British Journal of Rheumatology 1991 ;3O:37O-372
PRELIMINARY REPORT
EVENING PRIMROSE OIL IN PATIENTS WITH RHEUMATOID ARTHRITIS AND SIDE-EFFECTS OF NON-STEROIDAL ANTIINFLAMMATORY DRUGS BY M. BRZESKI, R. MADHOK AND H. A. CAPELL Centre for Rheumatic Diseases, University Department of Medicine, Roval Infirmary, Glasgow
SUMMARY Forty patients with rheumatoid arthritis and upper gastrointestinal lesions due to non-steroidal anti-inflammatory drugs entered a prospective 6-month double-blind placebo controlled study of dietary supplementation with gamma-linolenic acid 540 mg/day. Nineteen patients received active therapy (as evening primrose oil 6g/day) and 21 received placebo (olive oil 6g/day). No patient stopped non-steroidal anti-inflammatory therapy but three patients in each group reduced their dose. Other results showed a significant reduction in morning stiffness with gamma-linolenic acid at 3 months and reduction in pain and articular index at 6 months with olive oil. Whilst gamma-linolenic acid may produce mild improvement in rheumatoid arthritis, olive oil may itself have hitherto unrecognized benefits. KEY WORDS:
Rheumatoid arthritis. Inflammation, Evening primrose oil, NSAIDs, Peptic ulcer, Gastritis.
THE possibility that dietary supplements might improve clinical features of rheumatoid arthritis (RA) whilst reducing the need for potentially toxic drugs is both appealing and topical. Dietary fatty acids provide the direct substrates for biosynthesis of prostaglandins and leukotrienes and altering the type of dietary fatty acid changes the balance of prostaglandins produced during inflammation [1]. Gamma-linolenic acid (GLA), found in evening primrose oil (EPO) is metabolized to dihommogammalinolenic acid (DGLA) which results in an increase of 1-series prostaglandins (PGs) [1]. PGE, has important anti-inflammatory actions [2,3]. A previous study suggested that EPO supplements enable up to 60% of RA patients to reduce or stop their dose of non-steroidal anti-inflammatory drugs (NSAID) [4]. Since NSAIDs produce considerable morbidity and mortality [5], we studied EPO in patients with RA who already had upper gastrointestinal lesions attributable to NSAIDs. In these patients withdrawal of NSAID without worsening of joint symptoms would be particularly valuable.
6-month double-blind placebo controlled study. Patients received evening primrose oil 6g/day (i.e. 540 mg GLA/day) or placebo (olive oil, 6g/day) in identical capsules. EPO capsules contained 10 mg alpha-tocopherol each as antioxidant. At 0, 3 and 6 months the following assessments were performed: daily use of NSAID and analgesia (patient diary), morning stiffness (MS), 100 mm horizontal visual analogue scales for pain and well-being, Ritchie articular index and health assessment questionnaire (HAQ), Hb, platelets, ESR, C-reactive protein (CRP), globulins, and plasma fatty acid analysis by HPLC. After 3 months, patients were asked every week to attempt reduction of their NSAID and analgesics. Mann-Whitney t/-test was used for comparison between groups and Wilcoxon two-tailed paired test for sequential changes within a group. RESULTS Nineteen patients received EPO and 21 placebo. There were no statistically significant differences in TABLE I
PATIENTS AND METHODS Patients were aged 16-75 and had classical or definite RA. All had upper gastrointestinal lesions due to NSAIDs, i.e. peptic ulcer or gastritis at endoscopy or on barium meal, or symptoms strongly suggestive of these diagnoses. All patients continued routine medication, including NSAID, H2 blocker (if used) and analgesia. Patients on second-line therapy were randomized separately. Patients taking systemic corticosteroids were excluded. Local ethical committee approval was obtained and informed consent given by each patient. Forty patients were randomized into a prospective
BASELINE CLINICAL CHARACTERISTICS
Evening primrose oil Placebo (olive oil) n Female.male Age (years) Disease (years) Second-line therapy Peptic ulcer/gastritis H, antagonist Pain (mm) Well-being (mm) Al MS (min) HAQ Hb (g/dl) Platelets (xlOTl) ESR (mm/h) CRP(mg/l) Globulins (g/1)
Submitted 22 August: revised version accepted 31 December 1990. Correspondence to Dr M. Brzeski. Rheumatic Diseases Research Centre. Stoke Mandeville Hospital. Aylesburv. Bucks HP21 SAL.
19 17:2 60 (54-77) 8(6-22.8)
8 12/7 8 55 (40-65) 65 (50-85) 12 (5-21) 53 (14-75) 1.3 (1.0- 2.3) 11.9(10.4-13.1) 357 (270-428) 41 (19-59) 14 (10-^5) 27 (23-32)
Figures are medians (interquartile ranges). 370
21 15:6 61 (51-67) 8(4 5-12) 10 14/7 12 50 (3S-70) 60 (30-80) 14 (6-24) 60 (30-150) 2 1 (1.5- 2.4) 11.6(10.4-13.7) 338(241-490) 42 (11-69) 13 (10-53) 28 (24-30)
BRZESKI ETAL.:
EVENING PRIMROSE OIL AND RA
371
TABLE II CHANGE IN CLINICAL VARIABLES DURING THE STUDY
Evening primrose oil Time (months) Pain Articular index Mornine stiffness (mm)
0 3D
(40-65) 12 (5-21) 53 (14-75)
Placebo (olive oil)
3 40 (18-58)
6 55 (20-73)
3°
3°
(6-14)
(0-14) 10t (10-60)
10+ (5-45)
0 50 (38-70) 14 (6-24) 60 (30-150)
3 55 (30-70) 9 (2-16) 45 (20-60)
6
40U (25-50) 4V (2-12)
30f|
(25-60)
Figures are medians (interquartile ranges). *P = 0.2: IP = 0.04; +P = 0.14; \P = 0.02; H> P = 0.OO4; || P = 0.06 (Wilcoxon).
their baseline characteristics (Table I). Six of 19 EPO treated patients withdrew from the study—two with nausea, two for joint surgery, one with deteriorating RA and one due to 'flu-like' symptoms. Four of 21 in the placebo group withdrew—three with nausea and one leaving the area. Of the 13 completing EPO, 10 showed a significant rise in plasma DGLA (a metabolite of GLA) suggesting good compliance (mean at baseline 3.44% of total fatty acid, SD ± 0.95, mean at 6 months 4.85%, SD ±1.52; placebo group mean at baseline 3.62% ± 0.76, at 6 months 3.48 ± 0.57). No patients stopped NSAIDs but three in each group reduced the dose of NSAID—in all patients this was by only one tablet, e.g. ibuprofen 400 mg t.d.s. to b.d.—and one patient in the EPO group increased NSAID dosage. Four patients taking placebo and one taking EPO reduced analgesia dosage, and two in each group increased dosage. In the EPO group, MS was significantly reduced at 3 months with a trend to reduction at 6 months (Table II). The marked fall in AI did not reach significance. Pain score did not change. In the placebo group, there was significant reduction in AI and pain at 6 months and a trend to reduced morning stiffness at 6 months. There was no change in well-being or HAQ scores in either group. As expected, there were no changes in laboratory parameters of inflammation (data not shown). Separate analysis of the EPO group after excluding those showing no rise in plasma metabolites of GLA (i.e. excluding non-compliers) did not alter these results ( data not shown). DISCUSSION Evidence is accumulating that altering dietary fatty acid may affect activity of rheumatoid arthritis. Eicosapentaenoic acid (EPA) prolongs survival in experimental models of inflammation [6]. Clinical improvement occurs in RA after supplements offish oil rich in EPA, and is associated with reduced leukotriene B4 production [7]. Evening primrose oil is rich in GLA, whose metabolite 15-hydroxy DGLA blocks conversion of arachidonic acid (AA) to leukotrienes [8]. DGLA itself produces 1-series PGs, which are less inflammatory than those derived from AA [2], Animal studies have shown that GLA can suppress both acute and chronic inflammation [9]. Our study found that only 23% (3/13) of patients completing treatment with EPO could reduce their NSAID dose and none could stop, similar to that found
with placebo. This contrasts with a previous study [4] in which the same dose of EPO enabled 25% to stop and a further 38% to reduce NSAID dose after 6 months without clinical deterioration. Although continued for 1 year (as recommended by the manufacturer) most of the benefit was attained by 6 months. In that study, patients had less severe RA and none was on secondline therapy. EPO may thus be beneficial only in mild RA. In our study, EPO produced marked reduction in morning stiffness and articular index, although only the former reached statistical significance. We are unable to recommend EPO for severe RA. It is disappointing that we have been unable to confirm the possibility of substituting EPO for NSAIDs in patients with NSAID-induced upper gastrointestinal sideeffects. It is interesting that olive oil enabled as many patients to reduce NSAIDs as did EPO, and to produce improvement in more clinical parameters than EPO. Reduction in pain, articular index, and morning stiffness suggest a generalized improvement in clinical features of RA. Olive oil is thought to be neutral with respect to fatty acid metabolism, but Cleland et al. [10] found that, used as placebo, olive oil 18 g/day reduced morning stiffness and pain score. They interpreted this as a placebo effect on subjective indices of disease activity, which is a possible explanation for our findings. It is conceivable, however, that olive oil may displace DGLA from the cell membrane and exert an indirect anti-inflammatory effect. We are not aware of any accurate data on the incidence and severity of RA in countries where olive oil production and consumption is high. An unattributed comment in a major textbook [11] claims that severe RA is less frequent in Italy. Olive oil can no longer confidently be used as a placebo control, and should itself be investigated. ACKNOWLEDGEMENTS
We thank Miss A. Tierney for typing the manuscript, Sister S. Armstrong for metrology and Efamol Ltd for providing the materials, financial assistance and fatty acid analysis. Dr R. Madhok is an Arthritis and Rheumatism Council Lecturer. REFERENCES
1. Willis AL. Nutritional and pharmacological factors in eicosanoid biology. Nutr Rev 1981;39:289-301. 2. ZurierRB,QuagliataF. Effect of prostaglandin El on adjuvant arthritis. Nature 1971;234:304-5.
372
BRITISH JOURNAL OF RHEUMATOLOGY VOL. XXX NO. 5
3. Smith JW, Steiner AL, Parker CW. Human lymphocyte metabolism. Effects of cyclic and noncyclic nucleotides on stimulation by phytohemagglutinin. J Clin Invest 1971;50:442-8. 4. Belch JJF, Ansell D, Madhok R, O'Dowd A, Sturrock RD. Effects of altering dietary essential fatty acids on requirements for non-steroidal anti-inflammatory drugs in patients with rheumatoid arthritis: a double blind placebo controlled study. Ann Rheum Dis 1988;47:96-104. 5. Armstrong CP, Blower AL. Non-steroidal antiinflammatory drugs and life threatening complications of peptic ulceration. Gut 1987;28:527-32. 6. Prickett JD, Robinson DR, Steinberg AD. Dietary enrichment with polyunsaturated fatty acid eicosapentaenoic acid prevents proteinuria and prolongs survival in NZB x NZW Fl mice. J Clin Invest 1981;68:556-9. 7. Kremer J, Lawrence D, JubizW, DigiacomoR, Rynes
R, Bartholomew L. Different doses of fish-oil fatty acid supplementation in rheumatoid arthritis. A prospective double-blinded randomized study. Arthritis Rheum 1988;31:S30. 8. Voorhjees JJ. Leukotrienes and other lipoxygenase products in the pathogenesis and therapy of psoriasis and other dermatoses. Arch Dermatol 1983;119:541-7. 9. Tate G, Mandell BF, Laposta M, etal. Suppression of acute and chronic inflammation by dietary gamma linolenic acid. J Rheumatol 1989;16:729-33. 10. Cleland LG, French JK, Betts WH, Murphy GA, Elliott MJ. Clinical and biochemical effects of dietary fish oil supplements in rheumatoid arthritis. J Rheumatol 1988;15:1471-5. 11. Harris ED. The clinical features of rheumatoid arthritis. In: Kelley WN, Harris ED, Ruddy S, Sledge CD, eds. Textbook of rheumatology. 3rd ed. Philadelphia: WB Saunders, 1989:945.
PRELIMINARY REPORT
EVENING PRIMROSE OIL IN PATIENTS WITH RHEUMATOID ARTHRITIS AND SIDE-EFFECTS OF NON-STEROIDAL ANTIINFLAMMATORY DRUGS BY M. BRZESKI, R. MADHOK AND H. A. CAPELL Centre for Rheumatic Diseases, University Department of Medicine, Roval Infirmary, Glasgow
SUMMARY Forty patients with rheumatoid arthritis and upper gastrointestinal lesions due to non-steroidal anti-inflammatory drugs entered a prospective 6-month double-blind placebo controlled study of dietary supplementation with gamma-linolenic acid 540 mg/day. Nineteen patients received active therapy (as evening primrose oil 6g/day) and 21 received placebo (olive oil 6g/day). No patient stopped non-steroidal anti-inflammatory therapy but three patients in each group reduced their dose. Other results showed a significant reduction in morning stiffness with gamma-linolenic acid at 3 months and reduction in pain and articular index at 6 months with olive oil. Whilst gamma-linolenic acid may produce mild improvement in rheumatoid arthritis, olive oil may itself have hitherto unrecognized benefits. KEY WORDS:
Rheumatoid arthritis. Inflammation, Evening primrose oil, NSAIDs, Peptic ulcer, Gastritis.
THE possibility that dietary supplements might improve clinical features of rheumatoid arthritis (RA) whilst reducing the need for potentially toxic drugs is both appealing and topical. Dietary fatty acids provide the direct substrates for biosynthesis of prostaglandins and leukotrienes and altering the type of dietary fatty acid changes the balance of prostaglandins produced during inflammation [1]. Gamma-linolenic acid (GLA), found in evening primrose oil (EPO) is metabolized to dihommogammalinolenic acid (DGLA) which results in an increase of 1-series prostaglandins (PGs) [1]. PGE, has important anti-inflammatory actions [2,3]. A previous study suggested that EPO supplements enable up to 60% of RA patients to reduce or stop their dose of non-steroidal anti-inflammatory drugs (NSAID) [4]. Since NSAIDs produce considerable morbidity and mortality [5], we studied EPO in patients with RA who already had upper gastrointestinal lesions attributable to NSAIDs. In these patients withdrawal of NSAID without worsening of joint symptoms would be particularly valuable.
6-month double-blind placebo controlled study. Patients received evening primrose oil 6g/day (i.e. 540 mg GLA/day) or placebo (olive oil, 6g/day) in identical capsules. EPO capsules contained 10 mg alpha-tocopherol each as antioxidant. At 0, 3 and 6 months the following assessments were performed: daily use of NSAID and analgesia (patient diary), morning stiffness (MS), 100 mm horizontal visual analogue scales for pain and well-being, Ritchie articular index and health assessment questionnaire (HAQ), Hb, platelets, ESR, C-reactive protein (CRP), globulins, and plasma fatty acid analysis by HPLC. After 3 months, patients were asked every week to attempt reduction of their NSAID and analgesics. Mann-Whitney t/-test was used for comparison between groups and Wilcoxon two-tailed paired test for sequential changes within a group. RESULTS Nineteen patients received EPO and 21 placebo. There were no statistically significant differences in TABLE I
PATIENTS AND METHODS Patients were aged 16-75 and had classical or definite RA. All had upper gastrointestinal lesions due to NSAIDs, i.e. peptic ulcer or gastritis at endoscopy or on barium meal, or symptoms strongly suggestive of these diagnoses. All patients continued routine medication, including NSAID, H2 blocker (if used) and analgesia. Patients on second-line therapy were randomized separately. Patients taking systemic corticosteroids were excluded. Local ethical committee approval was obtained and informed consent given by each patient. Forty patients were randomized into a prospective
BASELINE CLINICAL CHARACTERISTICS
Evening primrose oil Placebo (olive oil) n Female.male Age (years) Disease (years) Second-line therapy Peptic ulcer/gastritis H, antagonist Pain (mm) Well-being (mm) Al MS (min) HAQ Hb (g/dl) Platelets (xlOTl) ESR (mm/h) CRP(mg/l) Globulins (g/1)
Submitted 22 August: revised version accepted 31 December 1990. Correspondence to Dr M. Brzeski. Rheumatic Diseases Research Centre. Stoke Mandeville Hospital. Aylesburv. Bucks HP21 SAL.
19 17:2 60 (54-77) 8(6-22.8)
8 12/7 8 55 (40-65) 65 (50-85) 12 (5-21) 53 (14-75) 1.3 (1.0- 2.3) 11.9(10.4-13.1) 357 (270-428) 41 (19-59) 14 (10-^5) 27 (23-32)
Figures are medians (interquartile ranges). 370
21 15:6 61 (51-67) 8(4 5-12) 10 14/7 12 50 (3S-70) 60 (30-80) 14 (6-24) 60 (30-150) 2 1 (1.5- 2.4) 11.6(10.4-13.7) 338(241-490) 42 (11-69) 13 (10-53) 28 (24-30)
BRZESKI ETAL.:
EVENING PRIMROSE OIL AND RA
371
TABLE II CHANGE IN CLINICAL VARIABLES DURING THE STUDY
Evening primrose oil Time (months) Pain Articular index Mornine stiffness (mm)
0 3D
(40-65) 12 (5-21) 53 (14-75)
Placebo (olive oil)
3 40 (18-58)
6 55 (20-73)
3°
3°
(6-14)
(0-14) 10t (10-60)
10+ (5-45)
0 50 (38-70) 14 (6-24) 60 (30-150)
3 55 (30-70) 9 (2-16) 45 (20-60)
6
40U (25-50) 4V (2-12)
30f|
(25-60)
Figures are medians (interquartile ranges). *P = 0.2: IP = 0.04; +P = 0.14; \P = 0.02; H> P = 0.OO4; || P = 0.06 (Wilcoxon).
their baseline characteristics (Table I). Six of 19 EPO treated patients withdrew from the study—two with nausea, two for joint surgery, one with deteriorating RA and one due to 'flu-like' symptoms. Four of 21 in the placebo group withdrew—three with nausea and one leaving the area. Of the 13 completing EPO, 10 showed a significant rise in plasma DGLA (a metabolite of GLA) suggesting good compliance (mean at baseline 3.44% of total fatty acid, SD ± 0.95, mean at 6 months 4.85%, SD ±1.52; placebo group mean at baseline 3.62% ± 0.76, at 6 months 3.48 ± 0.57). No patients stopped NSAIDs but three in each group reduced the dose of NSAID—in all patients this was by only one tablet, e.g. ibuprofen 400 mg t.d.s. to b.d.—and one patient in the EPO group increased NSAID dosage. Four patients taking placebo and one taking EPO reduced analgesia dosage, and two in each group increased dosage. In the EPO group, MS was significantly reduced at 3 months with a trend to reduction at 6 months (Table II). The marked fall in AI did not reach significance. Pain score did not change. In the placebo group, there was significant reduction in AI and pain at 6 months and a trend to reduced morning stiffness at 6 months. There was no change in well-being or HAQ scores in either group. As expected, there were no changes in laboratory parameters of inflammation (data not shown). Separate analysis of the EPO group after excluding those showing no rise in plasma metabolites of GLA (i.e. excluding non-compliers) did not alter these results ( data not shown). DISCUSSION Evidence is accumulating that altering dietary fatty acid may affect activity of rheumatoid arthritis. Eicosapentaenoic acid (EPA) prolongs survival in experimental models of inflammation [6]. Clinical improvement occurs in RA after supplements offish oil rich in EPA, and is associated with reduced leukotriene B4 production [7]. Evening primrose oil is rich in GLA, whose metabolite 15-hydroxy DGLA blocks conversion of arachidonic acid (AA) to leukotrienes [8]. DGLA itself produces 1-series PGs, which are less inflammatory than those derived from AA [2], Animal studies have shown that GLA can suppress both acute and chronic inflammation [9]. Our study found that only 23% (3/13) of patients completing treatment with EPO could reduce their NSAID dose and none could stop, similar to that found
with placebo. This contrasts with a previous study [4] in which the same dose of EPO enabled 25% to stop and a further 38% to reduce NSAID dose after 6 months without clinical deterioration. Although continued for 1 year (as recommended by the manufacturer) most of the benefit was attained by 6 months. In that study, patients had less severe RA and none was on secondline therapy. EPO may thus be beneficial only in mild RA. In our study, EPO produced marked reduction in morning stiffness and articular index, although only the former reached statistical significance. We are unable to recommend EPO for severe RA. It is disappointing that we have been unable to confirm the possibility of substituting EPO for NSAIDs in patients with NSAID-induced upper gastrointestinal sideeffects. It is interesting that olive oil enabled as many patients to reduce NSAIDs as did EPO, and to produce improvement in more clinical parameters than EPO. Reduction in pain, articular index, and morning stiffness suggest a generalized improvement in clinical features of RA. Olive oil is thought to be neutral with respect to fatty acid metabolism, but Cleland et al. [10] found that, used as placebo, olive oil 18 g/day reduced morning stiffness and pain score. They interpreted this as a placebo effect on subjective indices of disease activity, which is a possible explanation for our findings. It is conceivable, however, that olive oil may displace DGLA from the cell membrane and exert an indirect anti-inflammatory effect. We are not aware of any accurate data on the incidence and severity of RA in countries where olive oil production and consumption is high. An unattributed comment in a major textbook [11] claims that severe RA is less frequent in Italy. Olive oil can no longer confidently be used as a placebo control, and should itself be investigated. ACKNOWLEDGEMENTS
We thank Miss A. Tierney for typing the manuscript, Sister S. Armstrong for metrology and Efamol Ltd for providing the materials, financial assistance and fatty acid analysis. Dr R. Madhok is an Arthritis and Rheumatism Council Lecturer. REFERENCES
1. Willis AL. Nutritional and pharmacological factors in eicosanoid biology. Nutr Rev 1981;39:289-301. 2. ZurierRB,QuagliataF. Effect of prostaglandin El on adjuvant arthritis. Nature 1971;234:304-5.
372
BRITISH JOURNAL OF RHEUMATOLOGY VOL. XXX NO. 5
3. Smith JW, Steiner AL, Parker CW. Human lymphocyte metabolism. Effects of cyclic and noncyclic nucleotides on stimulation by phytohemagglutinin. J Clin Invest 1971;50:442-8. 4. Belch JJF, Ansell D, Madhok R, O'Dowd A, Sturrock RD. Effects of altering dietary essential fatty acids on requirements for non-steroidal anti-inflammatory drugs in patients with rheumatoid arthritis: a double blind placebo controlled study. Ann Rheum Dis 1988;47:96-104. 5. Armstrong CP, Blower AL. Non-steroidal antiinflammatory drugs and life threatening complications of peptic ulceration. Gut 1987;28:527-32. 6. Prickett JD, Robinson DR, Steinberg AD. Dietary enrichment with polyunsaturated fatty acid eicosapentaenoic acid prevents proteinuria and prolongs survival in NZB x NZW Fl mice. J Clin Invest 1981;68:556-9. 7. Kremer J, Lawrence D, JubizW, DigiacomoR, Rynes
R, Bartholomew L. Different doses of fish-oil fatty acid supplementation in rheumatoid arthritis. A prospective double-blinded randomized study. Arthritis Rheum 1988;31:S30. 8. Voorhjees JJ. Leukotrienes and other lipoxygenase products in the pathogenesis and therapy of psoriasis and other dermatoses. Arch Dermatol 1983;119:541-7. 9. Tate G, Mandell BF, Laposta M, etal. Suppression of acute and chronic inflammation by dietary gamma linolenic acid. J Rheumatol 1989;16:729-33. 10. Cleland LG, French JK, Betts WH, Murphy GA, Elliott MJ. Clinical and biochemical effects of dietary fish oil supplements in rheumatoid arthritis. J Rheumatol 1988;15:1471-5. 11. Harris ED. The clinical features of rheumatoid arthritis. In: Kelley WN, Harris ED, Ruddy S, Sledge CD, eds. Textbook of rheumatology. 3rd ed. Philadelphia: WB Saunders, 1989:945.
