclofenac or in those who have intrinsic liver disease or in whom it is suspected.

Disease-modifying antirheumatic drugs (DMARDs)

Hydroxychloroquine (HCQ). The major toxicity of antimalarial agents is retinal damage (rare), which can lead to visual impairment (26-31). Compared with other available DMARDs, HCQ has the least toxicity and is the least costly to monitor (2,32). Additional rare and usually less serious toxicities include GI symptoms, myopathy, blurred vision, accommodation difficulty, abnormal skin pigmentation, and peripheral neuropathy. The major risk factor for retinal toxicity appears to be the combination of cumulative dose >800 gm and age >70 years (presumably due to the increased prevalence of macular disease in the elderly) (33). A daily HCQ dosage of >6.0-6.5 mg/kg, particularly in patients with abnormal hepatic or renal function, may also be associated with an increased risk of retinal toxicity (26,34).

Patients taking HCQ should be cautioned to report any visual symptoms, particularly difficulty seeing entire words or faces, intolerance to glare, decreased night vision, or loss of peripheral vision. These symptoms of peripheral retinal toxicity should prompt drug discontinuation and ophthalmologic evaluation.

The goal of monitoring HCQ therapy is to detect early reversible retinal toxicity. A baseline eye evaluation is not routinely recommended in patients younger than age 40 and with no family history of eye disease. If a patient has had a clinical response to HCQ after 6 months, then a monitoring routine should be instituted. Patients with abnormal renal function or those who have received HCQ for more than 10 years require more frequent ophthalmologic evaluation. In the absence of risk factors, it is recommended that an ophthalmologic examination and central field testing be performed every 6-12 months. The central 10[degree] of the visual field is the initial site of antimalarial retinal toxicity. An Amsler test or a modified Amsler test can be used to screen for this early abnormality (35). This can be administered by self-testing if the patient is reliable, or by the patient's primary physician, to augment formal ophthalmologic testing.

Sulfasalazine (SSZ). Hematologic toxicities of SSZ, including leukopenia (1-3%), thrombocytopenia (rare), hemolysis in patients with glucose 6-phosphate dehydrogenase (G6PD) deficiency, agranulocytosis (rare), and aplastic anemia (rare), are the most serious potential side effects of SSZ (36). Except for G6PD deficiency and sulfa allergy, there are no known risk factors. Leukopenia is most likely to occur in the first 6 months of treatment (37,38), but may rarely occur later. Early dosage reduction and/or cessation may reverse leukopenia. More common but less serious toxicities include skin rashes, photosensitivity, headaches, mood alterations, and GI symptoms such as nausea, vomiting, anorexia, abdominal pain, dyspepsia, and indigestion (3). Patients should be questioned about previous allergies to sulfa drugs and cautioned about the development of possible oligospermia (low sperm count) (39,40). The main goal of monitoring is to detect the hematologic toxicities early. Some experts have recommended that liver enzyme levels be monitored in patients receiving SSZ, but supporting data for this recommendation are not available; nevertheless, a baseline assessment of aspartate aminotransferase or alanine aminotransferase is prudent in patients with known or suspected liver disease.

Methotrexate (MTX). The most serious toxicities of MTX include hepatic fibrosis (rare) and cirrhosis (rare), pneumonitis (uncommon), and myelosuppression. Independent risk factors for the development of serious liver disease (biopsy-proven cirrhosis or clinically evident liver disease such as ascites, esophageal varices, hepatic encephalopathy, etc.) in patients with RA include age and duration of therapy, as identified in a recent case-control study (41). Other potential risk factors for hepatic toxicity that have been suggested but were not identified in that small RA cohort study include obesity, diabetes, alcohol intake, and prior history of hepatitis B or C (41,42).

Prevention of hepatic fibrosis and cirrhosis includes the avoidance of MTX in patients with liver disease or another important risk factor. In patients with suspected liver disease, a pretreatment liver biopsy should be obtained. Prevention also includes advising the patient against alcohol consumption while taking MTX. Patients should report symptoms of jaundice or dark urine.

Routine surveillance liver biopsies are not recommended for RA patients receiving MTX in the recommended doses (43). Liver biopsy is not a cost-effective means of monitoring, at least for the first 10 years of therapy in patients with no abnormal ities identified on liver function tests (44). Liver bi opsy is recommended for patients with liver function abnormalities that persist during treatment with, or following discontinuation of, MTX (43).

Risk factors for myelosuppression include the use of antifolate agents such as trimethoprim, the presence of folate deficiency, and renal insufficiency (42). Severe myelosuppression is an uncommon complication of low-dose (5-20 mg/week) MTX therapy (42). The rationale for monitoring is to decrease the incidence and severity of severe myelosuppression and its complications, such as sepsis, severe anemia, and bleeding. The baseline evaluation consists of a complete blood cell count (CBC) with differential cell count. Monitoring consists of a CBC and platelet count performed every 4-8 weeks. Mean corpuscular volume >100 may indicate folate deficiency and predict myelosuppression (45). Because the kidneys are the primary route of excretion of MTX, renal insufficiency may lead to myelosuppressive levels of the drug. Routine monitoring of renal function every 4-8 weeks is therefore recommended (46).

Pneumonitis is an uncommon complication of long-term MTX therapy, with a frequency on the order of 2-6% (42). Precise risk factors for the development of pneumonitis are unknown. However, patients with preexisting lung damage have reduced pulmonary reserve and therefore have a greater likelihood of severe morbidity should this complication occur (47). Pneumonitis due to MTX can occur at any time during a course of therapy and at any dosage. Review of a radiograph obtained within 1 year prior to the initia tion of MTX therapy is recommended to determine if preexisting lung disease is present and to provide a baseline for future comparison (42,48). If evidence of significant lung disease is present, therapy with MTX should be reconsidered. Monitoring consists of assessing symptoms of pneumonitis, such as cough, dyspnea on exertion, or shortness of breath, at each followup visit.

Common but less serious toxicities of MTX include mucositis, mild alopecia, and GI disturbances, which may be caused by folate depletion (42). These toxicities are often treated or prevented with the use of folate supplementation, which should be considered in all patients taking MTX. Folic acid at a dosage of 1 mg per day or 7 mg once a week is less expensive and less complicated than the use of folinic acid. Neither low-dose folate (1 mg per day) nor folinic acid (<=5 mg per week) interferes with the beneficial effect of MTX (49,50).

Because of the teratogenic potential of MTX, pregnancy should be avoided if either partner is receiving the drug. Male patients should wait a minimum of 3 months after discontinuation of therapy. Female patients should wait at least 1 ovulatory cycle after discontinuation of MTX therapy before attempting conception (51,52).

Recent case reports suggest a possible association between MTX and lymphoma (53-55). However, a large retrospective study of 16,263 patients with RA showed no increased risk (56). In that study, only 12 of 39 patients who developed lymphoma were treated with MTX, and of those, there was no relationship with cumulative dose or duration of treatment (56). More studies are required; nevertheless, patients should be advised to report any lymph node swelling, and the lymph nodes should be routinely examined by the treating physician.

Gold compounds (aurothioglucose, aurothiomalate, auranofin). The major serious toxicities of gold compounds--hematologic, renal, and pulmonary--are rare (31,57-60). Other toxicities include oral ulcers (common), rash (common), pruritus without rash (uncommon), and vasomotor reactions (with parenteral gold, especially aurothiomalate) (60). The principal hematologic toxicities include thrombocytopenia (1- 3%) and aplastic anemia (<1%), which may occur suddenly and are believed to be idiosyncratic (60). The most common renal toxicity and the one which requires monitoring is membranous nephropathy, which is generally heralded by the development of proteinuria or hematuria. Isolated microscopic hematuria may occur in the course of gold therapy, or sometimes may be seen in RA in the absence of gold therapy and does not necessarily predict the development of serious renal disease. For patients who develop qualitative proteinuria, a 24-hour urinalysis should be obtained and cessation of the drug should be considered if protein excretion is >500 mg/24 hours.

Auranofin (oral gold) is associated with lower rates of both renal and hematologic toxicity than are parenteral gold compounds but may be less effective in controlling the disease (32). Its minor toxicities include diarrhea (common) and mucocutaneous reactions.

Hematologic and renal toxicities may occur at any time during the course of gold therapy. Except for the suggestion of genetic susceptibility, there are no known risk factors for gold toxicity. Patients need to be educated about the need for frequent monitoring and for prompt reporting of the development of rash, mucositis, hematuria or bleeding, or any new illness while receiving gold.

D-penicillamine (DP). The side effects of DP are rash (common), stomatitis (common), dysgeusia or metallic taste (common), myelosuppression (especially thrombocytopenia) (rare), and proteinuria (rare) (61). Other significant but rare toxicities include nephrotic syndrome or renal failure and induction of autoimmune syndromes such as systemic lupus erythematosus, myasthenia gravis, polymyositis, and Goodpasture's syndrome (61). Slowly increasing the dosage of DP by 125-250-mg increments every 3 months up to 750 mg per day seems to decrease the incidence of thrombocytopenia (61). Patients taking DP should report any new symptoms, especially rash, hematuria, or bleeding. As with gold, monitoring is directed at discontinuation of the medication in the presence of likely toxicity.

Azathioprine (AZA). AZA is a purine analog which is capable of inducing myelosuppression at dosages used to treat RA (1-2 mg/kg/day) (62). The rationale for monitoring is to decrease the incidence and severity of myelosuppression and its complications such as sepsis, severe anemia, and bleeding. Risk factors for myelosuppression include the use of concomitant allopurinol or ACE inhibitors and the presence of renal insufficiency. Prevention consists of reducing the dosage of AZA to one-fourth the usual dosage with concomitant allopurinol, avoiding the use of concomitant ACE inhibitors, and decreasing the dosage of AZA in patients with renal insufficiency. GI intolerance is the most common side effect of AZA therapy, resulting in discontinuation in [approx]10% of treated patients. Pancreatitis rarely may occur with AZA. The long-term risk of lymphoproliferative disorders due to AZA is debated, but does not appear to be significantly greater than that observed in RA patients not taking cytotoxic agents (62).

Glucocorticoids

The toxicities of low-dose systemic glucocor ticoids (<=10 mg prednisone daily or equivalent) in clude increased appetite, weight gain, fluid retention, acne, development of cushingoid facies, hypertension, diabetes, atherosclerosis, glaucoma and cataract formation, osteoporosis, a vascular necrosis, increased susceptibility to infection, and impaired wound healing. A decision to initiate or to increase the dosage of systemic steroids for the patient with RA should include an assessment of the patient's risk factors for adverse steroid effects, e.g., family history of diabetes, established hypertension or diabetes, preexisting cataract(s) or glaucoma, and documented low bone mineral density, history of osteoporotic fracture, or significant osteoporosis risk factors such as premature menopause. The patient should be informed about potential side effects, the importance of taking the medication only as directed, the importance of limiting the dosage and duration of glucocorticoid use, the potential difficulty of discontinuing prednisone in a patient with active RA, and the danger of abrupt cessation of the medication after long-term use. A medical alert bracelet should be worn by patients receiving long-term glucocorticoid therapy. Patients should be advised regarding smoking cessation and reduction of cholesterol intake to minimize cardiovascular risk factors.

The need for baseline studies to monitor glucocorticoid toxicity varies with the patient. Initial assessment may include measuring and recording weight and blood pressure, serum glucose and cholesterol levels (with high-density lipoprotein and low-density lipoprotein), and, in patients at high risk for osteoporosis, consideration of bone mineral density measurement and supplementation with calcium and vitamin D. Baseline eye examination and tonometry should be considered in patients over the age of 65 or with a family history of glaucoma (63).

Agents reserved for refractory RA or severe extraarticular complications

Cyclophosphamide, chlorambucil, and cyclosporin A are agents that are not Food and Drug Administration-approved for RA treatment (64). Their use is reserved for patients with refractory RA or with severe extraarticular complications such as vasculitis, corneal perforation, etc. Complicated RA is usually managed by a rheumatologist. However, since primary care physicians may participate in the care of patients taking these medications and may be required to monitor their toxicities, the guidelines for use of these agents are included in Table 1.

Antirheumatic agents and teratogenicity, lactation, and fertility

The majority of patients with RA are women, and many are in their reproductive years. Therefore, the effect of these drugs on fertility, their teratogenic potential, and their excretion in breast milk are important issues (52,65-68). Table 2 summarizes current information on this and is intended for use only as a guide. Decisions regarding the use of all medications in pregnancy require careful consideration of the risks and benefits to both mother and fetus (66,67).

Table 2. Antirheumatic drug therapy in pregnancy and lactation, and effects on fertility* Click here

Summary

Drugs used to treat RA may cause death, disability, and diseases, especially if the treatment continues in the setting of undetected toxicity. Prevention of toxicity may be enhanced by pretreatment assessment of individual risk factors for toxicity and by careful patient and physician education about safe use of the drug. Patients and their physicians must be alert to the signs and symptoms of toxicity that should prompt discontinuation of the drug and physician reassessment. Some drug toxicity may be discovered by appropriate laboratory monitoring before serious problems become clinically apparent.

The 3 major drug categories for the treatment of RA are the NSAIDs, DMARDs, and glucocorticoids. Most NSAIDs have common GI and renal toxicity that may be averted by careful patient selection and administration of the drug. The individual DMARDs have specific toxicities for which monitoring protocols have been developed. The serious side effects of systemic glucocorticoids are largely related to dose and duration of treatment. The recommendations summarized in Table 1 are for basic monitoring in patients with uncomplicated RA. Additional monitoring may be appropriate for patients with comorbid disease, concurrent medication, or other risk factors.

ACKNOWLEDGMENTS

The authors thank Drs. Doyt Conn, John Esdaile, Simon Helfgott, Herbert Kaplan, Donald Middleton, Daniel Rahn, Shaun Ruddy, Michael Schiff, Terence Starz, and Michael Weinblatt, and the American College of Rheumatology Committee on Rheumatologic Care. We also thank Donna Cosola, Steve Echard, Jacqueline Mazzie, and Mary Scamman for technical assistance.