DRUG INTERACTIONS AND REACTIONS UPDATE Edited by Edward A. Hartshorn
METHOTREXATE AND NONSTEROIDAL ANTIINFLAMMATORY DRUG INTERACTIONS Maureen L. Frenia and Kimberly S. Long
OBJECTIVE: To determine if the coadministration of methotrexate (MTX) and nonsteroidal antiinflanunatory drugs (NSAIDs) results in a clinically significant drug interaction. DATASOURCES: A case report of hematologic toxicity following the administration of MTX and flurbiprofen at our institution is presented. Six previously published case reports and five pharmacokinetic studies regarding MTX and NSAID interactions are available to assist in the evaluation of this potential interaction. DATASYNTHESIS: Cases of various clinical manifestations during concomitant MTX and NSAID administration, including acute renal failure and pancytopenia, have been reported. The exact mechanism of the interaction has not been fully elucidated. Suggested theories to explain the mechanism of MTX toxicity include reduction in MTX clearance secondary to renal capillary constriction induced by NSAIDs, displacement of MTX or its metabolite from plasma proteins, competition between MTX and NSAIDs for renal tubular excretion, or impairment of hepatic metabolism of MTX by NSAIDs. Studies comparing MTX pharmacokinetics with or without concurrent NSAID therapy have not shown statistical differences in the parameters evaluated. However, one study did demonstrate differences in the pharmacokinetics of7-hydroxymethotrexate, the active metabolite of MTX, when MTX was administered with aspirin. CONCLUSIONS: Although a clinically significant interaction does not occur in all patients, numerous case reports are available that demonstrate possible problems following the coadministration of MTX and NSAIDs. To date, the specific circumstances during which the reaction may occur have not been well defmed.
AnnPharmacother 1992;26:234-7. METHOTREXATE (MTX), a folic acid antagonist, is indicated for the treatment of rheumatoid arthritis, psoriasis, and ma-
MAUREEN L. FRENIA, Pharm.D .. is a Clinical Toxicology/Emergency Medicine Fellow, Department of Pharmacy, University Medical Center, Jacksonville, FL; at the time of writing, she was a Doctor of Pharmacy Student, College of Pharmacy, Rutgers-The State University of New Jersey; KIMBERLY S. LONG, Pharm.D., is an Assistant Professor, Department of Pharmacy Practice and Administration, College of Pharmacy, Rutgers-The State University of New Jersey, Piscataway, NJ, and a Clinical Pharmacist, The Medical Center at Princeton, Princeton, NJ. Reprints: Kimberly S. Long, Pharm.D., Department of Pharmacy Practice and Administration, College of Pharmacy, Busch Campus, Rutgers-The State University of New Jersey, P.O. Box 789, Piscataway, NJ 08855. EDWARD A, HARTSHORN, Ph.D., is a Clinical Professor of Pharmacy, University of Texas, and a Clinical Professor of Pharmacology, University of Texas Health Sciences Center, San Antonio, TX 78284.
234 •
The Annals ofPharmacotherapy •
lignancy. It is not uncommon to also fmd nonsteroidal antiinflammatory drugs (NSAIDs) incorporated into the treatment regimen of patients with the above diagnoses, A review of the literature suggests that an interaction between MTX and NSAIDs may occur when these drugs are administered concurrently. Complications have been observed with both high-dose MTX therapy used in oncology and low-dose MTX used to treat rheumatoid arthritis and psoriasis, Acute renal failure has been one of the most prominent manifestations reported. Other signs and symptoms apparently resulting from the interaction include: mucosal ulcerations, pancytopenia, diarrhea, vomiting, elevated liver transaminases, jaundice, pyrexia, and confusion. In our institution, we observed a case of hematologic toxicity when MTX and flurbiprofen were administered concurrently. CASE REPORT
A 74-year-old woman was hospitalized with hematemesis, diarrhea with melena, nausea, and weakness for one week. She had undergone a four-vessel coronary artery bypass graft three years prior and had a long history of rheumatoid arthritis that was being treated with MTX 2.5 mg three times a week for the last three years. One to two weeks prior to admission, flurbiprofen 100 mg/d had been added to her treatment regimen for rheumatoid arthritis. Her other medications at this time included folic acid, digoxin, gemfibrozil, iron supplements, and dipyridamole. On admission, her physical examination was unremarkable except for her abdominal examination, which revealed mild tenderness in the epigastric area. An upper endoscopy was consistent with acute gastrointestinal bleeding. Laboratory values on admission were: hemoglobin 52 gIL, hematocrit 0.14, white blood cells (WBC) 6.4 x 109/Lwith 0.02 segmented neutrophils and no bands,platelets 11 x ]()9/L, and red blood cells 1.5x 1()I2/L. Two weeks prior to admission her platelet count was 263 x 109/L. Her prothrombin time was slightly elevated A serum MTX concentration was not obtained. She was diagnosed with anemia, neutropenia, and thrombocytopenia secondary to the MTX therapy. Knowing that the patient had previously received MTX for three years without complications, it was hypothesized that the addition of flurbiprofen may have exacerbated the effects of
MTX. Treatment consisted of packed red blood cell and platelet transfusions, folic acid, vitamin K, sucralfate, and ranitidine. During her hospital stay, her WBC count decreased to 4.0 x 109/L with zero neutrophils. On hospital day 5 the patient developed a fever and was started on mezlocillin and gentamicin for pre-
1992 February, Volume 26
Downloaded from aop.sagepub.com at CMU Libraries - library.cmich.edu on September 11, 2015
sumed leukopenic sepsis, although all blood cultures were negative. A full recovery was eventually made and the patient was discharged on day 12. All laboratory values at this time were within normal limits.
Our patient's hematologic complications were attributed to the concomitant administration of flurbiprofen and MTX. Case reports have demonstrated similar toxicity with MTX and a variety of NSAIDs; however, to date no reports involve the combination ofMTX and flurbiprofen.l" Published reports describing the morbidity and mortality associated with concomitant administration of MTX and NSAIDs are limited. In a retrospective review of 36 patients who received a total of 118 cycles of high-dose MTX therapy for the treatment of malignancy, four of nine cases of severe MTX toxicity occurred in patients receiving concurrent ketoprofen. Simultaneous administration of MTX and ketoprofen resulted in renal impairment, bone marrow suppression, stomatitis, and eventually death in 3 of the 4 patients studied. MTX concentrations were elevated after ketoprofen administration; however, no MTX toxicity was demonstrated when ketoprofen was administered at least 12 hours after the MTX infusions were completed. In the same study, another patient receiving MTX developed severe MTX toxicity following the concurrent use of diclofenac.' In another report, a 30-year-old woman experienced reversible bone marrow depression and mucosal ulceration in the oral and genital regions after receiving MTX and apazone (a nonsteroidal agent used in Europe) for the treatment of psoriasis. The patient had been taking MTX 25 mg/wk for four years without adverse effects. Ulcerations, sore throat, and bloody diarrhea occurred ten days after apazone therapy was initiated for joint pain.' Gabrielli et al. described a fatal outcome after the concomitant administration of MTX and NSAIDs. The patient, a 71-year-old woman with a history of rheumatoid arthritis, was prescribed indomethacin 50 mg/d and diclofenac 100 mg/d. After eight days of NSAID therapy, one dose ofMTX 10 mg was administered intramuscularly with a second dose given 24 hours later. Eight hours after the first MTX dose the patient became confused and, over the next few days, became comatose. She showed clinical signs of leukopenia and anemia by day 12 and continued to deteriorate. She died on day 15.3 Ellison and Servi reported the deaths of two oncology patients who received intermediate-dose MTX and fluorouracil regimens, despite the administration of leucovorin rescue. Both patients had increased MTX concentrations when indomethacin and MTX were given concomitantly. This resulted in severe mucositis, anemia, leukopenia, and renal failure after seven to ten days of concurrent indomethacin and MTX therapy.' In another report, a patient receiving low-dose, intermittent MTX for rheumatoid arthritis developed MTX toxicity when naproxen was added to her regimen. The patient was hospitalized following seven days of fever and diarrhea that originated 24 hours after concurrent MTX and naproxen therapy. On admission the patient had diarrhea, melena, vomiting, hematemesis, epistaxis, mucosal ulcerations, and pancytopenia, which eventually led to her death. It was later noted that the patient had taken 27.5 mg of MTX that week instead of the prescribed dose of 7.5 rug/wk.'
Maiche reported on a 65-year-old woman with nonHodgkin's lymphoma in whom acute renal failure occurred secondary to concurrent administration of indomethacin 90 mg/d with MTX 5 g followed 12 hours later by leucovorin rescue. Prior to MTX therapy the patient's serum creatinine was within normal limits; however, two days after concurrent administration of indomethacin and MTX her serum creatinine increased and the serum MTX concentration was higher than expected. At this time the patient had severe nausea and vomiting, a tender abdomen with guarding, severe mucositis, weakness, and confusion. After discontinuation of indomethacin, the patient recovered and was later released from the hospital on an intermediatedose MTX regimen/ The exact mechanism of the interaction between MTX and NSAIDs has not been determined. The most likely theories explaining the interaction include a decrease in MTX excretion secondary to renal capillary constriction induced by NSAIDs, displacement of MTX or its metabolite from plasma proteins, competition between MTX and NSAIDs for renal tubular secretion, or impairment of hepatic metabolism of MTX by NSAIDS.l,2,4,~ One of the most widely accepted theories is related to the fact that NSAIDs prevent the production of renal prostaglandins such as prostaglandin E 2 and prostacyclin. These prostaglandins are important in increasing renal blood flow and maintaining renal function. Renal prostaglandin inhibition results in a constriction of the renal capillaries, a decrease in renal blood flow, a reduction in MTX clearance, and a potentially increased serum MTX concentration, which may lead to life-threatening adverse effects. Most clinicians believe that this mechanism is involved in at least part of the interaction.w Another possible mechanism is the displacement of MTX from plasma proteins." Displacement from plasma proteins is most significant when a drug is more than 90 percent protein bound and has a volume of distribution that does not exceed the volume of extracellular water.' As MTX demonstrates low plasma protein binding (30-70 percent), it is unlikely that an increase in MTX concentrations and toxicity can be attributed to displacement by NSAIDS.l,2,4,8,lO-12 However, the major metabolite of MTX, 7-hydroxymethotrexate (7-0H-MTX), is highly bound (91-93 percent) to plasma proteins and its displacement may playa role in the clinical signs of MTX toxicity.t-" Reports in the literature have indicated that 7-0H-MTX has cytotoxic properties in vitro and has caused renal toxicity in humans and animals." Because both MTX and NSAIDs are eliminated largely by the kidneys, competition for secretion in the renal tubule has also been suggested. l,2,?12 Serum concentrations of MTX in the circulation may increase if NSAIDs, rather than MTX, are selectively excreted, resulting in a decrease in MTX clearance and an increase in serum concentrations. Ibuprofen and indomethacin decreased the clearance of MTX in two patients with bladder and esophageal cancer, respectively. The patients' renal function did not change; therefore, the authors postulated that these weak acids competed with MTX for excretion.' Another possible mechanism proposed to account for the interaction between MTX and NSAIDs is through altered hepatic metabolism of MTX and NSAIDs.2 In this case, the interaction would result in a decrease in the me-
The Annals ofPharmacotherapy
•
1992 February, Volume 26 •
Downloaded from aop.sagepub.com at CMU Libraries - library.cmich.edu on September 11, 2015
235
tabolism and clearance of MTX, thereby prolonging plasma MTX concentrations. However, renal excretion of unchanged drug accounts for the majority of MTX elimination. Only a small percentage of MTX undergoes hepatic metabolism to 7-0H-MTX, which makes the significance of this mechanism appear less likely. As a result of the increasing concern regarding this interaction, pharmacokinetic studies have begun to address this potential drug interaction. To date, no study has been able to document a consistent pharmacokinetic interaction between NSAIDs and MTX.II.13-16 Ahern et al. monitored 13 rheumatoid arthritis patients who received low-dose MTX (15 mg po) with and without NSAID therapy. The individual NSAID and dose varied for each patient. More than half of the patients received sulindac. Naproxen, indomethacin, diclofenac, and diflunisal were also evaluated. Overall, pharmacokinetic variables did not differ significantly, regardless of whether patients were receiving NSAIDs. However, individual patients did have noticeable differences in their pharmacokinetic parameters with and without concomitant administration of NSAIDs. No specific risk factors or NSAIDs common to these patients were observed." Dupuis et al. also conducted a study to evaluate the pharmacokinetic interaction between MTX and various NSAIDs (e.g., tolmetin, indomethacin, naproxen, aspirin). The pharmacokinetics of MTX, both in the presence and absence of NSAIDs, were studied in seven children who had been maintained on weekly oral doses of MTX for control of their chronic inflammatory arthritis. More than one NSAID was administered to six of the children. Increases in half-life and area under the curve (AVe) and decreases in apparent MTX clearance were noted in six of the patients while they received NSAIDs. However, the only MTX variable that had a statistically significant difference was the half-life, which increased in the presence of NSAIDs (p=O.03). Overall, the half-life of MTX in these children was shorter than that usually seen in adults." This may reduce the potential for the interaction between MTX and NSAIDs in children; however, more studies in this area are warranted. Stewart et al. conducted a study to determine the effects of naproxen on MTX pharmacokinetics in 12 patients with rheumatoid arthritis and "adequate" renal function (serum creatinine ~1.4 mg/dL or measured creatinine clearance ;80 mLlmin). The study consisted of four treatment groups: oral MTX alone, oral MTX with naproxen, intravenous MTX with naproxen, and intravenous MTX alone. Naproxen 500 mg bid was administered for 38 days with MTX 15 mg given orally on days 0 and 14 and parenterallyon day 28. Serum and urine samples were collected at various intervals for 24 hours after coadministration of the agents and pharmacokinetic parameters were calculated. No statistically significant difference was found between the groups when bioavailability, renal clearance, apparent oral clearance, systemic clearance, and MTX protein binding were assessed." Skeith et al. conducted a crossover study to determine the effect of flurbiprofen and ibuprofen on MTX pharmacokinetics. Six patients with arthritis and normal renal function were studied. All patients had been maintained on MTX for a minimum of six months prior to the study. MTX 10-25 mg/dose therapy was continued during the 236 •
The Annals ofPharmacotherapy
•
study. Patients received flurbiprofen 300 mg/d or ibuprofen 2400 mg/d with and without intramuscular and oral MTX in a six-way crossover design. Serum MTX concentrations were obtained following NSAID administration. The pharmacokinetic parameters of MTX evaluated included AVC per unit dose, maximum concentration, and serum half-life. No significant differences were found in any of the pharmacokinetic indices when MTX was administered with or without NSAIDs. The authors postulated that toxicities previously documented with the administration of these two agents may be a result of idiosyncratic reactions. IS Furst et al. evaluated the effects of sulindac and aspirin on the pharmacokinetics of MTX and 7-0H-MTX in 12 patients with rheumatoid arthritis. The patients had been maintained on low-dose MTX therapy (5-10 mg/mvwk) for a minimum of one year and had normal renal function. During the study all patients received MTX 10 mg/malone or with aspirin (45-50 mg/kg/d), or sulindac 400 mg/d in a three-way crossover design. Serum samples were obtained for 48 hours following the concurrent administration of the agents and HPLC was used to determine the concentrations of MTX and 7-0H-MTX. Clearance and AVC for the drug and metabolite were calculated. No significant differences in AVC, maximum concentration, or clearance of MTX were observed when the drug was administered alone or concomitantly with aspirin or sulindac. However, one subject did have very low MTX clearance when the drug was administered alone. If this patient had been excluded there would have been a statistically significant difference in MTX clearance when aspirin or sulindac was included in the regimen. Differences were noted when the pharmacokinetics of 7-0H-MTX were evaluated as there was a significant difference in the AVC and a nonsignificant increase in the maximum concentration between the control and the aspirin groups. An increase in AVC and maximum concentration of7-0H-MTX was also observed in the sulindac group; however, the difference was not significant." The probability of a type II error in the study was 18 percent, which makes the true significance of this study unknown. The pharmacokinetic studies discussed here had several limitations in their study design. The number of patients involved was small in all investigations. These studies may not have detected differences that could become significant if larger numbers of patients had been used. In addition, most patients had normal renal function. Patients with preexisting renal impairment (secondary to age, disease, or drugs) may be at higher risk because of the effects of NSAIDs on renal function and should also be evaluated. Furst, II Ahern," and Stewart" administered only one dose of MTX in their trials. Therefore, one cannot accurately predict the effect of NSAIDs on MTX following multipledose regimens. Finally, patients were monitored only for short periods of time (6-48 h) following the coadministration of NSAIDs and MTX. This may be appropriate for NSAIDs with short half-lives (i.e., aspirin, ibuprofen); however, agents with longer half-lives (i.e., sulindac, piroxicam) need to be evaluated for greater periods of time. Summary
Coadministration of MTX and NSAIDs occurs frequently and may result in significant toxicity. Clinical
1992 February, Volume 26
Downloaded from aop.sagepub.com at CMU Libraries - library.cmich.edu on September 11, 2015
Methotrexate and NSAIDs
manifestations, including pancytopenia and renal failure, have been demonstrated when patients received either high-dose or low-dose MTX together with NSAIDs. The exact mechanism of the interaction has not been determined; however, it is thought to occur at the level of the kidney. It is widely accepted that the reaction is related to a reduction in MTX clearance as a result of renal capillary constriction induced by NSAIDs. More recent data suggest a change in the pharmacokinetic parameters of the active metabolite of MTX, 7-0H-MTX, may be responsible. The possibility that the adverse effects reported in the literature may be attributed to idiosyncratic reactions cannot be ruled out. A combination of these factors may also be involved in the reaction. Although this interaction is not seen in all patients, numerous case reports describe the potential toxicity. In the case report presented here, signs and symptoms of the interaction occurred after one to two weeks of concomitant administration of MTX and NSAIDs. Large, well-controUedtrials to evaluate this reaction have not been conducted; as a result, the specific circumstances during which the reaction may occur have not been defined. Renal function should be monitored closely as increased MTX concentrations may be a result of NSAID -induced renal dysfunction in certain patients. Because the literature indicates a potential interaction between MTX and NSAIDs, caution should be used when these agents are administered concurrently. Clinicians should not be deterred from concomitant use of these drugs; however, judicious monitoring is warranted. ::::::
References I. Thyss A, Milano G, Kubar J, Namer M, Schneider M. Clinical and phannacokinetic evidence of a life-threatening interaction between methotrexate and ketoprofen. Lancet 1986;I:256-8. 2. Daly HM, Scott GL, Boyle J, Roberts CJc. Methotrexate toxicity precipitated by azapropazone. Br J Dermato/ 1986;114:733-5. 3. Gabrielli A, Leoni P, Danieli G. Methotrexate and nonsteroidal antiinflammatory drugs (letter). Br Med J 1987;294:776. 4. Ellison NM, Servi RJ. Acute renal failure and death following sequential intermediate-dose methotrexate and 5-FU: a possible adverse effect due to concomitant indomethacin administration. Cancer Treat Rep 1985;69:342-3. 5. Singh RR, Malaviya AN, Pandey IN, Guleria JS. Fatal interaction between methotrexate and naproxen (letter). Lancet 1986;I: 1390. 6. Maiche AG. Acute renal failure due to concomitant action of methotrexate and indomethacin (letter). Lancet 1986;I: 1390. 7. Bloom EJ, IgnotTo RJ, Reis CA, Cadman E. Delayed clearance of methotrexate associated with antibiotics and antiinflammatory agents (abstract). C/in Res 1986;34:560A. 8. Taylor JR, Halprin KM. Effect of sodium salicylate and indomethacin on methotrexate-serum albumin binding. Arch Dermato/ 1977; 113: 588-91. 9. Dettli L. Phannacokinetic aspects of drug interactions. In: Cluff LE, Petrie lC, eds. Clinical effects of interactions between drugs. New York: Elsevier, 1974:39-68. 10. Siordal L, Sager G, Aarbakke J. Phannacokinetic interactions with methotrexate: is 7-hydroxymethotrexate the culprit? (letter). Lancet 1988;1:591-2. II. Furst DE, Herman RA, Koehnke R, Ericksen N, Hash L, Riggs CEo et aI. Effect of aspirin and sulindac on methotrexate clearance. J Pharm Sci 1990;79:782-6. 12. Shen DD, AzarnotT DL. Clinical pharmacokinetics of methotrexate. Clin Pharmacokinet 1978;3:1-13. 13. Ahem M, Booth J, Loxton A, McCarthy P, Mellin P, Kevat S. Meth-
otrexate kinetics in rheumatoid arthritis: is there an interaction with nonsteroidal anti-inflammatory drugs? J Rheumatol 1988;15: 1356-60. 14. Stewart CF, Fleming RA, Arkin CR, Evans WE. Coadministrationof naproxen and low-dose methotrexate in patients with rheumatoid arthritis. C/in Pharmacol Ther 1990;47:540-6. 15. Skeith KJ, Russell AS, Jamali F, Coates J, Friedman H. Lack of significant interaction between low-dose methotrexate and ibuprofen or flurbiprofen in patients with arthritis. J Rheumatol 1990;17:1008-10. 16. Dupuis LL, Koren G, Shore A, Silverman ED, Laxer RM. Methotrexate-nonsteroidal antiinflammatory drug interaction in children with arthritis. J Rheumato/ 1990;17:1469-73.
EXTRACfO Metotrexato (MTX) y los agentes antiinfiamatorios no esteroidales (AANE) son frecuentemente administrados concomitantemente para el tratamiento de artritis reumatoidea, psoriasis, y malignidad. Evidencia reciente indica que la administraci6n conjunta de estos dos agentes podrfa ser potencialmente peligrosa. Reportes de casas de varias manifestaciones clfnicas, incluyendo fallo renal agudo y pancitopenia, han sido publicados. Los autores observaron una paciente de 74 aiios de edad que desarroll6 toxicidad hematol6gica subsiguiente a la administraci6n de MTX con flurbiprofen. EI mecanismo exacto de la interacci6n no se ha elucidado por completo. Teorias que han sido sugeridas para explicar el mecanismo de toxicidad de MTX incluyen reducci6n en la excresi6n de MTX secundaria a constricci6n de los capilares renales inducida por AANE, desplazamiento de MTX 0 su metabolito de las proteinas plasmaticas, competencia entre MTX y AANE por la excresi6n tubular renal 0 deterioro del metabolismo hepatico de MTX por AANE. Estudios que comparan la farmacocinetica de MTX durante la coadministraci6n de MTX con AANE y MTX solo no han demostrado diferencias estadfsticas en los parametres evaluados. Sin embardo, un estudio demostr6 diferencias en la farmacocinetica de 7-hidroximetotrexato, el metabolito activo de MTX, aI administrar MTX con sulindac 0 aspirina. Aunque la reacci6n no occurre en todos los pacientes, existen un mirnero de reportes de casos que demuestran posibles problemas subsiguientes a la coadministraci6n de MTX y AANE. La literatura indica el potencial para una interacci6n entre MTX y AANE, sin embargo, las circunstancias especfficas durante las cuales la reacci6n podrfa verse no han sido bien defmidas. BRENDA R. MORAND
RESUME
La pharmacotherapie de I'arthrite rhumatoide, du psoriasis ou de certaines tumeurs mal ignes implique frequemrnent la coadministration du methotrexate (MTX) et d'un antiinfiammatoire non-steroidien (AINS). De nouvelles evidences indiquent que I'utilisation concomitante de ces deux agents pourrait causer certaines reactions indesirables telles que pancytopenic, insuffisance renale aigue et toxicite hematologique. Le mecanisme exact de cette interaction medicamenteuse n' a pas ete clairement etabli, Plusieurs hypotheses ont cependant ete suggerees et incluent (I) reduction de l'elimination urinaire du MTX suite a une vasoconstriction renale induite par les AINS, (2) deplacement du MTX de ses sites de liaison aux proteines plasmatiques, (3) competition entre Ie MTX et les AINS pour les sites de secretion tubulaire, (4) alteration du metabolisme hepatique du MTX par les AINS. Aucune etude n' a demontre une difference dans la pharmacocinetique du MTX en presence d'un AINS. Cependant, une etude rapporte que I' administration de sulindac ou d' aspirine engendrerait des changements dans la pharmacocinetique du metabolite actif du MTX, Ie 7-hydroxyMTX. Considerant qu' aucun facteur pouvant predisposer a une telle interaction rnedicamenteuse n' a ete identifie ace jour, un monitoring pharmacotherapeutique etroit devrait etre assure par tout clinicien utilisant la combinaison MTX et AINS.
The Annals ofPharmacotherapy
SYLVIE ROBERT
•
Downloaded from aop.sagepub.com at CMU Libraries - library.cmich.edu on September 11, 2015
1992 February, Volume 26 •
237
METHOTREXATE AND NONSTEROIDAL ANTIINFLAMMATORY DRUG INTERACTIONS Maureen L. Frenia and Kimberly S. Long
OBJECTIVE: To determine if the coadministration of methotrexate (MTX) and nonsteroidal antiinflanunatory drugs (NSAIDs) results in a clinically significant drug interaction. DATASOURCES: A case report of hematologic toxicity following the administration of MTX and flurbiprofen at our institution is presented. Six previously published case reports and five pharmacokinetic studies regarding MTX and NSAID interactions are available to assist in the evaluation of this potential interaction. DATASYNTHESIS: Cases of various clinical manifestations during concomitant MTX and NSAID administration, including acute renal failure and pancytopenia, have been reported. The exact mechanism of the interaction has not been fully elucidated. Suggested theories to explain the mechanism of MTX toxicity include reduction in MTX clearance secondary to renal capillary constriction induced by NSAIDs, displacement of MTX or its metabolite from plasma proteins, competition between MTX and NSAIDs for renal tubular excretion, or impairment of hepatic metabolism of MTX by NSAIDs. Studies comparing MTX pharmacokinetics with or without concurrent NSAID therapy have not shown statistical differences in the parameters evaluated. However, one study did demonstrate differences in the pharmacokinetics of7-hydroxymethotrexate, the active metabolite of MTX, when MTX was administered with aspirin. CONCLUSIONS: Although a clinically significant interaction does not occur in all patients, numerous case reports are available that demonstrate possible problems following the coadministration of MTX and NSAIDs. To date, the specific circumstances during which the reaction may occur have not been well defmed.
AnnPharmacother 1992;26:234-7. METHOTREXATE (MTX), a folic acid antagonist, is indicated for the treatment of rheumatoid arthritis, psoriasis, and ma-
MAUREEN L. FRENIA, Pharm.D .. is a Clinical Toxicology/Emergency Medicine Fellow, Department of Pharmacy, University Medical Center, Jacksonville, FL; at the time of writing, she was a Doctor of Pharmacy Student, College of Pharmacy, Rutgers-The State University of New Jersey; KIMBERLY S. LONG, Pharm.D., is an Assistant Professor, Department of Pharmacy Practice and Administration, College of Pharmacy, Rutgers-The State University of New Jersey, Piscataway, NJ, and a Clinical Pharmacist, The Medical Center at Princeton, Princeton, NJ. Reprints: Kimberly S. Long, Pharm.D., Department of Pharmacy Practice and Administration, College of Pharmacy, Busch Campus, Rutgers-The State University of New Jersey, P.O. Box 789, Piscataway, NJ 08855. EDWARD A, HARTSHORN, Ph.D., is a Clinical Professor of Pharmacy, University of Texas, and a Clinical Professor of Pharmacology, University of Texas Health Sciences Center, San Antonio, TX 78284.
234 •
The Annals ofPharmacotherapy •
lignancy. It is not uncommon to also fmd nonsteroidal antiinflammatory drugs (NSAIDs) incorporated into the treatment regimen of patients with the above diagnoses, A review of the literature suggests that an interaction between MTX and NSAIDs may occur when these drugs are administered concurrently. Complications have been observed with both high-dose MTX therapy used in oncology and low-dose MTX used to treat rheumatoid arthritis and psoriasis, Acute renal failure has been one of the most prominent manifestations reported. Other signs and symptoms apparently resulting from the interaction include: mucosal ulcerations, pancytopenia, diarrhea, vomiting, elevated liver transaminases, jaundice, pyrexia, and confusion. In our institution, we observed a case of hematologic toxicity when MTX and flurbiprofen were administered concurrently. CASE REPORT
A 74-year-old woman was hospitalized with hematemesis, diarrhea with melena, nausea, and weakness for one week. She had undergone a four-vessel coronary artery bypass graft three years prior and had a long history of rheumatoid arthritis that was being treated with MTX 2.5 mg three times a week for the last three years. One to two weeks prior to admission, flurbiprofen 100 mg/d had been added to her treatment regimen for rheumatoid arthritis. Her other medications at this time included folic acid, digoxin, gemfibrozil, iron supplements, and dipyridamole. On admission, her physical examination was unremarkable except for her abdominal examination, which revealed mild tenderness in the epigastric area. An upper endoscopy was consistent with acute gastrointestinal bleeding. Laboratory values on admission were: hemoglobin 52 gIL, hematocrit 0.14, white blood cells (WBC) 6.4 x 109/Lwith 0.02 segmented neutrophils and no bands,platelets 11 x ]()9/L, and red blood cells 1.5x 1()I2/L. Two weeks prior to admission her platelet count was 263 x 109/L. Her prothrombin time was slightly elevated A serum MTX concentration was not obtained. She was diagnosed with anemia, neutropenia, and thrombocytopenia secondary to the MTX therapy. Knowing that the patient had previously received MTX for three years without complications, it was hypothesized that the addition of flurbiprofen may have exacerbated the effects of
MTX. Treatment consisted of packed red blood cell and platelet transfusions, folic acid, vitamin K, sucralfate, and ranitidine. During her hospital stay, her WBC count decreased to 4.0 x 109/L with zero neutrophils. On hospital day 5 the patient developed a fever and was started on mezlocillin and gentamicin for pre-
1992 February, Volume 26
Downloaded from aop.sagepub.com at CMU Libraries - library.cmich.edu on September 11, 2015
sumed leukopenic sepsis, although all blood cultures were negative. A full recovery was eventually made and the patient was discharged on day 12. All laboratory values at this time were within normal limits.
Our patient's hematologic complications were attributed to the concomitant administration of flurbiprofen and MTX. Case reports have demonstrated similar toxicity with MTX and a variety of NSAIDs; however, to date no reports involve the combination ofMTX and flurbiprofen.l" Published reports describing the morbidity and mortality associated with concomitant administration of MTX and NSAIDs are limited. In a retrospective review of 36 patients who received a total of 118 cycles of high-dose MTX therapy for the treatment of malignancy, four of nine cases of severe MTX toxicity occurred in patients receiving concurrent ketoprofen. Simultaneous administration of MTX and ketoprofen resulted in renal impairment, bone marrow suppression, stomatitis, and eventually death in 3 of the 4 patients studied. MTX concentrations were elevated after ketoprofen administration; however, no MTX toxicity was demonstrated when ketoprofen was administered at least 12 hours after the MTX infusions were completed. In the same study, another patient receiving MTX developed severe MTX toxicity following the concurrent use of diclofenac.' In another report, a 30-year-old woman experienced reversible bone marrow depression and mucosal ulceration in the oral and genital regions after receiving MTX and apazone (a nonsteroidal agent used in Europe) for the treatment of psoriasis. The patient had been taking MTX 25 mg/wk for four years without adverse effects. Ulcerations, sore throat, and bloody diarrhea occurred ten days after apazone therapy was initiated for joint pain.' Gabrielli et al. described a fatal outcome after the concomitant administration of MTX and NSAIDs. The patient, a 71-year-old woman with a history of rheumatoid arthritis, was prescribed indomethacin 50 mg/d and diclofenac 100 mg/d. After eight days of NSAID therapy, one dose ofMTX 10 mg was administered intramuscularly with a second dose given 24 hours later. Eight hours after the first MTX dose the patient became confused and, over the next few days, became comatose. She showed clinical signs of leukopenia and anemia by day 12 and continued to deteriorate. She died on day 15.3 Ellison and Servi reported the deaths of two oncology patients who received intermediate-dose MTX and fluorouracil regimens, despite the administration of leucovorin rescue. Both patients had increased MTX concentrations when indomethacin and MTX were given concomitantly. This resulted in severe mucositis, anemia, leukopenia, and renal failure after seven to ten days of concurrent indomethacin and MTX therapy.' In another report, a patient receiving low-dose, intermittent MTX for rheumatoid arthritis developed MTX toxicity when naproxen was added to her regimen. The patient was hospitalized following seven days of fever and diarrhea that originated 24 hours after concurrent MTX and naproxen therapy. On admission the patient had diarrhea, melena, vomiting, hematemesis, epistaxis, mucosal ulcerations, and pancytopenia, which eventually led to her death. It was later noted that the patient had taken 27.5 mg of MTX that week instead of the prescribed dose of 7.5 rug/wk.'
Maiche reported on a 65-year-old woman with nonHodgkin's lymphoma in whom acute renal failure occurred secondary to concurrent administration of indomethacin 90 mg/d with MTX 5 g followed 12 hours later by leucovorin rescue. Prior to MTX therapy the patient's serum creatinine was within normal limits; however, two days after concurrent administration of indomethacin and MTX her serum creatinine increased and the serum MTX concentration was higher than expected. At this time the patient had severe nausea and vomiting, a tender abdomen with guarding, severe mucositis, weakness, and confusion. After discontinuation of indomethacin, the patient recovered and was later released from the hospital on an intermediatedose MTX regimen/ The exact mechanism of the interaction between MTX and NSAIDs has not been determined. The most likely theories explaining the interaction include a decrease in MTX excretion secondary to renal capillary constriction induced by NSAIDs, displacement of MTX or its metabolite from plasma proteins, competition between MTX and NSAIDs for renal tubular secretion, or impairment of hepatic metabolism of MTX by NSAIDS.l,2,4,~ One of the most widely accepted theories is related to the fact that NSAIDs prevent the production of renal prostaglandins such as prostaglandin E 2 and prostacyclin. These prostaglandins are important in increasing renal blood flow and maintaining renal function. Renal prostaglandin inhibition results in a constriction of the renal capillaries, a decrease in renal blood flow, a reduction in MTX clearance, and a potentially increased serum MTX concentration, which may lead to life-threatening adverse effects. Most clinicians believe that this mechanism is involved in at least part of the interaction.w Another possible mechanism is the displacement of MTX from plasma proteins." Displacement from plasma proteins is most significant when a drug is more than 90 percent protein bound and has a volume of distribution that does not exceed the volume of extracellular water.' As MTX demonstrates low plasma protein binding (30-70 percent), it is unlikely that an increase in MTX concentrations and toxicity can be attributed to displacement by NSAIDS.l,2,4,8,lO-12 However, the major metabolite of MTX, 7-hydroxymethotrexate (7-0H-MTX), is highly bound (91-93 percent) to plasma proteins and its displacement may playa role in the clinical signs of MTX toxicity.t-" Reports in the literature have indicated that 7-0H-MTX has cytotoxic properties in vitro and has caused renal toxicity in humans and animals." Because both MTX and NSAIDs are eliminated largely by the kidneys, competition for secretion in the renal tubule has also been suggested. l,2,?12 Serum concentrations of MTX in the circulation may increase if NSAIDs, rather than MTX, are selectively excreted, resulting in a decrease in MTX clearance and an increase in serum concentrations. Ibuprofen and indomethacin decreased the clearance of MTX in two patients with bladder and esophageal cancer, respectively. The patients' renal function did not change; therefore, the authors postulated that these weak acids competed with MTX for excretion.' Another possible mechanism proposed to account for the interaction between MTX and NSAIDs is through altered hepatic metabolism of MTX and NSAIDs.2 In this case, the interaction would result in a decrease in the me-
The Annals ofPharmacotherapy
•
1992 February, Volume 26 •
Downloaded from aop.sagepub.com at CMU Libraries - library.cmich.edu on September 11, 2015
235
tabolism and clearance of MTX, thereby prolonging plasma MTX concentrations. However, renal excretion of unchanged drug accounts for the majority of MTX elimination. Only a small percentage of MTX undergoes hepatic metabolism to 7-0H-MTX, which makes the significance of this mechanism appear less likely. As a result of the increasing concern regarding this interaction, pharmacokinetic studies have begun to address this potential drug interaction. To date, no study has been able to document a consistent pharmacokinetic interaction between NSAIDs and MTX.II.13-16 Ahern et al. monitored 13 rheumatoid arthritis patients who received low-dose MTX (15 mg po) with and without NSAID therapy. The individual NSAID and dose varied for each patient. More than half of the patients received sulindac. Naproxen, indomethacin, diclofenac, and diflunisal were also evaluated. Overall, pharmacokinetic variables did not differ significantly, regardless of whether patients were receiving NSAIDs. However, individual patients did have noticeable differences in their pharmacokinetic parameters with and without concomitant administration of NSAIDs. No specific risk factors or NSAIDs common to these patients were observed." Dupuis et al. also conducted a study to evaluate the pharmacokinetic interaction between MTX and various NSAIDs (e.g., tolmetin, indomethacin, naproxen, aspirin). The pharmacokinetics of MTX, both in the presence and absence of NSAIDs, were studied in seven children who had been maintained on weekly oral doses of MTX for control of their chronic inflammatory arthritis. More than one NSAID was administered to six of the children. Increases in half-life and area under the curve (AVe) and decreases in apparent MTX clearance were noted in six of the patients while they received NSAIDs. However, the only MTX variable that had a statistically significant difference was the half-life, which increased in the presence of NSAIDs (p=O.03). Overall, the half-life of MTX in these children was shorter than that usually seen in adults." This may reduce the potential for the interaction between MTX and NSAIDs in children; however, more studies in this area are warranted. Stewart et al. conducted a study to determine the effects of naproxen on MTX pharmacokinetics in 12 patients with rheumatoid arthritis and "adequate" renal function (serum creatinine ~1.4 mg/dL or measured creatinine clearance ;80 mLlmin). The study consisted of four treatment groups: oral MTX alone, oral MTX with naproxen, intravenous MTX with naproxen, and intravenous MTX alone. Naproxen 500 mg bid was administered for 38 days with MTX 15 mg given orally on days 0 and 14 and parenterallyon day 28. Serum and urine samples were collected at various intervals for 24 hours after coadministration of the agents and pharmacokinetic parameters were calculated. No statistically significant difference was found between the groups when bioavailability, renal clearance, apparent oral clearance, systemic clearance, and MTX protein binding were assessed." Skeith et al. conducted a crossover study to determine the effect of flurbiprofen and ibuprofen on MTX pharmacokinetics. Six patients with arthritis and normal renal function were studied. All patients had been maintained on MTX for a minimum of six months prior to the study. MTX 10-25 mg/dose therapy was continued during the 236 •
The Annals ofPharmacotherapy
•
study. Patients received flurbiprofen 300 mg/d or ibuprofen 2400 mg/d with and without intramuscular and oral MTX in a six-way crossover design. Serum MTX concentrations were obtained following NSAID administration. The pharmacokinetic parameters of MTX evaluated included AVC per unit dose, maximum concentration, and serum half-life. No significant differences were found in any of the pharmacokinetic indices when MTX was administered with or without NSAIDs. The authors postulated that toxicities previously documented with the administration of these two agents may be a result of idiosyncratic reactions. IS Furst et al. evaluated the effects of sulindac and aspirin on the pharmacokinetics of MTX and 7-0H-MTX in 12 patients with rheumatoid arthritis. The patients had been maintained on low-dose MTX therapy (5-10 mg/mvwk) for a minimum of one year and had normal renal function. During the study all patients received MTX 10 mg/malone or with aspirin (45-50 mg/kg/d), or sulindac 400 mg/d in a three-way crossover design. Serum samples were obtained for 48 hours following the concurrent administration of the agents and HPLC was used to determine the concentrations of MTX and 7-0H-MTX. Clearance and AVC for the drug and metabolite were calculated. No significant differences in AVC, maximum concentration, or clearance of MTX were observed when the drug was administered alone or concomitantly with aspirin or sulindac. However, one subject did have very low MTX clearance when the drug was administered alone. If this patient had been excluded there would have been a statistically significant difference in MTX clearance when aspirin or sulindac was included in the regimen. Differences were noted when the pharmacokinetics of 7-0H-MTX were evaluated as there was a significant difference in the AVC and a nonsignificant increase in the maximum concentration between the control and the aspirin groups. An increase in AVC and maximum concentration of7-0H-MTX was also observed in the sulindac group; however, the difference was not significant." The probability of a type II error in the study was 18 percent, which makes the true significance of this study unknown. The pharmacokinetic studies discussed here had several limitations in their study design. The number of patients involved was small in all investigations. These studies may not have detected differences that could become significant if larger numbers of patients had been used. In addition, most patients had normal renal function. Patients with preexisting renal impairment (secondary to age, disease, or drugs) may be at higher risk because of the effects of NSAIDs on renal function and should also be evaluated. Furst, II Ahern," and Stewart" administered only one dose of MTX in their trials. Therefore, one cannot accurately predict the effect of NSAIDs on MTX following multipledose regimens. Finally, patients were monitored only for short periods of time (6-48 h) following the coadministration of NSAIDs and MTX. This may be appropriate for NSAIDs with short half-lives (i.e., aspirin, ibuprofen); however, agents with longer half-lives (i.e., sulindac, piroxicam) need to be evaluated for greater periods of time. Summary
Coadministration of MTX and NSAIDs occurs frequently and may result in significant toxicity. Clinical
1992 February, Volume 26
Downloaded from aop.sagepub.com at CMU Libraries - library.cmich.edu on September 11, 2015
Methotrexate and NSAIDs
manifestations, including pancytopenia and renal failure, have been demonstrated when patients received either high-dose or low-dose MTX together with NSAIDs. The exact mechanism of the interaction has not been determined; however, it is thought to occur at the level of the kidney. It is widely accepted that the reaction is related to a reduction in MTX clearance as a result of renal capillary constriction induced by NSAIDs. More recent data suggest a change in the pharmacokinetic parameters of the active metabolite of MTX, 7-0H-MTX, may be responsible. The possibility that the adverse effects reported in the literature may be attributed to idiosyncratic reactions cannot be ruled out. A combination of these factors may also be involved in the reaction. Although this interaction is not seen in all patients, numerous case reports describe the potential toxicity. In the case report presented here, signs and symptoms of the interaction occurred after one to two weeks of concomitant administration of MTX and NSAIDs. Large, well-controUedtrials to evaluate this reaction have not been conducted; as a result, the specific circumstances during which the reaction may occur have not been defined. Renal function should be monitored closely as increased MTX concentrations may be a result of NSAID -induced renal dysfunction in certain patients. Because the literature indicates a potential interaction between MTX and NSAIDs, caution should be used when these agents are administered concurrently. Clinicians should not be deterred from concomitant use of these drugs; however, judicious monitoring is warranted. ::::::
References I. Thyss A, Milano G, Kubar J, Namer M, Schneider M. Clinical and phannacokinetic evidence of a life-threatening interaction between methotrexate and ketoprofen. Lancet 1986;I:256-8. 2. Daly HM, Scott GL, Boyle J, Roberts CJc. Methotrexate toxicity precipitated by azapropazone. Br J Dermato/ 1986;114:733-5. 3. Gabrielli A, Leoni P, Danieli G. Methotrexate and nonsteroidal antiinflammatory drugs (letter). Br Med J 1987;294:776. 4. Ellison NM, Servi RJ. Acute renal failure and death following sequential intermediate-dose methotrexate and 5-FU: a possible adverse effect due to concomitant indomethacin administration. Cancer Treat Rep 1985;69:342-3. 5. Singh RR, Malaviya AN, Pandey IN, Guleria JS. Fatal interaction between methotrexate and naproxen (letter). Lancet 1986;I: 1390. 6. Maiche AG. Acute renal failure due to concomitant action of methotrexate and indomethacin (letter). Lancet 1986;I: 1390. 7. Bloom EJ, IgnotTo RJ, Reis CA, Cadman E. Delayed clearance of methotrexate associated with antibiotics and antiinflammatory agents (abstract). C/in Res 1986;34:560A. 8. Taylor JR, Halprin KM. Effect of sodium salicylate and indomethacin on methotrexate-serum albumin binding. Arch Dermato/ 1977; 113: 588-91. 9. Dettli L. Phannacokinetic aspects of drug interactions. In: Cluff LE, Petrie lC, eds. Clinical effects of interactions between drugs. New York: Elsevier, 1974:39-68. 10. Siordal L, Sager G, Aarbakke J. Phannacokinetic interactions with methotrexate: is 7-hydroxymethotrexate the culprit? (letter). Lancet 1988;1:591-2. II. Furst DE, Herman RA, Koehnke R, Ericksen N, Hash L, Riggs CEo et aI. Effect of aspirin and sulindac on methotrexate clearance. J Pharm Sci 1990;79:782-6. 12. Shen DD, AzarnotT DL. Clinical pharmacokinetics of methotrexate. Clin Pharmacokinet 1978;3:1-13. 13. Ahem M, Booth J, Loxton A, McCarthy P, Mellin P, Kevat S. Meth-
otrexate kinetics in rheumatoid arthritis: is there an interaction with nonsteroidal anti-inflammatory drugs? J Rheumatol 1988;15: 1356-60. 14. Stewart CF, Fleming RA, Arkin CR, Evans WE. Coadministrationof naproxen and low-dose methotrexate in patients with rheumatoid arthritis. C/in Pharmacol Ther 1990;47:540-6. 15. Skeith KJ, Russell AS, Jamali F, Coates J, Friedman H. Lack of significant interaction between low-dose methotrexate and ibuprofen or flurbiprofen in patients with arthritis. J Rheumatol 1990;17:1008-10. 16. Dupuis LL, Koren G, Shore A, Silverman ED, Laxer RM. Methotrexate-nonsteroidal antiinflammatory drug interaction in children with arthritis. J Rheumato/ 1990;17:1469-73.
EXTRACfO Metotrexato (MTX) y los agentes antiinfiamatorios no esteroidales (AANE) son frecuentemente administrados concomitantemente para el tratamiento de artritis reumatoidea, psoriasis, y malignidad. Evidencia reciente indica que la administraci6n conjunta de estos dos agentes podrfa ser potencialmente peligrosa. Reportes de casas de varias manifestaciones clfnicas, incluyendo fallo renal agudo y pancitopenia, han sido publicados. Los autores observaron una paciente de 74 aiios de edad que desarroll6 toxicidad hematol6gica subsiguiente a la administraci6n de MTX con flurbiprofen. EI mecanismo exacto de la interacci6n no se ha elucidado por completo. Teorias que han sido sugeridas para explicar el mecanismo de toxicidad de MTX incluyen reducci6n en la excresi6n de MTX secundaria a constricci6n de los capilares renales inducida por AANE, desplazamiento de MTX 0 su metabolito de las proteinas plasmaticas, competencia entre MTX y AANE por la excresi6n tubular renal 0 deterioro del metabolismo hepatico de MTX por AANE. Estudios que comparan la farmacocinetica de MTX durante la coadministraci6n de MTX con AANE y MTX solo no han demostrado diferencias estadfsticas en los parametres evaluados. Sin embardo, un estudio demostr6 diferencias en la farmacocinetica de 7-hidroximetotrexato, el metabolito activo de MTX, aI administrar MTX con sulindac 0 aspirina. Aunque la reacci6n no occurre en todos los pacientes, existen un mirnero de reportes de casos que demuestran posibles problemas subsiguientes a la coadministraci6n de MTX y AANE. La literatura indica el potencial para una interacci6n entre MTX y AANE, sin embargo, las circunstancias especfficas durante las cuales la reacci6n podrfa verse no han sido bien defmidas. BRENDA R. MORAND
RESUME
La pharmacotherapie de I'arthrite rhumatoide, du psoriasis ou de certaines tumeurs mal ignes implique frequemrnent la coadministration du methotrexate (MTX) et d'un antiinfiammatoire non-steroidien (AINS). De nouvelles evidences indiquent que I'utilisation concomitante de ces deux agents pourrait causer certaines reactions indesirables telles que pancytopenic, insuffisance renale aigue et toxicite hematologique. Le mecanisme exact de cette interaction medicamenteuse n' a pas ete clairement etabli, Plusieurs hypotheses ont cependant ete suggerees et incluent (I) reduction de l'elimination urinaire du MTX suite a une vasoconstriction renale induite par les AINS, (2) deplacement du MTX de ses sites de liaison aux proteines plasmatiques, (3) competition entre Ie MTX et les AINS pour les sites de secretion tubulaire, (4) alteration du metabolisme hepatique du MTX par les AINS. Aucune etude n' a demontre une difference dans la pharmacocinetique du MTX en presence d'un AINS. Cependant, une etude rapporte que I' administration de sulindac ou d' aspirine engendrerait des changements dans la pharmacocinetique du metabolite actif du MTX, Ie 7-hydroxyMTX. Considerant qu' aucun facteur pouvant predisposer a une telle interaction rnedicamenteuse n' a ete identifie ace jour, un monitoring pharmacotherapeutique etroit devrait etre assure par tout clinicien utilisant la combinaison MTX et AINS.
The Annals ofPharmacotherapy
SYLVIE ROBERT
•
Downloaded from aop.sagepub.com at CMU Libraries - library.cmich.edu on September 11, 2015
1992 February, Volume 26 •
237
