oral iloprost in raynaud's phenomenon secondary ... - Semantic Scholar

British Journal of Rheumatology 1998;37:952–960

ORAL ILOPROST IN RAYNAUD’S PHENOMENON SECONDARY TO SYSTEMIC SCLEROSIS: A MULTICENTRE, PLACEBO-CONTROLLED, DOSE-COMPARISON STUDY C. M. BLACK, L. HALKIER-SØRENSEN,* J. J. F. BELCH,† S. ULLMAN,‡ R. MADHOK,§ A. J. SMIT,¶ J.-D. BANGAd and H. R. WATSON** ON BEHALF OF THE STUDY GROUP Royal Free Hospital, London NW3 2QG, *Marselisborg Hospital, 8000 Aarhus, Denmark, †Ninewells Hospital and Medical School, Dundee DD1 9SY, ‡Bispebjerg Hospital, 2400 Copenhagen NV, Denmark, §Royal Infirmary, Glasgow G4 0SF, ¶University Hospital, 9713 EZ Groningen, dUniversity Hospital, 3508 Utrecht, The Netherlands and **Schering AG, 13324 Berlin, Germany SUMMARY Objective. To identify the optimal dose of oral iloprost on the basis of efficacy and tolerability in patients with Raynaud’s phenomenon secondary to systemic sclerosis. Design. Multicentre, randomized, parallel-group comparison of two different doses of oral iloprost and placebo. Setting. European university hospitals. Patients. A total of 103 patients with Raynaud’s phenomenon secondary to systemic sclerosis. Intervention. Patients received one of three treatments for 6 weeks: placebo, oral iloprost 50 mg or oral iloprost 100 mg. Each treatment was taken twice daily, giving total daily doses of iloprost of 100 and 200 mg. Measurements. The frequency, total daily duration and severity of Raynaud’s attacks were recorded in a specially designed patient diary; physician’s global assessment and adverse events were recorded at visits to the clinic. Analysis was performed on an intention-to-treat population. Results. A total of 103 patients were recruited, 89 completed the assessments throughout the treatment period and 82 completed an additional 6 weeks of follow-up after treatment. Thirty-five patients received placebo, 33 received iloprost 50 mg and 35 received iloprost 100 mg. The mean percentage reductions in the frequency, total daily duration and severity of Raynaud’s attacks were numerically greater in the iloprost groups at the end of treatment and at the end of follow-up. At the end of treatment (6 weeks), there were significant treatment differences in the total daily duration of attacks (P = 0.03), but not in the severity (P = 0.07) or the frequency of attacks (P = 0.37). At the end of follow-up (12 weeks), there were significant treatment differences in the total daily duration of attacks (P = 0.001) and in the severity of attacks (P = 0.007), but not in the frequency of attacks (P = 0.07). Percentages of patients improved at the end of treatment as assessed by the physician were 44% placebo, 57% iloprost 50 mg and 64% iloprost 100 mg (not significant). Side-effects were reported by 80% of patients on placebo, 85% on oral iloprost 50 mg and 97% on oral iloprost 100 mg. Premature discontinuations of treatment in each group were 9, 30 and 51%, respectively, with 6, 27 and 51% being due to adverse events. Conclusion. The results on the daily duration of Raynaud’s attacks suggest that both 50 and 100 mg oral iloprost twice daily may be effective in the treatment of Raynaud’s phenomenon secondary to systemic sclerosis. The 50 mg iloprost dose was better tolerated in this patient group. K : Iloprost, Raynaud’s phenomenon, Systemic sclerosis, Optimal dose.

R’ phenomenon, which occurs in over 90% of patients with systemic sclerosis, is often difficult to treat [1]. In addition to very troublesome vasospasm, which appears to be the result of an imbalance of vascular factors normally maintaining vessel tone, such patients have a damaging proliferative structural vascular lesion [2]. The endothelial lesion, characteristic of scleroderma, is associated with evidence of immunemediated damage [3], in vivo platelet activation [4] and release of the vasoconstrictors serotonin and thromboxane [5, 6 ], a decrease in the vasodilators nitric oxide [7] and calcitonin gene-related peptide [8], increased resistance to the vasodilator prostacyclin [9, 10], free-radical generation [11, 12] and a leaky endothelium which permits the egress of inflammatory cells into the perivascular space [13].

Iloprost is a chemically stable prostacyclin analogue with potent vasodilatory effects [14]. It also has additional potentially useful properties of inhibiting platelet adhesion and aggregation, altering neutrophil function, including free-radical production by neutrophils and neutrophil adhesion to endothelial cells [15], and it has been hypothesized that it may speed the repair of damaged endothelium. Iloprost has been shown in previous studies to have immediate and long-term benefits in the treatment of Raynaud’s when compared with placebo [16, 17], and also to be as effective as nifedipine in a parallel group trial [18]. It has also been shown to be effective against vasospasm even when given in low dose [19] and can produce significant healing of digital ulcers [20]. Its use in the treatment of Raynaud’s has been restricted to date to patients with very severe disease because of the need for hospital admissions and i.v. infusions. An orally active preparation of iloprost is now available and doses were chosen from a dose titration study [21]. We describe the results of the first multicentre,

Submitted 22 September 1997; revised version accepted 19 April 1998. Correspondence to: C. M. Black, Department of Rheumatology, Royal Free Hospital, Pond Street, London NW3 2OG.

© 1998 British Society for Rheumatology 952

BLACK ET AL.: ORAL ILOPROST IN SECONDARY RAYNAUD’S

double-blind, placebo-controlled, parallel-group study of two doses of oral iloprost in 103 patients with Raynaud’s phenomenon secondary to systemic sclerosis. METHODS Patients Patients with Raynaud’s phenomenon secondary to systemic sclerosis were recruited from out-patient clinics at seven university hospitals in Denmark (two centres), Great Britain (three centres) and The Netherlands (two centres). Raynaud’s phenomenon was diagnosed on the basis of a history of episodic digital pallor and cyanosis. Systemic sclerosis was diagnosed according to the criteria of the American College of Rheumatology [22] with the addition that abnormal nailfold capillaries were permitted as one of the signs. In order to be randomized into the study, patients had to have an average of at least six Raynaud’s attacks per week in the 2 week pretreatment assessment period, documented by patient diary. Males and females were included and all patients were at least 18 yr old. Women of child-bearing age were required to practise a medically acceptable method of birth control, and to have a negative pregnancy test at the beginning and end of the pretreatment assessment period. Patients were excluded from participation if they fulfilled any of the following criteria: treatment with any prostanoid for vascular disease in the previous 2 months; surgical sympathectomy performed in the previous 12 months; bleeding diathesis or a platelet disorder; stroke, transient ischaemic attack or myocardial infarction within the past 6 months; or onset of renal insufficiency in the previous 4 weeks. Administration of other drugs for the treatment of Raynaud’s phenomenon was not allowed for the duration of the study. Drugs for the treatment of the underlying systemic sclerosis were not initiated or terminated during the study period. Vasodilators for the treatment of hypertension were maintained at a constant dose throughout the study. The study protocol was approved by the appropriate ethics committees for each trial centre and all patients provided written informed consent to their participation in the study. Treatments Iloprost was supplied as an extended-release formulation consisting of hard gelatin capsules containing 50 mg iloprost in the form of iloprost-b-cyclodextrin clathrate. Placebo capsules were identical in appearance and contained the same excipients. Patients took one of three treatments: placebo, oral iloprost 50 mg or oral iloprost 100 mg twice daily for 6 weeks, giving total daily doses of 0, 100 and 200 mg iloprost, respectively, in the three treatment groups. Capsules were taken twice daily with food with a 5 h interval between doses. The placebo group took two placebo capsules at breakfast and two more at lunch-

953

time. The 50 mg group took one 50 mg iloprost capsule and one placebo capsule twice daily. The dose in the oral iloprost 100 mg group was titrated so that patients took one 50 mg iloprost capsule and one placebo capsule twice daily for the first week, and the full dose of two 50 mg iloprost capsules twice daily for the remaining 5 weeks. Thus, all patients took two capsules twice daily for 6 weeks, maintaining the double-blind nature of the study. Allocation to a treatment group was carried out by assigning to the patient the next available randomization number in the sequence given to the centre. Each randomization number corresponded to a numbered set of study medication. The sequence of treatments in the randomization list was determined by a computer-generated random sequence in balanced blocks of four. Each centre received separate sequences of randomization numbers for patients with ulcers and patients without ulcers in order to achieve a balanced randomization stratified by centre and presence of digital cutaneous lesions. Study design The study consisted of three periods: a 2 week pretreatment period, a 6 week treatment period and a 6 week post-treatment follow-up. Randomization to one of three treatments occurred on satisfactory completion of the pre-treatment period. The treatment and followup periods were double blind. Patients visited the clinics 1, 2, 4, 6, 8 and 12 weeks after the start of treatment. Patients were assessed for the full 12 weeks after the start of treatment irrespective of whether the treatment was taken for the full 6 weeks or not. All patients were recruited during the months of January and February. Clinical outcome variables Patients were given specially designed diaries to be completed daily throughout the study even if the treatment was discontinued prematurely ( Fig. 1). Diaries were provided in the native language of the patient: Danish, Dutch or English, as appropriate. Each day, patients recorded the number of Raynaud’s attacks, the duration of each attack and an assessment of the overall severity of their condition on a scale of 0–10, the Raynaud’s Condition Score ( RCS). An attack was defined as an episode of pallor followed by cyanosis with or without associated pain. The mean daily number of Raynaud’s attacks for each patient was calculated for each 2 week interval throughout the study. The mean total daily duration of attacks for each 2 week interval was calculated for each patient by adding the individual attack durations for the interval and dividing by the number of diary days. The mean of the daily RCS was also calculated for each 2 week interval. Mean weekly outdoor temperatures were determined for each patient from data provided by the local meteorological centres for each participating centre. Mean changes were calculated for each patient group. The physician’s global assessment of the patient’s

954

BRITISH JOURNAL OF RHEUMATOLOGY VOL. 37 NO. 9

F. 1.—Specimen treatment diary. This diary record was completed by all subjects to assess the severity and frequency of their Raynaud’s symptoms, as outlined in the text.

condition compared to pre-treatment was recorded 6 weeks after the start of treatment. Each patient was classified as being worse, unchanged or improved compared to baseline. Digital cutaneous lesions were recorded before the start of treatment and at 6 and 12 weeks after the start of treatment. The total number of lesions on the hands was recorded, and the number of each type classified as ulcers, fissures and paronychiae. Tolerability and safety Patients were asked at each visit to the clinic to report any adverse events occurring in the interval since the previous visit. Systolic and diastolic blood pressure and heart rate were measured at each visit while the patient was in a supine position. Laboratory parameters were investigated before the start of treatment and at the end of treatment. Haematology parameters measured were haematocrit, haemoglobin, erythrocyte count, leucocyte count and platelet count. Serum chemistry consisted of sodium, potassium, creatinine, alanine transaminase (ALT ) and aspartate transaminase (AST ). Urinalysis consisted of dipstick analysis for protein, glucose and blood. Statistical analysis The results presented derive from an observed case analysis in which all patients randomized and all data recorded were included, but no attempt was made to impute values where data were missing. However, a secondary analysis of the diary parameters (total daily duration, frequency and severity of Raynaud’s attacks) was also performed using the last observation carried

forward (LOCF ) principle in order to investigate possible bias due to missing information. Treatment differences in the efficacy parameters total daily duration, frequency and severity of Raynaud’s attacks were determined by use of a two-way analysis of variance on the percentage change from baseline. The two-way model incorporated treatment, baseline lesion status and treatment × baseline lesion status as an interaction term. Ordinal variables, such as the physician’s global assessment of efficacy, were tested for treatment differences using the Mantel–Haenszel x2 test. Differences in frequencies of adverse events were tested by the Mantel–Haenszel test for trend in proportions across the increasing doses. All tests were two tailed and run with a = 0.05. RESULTS Of the 103 patients enrolled in the study, 35 were randomized to placebo, 33 to iloprost 50 mg and 35 to iloprost 100 mg. Eighty-nine patients (86%) completed diaries for the whole 6 week treatment period and 82 patients (80%) for the 12 weeks of treatment plus follow-up. Twenty-one patients failed to complete all of the assessments up to 12 weeks: five on placebo, seven on iloprost 50 mg and nine on iloprost 100 mg. Study medication was discontinued prematurely in 31 patients: placebo, three patients (9%); iloprost 50 mg, 10 patients (30%); iloprost 100 mg, 18 patients (51%). Of the patients who discontinued treatment early, 58% discontinued in the first 2 weeks: placebo, three patients; iloprost 50 mg, seven patients; iloprost 100 mg, eight patients. All but two of the treatment discontinuations were due to adverse events.


The demographic characteristics of the patients were similar in the three treatment groups ( Table I ). Seventy-seven patients (75%) were classified as having systemic sclerosis with limited skin involvement and 24 patients (22%) with diffuse skin involvement. Two patients were unclassifiable in this respect. The distribution of the systemic sclerosis types was comparable in the three treatment groups. Twenty-eight patients had some kind of digital cutaneous lesions at the start of treatment, 22 of them had ulcers. These were also evenly distributed between the three treatment groups. The prevalence of gastrointestinal and pulmonary involvement was similar in each group. Over 70% of the patients in each group had received other medication for systemic sclerosis or Raynaud’s phenomenon in the previous 12 months. None had received prostanoids in the previous 2 months and only three patients has received prostanoids in the previous 3 months (iloprost 50 mg, two patients; iloprost 100 mg, one patient). Nine patients distributed across the three treatment groups received vasoactive treatment for hypertension during the study, but doses were kept constant throughout the pre-treatment and followup periods. Total daily duration of Raynaud’s attacks During the pre-treatment period, the total duration of attacks was lowest in the placebo group and highest with oral iloprost 100 mg ( Table II ), but the difference was not significant. The two iloprost groups showed similar reductions in the total duration at all time points ( Fig. 2) which were markedly greater than those in the placebo group at weeks 5–6 (P = 0.03) and weeks 11–12 (P = 0.001) ( Table II ). Frequency of Raynaud’s attacks Differences in the frequency of attacks between the treatment groups at baseline were not significant. The variability of the frequency of attacks was very high and, despite greater mean reductions in the two iloprost groups, differences between the groups were not significant at weeks 5–6 (P = 0.37) and showed a trend at weeks 11–12 (P = 0.07) ( Fig. 3 and Table III ).

TABLE I Baseline characteristics of patients

Variable Median age (range) (yr) Females Current smokers Median body mass index (range) Digital skin lesions Limited SSc Diffuse SSc SSc, systemic sclerosis.

Placebo (n = 35)

Iloprost 50 mg (n = 33)

Iloprost 100 mg (n = 35)

47 (24–73) 28 (80%) 8 (23%)

49 (31–75) 22 (67%) 10 (30%)

47 (23–72) 32 (91%) 8 (23%)

23 9 26 9

24 (17–32) 10 (30%) 27 (82%) 6 (18%)

24 (18–31) 9 (26%) 25 (71%) 8 (23%)

(17–28) (26%) (74%) (26%)

955

Raynaud’s condition score The RCS was very similar in all three groups before treatment. During and after treatment, reductions in the RCS appeared to show a dose relationship, with the smallest reductions with placebo and the greatest with iloprost 100 mg (Fig. 4). There was a strong statistical trend towards a difference between the groups at weeks 5–6 (P = 0.07) and a significant difference at weeks 11–12 (P = 0.007) (Table IV ). Statistical power and the effect of missing data The statistical power of the observed case analyses of duration, frequency and RCS, calculated on the data obtained in the study, was 65, 30 and 72%, respectively, at weeks 5–6, and 97, 76 and 96%, respectively, at weeks 11–12. Analysis of the data from patient diaries was repeated using the LOCF principle in the case of missing data at the end of the planned treatment period. Using this method, the mean percentage changes in the total duration of attacks at weeks 5–6 were placebo 11%, iloprost 50 mg −40% and iloprost 100 mg −39%, compared to mean percentage changes of 10, −40 and −35%, respectively, in the observed case analysis. Other analyses also produced essentially similar results to those already described, suggesting that no significant bias resulted from the failure of some patients to complete the diaries. Physician’s global assessment Improvement in the patient’s condition tended to be more likely with increasing dose of oral iloprost at the end of the treatment period, placebo 44%, iloprost 50 mg 57%, and iloprost 100 mg 64%, but betweengroup differences were not significant. By the end of the follow-up period, at least 50% of the patients were considered by the physician to have improved in each treatment group, including placebo, and there were no significant between-group differences. Digital cutaneous lesions Twenty-eight patients had digital cutaneous lesions on entry into the study (nine placebo, 10 iloprost 50 mg and nine iloprost 100 mg). Twenty-two of these patients had ulcers. The numbers of patients in whom ulcers healed for placebo, iloprost 50 mg and iloprost 100 mg were 1/7, 1/8 and 3/7, respectively. The numbers of patients in each group with ulcers were thus too small to draw any conclusions on an effect of treatment on this parameter. No association of the presence of lesions with the effect of treatment on Raynaud’s attacks was identified. Adverse events At least one adverse event was reported by 28 of the 35 patients receiving placebo (80%), 28 of the 33 patients receiving iloprost 50 mg (85%) and 34 of the 35 patients receiving iloprost 100 mg (97%) (P = 0.08). Significant differences were seen in the frequency of five types of adverse events. Headache, flushing, nausea and trismus were all more common with increasing

956

BRITISH JOURNAL OF RHEUMATOLOGY VOL. 37 NO. 9

TABLE II Total daily duration of Raynaud’s attacks expressed as mean and standard deviations. The P value represents comparison of all three groups by ANOVA Placebo Variable Pre-treatment Mean value Weeks 5 and 6 Mean value Change from baseline % change from baseline 95% CI for % change Weeks 11 and 12 Mean value Change from baseline % change from baseline 95% CI for % change

Iloprost 50 mg

Iloprost 100 mg

n

Duration (min)

n

Duration (min)

n

Duration (min)

35

69 (73)

33

83 (83)

35

95 (89)

33 33 33

81 9 10 53

(113) (72) (125) to −33

28 28 28

46 −32 −40 −25

(43) (63) (41) to −55

28 28 28

51 −34 −35 −19

(57) (38) (44) to −51

30 30 30

70 2 −9 19

(101) (52) (79) to −37

26 26 26

32 −46 −60 −50

(32) (66) (27) to −70

26 26 26

27 −49 −60 −49

(32) (54) (29) to −71

F. 2.—Effect of oral iloprost on total Raynaud’s duration. Total Raynaud’s duration was calculated as outlined in the text. The effects of placebo and two doses of oral iloprost on this variable are shown (mean ± ...).

iloprost dose, while flu syndrome was most commonly reported in the placebo group ( Table V ). Assessment of vital signs before and during treatment showed no clinically significant effects of either dose of iloprost on blood pressure or heart rate. Premature discontinuation due to adverse events appeared to be dose related: placebo, two patients (6%); iloprost 50 mg, nine patients (27%); iloprost 100 mg, 18 patients (51%) (P < 0.001). There were no deaths during the study, but serious adverse events were reported in two patients. These were one patient on placebo who experienced an episode of faintness and palpitations due to atrial fibrillation, and one patient on oral iloprost 100 mg who experienced shortness of breath, lethargy, loss of weight and cough due to a pulmonary infection which was treated with a course of antibiotics. Neither of these were likely to have been caused by study treat-

P

0.03

0.001

F. 3.—Effect of oral iloprost on Raynaud’s attack frequency. Raynaud’s attack frequency was calculated as outlined in the text. The effects of placebo and two doses of oral iloprost on this variable are shown (mean ± ...).

ment and it was not felt necessary to open the treatment codes prematurely. Laboratory tests No trends were seen in any of the laboratory parameters and no individual changes during the study were attributed to the study medication. Outdoor temperatures The minimum and maximum outdoor temperatures experienced by the patients in each treatment group were comparable during the pre-treatment period. The changes in minimum and maximum daily temperatures recorded from baseline to the end of the 6 week treatment (mean +1°C and +4°C, respectively) and from baseline to the end of 12 weeks follow-up (mean


957

TABLE III Frequency of Raynaud’s attacks expressed as mean and standard deviation. The P value represents comparison of all three groups by ANOVA Placebo Variable Pre-treatment Mean value Weeks 5 and 6 Mean value Change from baseline % change from baseline 95% CI for % change Weeks 11 and 12 Mean value Change from baseline % change from baseline 95% CI for % change

Iloprost 50 mg

Iloprost 100 mg

n

Frequency (per day)

n

Frequency (per day)

n

Frequency (per day)

35

3.3 (2.5)

33

3.6 (1.7)

35

4.1 (2.8)

33 33 33

3.0 (3.4) −0.3 (1.7) −13 (58) 7 to −33

28 28 28

2.5 (1.8) −1.0 (1.0) −31 (36) −18 to −44

28 28 28

2.9 (2.5) −1.5 (1.7) −34 (40) −19 to −49

30 30 30

3.0 (3.5) −0.3 (1.9) −15 (66) −9 to −39

26 26 26

2.1 (1.6) −1.5 (0.9) −46 (29) −35 to −57

26 26 26

2.3 (2.3) −1.8 (1.4) −50 (32) −38 to −62

P

0.37

0.07

+5°C and +8°C, respectively) were almost identical (±