- Email: [email protected]
Intravenous pulse methylprednisolone therapy in eye disease
Intravenous Pulse Methylprednisolone Therapy in Eye Disease Effect on Glucose Tolerance Sylvie Feldman-Billard, MD, Bruno Lissak, MD, Rabah Benrabah, MD, Roxana Kassaei, MD, Emmanuel He´ron, MD Purpose: To determine which subjects need close glycemic monitoring during intravenous pulse methylprednisolone therapy for eye disease. Design: Retrospective study in a national eye center. Participants: Two hundred twenty-four subjects who received over a one-year period 250 to 1000 mg daily intravenous methylprednisolone over 3 consecutive days for ophthalmologic conditions. Methods: Blood glucose monitoring during pulse methylprednisolone therapy followed a protocol written in 1995. We analyzed the effects of 3 days of pulse methylprednisolone therapy on glucose tolerance and their clinical implications in diabetic and nondiabetic subjects treated during 2000. Main Outcome Measure: Serial morning fasting blood glucose; that is, before the first pulse and the day after each pulse, blood glucose self-monitoring for diabetic subjects and specific hypoglycemic drug interventions were recorded. Results: All subjects showed a median 50% increase in fasting glucose after the first steroid infusion, without a significant difference between diabetic and nondiabetic subjects. Thereafter, the 196 nondiabetic subjects showed spontaneous decreases of their fasting glucose toward baseline values despite the following infusions, whereas the 28 diabetic subjects (all type 2) demonstrated further increases of blood glucose, and 7 received rapid-release insulin to maintain blood glucose lower than 14 mmol/l. All 5 diabetic subjects with baseline glycosylated hemoglobin ⱖ 8.3% required insulin therapy. Conclusions: Close glycemic monitoring seems necessary only for subjects with diabetes during intravenous pulse methylprednisolone therapy for ophthalmologic conditions. The probability of subjects with type 2 diabetes requiring insulin during this therapy does seem to be positively related to the level of pretreatment glycosylated hemoglobin. Ophthalmology 2003;110:2369 –2371 © 2003 by the American Academy of Ophthalmology.
Short-term, high-dose intravenous methylprednisolone therapy, also called “pulse methylprednisolone” therapy, is widely used for a variety of inflammatory and immunologic disorders.1,2 In particular, it is used in ophthalmologic conditions such as acute optic neuritis,3 severe uveitis,4 corneal graft rejection,5 and giant cell arteritis.6 Because of the well-known hyperglycemic effects of glucocorticosteroids,7–9 the monitoring of blood glucose during pulse methylprednisolone therapy is recommended.2,10 However, whether and which type of glucose monitoring should be applied according to a subject’s characteristics, particularly the presence of diabetes, have not been evaluated. To determine which subjects need close glycemic monitoring Originally received: October 28, 2002. Accepted: April 16, 2003. Manuscript no. 220869. From Service de Me´decine Interne, Centre Hospitalier National d’Ophtalmologie des Quinze-Vingts, Paris, France. Correspondence and reprint requests to Sylvie Feldman-Billard, MD, Service de Me´decine Interne, Centre Hospitalier National d’Ophtalmologie des Quinze-Vingts, 28 rue de Charenton, 75571 Paris Cedex 12, France. E-mail: [email protected]. © 2003 by the American Academy of Ophthalmology Published by Elsevier Inc.
during pulse methylprednisolone therapy, we analyzed the results of serial fasting blood glucose measurements in subjects who received a three-day course of pulse methylprednisolone therapy for ophthalmologic conditions in our eye center over a one-year period.
Subjects and Methods The Centre Hospitalier National d’Ophtalmologie des QuinzeVingts in Paris, France, is a single-specialty eye center. Before pulse methylprednisolone therapy, every subject is seen by an internist who verifies the application of a standardized protocol for blood glucose monitoring written in 1995. According to this protocol, morning fasting blood glucose was measured in all subjects. In subjects with diabetes, the self-monitoring of capillary blood glucose was performed at least 3 times per day before each meal. Glycosylated hemoglobin was measured in subjects with known diabetes or initial fasting glucose ⱖ 7 mmol/l. A blood glucose of 14 mmol/l or higher led to hypoglycemic interventions, based on the administration of rapid-release insulin with a sliding scale of blood glucose values: 4 IU of rapid-release insulin for blood glucose values between 14 and 17 mmol/l, 6 IU for blood glucose ISSN 0161-6420/03/$–see front matter doi:10.1016/S0161-6420(03)00818-2
2369
Ophthalmology Volume 110, Number 12, December 2003 Table 1. Main Characteristics of the Study Population According to Diabetic Status Total Population (N ⴝ 224) Age, mean (SD), years Number of women, n (%) Baseline fasting blood glucose, mean (SD), mmol/l Baseline fasting blood glucose ⬎ 7 mmol/l, n (%)
Subjects without Subjects with Known Diabetes Known Diabetes (N ⴝ 196) (N ⴝ 28) 46.1 (19.8) 130 (66) 5.2 (0.8) 7 (3.6)
P
61.4 (20.2) 12 (43) 8.3 (4.1)
⬍0.001 0.02 ⬍0.001
14 (50)
0.01
SD ⫽ standard deviation.
values between 17 and 20 mmol/l, 8 IU for blood glucose values between 20 and 23 mmol/l, and so on. The medical reports of all subjects who received pulse methylprednisolone therapy during 2000 were analyzed. The steroid dosages, the main characteristics of subjects, the serial morning fasting blood glucose (i.e., before the first pulse and the day after each pulse), and the number and characteristics of subjects requiring insulin during steroid therapy were recorded. Statistical analysis was performed with the SAS statistical software for Windows (SAS Institute Inc., Cary, NC). Significance was considered for P values ⬍ 0.05. To evaluate the influence of initial fasting blood glucose and age on the glucose tolerance of pulse methylprednisolone therapy, these parameters were considered as continuous variables or separated in different categories (⬍5 mmol/l, 5–5.9 mmol/l, 6 – 6.9 mmol/l, and ⱖ7 mmol/l; ⬍30 years, 30 – 49 years, 50 – 69 years, and ⱖ70 years).
Results Two hundred seventy-one subjects received pulse methylprednisolone therapy at our institution during the year 2000. Fortyseven subjects received fewer than three methylprednisolone infusions or had incomplete blood glucose data. Ultimately, the analysis was completed in 224 subjects (mean age ⫽ 46.1 years;
56% women) who received 250 mg (n ⫽ 20), 500 mg (n ⫽ 109), or 1000 mg (n ⫽ 95) pulse methylprednisolone once a day over 3 consecutive days. The indications of pulse methylprednisolone therapy were acute optic neuritis (n ⫽ 91), severe uveitis (n ⫽ 35), ocular infectious diseases (n ⫽ 22), corneal graft rejection (n ⫽ 17), and miscellaneous conditions (n ⫽ 59). Table 1 shows the main characteristics of the subjects according to diabetic status. No subject had type 1 diabetes, and 28 had type 2 diabetes, defined as an age of diabetes onset of 30 years or older. All but 1 had been treated with oral hypoglycemic agents for 10.6 ⫾ 8.4 years (range ⫽ 4 –28) or with insulin (n ⫽ 2); these treatments were maintained during methylprednisolone therapy. The fasting blood glucose curves during pulse methylprednisolone therapy according to diabetic status are shown in Figure 1. The 196 nondiabetic subjects showed a median 50% increase of fasting blood glucose after the first pulse and a subsequent spontaneous decay toward baseline values, despite the second and third pulses; their fasting blood glucose after the third pulse was a median of 23% higher than baseline values. Among these 196 subjects without known diabetes, 17 (9%) had fasting glucose values ⬎ 9 mmol/l after the first pulse, and 6 (3%) had fasting glucose values ⬎ 10 mmol/l, but none of these subjects reached a fasting blood glucose value of 14 mmol/l or higher, and none required hypoglycemic drugs. Older age was associated with higher baseline values of fasting blood glucose in nondiabetic subjects (r ⫽ 0.33, P ⬍ 0.01). Moreover, compared with younger subjects, the 33 subjects aged 70 years or older showed slower fasting blood glucose decay curves after the hyperglycemic peaks induced by the first pulse (Fig 1). But none of these subjects required hypoglycemic intervention. Conversely, the 28 subjects with type 2 diabetes showed continuous increases of fasting blood glucose during pulse methylprednisolone therapy, after a median 44% increased value after the first pulse. Seven diabetic subjects (25%) received rapid-release insulin during pulse methylprednisolone therapy to maintain blood glucose values ⬍ 14 mmol/l, which was always triggered by the results of the self-monitoring of blood glucose in these subjects. They included all 5 subjects with glycosylated hemoglobin values ⱖ 8.3% and 2 of the 13 subjects with glycosylated hemoglobin values between 7.0% and 8.2%. No diabetic subject with a glyco-
Figure 1. Fasting blood glucose before methylprednisolone therapy and the day after each of three pulses according to diabetic status or subject’s age for nondiabetic subjects.
2370
Feldman-Billard et al 䡠 Intravenous Methylprednisolone Therapy and Glucose Tolerance sylated hemoglobin value ⬍ 7.0% required rapid-release insulin. In our series, there were no episodes of ketoacidosis or hyperosmolar hyperglycemia associated with pulse methylprednisolone therapy.
Discussion To our knowledge, this study is the first that shows the fasting blood glucose curves in a large nondiabetic and diabetic population of subjects over a three-day, high-dose methylprednisolone treatment period. These data may greatly improve daily practice because they demonstrate that nondiabetic subjects tolerate very well the hyperglycemic effects of methylprednisolone infusions, whereas diabetic subjects are dramatically sensitive to those effects and require close glucose monitoring during the treatment period. Glucocorticosteroids decrease glucose utilization; increase hepatic glucose production, lipolysis, proteolysis, and gluconeogenesis; and induce peripheral insulin resistance through a receptor and postreceptor steroid effect.7 These hyperglycemic effects are responsible for the socalled steroid diabetes observed in 6% to 25% of subjects treated with long-term oral corticosteroid therapy.8,9 Contrary to oral treatment, quantitative data are lacking on the hyperglycemic effects of short-term, high-dose intravenous steroid therapy and their therapeutic implications. In a series of 84 subjects who received 275 high-dose intravenous steroid infusions for systemic rheumatic disease, hyperglycemia was recorded in 16 subjects (19%) and blood glucose values of ⬎11.1 mmol/l in only 5 subjects (6%).10 In this study, one insulin-dependent subject required hypoglycemic therapeutic adjustments. In an other series, only 1 of 17 subjects (6%) experienced a raised blood glucose level (15.3 mmol/l).4 However, none of these studies reported detailed glycemic data for all subjects, whether diabetic or not, during intravenous steroid therapy. Our study shows that there is systematically a median 50% increase of morning fasting blood glucose values from baseline in diabetic and nondiabetic subjects after one pulse of methylprednisolone. During a three-day course of repeated pulses, the main points in nondiabetic subjects are as follows: (1) the hyperglycemic effects of steroids are at their maximum after the first pulse, with rapid physiologic regulatory mechanisms limiting the effects of the next pulses on blood glucose; (2) raised fasting blood glucose values, measured the day after the first, second, and third pulses, decrease more slowly toward baseline values with older age, probably because of the impairment of glucose-induced insulin release with aging; and (3) the glucose tolerance of the treatment was excellent in our subjects, who did not
require any hypoglycemic drug administration. In parallel, the main points in the diabetic population of our study are as follows: (1) the hyperglycemic effects of repeated pulses seem to be cumulative, and fasting blood glucose tends to increase during the treatment period, and (2) one fourth of diabetic subjects needed rapid-release insulin to maintain blood glucose values of ⬍14 mmol/l, including all 5 patients with hemoglobin glycosylated values ⱖ 8.3% at the time of treatment. We conclude that close glycemic monitoring is not necessary during pulse methylprednisolone therapy for ophthalmologic conditions in nondiabetic subjects with fasting baseline blood glucose values of ⬍7 mmol/l. Subjects with type 2 diabetes need careful blood glucose monitoring and have a high probability of requiring therapeutic adjustments if their glycosylated hemoglobin level is 8.3% or higher.
References 1. Boumpas DT, Chrousos GP, Wilder RL, et al. Glucocorticoid therapy for immune-mediated diseases: basic and clinical correlates. Ann Intern Med 1993;119:1198 –208. 2. Sabir S, Werth VP. Pulse glucocorticoids. Dermatol Clin 2000;18:437– 46. 3. Beck RW, Cleary PA, Anderson MM, et al, the Optic Neuritis Study Group. A randomized, controlled trial of corticosteroids in the treatment of acute optic neuritis. N Engl J Med 1992; 326:581– 8. 4. Wakefield D, McCluskey P, Penny R. Intravenous pulse methylprednisolone therapy in severe inflammatory eye disease. Arch Ophthalmol 1986;104:847–51. 5. Lam DS, Wong AK, Tham CC, Leung AT. The use of combined intravenous pulse methylprednisolone and oral cyclosporin A in the treatment of corneal graft rejection: a preliminary study. Eye 1998;12:615– 8. 6. Chevalet P, Barrier JH, Pottier P, et al. A randomized, multicenter, controlled trial using intravenous pulses of methylprednisolone in the initial treatment of simple forms of giant cell arteritis: a one year followup study of 164 patients. J Rheumatol 2000;27:1484 –91. 7. McMahon M, Gerich J, Rizza R. Effects of glucocorticoids on carbohydrate metabolism. Diabetes Metab Rev 1988;4:17–30. 8. Pandit MK, Burke J, Gustafson AB, et al. Drug-induced disorders of glucose tolerance. Ann Intern Med 1993;118: 529 –39. 9. Gurwitz JH, Bohn RL, Glynn RJ, et al. Glucocorticoids and the risk for initiation of hypoglycemic therapy. Arch Intern Med 1994;154:97–101. 10. Baethge BA, Lidsky MD, Goldberg JW. A study of adverse effects of high-dose intravenous (pulse) methylprednisolone therapy in patients with rheumatic disease. Ann Pharmacother 1992;26:316 –20.
2371
Short-term, high-dose intravenous methylprednisolone therapy, also called “pulse methylprednisolone” therapy, is widely used for a variety of inflammatory and immunologic disorders.1,2 In particular, it is used in ophthalmologic conditions such as acute optic neuritis,3 severe uveitis,4 corneal graft rejection,5 and giant cell arteritis.6 Because of the well-known hyperglycemic effects of glucocorticosteroids,7–9 the monitoring of blood glucose during pulse methylprednisolone therapy is recommended.2,10 However, whether and which type of glucose monitoring should be applied according to a subject’s characteristics, particularly the presence of diabetes, have not been evaluated. To determine which subjects need close glycemic monitoring Originally received: October 28, 2002. Accepted: April 16, 2003. Manuscript no. 220869. From Service de Me´decine Interne, Centre Hospitalier National d’Ophtalmologie des Quinze-Vingts, Paris, France. Correspondence and reprint requests to Sylvie Feldman-Billard, MD, Service de Me´decine Interne, Centre Hospitalier National d’Ophtalmologie des Quinze-Vingts, 28 rue de Charenton, 75571 Paris Cedex 12, France. E-mail: [email protected]. © 2003 by the American Academy of Ophthalmology Published by Elsevier Inc.
during pulse methylprednisolone therapy, we analyzed the results of serial fasting blood glucose measurements in subjects who received a three-day course of pulse methylprednisolone therapy for ophthalmologic conditions in our eye center over a one-year period.
Subjects and Methods The Centre Hospitalier National d’Ophtalmologie des QuinzeVingts in Paris, France, is a single-specialty eye center. Before pulse methylprednisolone therapy, every subject is seen by an internist who verifies the application of a standardized protocol for blood glucose monitoring written in 1995. According to this protocol, morning fasting blood glucose was measured in all subjects. In subjects with diabetes, the self-monitoring of capillary blood glucose was performed at least 3 times per day before each meal. Glycosylated hemoglobin was measured in subjects with known diabetes or initial fasting glucose ⱖ 7 mmol/l. A blood glucose of 14 mmol/l or higher led to hypoglycemic interventions, based on the administration of rapid-release insulin with a sliding scale of blood glucose values: 4 IU of rapid-release insulin for blood glucose values between 14 and 17 mmol/l, 6 IU for blood glucose ISSN 0161-6420/03/$–see front matter doi:10.1016/S0161-6420(03)00818-2
2369
Ophthalmology Volume 110, Number 12, December 2003 Table 1. Main Characteristics of the Study Population According to Diabetic Status Total Population (N ⴝ 224) Age, mean (SD), years Number of women, n (%) Baseline fasting blood glucose, mean (SD), mmol/l Baseline fasting blood glucose ⬎ 7 mmol/l, n (%)
Subjects without Subjects with Known Diabetes Known Diabetes (N ⴝ 196) (N ⴝ 28) 46.1 (19.8) 130 (66) 5.2 (0.8) 7 (3.6)
P
61.4 (20.2) 12 (43) 8.3 (4.1)
⬍0.001 0.02 ⬍0.001
14 (50)
0.01
SD ⫽ standard deviation.
values between 17 and 20 mmol/l, 8 IU for blood glucose values between 20 and 23 mmol/l, and so on. The medical reports of all subjects who received pulse methylprednisolone therapy during 2000 were analyzed. The steroid dosages, the main characteristics of subjects, the serial morning fasting blood glucose (i.e., before the first pulse and the day after each pulse), and the number and characteristics of subjects requiring insulin during steroid therapy were recorded. Statistical analysis was performed with the SAS statistical software for Windows (SAS Institute Inc., Cary, NC). Significance was considered for P values ⬍ 0.05. To evaluate the influence of initial fasting blood glucose and age on the glucose tolerance of pulse methylprednisolone therapy, these parameters were considered as continuous variables or separated in different categories (⬍5 mmol/l, 5–5.9 mmol/l, 6 – 6.9 mmol/l, and ⱖ7 mmol/l; ⬍30 years, 30 – 49 years, 50 – 69 years, and ⱖ70 years).
Results Two hundred seventy-one subjects received pulse methylprednisolone therapy at our institution during the year 2000. Fortyseven subjects received fewer than three methylprednisolone infusions or had incomplete blood glucose data. Ultimately, the analysis was completed in 224 subjects (mean age ⫽ 46.1 years;
56% women) who received 250 mg (n ⫽ 20), 500 mg (n ⫽ 109), or 1000 mg (n ⫽ 95) pulse methylprednisolone once a day over 3 consecutive days. The indications of pulse methylprednisolone therapy were acute optic neuritis (n ⫽ 91), severe uveitis (n ⫽ 35), ocular infectious diseases (n ⫽ 22), corneal graft rejection (n ⫽ 17), and miscellaneous conditions (n ⫽ 59). Table 1 shows the main characteristics of the subjects according to diabetic status. No subject had type 1 diabetes, and 28 had type 2 diabetes, defined as an age of diabetes onset of 30 years or older. All but 1 had been treated with oral hypoglycemic agents for 10.6 ⫾ 8.4 years (range ⫽ 4 –28) or with insulin (n ⫽ 2); these treatments were maintained during methylprednisolone therapy. The fasting blood glucose curves during pulse methylprednisolone therapy according to diabetic status are shown in Figure 1. The 196 nondiabetic subjects showed a median 50% increase of fasting blood glucose after the first pulse and a subsequent spontaneous decay toward baseline values, despite the second and third pulses; their fasting blood glucose after the third pulse was a median of 23% higher than baseline values. Among these 196 subjects without known diabetes, 17 (9%) had fasting glucose values ⬎ 9 mmol/l after the first pulse, and 6 (3%) had fasting glucose values ⬎ 10 mmol/l, but none of these subjects reached a fasting blood glucose value of 14 mmol/l or higher, and none required hypoglycemic drugs. Older age was associated with higher baseline values of fasting blood glucose in nondiabetic subjects (r ⫽ 0.33, P ⬍ 0.01). Moreover, compared with younger subjects, the 33 subjects aged 70 years or older showed slower fasting blood glucose decay curves after the hyperglycemic peaks induced by the first pulse (Fig 1). But none of these subjects required hypoglycemic intervention. Conversely, the 28 subjects with type 2 diabetes showed continuous increases of fasting blood glucose during pulse methylprednisolone therapy, after a median 44% increased value after the first pulse. Seven diabetic subjects (25%) received rapid-release insulin during pulse methylprednisolone therapy to maintain blood glucose values ⬍ 14 mmol/l, which was always triggered by the results of the self-monitoring of blood glucose in these subjects. They included all 5 subjects with glycosylated hemoglobin values ⱖ 8.3% and 2 of the 13 subjects with glycosylated hemoglobin values between 7.0% and 8.2%. No diabetic subject with a glyco-
Figure 1. Fasting blood glucose before methylprednisolone therapy and the day after each of three pulses according to diabetic status or subject’s age for nondiabetic subjects.
2370
Feldman-Billard et al 䡠 Intravenous Methylprednisolone Therapy and Glucose Tolerance sylated hemoglobin value ⬍ 7.0% required rapid-release insulin. In our series, there were no episodes of ketoacidosis or hyperosmolar hyperglycemia associated with pulse methylprednisolone therapy.
Discussion To our knowledge, this study is the first that shows the fasting blood glucose curves in a large nondiabetic and diabetic population of subjects over a three-day, high-dose methylprednisolone treatment period. These data may greatly improve daily practice because they demonstrate that nondiabetic subjects tolerate very well the hyperglycemic effects of methylprednisolone infusions, whereas diabetic subjects are dramatically sensitive to those effects and require close glucose monitoring during the treatment period. Glucocorticosteroids decrease glucose utilization; increase hepatic glucose production, lipolysis, proteolysis, and gluconeogenesis; and induce peripheral insulin resistance through a receptor and postreceptor steroid effect.7 These hyperglycemic effects are responsible for the socalled steroid diabetes observed in 6% to 25% of subjects treated with long-term oral corticosteroid therapy.8,9 Contrary to oral treatment, quantitative data are lacking on the hyperglycemic effects of short-term, high-dose intravenous steroid therapy and their therapeutic implications. In a series of 84 subjects who received 275 high-dose intravenous steroid infusions for systemic rheumatic disease, hyperglycemia was recorded in 16 subjects (19%) and blood glucose values of ⬎11.1 mmol/l in only 5 subjects (6%).10 In this study, one insulin-dependent subject required hypoglycemic therapeutic adjustments. In an other series, only 1 of 17 subjects (6%) experienced a raised blood glucose level (15.3 mmol/l).4 However, none of these studies reported detailed glycemic data for all subjects, whether diabetic or not, during intravenous steroid therapy. Our study shows that there is systematically a median 50% increase of morning fasting blood glucose values from baseline in diabetic and nondiabetic subjects after one pulse of methylprednisolone. During a three-day course of repeated pulses, the main points in nondiabetic subjects are as follows: (1) the hyperglycemic effects of steroids are at their maximum after the first pulse, with rapid physiologic regulatory mechanisms limiting the effects of the next pulses on blood glucose; (2) raised fasting blood glucose values, measured the day after the first, second, and third pulses, decrease more slowly toward baseline values with older age, probably because of the impairment of glucose-induced insulin release with aging; and (3) the glucose tolerance of the treatment was excellent in our subjects, who did not
require any hypoglycemic drug administration. In parallel, the main points in the diabetic population of our study are as follows: (1) the hyperglycemic effects of repeated pulses seem to be cumulative, and fasting blood glucose tends to increase during the treatment period, and (2) one fourth of diabetic subjects needed rapid-release insulin to maintain blood glucose values of ⬍14 mmol/l, including all 5 patients with hemoglobin glycosylated values ⱖ 8.3% at the time of treatment. We conclude that close glycemic monitoring is not necessary during pulse methylprednisolone therapy for ophthalmologic conditions in nondiabetic subjects with fasting baseline blood glucose values of ⬍7 mmol/l. Subjects with type 2 diabetes need careful blood glucose monitoring and have a high probability of requiring therapeutic adjustments if their glycosylated hemoglobin level is 8.3% or higher.
References 1. Boumpas DT, Chrousos GP, Wilder RL, et al. Glucocorticoid therapy for immune-mediated diseases: basic and clinical correlates. Ann Intern Med 1993;119:1198 –208. 2. Sabir S, Werth VP. Pulse glucocorticoids. Dermatol Clin 2000;18:437– 46. 3. Beck RW, Cleary PA, Anderson MM, et al, the Optic Neuritis Study Group. A randomized, controlled trial of corticosteroids in the treatment of acute optic neuritis. N Engl J Med 1992; 326:581– 8. 4. Wakefield D, McCluskey P, Penny R. Intravenous pulse methylprednisolone therapy in severe inflammatory eye disease. Arch Ophthalmol 1986;104:847–51. 5. Lam DS, Wong AK, Tham CC, Leung AT. The use of combined intravenous pulse methylprednisolone and oral cyclosporin A in the treatment of corneal graft rejection: a preliminary study. Eye 1998;12:615– 8. 6. Chevalet P, Barrier JH, Pottier P, et al. A randomized, multicenter, controlled trial using intravenous pulses of methylprednisolone in the initial treatment of simple forms of giant cell arteritis: a one year followup study of 164 patients. J Rheumatol 2000;27:1484 –91. 7. McMahon M, Gerich J, Rizza R. Effects of glucocorticoids on carbohydrate metabolism. Diabetes Metab Rev 1988;4:17–30. 8. Pandit MK, Burke J, Gustafson AB, et al. Drug-induced disorders of glucose tolerance. Ann Intern Med 1993;118: 529 –39. 9. Gurwitz JH, Bohn RL, Glynn RJ, et al. Glucocorticoids and the risk for initiation of hypoglycemic therapy. Arch Intern Med 1994;154:97–101. 10. Baethge BA, Lidsky MD, Goldberg JW. A study of adverse effects of high-dose intravenous (pulse) methylprednisolone therapy in patients with rheumatic disease. Ann Pharmacother 1992;26:316 –20.
2371