115
CORRELATIONS BETWEEN TRANSAMINASE CONCENTRATIONS AND SERUM SALICYLATE CONCENTRATION IN JUVENILE RHEUMATOID ARTHRITIS JOHN J. MILLER 111 and DAVID B. WEISSMAN Simultaneous measurements of serum salicylate, SGPT, and SGOT concentrations were made on 92 children receiving salicylates for arthritis during a 13-month period. Statistical correlations were found between salicylate and transaminase concentrations in patients with juvenile rheumatoid arthritis when in active stages. However elevated transaminase levels occurred with low salicylate levels and during inactive stages, and were inconstant in individual patients. Marked transaminase abnormalities were transient and appeared unpredictably in individual patients. Some factor in addition to salicylate concentration must be involved in raising serum transaminase levels. No evidence of chronic hepatotoxicity was noted. Elevated concentrations of serum glutamic oxaloacetic transaminase (SGOT) and serum glutamic pyruvic transaminase (SGPT) have been reported in association with salicylate therapy of rheumatic fever (1,2), systemic lupus erythematosus (SLE) (3-5), and juvenile rheumatoid arthritis (JRA) (2,6-8). Although From the Children’s Hospital at Stanford and Stanford University School of Medicine, Stanford, California 94305. Presented in part to the Society for Pediatric Research, Denver, Colorado, April 19, 1975. John J . Miller 111, M.D.: Director, Rheumatic Disease Service, Children’s Hospital at Stanford, and Senior Attending Physician, Stanford University School of Medicine; David B. Weissman, Ph.D.: Medical Student, University of California, Irvine. Address reprint requests to Dr. Miller. Submitted for publication February 24, 1975; accepted May 29, 1975. Arthritis and Rheumatism, Vol. 19, No. 1 (January-February 1976)
the first account of this association was in 1956 (l), most reports are more recent and involve small groups of patients (2-8). The absence of data regarding incidence of changes i n SGOT and SGPT in large groups of patients prompted this prospective study of.children receiving salicylates in the pediatric arthritis clinic of the Children’s Hospital at Stanford.
MATERIALS AND METHODS All outpatients with arthritis receiving aspirin, choline salicylate, or sodium salicylate seen at Children’s Hospital at Stanford during a 13-month period were studied. Approximately 10% of patients and patient visits were lost from the study because of random clerical or laboratory omissions or errors. The evaluated children included 7 with systemic JRA, 49 with polyarticular JRA, 32 with pauciarticular JRA (9), 1 with ankylosing spondylitis, and 3 who were eventually proved to have no rheumatic disease. There were 60 girls and 32 boys. Ages ranged from 1.5 years to 19.25 years. The number of visits made by individual patients varied between 1 and 25 and totaled 335. At each visit the disease was classified as active or inactive depending on the presence of joint swelling or elevated sedimentation rate. Blood was taken for simultaneous determination of serum salicylate (lo), SGOT (1 1) , and SGPT (12) concentrations and of the blood eosinophil concentration. Normal values for the SGOT and SGPT concentrations were 5-19 and 4-25 1U/I respectively using reagents and standards from Worthington Biochemical Corporation, Freehold, New Jersey. Other liver function tests were performed too infrequently to be correlated with the data regarding salicylates or transaminases. Statistical analysis of the data was performed on Stanford University’s IBM 360/67 computer.
MILLER AND WEISSMAN
I16
RESULTS As expected serum salicylate concentrations varied widely, from 0 to 53 mg/dl, but the mean and standard deviation of all values was 21 & 1 1 mg/dl indicating a high degree of patient compliance. The 3 patients with no rheumatic disease had absent or low salicylate levels and normal SGOT and SGPT concentrations. The patient with ankylosing spondylitis maintained salicylate levels between 20 and 48 mg/dl and had normal SGOT and SGPT concentrations on 6 visits. Of the remaining 88 patients, all with JRA, 41%
a
a
SERUM SALICYLATE rng/dl Fig 1. Graphs showing plots of all the values for salicylate and SGPT concentrations of all visits of patients with JRA not receiving gold, classifed by type of disease and activity of disease at the time tested. The range of normal values of SGPT concentrations in this laboratory (dashed lines) is 4-25 l l J / l .The valuesfor r and p in the upper left corner of each graph are the correlation coeflcient and the value of probability that a correlation exists between the SGPT and salicylate concentrations. The plots show wide distribution of values even when correlations are statistically present.
had an elevated SGPT concentration and 59% an elevated SGOT concentration at some point during the study period. All patients with an elevated SGPT had an elevated SGOT, but the latter was frequently slightly elevated when the SGPT remained normal. There was no statistical difference by x2 test between the incidence of SGOT or SGPT elevations in boys or girls. The proportion of children younger than 11 years who had an increased SGOT at some time, 24/3 1, was greater than this proportion in patients over 10 years of age, 31/57 (0.05 > P > 0.01). This difference was less marked in SGPT elevations with incidences of 15/30 and 20/56 respectively ( P > 0.05). Concentrations of either SGPT or SGOT greater than 100 IU/I occurred in 10 (1 1%) of the patients on one or two occasions. These had always decreased by the time the patient was retested with no greater interval than I week. In 4 children these high elevations appeared shortly after salicylates were started and decreased even when serum salicylate levels were maintained. In the other 6 instances the patients had been taking salicylates for prolonged periods and no particular event associated with the transaminase increase was evident. Only 1 patient had associated symptoms, namely nausea and vomiting. The salicylates were maintained or briefly stopped and then resumed without further transaminase elevation in each instance. The distribution of SGPT values greater than normal is graphically shown in Figure 1. The data from individual visits were grouped by type of disease and by activity of disease and were examined for correlations between salicylate level and SGPT and SGOT concentrations. Data from patients receiving gold (aurothiomalate) as well as salicylate were considered separately. The data with respect to SGPT are shown in Figure 1 . Greater than normal concentrations of SGPT occurred with all types of JRA, and whether the disease was active or inactive. Elevated SGPT concentrations occurred with low salicylate concentrations and, conversely, normal levels were present with high salicylate concentrations. However a positive statistical correlation ( P < 0.02) between concentrations of salicylate and SGPT was found in association with all three types of active JRA and in inactive, pauciarticular disease as well. The data regarding salicylate and SGPT concentrations produced similar correlations and differed only in that elevated SGOT concentrations occurred more frequently in all situations. The x2 test was used to analyze the differences in incidence of SGPT elevations between active and inactive phases of each type of JRA. The proportion of
TRANSAMINASE AND SERUM SALICYLATE CONCENTRATIONS
elevated SGPT values was always slightly greater in active phases as can be seen in Figure 1, but the differences were never statistically significant. To determine whether 25 mg/dl represents a critical concentration of salicylate in respect to transaminase levels, the values of the correlation coefficients ( r ) and of the significance of the correlation (P)were determined for those instances when the concentration of salicylate exceeded this level. The values of the correlation coefficient and the significance of correlation for these data were greater than those for all salicylate concentrations only in patients with active polyarticular disease who were not receiving gold (r: 0.24-0.37; and P: 0.005-0.003). Further, xz tests of the difference in incidence of elevated SGPT values at salicylate levels less than or greater than 25 mg/dl as shown in Figure 1 showed no statistical difference unless patients with active, pauciarticular disease were considered as an isolated group (0.01 > P > 0.001). The simultaneous administration of aurothiomalate produced differences neither in SGPT and SGOT concentrations nor in the correlations of these concentrations with salicylate levels. N o significant correlation wa6 found between salicylate concentration and blood eosinophil concentration in any circumstance. A liver biopsy was performed in 1 patient who had been receiving salicylates for 7 years, at a time when he had raised serum concentrations of SGOT, lactic dehydrogenase, and alkaline phosphatase, a normal serum bilirubin concentration, and a serum salicylate concentration of 38 mg/d1. There were an increased number of individual necrotic hepatocytes undergoing phagocytosis but no fibrosis was evident (Figure 2).
DISCUSSION These data confirm a statistical correlation between salicylate concentration and transaminase concentration in patients with JRA. However, correlations in individual patients were inconstant, and increased SGOT and SGPT levels were found at low salicylate concentrations as well as at normal transaminase concentrations with high serum levels of salicylate. Therefore no single factor can be involved in the relationship. It was expected that elevated transaminases might be associated particularly with systemic disease, but it was not possible to distinguish differences between the effects of acute, active, systemic JRA and the localized, nonsystemic, pauciarticular form. The degree of activity of the disease may be a factor. Although not statistically
117
Fig 2. Photomicrograph of a section of liver from a child who had received salicylates f o r 7 years and who had elevated hepatic enzyme serum concentrations and a salicylate level of 38 mgldl at the time of biopsy. A focal area of hepatocyte necrosis and phagocytosis is seen. Such areas were increased in number. The biopsy was otherwise normal. H & E. X500.
demonstrable, there was a tendency toward increased transarninase concentrations in active compared to inactive phases of JRA that was similar to a relation that has been seen in SLE (3). The situation is obviously confused by the presence of hepatic changes that are part of JRA itself (13,14). However the hepatitis of JRA appears to be clinically evident primarily in patients with severe systemic disease ( 1 3,14), and transaminase abnormalities were found in pauciarticular disease and even when all disease appeared inactive. In one series of 32 children with JRA, elevated transaminase levels were found only with salicylate concentrations greater than 35 mg/dl (6). In another study of 6 adults and children with JRA, abnormal transaminase concentrations appeared only with salicylate concentrations greater than 25 mg/dl(7). In contrast the present study revealed high SGPT and SGOT concentrations with low levels of salicylate. Further evidence against a “threshold” toxic salicylate serum concentration was the present authors’ failure to find changes in correlation coefficients or statistical increases i n incidence of transaminase elevations above this level in most of the patients. Rabbits fed salicylates have been reported to de-
MILLER AND WEISSMAN
118
velop serum transaminase elevations (1 5 ) and reversible alterations in hepatocellular morphology (16,17). In the only long-term study chronic focal interstitial inflammation was found in 3 of 24 rabbits treated for 3 months, but this pathologic change was also seen in other organs and was not proved t o b e related to t h e salicylates (16). The rabbit experiments point o u t that salicylate effect o n liver enzyme concentrations is probably not determined by t h e disease for which the salicylates are being administered. A relationship between salicylate level and circulating eosinophils has been reported in a study of 6 patients (7). Such a relationship was not evident in the more extensive but less linearly time-oriented study reported here. The authors’ experience does not disprove the possibility that acute hepatotoxicity from salicylates may become irreversible, but the transient nature of the t.ransaminase elevations observed in this study suggests that this will be unusual. T h e high incidence of minor transaminase elevations in contrast with the very low incidence of symptomatic hepatic disease argues that the effect of salicylates is relatively benign despite the measurable abnormalities. A further question is whether chronic and irreparable liver damage occurs as part of the process causing the transaminase elevations. In the present study a single biopsy from a child who had been receiving salicylates for years and who had abnormal liver enzyme concentrations in his serum was similar to previous reports of only minor chronic changes in other humans (3,7) or in rabbits (16). T h e absence of reports of chronic hepatitis or cirrhosis associated with the widespread use of salicylates is impressive ( I 8). The known chronic toxicity of other drugs available for use in children appears to be more significant than the transaminase elevations seen in these studies. The authors believe salicylates remain the best therapy for most children with JRA, but suggest that the SGPT be followed in all children receiving chronic high doses,
ACKNOWLEDGMENTS The authors are grateful to Dr. John H. Vaughan, who provided helpful criticism of the manuscript.
REFERENCES I . Manso C, Taranta A, Mydick I: Effect of aspirin administration on serum glutamic oxaloacetic and glutamic pyruvic transaminases in children. Proc SOCExp Biol Med 93~84-88, 1956 2. Iancu T: Serum transaminases and salicylate therapy. Br Med J 2: 167, 1972 3. Seaman WE, Ishak KG, Plotz PH: Aspirin-induced hepatotoxicity in patients with systemic lupus erythematosus. Ann Intern Med 8O:l-8, 1974 4. Wolfe JD, Metzger AL, Goldstein RC: Aspirin hepatitis. Ann Intern Med 80:74-76, 1974 5. Goldenberg DL: Aspirin hepatotoxicity. Ann lntern Med 80773, 1974 6. Russell AS, Sturge RA, Smith MA: Serum transaminases during salicylate therapy. Br Med J 2:428-429, 1971 7. Rich RR, Johnson JS: Salicylate hepatotoxicity in patients with juvenile rheumatoid arthritis. Arthritis Rheum 161-9, 1973 8. Athreya BH, Moser G, Myers AR: Effects of salicylates on liver function in juvenile rheumatoid arthritis. Arthritis Rheum 17:495, 1974 9. Schaller J, Wedgewood RJ: Juvenile rheumatoid arthritis: a review. Pediatrics 50:940-953, 1972 10. Trinder P Rapid determination of salicylate in biological fluids. Biochem J 57:301-303, 1954 1 I . Amador E, Wacher WEC: Serum glutamic-oxaloacetic transaminase activity. Clin Chem 8:343-350, 1962 12. Henry RJ, Chiamori N, Golub DJ, et al: Revised spectrophotometic methods for the determination of glutamicoxaloacetic transaminase, glutamic-pyruvic transaminase and lactic dehydrogenase. Am J Clin Pathol 34:381-398, 1960 13. Schaller J, Beckwith B, Wedgewood RJ: Hepatic involvement in juvenile rheumatoid arthritis. J Pediatr 771203-210, 1970 14. Kornreich H, Malouf NN,Hanson V: Acute hepatic dysfunction in juvenile rheumatoid arthritis. J Pediatr 79:27-35, 1971 15. Janota I, Wincey CW, Sandiford M, et al: Effect of salicylate on the activity of plasma enzymes in the rabbit. Nature 185:935-936, 1960 16. Kalczak M, Gutavska-Grzegorczyk G, Maldyk E: The effect of chronic administration of acetylsalicylic acid on the rabbit’s liver. Pol Med J 9:128-134, 1970 17. Hruban Z, Gotoh M, Slesers A, et al: Structure of hepatic microbodies in rats treated with acetylsalicylic acid, clofibrate, and dimethrin. Lab Invest 3064-75, 1974 18. Zimmerman HJ: Aspirin-induced hepatic injury. Ann Intern Med 80103-105, 1974
CORRELATIONS BETWEEN TRANSAMINASE CONCENTRATIONS AND SERUM SALICYLATE CONCENTRATION IN JUVENILE RHEUMATOID ARTHRITIS JOHN J. MILLER 111 and DAVID B. WEISSMAN Simultaneous measurements of serum salicylate, SGPT, and SGOT concentrations were made on 92 children receiving salicylates for arthritis during a 13-month period. Statistical correlations were found between salicylate and transaminase concentrations in patients with juvenile rheumatoid arthritis when in active stages. However elevated transaminase levels occurred with low salicylate levels and during inactive stages, and were inconstant in individual patients. Marked transaminase abnormalities were transient and appeared unpredictably in individual patients. Some factor in addition to salicylate concentration must be involved in raising serum transaminase levels. No evidence of chronic hepatotoxicity was noted. Elevated concentrations of serum glutamic oxaloacetic transaminase (SGOT) and serum glutamic pyruvic transaminase (SGPT) have been reported in association with salicylate therapy of rheumatic fever (1,2), systemic lupus erythematosus (SLE) (3-5), and juvenile rheumatoid arthritis (JRA) (2,6-8). Although From the Children’s Hospital at Stanford and Stanford University School of Medicine, Stanford, California 94305. Presented in part to the Society for Pediatric Research, Denver, Colorado, April 19, 1975. John J . Miller 111, M.D.: Director, Rheumatic Disease Service, Children’s Hospital at Stanford, and Senior Attending Physician, Stanford University School of Medicine; David B. Weissman, Ph.D.: Medical Student, University of California, Irvine. Address reprint requests to Dr. Miller. Submitted for publication February 24, 1975; accepted May 29, 1975. Arthritis and Rheumatism, Vol. 19, No. 1 (January-February 1976)
the first account of this association was in 1956 (l), most reports are more recent and involve small groups of patients (2-8). The absence of data regarding incidence of changes i n SGOT and SGPT in large groups of patients prompted this prospective study of.children receiving salicylates in the pediatric arthritis clinic of the Children’s Hospital at Stanford.
MATERIALS AND METHODS All outpatients with arthritis receiving aspirin, choline salicylate, or sodium salicylate seen at Children’s Hospital at Stanford during a 13-month period were studied. Approximately 10% of patients and patient visits were lost from the study because of random clerical or laboratory omissions or errors. The evaluated children included 7 with systemic JRA, 49 with polyarticular JRA, 32 with pauciarticular JRA (9), 1 with ankylosing spondylitis, and 3 who were eventually proved to have no rheumatic disease. There were 60 girls and 32 boys. Ages ranged from 1.5 years to 19.25 years. The number of visits made by individual patients varied between 1 and 25 and totaled 335. At each visit the disease was classified as active or inactive depending on the presence of joint swelling or elevated sedimentation rate. Blood was taken for simultaneous determination of serum salicylate (lo), SGOT (1 1) , and SGPT (12) concentrations and of the blood eosinophil concentration. Normal values for the SGOT and SGPT concentrations were 5-19 and 4-25 1U/I respectively using reagents and standards from Worthington Biochemical Corporation, Freehold, New Jersey. Other liver function tests were performed too infrequently to be correlated with the data regarding salicylates or transaminases. Statistical analysis of the data was performed on Stanford University’s IBM 360/67 computer.
MILLER AND WEISSMAN
I16
RESULTS As expected serum salicylate concentrations varied widely, from 0 to 53 mg/dl, but the mean and standard deviation of all values was 21 & 1 1 mg/dl indicating a high degree of patient compliance. The 3 patients with no rheumatic disease had absent or low salicylate levels and normal SGOT and SGPT concentrations. The patient with ankylosing spondylitis maintained salicylate levels between 20 and 48 mg/dl and had normal SGOT and SGPT concentrations on 6 visits. Of the remaining 88 patients, all with JRA, 41%
a
a
SERUM SALICYLATE rng/dl Fig 1. Graphs showing plots of all the values for salicylate and SGPT concentrations of all visits of patients with JRA not receiving gold, classifed by type of disease and activity of disease at the time tested. The range of normal values of SGPT concentrations in this laboratory (dashed lines) is 4-25 l l J / l .The valuesfor r and p in the upper left corner of each graph are the correlation coeflcient and the value of probability that a correlation exists between the SGPT and salicylate concentrations. The plots show wide distribution of values even when correlations are statistically present.
had an elevated SGPT concentration and 59% an elevated SGOT concentration at some point during the study period. All patients with an elevated SGPT had an elevated SGOT, but the latter was frequently slightly elevated when the SGPT remained normal. There was no statistical difference by x2 test between the incidence of SGOT or SGPT elevations in boys or girls. The proportion of children younger than 11 years who had an increased SGOT at some time, 24/3 1, was greater than this proportion in patients over 10 years of age, 31/57 (0.05 > P > 0.01). This difference was less marked in SGPT elevations with incidences of 15/30 and 20/56 respectively ( P > 0.05). Concentrations of either SGPT or SGOT greater than 100 IU/I occurred in 10 (1 1%) of the patients on one or two occasions. These had always decreased by the time the patient was retested with no greater interval than I week. In 4 children these high elevations appeared shortly after salicylates were started and decreased even when serum salicylate levels were maintained. In the other 6 instances the patients had been taking salicylates for prolonged periods and no particular event associated with the transaminase increase was evident. Only 1 patient had associated symptoms, namely nausea and vomiting. The salicylates were maintained or briefly stopped and then resumed without further transaminase elevation in each instance. The distribution of SGPT values greater than normal is graphically shown in Figure 1. The data from individual visits were grouped by type of disease and by activity of disease and were examined for correlations between salicylate level and SGPT and SGOT concentrations. Data from patients receiving gold (aurothiomalate) as well as salicylate were considered separately. The data with respect to SGPT are shown in Figure 1 . Greater than normal concentrations of SGPT occurred with all types of JRA, and whether the disease was active or inactive. Elevated SGPT concentrations occurred with low salicylate concentrations and, conversely, normal levels were present with high salicylate concentrations. However a positive statistical correlation ( P < 0.02) between concentrations of salicylate and SGPT was found in association with all three types of active JRA and in inactive, pauciarticular disease as well. The data regarding salicylate and SGPT concentrations produced similar correlations and differed only in that elevated SGOT concentrations occurred more frequently in all situations. The x2 test was used to analyze the differences in incidence of SGPT elevations between active and inactive phases of each type of JRA. The proportion of
TRANSAMINASE AND SERUM SALICYLATE CONCENTRATIONS
elevated SGPT values was always slightly greater in active phases as can be seen in Figure 1, but the differences were never statistically significant. To determine whether 25 mg/dl represents a critical concentration of salicylate in respect to transaminase levels, the values of the correlation coefficients ( r ) and of the significance of the correlation (P)were determined for those instances when the concentration of salicylate exceeded this level. The values of the correlation coefficient and the significance of correlation for these data were greater than those for all salicylate concentrations only in patients with active polyarticular disease who were not receiving gold (r: 0.24-0.37; and P: 0.005-0.003). Further, xz tests of the difference in incidence of elevated SGPT values at salicylate levels less than or greater than 25 mg/dl as shown in Figure 1 showed no statistical difference unless patients with active, pauciarticular disease were considered as an isolated group (0.01 > P > 0.001). The simultaneous administration of aurothiomalate produced differences neither in SGPT and SGOT concentrations nor in the correlations of these concentrations with salicylate levels. N o significant correlation wa6 found between salicylate concentration and blood eosinophil concentration in any circumstance. A liver biopsy was performed in 1 patient who had been receiving salicylates for 7 years, at a time when he had raised serum concentrations of SGOT, lactic dehydrogenase, and alkaline phosphatase, a normal serum bilirubin concentration, and a serum salicylate concentration of 38 mg/d1. There were an increased number of individual necrotic hepatocytes undergoing phagocytosis but no fibrosis was evident (Figure 2).
DISCUSSION These data confirm a statistical correlation between salicylate concentration and transaminase concentration in patients with JRA. However, correlations in individual patients were inconstant, and increased SGOT and SGPT levels were found at low salicylate concentrations as well as at normal transaminase concentrations with high serum levels of salicylate. Therefore no single factor can be involved in the relationship. It was expected that elevated transaminases might be associated particularly with systemic disease, but it was not possible to distinguish differences between the effects of acute, active, systemic JRA and the localized, nonsystemic, pauciarticular form. The degree of activity of the disease may be a factor. Although not statistically
117
Fig 2. Photomicrograph of a section of liver from a child who had received salicylates f o r 7 years and who had elevated hepatic enzyme serum concentrations and a salicylate level of 38 mgldl at the time of biopsy. A focal area of hepatocyte necrosis and phagocytosis is seen. Such areas were increased in number. The biopsy was otherwise normal. H & E. X500.
demonstrable, there was a tendency toward increased transarninase concentrations in active compared to inactive phases of JRA that was similar to a relation that has been seen in SLE (3). The situation is obviously confused by the presence of hepatic changes that are part of JRA itself (13,14). However the hepatitis of JRA appears to be clinically evident primarily in patients with severe systemic disease ( 1 3,14), and transaminase abnormalities were found in pauciarticular disease and even when all disease appeared inactive. In one series of 32 children with JRA, elevated transaminase levels were found only with salicylate concentrations greater than 35 mg/dl (6). In another study of 6 adults and children with JRA, abnormal transaminase concentrations appeared only with salicylate concentrations greater than 25 mg/dl(7). In contrast the present study revealed high SGPT and SGOT concentrations with low levels of salicylate. Further evidence against a “threshold” toxic salicylate serum concentration was the present authors’ failure to find changes in correlation coefficients or statistical increases i n incidence of transaminase elevations above this level in most of the patients. Rabbits fed salicylates have been reported to de-
MILLER AND WEISSMAN
118
velop serum transaminase elevations (1 5 ) and reversible alterations in hepatocellular morphology (16,17). In the only long-term study chronic focal interstitial inflammation was found in 3 of 24 rabbits treated for 3 months, but this pathologic change was also seen in other organs and was not proved t o b e related to t h e salicylates (16). The rabbit experiments point o u t that salicylate effect o n liver enzyme concentrations is probably not determined by t h e disease for which the salicylates are being administered. A relationship between salicylate level and circulating eosinophils has been reported in a study of 6 patients (7). Such a relationship was not evident in the more extensive but less linearly time-oriented study reported here. The authors’ experience does not disprove the possibility that acute hepatotoxicity from salicylates may become irreversible, but the transient nature of the t.ransaminase elevations observed in this study suggests that this will be unusual. T h e high incidence of minor transaminase elevations in contrast with the very low incidence of symptomatic hepatic disease argues that the effect of salicylates is relatively benign despite the measurable abnormalities. A further question is whether chronic and irreparable liver damage occurs as part of the process causing the transaminase elevations. In the present study a single biopsy from a child who had been receiving salicylates for years and who had abnormal liver enzyme concentrations in his serum was similar to previous reports of only minor chronic changes in other humans (3,7) or in rabbits (16). T h e absence of reports of chronic hepatitis or cirrhosis associated with the widespread use of salicylates is impressive ( I 8). The known chronic toxicity of other drugs available for use in children appears to be more significant than the transaminase elevations seen in these studies. The authors believe salicylates remain the best therapy for most children with JRA, but suggest that the SGPT be followed in all children receiving chronic high doses,
ACKNOWLEDGMENTS The authors are grateful to Dr. John H. Vaughan, who provided helpful criticism of the manuscript.
REFERENCES I . Manso C, Taranta A, Mydick I: Effect of aspirin administration on serum glutamic oxaloacetic and glutamic pyruvic transaminases in children. Proc SOCExp Biol Med 93~84-88, 1956 2. Iancu T: Serum transaminases and salicylate therapy. Br Med J 2: 167, 1972 3. Seaman WE, Ishak KG, Plotz PH: Aspirin-induced hepatotoxicity in patients with systemic lupus erythematosus. Ann Intern Med 8O:l-8, 1974 4. Wolfe JD, Metzger AL, Goldstein RC: Aspirin hepatitis. Ann Intern Med 80:74-76, 1974 5. Goldenberg DL: Aspirin hepatotoxicity. Ann lntern Med 80773, 1974 6. Russell AS, Sturge RA, Smith MA: Serum transaminases during salicylate therapy. Br Med J 2:428-429, 1971 7. Rich RR, Johnson JS: Salicylate hepatotoxicity in patients with juvenile rheumatoid arthritis. Arthritis Rheum 161-9, 1973 8. Athreya BH, Moser G, Myers AR: Effects of salicylates on liver function in juvenile rheumatoid arthritis. Arthritis Rheum 17:495, 1974 9. Schaller J, Wedgewood RJ: Juvenile rheumatoid arthritis: a review. Pediatrics 50:940-953, 1972 10. Trinder P Rapid determination of salicylate in biological fluids. Biochem J 57:301-303, 1954 1 I . Amador E, Wacher WEC: Serum glutamic-oxaloacetic transaminase activity. Clin Chem 8:343-350, 1962 12. Henry RJ, Chiamori N, Golub DJ, et al: Revised spectrophotometic methods for the determination of glutamicoxaloacetic transaminase, glutamic-pyruvic transaminase and lactic dehydrogenase. Am J Clin Pathol 34:381-398, 1960 13. Schaller J, Beckwith B, Wedgewood RJ: Hepatic involvement in juvenile rheumatoid arthritis. J Pediatr 771203-210, 1970 14. Kornreich H, Malouf NN,Hanson V: Acute hepatic dysfunction in juvenile rheumatoid arthritis. J Pediatr 79:27-35, 1971 15. Janota I, Wincey CW, Sandiford M, et al: Effect of salicylate on the activity of plasma enzymes in the rabbit. Nature 185:935-936, 1960 16. Kalczak M, Gutavska-Grzegorczyk G, Maldyk E: The effect of chronic administration of acetylsalicylic acid on the rabbit’s liver. Pol Med J 9:128-134, 1970 17. Hruban Z, Gotoh M, Slesers A, et al: Structure of hepatic microbodies in rats treated with acetylsalicylic acid, clofibrate, and dimethrin. Lab Invest 3064-75, 1974 18. Zimmerman HJ: Aspirin-induced hepatic injury. Ann Intern Med 80103-105, 1974
