DIAGN MICROBIOLINFECTDIS 1992;15:111S-117S
111S
Clinical Evaluation of Roxithromycin 300 mg Once Daily as an Alternative to 150 mg Twice Daily Jean-Claude
Pech~re
A multicenter, double-blind, randomized, two-arm study was conducted to compare two dosage regimens of roxithromycin-150 mg b.i.d, and 300 mg once daily--in 1588 patients with infectious disease, 1573 of whom were evaluable for safety and 1391 for efficacy. There were no significant differences in demographic or disease characteristics at presentation. Treatment duration averaged 10 days in both groups. Clinical response rates (% satisfactory response) were similar in both arms (b.i.d. versus once daily, respectively): upper respiratory tract,
95.3 and 95.0; lower respiratory tract, 93.1 and 93.6; and skin and soft tissue, 93.1 and 93.0. Bacteriologic response was also similar. The overall incidences of adverse events were 8.5% for daily dosing and 7.2% for b.i.d, dosing, but there was a slightly greater frequency of gastrointestinal events (p < 0.03) was observed with the once-daily regimen. Thus roxithromycin 300 mg once daily can be recommended as an effective and well-tolerated alternative to the conventional b.i.d. regimen.
INTRODUCTION
fort, the present study was designed to evaluate oncedaily dosing with 300 mg roxithromycin in comparison with the proven efficacy of the same dosage given in two divided doses of 150 mg. Consideration was given to biostatistical aspects, particularly to avoid [~-type error, which is so important for this kind of comparative study. In addition, to ensure extrapolation of the results to a large patient population, the study was carried out in 12 cotmtries on three continents.
Roxithromycin is an acid-stable, orally administered macrolide, structurally related to erythromycin. In vitro, its antibacterial profile is similar to that of erythromycin (Pech~re and Auckenthaler, 1987; Barlam and Neu, 1984; Ridgway, 1987), with a few exceptions [for example, roxithromycin seems more active than erythromycin against Toxoplasma gondii (Chang and Pech~re, 1987)]. Roxithromycin has a distinctive pharmacokinetic profile characterized by excellent enteral absorption, prolonged serum h a l f life (~12 hr), and high concentrations in most tissues and body fluids (Periti et al., 1989; Purl and Lassman, 1987; Ridgway, 1987). This profile could in theory be compatible with once-daily dosing in infections for which roxithromycin is indicated. On the basis of these pharmacokinetic characteristics, and to enhance compliance and patient cornFrom the Facultyof Medicine, Department of Microbiology, CMU Geneva, Switz~*rland. Additional copies of this supplement are availablefrom Roussel UCLAF, DomaineTh~rapeutiqueAntibioth(~rapie,35 Boulevard des Invalides, 75007Paris, France. Received 8 October 1991;revised and accepted 20 December 1991. © 1992ElsevierSciencePublishingCo., Inc. 655 Avenue of the Americas, New York, NY 10010 0732-8893192/$5.00
MATERIALS AND METHODS Study Design This was a multicenter, double-blind, randomized, two-arm study. The target was to enroll 1600 patients to ensure that at least 1240 patients, that is, 620 patients per group, would be evaluable for clinical efficacy. This sample size enables the detection of a 5% difference between the two dosage regimens, with an ot = 0.05 and [3 = 0.01 (90% power), assuming a success rate on both regimens of 90%. Randomization was performed by a random-number generation, with a separate randomization list for each country. Randomization was one to one in blocks of four.
112S
Treatment Roxithromycin was supplied as white film-coated tablets, each containing 300 mg or 150 mg of active drug. A double-dummy technique was used in order to ensure the double-blind design: two tablets had to be taken every morning and one table every evening, an hour before meals with a glass of water. This allowed administration of either 300 mg once daffy or 150 mg twice daily. Five days were considered as the minimal treatment duration for a patient to be evaluable for clinical efficacy, except in the case of clinical failure (3 days). The treatment could be continued as long as necessary according to the investigator's opinion, but was not to exceed 20 days. It could be discontinued at any time if (a) the pathogen was shown to be resistant to roxithromycin in vitro; (b) there was no clinical response, despite susceptibility of the isolate to roxithromycin in vitro; or (c) a severe adverse effect was noted that was thought to be related to roxithromycin. Antimicrobial therapy other than roxithromycin was prohibited, as were ergot derivatives, cyclosporin, and any drug affecting the absorption of roxithromycin.
Patients, Surveillance, and Criteria Eligible patients were men and women 18 years of age or older, with an infectious disease caused by susceptible pathogens that usually respond to oral macrolide therapy, including upper and lower respiratory tract infection, and skin and soft tissue infections. Patients were entered before bacteriologic results were available and their evaluability subsequently determined. The following patients were excluded: pregnant or nursing women; patients with a history of hypersensitivity to macrolides; patients with a known severe hepatic or renal insufficiency; patients receiving any investigational drug within the 2 previous weeks; patients receiving any other antibiotic within 3 days prior to entering the study, unless the microorganism responsible for the infection was shown to be resistant to the antibiotic previously used; patients receiving other concurrent antibiotic therapy; patients with abnormal findings that could affect interpretation of the results; and patients unable to comply with the study protocol. A full clinical examination was performed at the beginning, middle, and end of the study. Chest radiography was required for patients with lower respiratory tract infections, within 48 hr of enrollment and, in principle, 48 hr after completion of treatment. Standard hematology and biochemistry tests were performed within 48 hr prior to initiating therapy, at midtreat-
J.-C. Pech~re
ment, and within 48 hr of completing therapy. Specimens of the infected site material were cultured and tested for sensitivity. A minimum of two cultures had to be performed, one prior to treatment and another within 48 hr of the end of treatment, whenever technically possible. Minimum inhibitory concentration (MIC) break points were 1 and 4 p~g/ml (Acar et al., 1987) for susceptible and resistant, respectively. Adverse events were categorized, according to the investigator's opinion of their severity, as mild, moderate, or severe. Their relationship to the study drug was assessed by the investigator as nil, unlikely, possible, probable, clinically nonsignificant, or insufficiently documented. A three-scale assessment of clinical response was used: "satisfactory" = clinical improvement or resolution; "unsatisfactory" = deterioration or no change; and "unevaluable." Bacteriologic response was assessed as "satisfactory" = susceptible pathogen eradicated and/or new isolate without clinical symptoms, or clinical cure, making follow-up culture impossible; "unsatisfactory" = persistence of susceptible pathogen with persistence or reappearance of clinical signs, or partial eradication in the event of multiple susceptible pathogens; and "nonevaluable." Statistical Methods We used Student's t-test for continuous variables, the chi-square test for categorical variables (using Fisher's exact test for 2 x 2 tables or the likehl~ood ratio chi-square for larger tables whenever suitable), and Wilcoxon's nonparametric test for ordinal parameters. Overall clinical and bacteriologic response rates were summarized by treatment groups and compared using Mantel-Haenszel statistics with respect to diagnosis. RESULTS A total of 1588 patients from 62 centers and 12 countries entered the study: 793 were randomized to 300mg once-daily group and 795 to 150 mg b.i.d. Fifteen patients (nine in the once-daily group and six in the b.i.d, group) were included, but did not receive any treatment and were lost to follow-up just after inclusion. In the remaining 1573 patients, demographic parameters and base-line data were similar in the two dosage groups (Table 1). In addition, we found no significant difference between the two treatment groups in symptom duration, prior antibiotic therapy, underlying disease, body temperature, blood pressure and pulse, or the findings of physical examination and chest radiography (for lower respiratory tract infection).
113S
Roxithromycin 300 mg Daily vs 150 mg b.i.d.
TABLE 1 Demographic and Disease Characteristics at Presentation Patient Group
Men Outpatients (%) Inpatients (%) Race Caucasian Black Oriental Other Age (years) (m - SD) Weight (kg) (m -+ SD) Height (cm)(m -+ SD) Overall status (%) Asymptomatic Fair Serious Critical History of infection Acute Chronic Recurrent Diagnosis of infection Upper respiratory tract Lower respiratory tract Skin and soft tissue
~ 0 mg Once Daily (n = 784)
150 mg Twice Daily (n = 789)
52.0 60.5 39.4
51.1 59.8 40.2
87.5 2.9 0.1 9.4 46.1 ± 19.5 67.8 -- 13.9 168.1 -+ 9.6
86.9 3 0 10.0 45.8 -+ 19.6 67.6 -+ 13.8 168.1 -+ 9.5
14.8 71.6 13.1 0.5
15.3 68.3 16 2 0
86.0 3.6 10.5
87.8 2.7 9.4
37.5 54.2 8.3
39.5 52.0 8.5
p > 0.05 for all comparisons.
Patients were excluded from the analysis of clinical efficacy (n = 182) either because a resistant (or intermediate) pathogen was identified on pretreatment bacteriology (86 cases) or because of loss to follow-up (45 patients). A further 17 patients were excluded for protocol violations and 14 for diagnostic errors including active tuberculosis. A total of 872 patients were excluded from the analysis of bacteriologic efficacy either because no causative pathogen was identified in the pretreatment bacteriologic specimen (637 cases), susceptibility was not tested on the isolate (130 cases), or no bacteriology was performed at all (126 cases). Treatment duration was 9.8-+ 2.9 days in the oncedaily group and 9.9+2.9 days in the b.i.d, group. During the trial, 754 patients, 377 per group, reported one or more concomitant or intercurrent conditions and received concomitant treatments allowed by the protocol. Statistical analysis of clinical efficacy in 1391 patients showed no significant difference between regimens either in the total population or per diagnosis (see Table 2). Success rates differed by 0.1% between treatment groups (95% conference interval, - 2 . 4 %
to + 2.6%). The Mantel-Haenszel probability, stratified by diagnosis, was 0.90. The clinical efficacy in patients over 65 years of age, w h o accounted for 20.7% of the total population, was separately analyzed a posteriori. Overall clinical efficacy was 88.9% a n d 90.3% with oncedaily versus b.i.d, dosing, respectively (hiS), compared with 95.5% and 94.9% in younger patients. Statistical analysis in 519 patients showed no significant difference between regimens in bacteriologic efficacy either in the total population or per diagnosis (Table 3). Of these 519 patients, clinical cure was achieved in 128, making posttreatment bacteriology impossible. The difference between success rates in the two groups was 1.9% (95% confidence interval, - 1 . 5 % to 5.2%). The Mantel-Haenszel probability, stratified by diagnosis, was 0.345. Eradication rates per pathogen are shown in Table 4. As allowed by the protocol, treatment was started before receiving the bacteriologic results. However, investigators were allowed to continue treatment if the pa'tient was improving clinically. Pretreatment bacteriology identified 69 resistant isolates. This was followed up after treatment in 26 cases: 16 isolates
114S
J.-C. Pech~re
TABLE 2 Clinical Response per Diagnosis in 1391 Patients Treated with Roxithromycin 300 mg Once Daily or 150 mg Twice Daily Satisfactory Clinical Response (%) by GroulY' 300 mg Once Daily (n = 694)
150 mg Twice Daily (n = 697)
Pharyngitis (84) Tonsillitis (139) Sinusitis (262; Otitis (39) Pneumonia (344) Acute bi'ollChitis(222) Acute episode of chronic bronchitis {134) Other respiratory infections (52) Skin and soft tissue infections (115)
95.7 98.4 93.4 92.3 93.3 92.1 97.1
100 98.7 92.9 92.3 89.8 98.1 90.9
95.0
96.9
93.0
93.1
Total patients (1391)
94.1
94.0
Diagnosis (No. of Patients)
"No significantdifferencebetween the two groups. (61.5%) were found to have been eradicated, and clinical cure precluded bacteriology in six cases. An unsatisfactory clinical and/or bacteriologic response was seen in 93 (6.6%) of 1391 patients, 47 (6.7%) in the once-daily group and 46 (6.5%) in the b.i.d, group. Superinfection occurred in 19 patients (six and 13 cases in the orlce-daiiy and b.i.d, groups, respectively), associated in nine cases with persistent clinical infection. The pathogens associated with superinfection were streptococci (including one
Streptococcus pneumoniae), Staphylococcus aureus, Moraxella catarrhalis, Enterobacter aerogenes, and Candida albicans. The incidence of adverse events classified as being "unlikely," "possibly," or "probably" related to roxithromycin therapy is presented in Table 5. Gastrointestinal events were commoner in the once-daily group (p = 0.03), particularly nausea and/or vomiting and abdominal pain. Most gastrointestinal upsets were mild or moderate. Only four events in the oncedaily group and one in the b.i.d, group were assessed as severe by the investigator. Neurologic adverse events, mostly dizziness, vertigo, or headache, were mild or moderate, and thought unlikely to be due to the test drug in 12 of 17 cases. Skin reactions consisted of rash, pruritis, and urticaria, and were possibly due to roxithromycin in eight of 13 cases. Liver function changes consisted of transient enzyme elevation in six of seven cases, and one case of cholecystitis, which was thought unlikely to be due to roxithromycin. Also described were 19 "other" events, of which 18 were thought unlikely to be due to roxithromycin. Ten patients in the once-daily group and seven in the
b.i.d, group were withdrawn due to an adverse event possibly related to the test drug: 12 cases of gastrointestinal upset and five skin reactions. The figures for the events both possibly related to the study drug and defined as severe by the investigators were five in the once-daily group and one in the b.i.d. group. Thirteen subjects (six and seven per group, respectively) developed clinically significant laboratory changes possibly caused by the study drug, including decreases in hemoglobin or platelet count, serum creatinine, SGPT, and/or SGOT. All were relatively mild and transient.
DISCUSSION The ability to offer treatment on once-a-day basis is an appreciable benefit in terms of patient comfort, especially for the ambulatory patients who make up the majority of those likely to be treated with macrolides. Compliance is also likely to be enhanced by once-daily rather than more frequent dosing. The sample size of the present study was designed to be able to detect a 5% difference between the two dosage regimens. The effective power of the study, computed on the basis of the number of evaluable patients, was >99% for the analysis of clinical response and -92% for the analysis of bacteriologic response. The randomization list was adhered to, as evidenced by final results, and no difference was observed between the two groups in base-line variables or treatment duration. No statistically significant difference between the regimens was found in clinical or bacteriologic efficacy, and the overall incidence of sat-
Roxithromycin 300 mg Daily vs 150 mg b.i.d.
TABLE 3
115S
Bacteriologic Efficacy per Diagnosis in 519 Patients Treated with Roxithromycin 300 mg Once Daily or 150 mg b.i.d. Satisfactory Bacteriologic Response (%) by Group~
Diagnosis (No. of Patients)
300 mg Once Daily (n = 257)
150 mg Twice Daily (n = 262)
Pharyngitis (31) Tonsillitis (76) Sinusitis (34) Otitis (19) Pneumonia (113) Acute bronchitis (91) Acute episode of chronic bronchitis (65) Other respiratory infections (15) Skin and soft tissue infections (75)
92.9 89.2 100 100 98.3 95.8 100
88.2 89.7 100 93.3 92.6 100 100
100
100
Total of patients (519)
100
94.2
96.9
95.0
aNo significantdifferencebetween the twe groups.
TABLE 4
Bacteriologic Efficacy per Pathogen in 403 Isolates Percent Eradicated in Groul~ 300 mg Once Daily
150 mg Twice Daily
Other (6)
90.7 98.7 100 94.4 100 100
91.8 97.0 95 91.7 100 100
Total (403)
96.6
94.9
Pathogen (No. of Strain) Streptococcus pyogenes (!03) Streptococcus pneumoniae (146) Staphylococcus spp. (106)b Haemophilus influenzae (30) MoraxeUa catarralis (12)
aNo significantdifferencebetween the two groups. bS. aureus = 105 and S. epidermidis = 1. isfactory responses was high, confirming that roxithromycin is active against the infections selected in this study (Lachat et al., 1986; Gentry, 1987; Champetier de Ribes et al., 1989; Young et al., 1989). This success is due to its antibacterial spectrum, combined with the high concentrations achieved in tissue and body fluids, for example, tonsil, middleear exudate, and bronchial secretions (Periti et al., 1989). In particular, roxithromycin concentrations in bronchoalveolar lavage cells are two and ten times higher than concomitant levels in plasma and epithelial fluid, respectively (Chastre et al., 1987). In individual infections, both regimens produced very similar results in tonsillitis, sinusitis, otitis, and skin and soft tissue infections. The once-daily dose
gave (nonsignificantly) better results in pneumonia (93.3% vs 89.9%) and exacerbations of chronic bronchitis (97.1% vs 90.9%), suggesting that this r e d , h e n could provide improved pharmacokinetic conditions in lung tissue and bronchial secretion. Conversely, the twice-daily regimen was associated with a higher rate of satisfactory clinical response (nonsignificant) in pharyngitis (100% vs 95.7%) and acute bronchitis (98.1% vs 92.1%). In some of the diseases treated in the present study, however, particularly pharyngitis/tonsillitis and acute bronchitis, the efficacy of cure was difficult to evaluate because of the propensity for spontaneous cure and the frequent viral involvement. It was therefore important to consider bacteriologic ef-
1165
J.-C. Pech6re
TABLE 5 Incidence of Reported Adverse Events on Therapy with Roxithromycin 300 mg Once Daily or 150 mg b.i.d. Treatment Group 300 mg Once Daily (n = 784)
150 mg Twice Daily (n = 789)
Gastrointestinal Nausea Vo~mting Abdominal pian Diarrhea/loose stools Flatulence Other Neumlogic Skin reaction Liver Other
54 (7.0%) 19 7 14 11 2 1 10 (1.3%) 5 (0.6%) 4 (0.5%) 7 (0.9%)
35 (4.4%) 12 2 9 10 1 1 7 (0.9%) 7 (0.9%) 3 (0.4%) 12 (1.5%)
0.46 0.57 0.69 0.25
Total of patients with one or more adverse effects
67 (8.5%)
57 (7.2%)
0.33
Adverse Event
0.03a
"Significantdifference,p < 0.05. ficacy. High activity against streptococci was found, as expected from previous reports (Herron, 1987) and low MICs (Pech6re and Auckenthaler, 1987), but the activity of roxithromycin against staphylococci and Haemophilus influenzae, where MICs are higher (Pech6re and Auckenthaler, 1987), deserves spedal mention. Anti-Haemophilus activity by roxithromycin in pneumonia, despite often borderline in vitro potency, agrees with previous fndings (Brfickner et al., 1989) and is probably the result of the drug's favorable pharmacokinetics (Nilsen et al., 1992). The study comprised 61 cases of atypical pneumonia, a critical indication. Clinical cure was achieved in 58 cases. In three patients, including two documented mycoplasma infections, the clinical response was unsatisfactory. All three werc in the twice-daily group. The favorable overall results in atypical pneumonia confirmed those obtained elsewhere (Fernandez-MacLoughli et al., 1988; Breux et al., 1989).
Skin and soft tissue infections consisted mainly of erysipelas and, as expected, most of the isolates were S. aureus and streptococci, evenly distributed in the two dosage groups. The clinical and bacteriologic responses were excellent, with notably a 100% bacteriologic cure rate in the once-daily group. The adverse events potentially due to roxithromycin were generally those expected for this type of drug. Overall, the incidence of side effects was similar in both groups, but with more gastrointestinal events on the once-daily group. According to pharmacokinetic studies, this difference is not due to drug accumulation in tissues on once-daily dosing (Nilsen et al., 1992). On macrolide therapy, gastrointestinal symptoms are due to local irritation, probably by direct action on smooth muscle. Thus, a higher incidence of gastrointestinal symptoms could be expected in the group receiving the higher unit dose (Keller and Follath, 1988).
REFERENCES Acar J, Bergogne-Berezin E, Chabbert Y, Ouzel R, Courtieu A, Courvalin P (1987) Comminqu~s 1987du Comit~ de I'Antlbiogramme de la Soci~t(~Fran(~aise de Microbioiogie. Pathol Biol 35:1101-1004. Barlam T, Neu NC (1984) In-vitro comparison of the activity of RU 28965, a new macrolide, with that of erythromycin against aerobic and anaerobic bacteria. Antimicrob Agents Chemot~r 25:529-531. Breux JP, Fort E, Agius G, Grollier G, Cazenave-Roblot F, Casters M (1989) Empirical therapy for atypical pneu-
monia with roxithromycin. In Abstracts, 16th International Congress of Chemotherapy, Jerusalem, Israel. International Society of Chemotherapy. Briickner O, Trautmann M, Kemmerich B (1989) Roxithromydn compared to a new cephalosporin, cefixime, and dprofloxadn in purulent chest disease and pneumonia. In Abstracts, 16th International Congress of Chemotherapy, Jerusalem, Israel. International Society of Chemotherapy. Champetier de Ribes D, Jockey C, Gaudouen G (1989)
Roxithromycin 300 m g Daily vs 150 ,rig b.i.d.
Etude de refficacit~ et de la tolerance de la roxithromycine dans les infections broncho-pulmonaires. Sem Hop Paris 65:2819-2821. Chang HR Pech~re JC (1987) Effect of roxithromycin on acute toxoplasmosis in mice. Antimicrob Agents Chemother 31:1147-1149. Chastre J, Brun B, Fourtillan JB, Soler P, Basset G, Manuel C, Trouillet JL, Gibert C (1987) Pulmonary disposition of roxithromycin (RU 28965), a new macrolide antibiotic. Antimicrob Agents Chemother 31:1321-1316. Fernandez-MacLoughli GF, Lanoel JL, Staraboulian D, Sarachian B (1988) Roxithromycin in the treatment of atypical pneumonia. Br J Clin Pract 42(Supp155):92-93. Gentry LO (1987) Roxithromycin, a new macrolide antibiotic, in the treatment of infections in the lower respiratory tract: an overview. J Antimicrob Chemother 20(Suppi B):145-152. Herron JM (1987) Roxithromycin in the therapy of Streptococcus pyogenes throat infections. J Antimicrob Chemother 20(Suppl B):139-144. Keller H, Follath F (1988) Meyler's Side-Effects of Drugs, l l t h ed. Ed, MNG Dukes. Elsevier: Amsterdam, pp 543566.
117S
Lachat JM, Ane× JF, Regamey C (1986) Ro×ithromycin, a new macrolide for pulmonary infections. Schweiz Med Wochenschr 116:1739-1741. Nilsen OG, Aamo T, Zahlsen K, Svarva P (1992) Macrolide pharmacokinetics and dose scheduling of roxithromycin. Diagn Microbiol Inf Dis 15(Suppl):71S--76S [this supplement]. Pech~re JC, Auckenthaler R (1987) In-vitro activity of ro×ithromycin against respiratory and skin pathogens. J Antimicrob Chemother 20(Suppl B):l-5. Periti P, Mazzei T, Mini E, Nove|ii A (1989) Clinical pharmacokinetic properties of the macrolide antibiotics. Clin Pharmacokinet 16:261-282. Purl SK, Lassman HB (1987) Roxithromycin: a pharmacokinetic review of a macrolide. J Antimicrob Chemother 20(Suppl B):89-110. Ridgway GL (1987) A review of the in-vitro activity of roxithromycin against genital pathogens. J Antimicrob Chemother 20(Suppl B):7-11. Young RA, Gonzales JP, Sorkin EM (1989) Roxithromycin: a review of its antibacterial activity, pharmacokinetic properties and clinical efficacy. Drugs 37:8-41.
111S
Clinical Evaluation of Roxithromycin 300 mg Once Daily as an Alternative to 150 mg Twice Daily Jean-Claude
Pech~re
A multicenter, double-blind, randomized, two-arm study was conducted to compare two dosage regimens of roxithromycin-150 mg b.i.d, and 300 mg once daily--in 1588 patients with infectious disease, 1573 of whom were evaluable for safety and 1391 for efficacy. There were no significant differences in demographic or disease characteristics at presentation. Treatment duration averaged 10 days in both groups. Clinical response rates (% satisfactory response) were similar in both arms (b.i.d. versus once daily, respectively): upper respiratory tract,
95.3 and 95.0; lower respiratory tract, 93.1 and 93.6; and skin and soft tissue, 93.1 and 93.0. Bacteriologic response was also similar. The overall incidences of adverse events were 8.5% for daily dosing and 7.2% for b.i.d, dosing, but there was a slightly greater frequency of gastrointestinal events (p < 0.03) was observed with the once-daily regimen. Thus roxithromycin 300 mg once daily can be recommended as an effective and well-tolerated alternative to the conventional b.i.d. regimen.
INTRODUCTION
fort, the present study was designed to evaluate oncedaily dosing with 300 mg roxithromycin in comparison with the proven efficacy of the same dosage given in two divided doses of 150 mg. Consideration was given to biostatistical aspects, particularly to avoid [~-type error, which is so important for this kind of comparative study. In addition, to ensure extrapolation of the results to a large patient population, the study was carried out in 12 cotmtries on three continents.
Roxithromycin is an acid-stable, orally administered macrolide, structurally related to erythromycin. In vitro, its antibacterial profile is similar to that of erythromycin (Pech~re and Auckenthaler, 1987; Barlam and Neu, 1984; Ridgway, 1987), with a few exceptions [for example, roxithromycin seems more active than erythromycin against Toxoplasma gondii (Chang and Pech~re, 1987)]. Roxithromycin has a distinctive pharmacokinetic profile characterized by excellent enteral absorption, prolonged serum h a l f life (~12 hr), and high concentrations in most tissues and body fluids (Periti et al., 1989; Purl and Lassman, 1987; Ridgway, 1987). This profile could in theory be compatible with once-daily dosing in infections for which roxithromycin is indicated. On the basis of these pharmacokinetic characteristics, and to enhance compliance and patient cornFrom the Facultyof Medicine, Department of Microbiology, CMU Geneva, Switz~*rland. Additional copies of this supplement are availablefrom Roussel UCLAF, DomaineTh~rapeutiqueAntibioth(~rapie,35 Boulevard des Invalides, 75007Paris, France. Received 8 October 1991;revised and accepted 20 December 1991. © 1992ElsevierSciencePublishingCo., Inc. 655 Avenue of the Americas, New York, NY 10010 0732-8893192/$5.00
MATERIALS AND METHODS Study Design This was a multicenter, double-blind, randomized, two-arm study. The target was to enroll 1600 patients to ensure that at least 1240 patients, that is, 620 patients per group, would be evaluable for clinical efficacy. This sample size enables the detection of a 5% difference between the two dosage regimens, with an ot = 0.05 and [3 = 0.01 (90% power), assuming a success rate on both regimens of 90%. Randomization was performed by a random-number generation, with a separate randomization list for each country. Randomization was one to one in blocks of four.
112S
Treatment Roxithromycin was supplied as white film-coated tablets, each containing 300 mg or 150 mg of active drug. A double-dummy technique was used in order to ensure the double-blind design: two tablets had to be taken every morning and one table every evening, an hour before meals with a glass of water. This allowed administration of either 300 mg once daffy or 150 mg twice daily. Five days were considered as the minimal treatment duration for a patient to be evaluable for clinical efficacy, except in the case of clinical failure (3 days). The treatment could be continued as long as necessary according to the investigator's opinion, but was not to exceed 20 days. It could be discontinued at any time if (a) the pathogen was shown to be resistant to roxithromycin in vitro; (b) there was no clinical response, despite susceptibility of the isolate to roxithromycin in vitro; or (c) a severe adverse effect was noted that was thought to be related to roxithromycin. Antimicrobial therapy other than roxithromycin was prohibited, as were ergot derivatives, cyclosporin, and any drug affecting the absorption of roxithromycin.
Patients, Surveillance, and Criteria Eligible patients were men and women 18 years of age or older, with an infectious disease caused by susceptible pathogens that usually respond to oral macrolide therapy, including upper and lower respiratory tract infection, and skin and soft tissue infections. Patients were entered before bacteriologic results were available and their evaluability subsequently determined. The following patients were excluded: pregnant or nursing women; patients with a history of hypersensitivity to macrolides; patients with a known severe hepatic or renal insufficiency; patients receiving any investigational drug within the 2 previous weeks; patients receiving any other antibiotic within 3 days prior to entering the study, unless the microorganism responsible for the infection was shown to be resistant to the antibiotic previously used; patients receiving other concurrent antibiotic therapy; patients with abnormal findings that could affect interpretation of the results; and patients unable to comply with the study protocol. A full clinical examination was performed at the beginning, middle, and end of the study. Chest radiography was required for patients with lower respiratory tract infections, within 48 hr of enrollment and, in principle, 48 hr after completion of treatment. Standard hematology and biochemistry tests were performed within 48 hr prior to initiating therapy, at midtreat-
J.-C. Pech~re
ment, and within 48 hr of completing therapy. Specimens of the infected site material were cultured and tested for sensitivity. A minimum of two cultures had to be performed, one prior to treatment and another within 48 hr of the end of treatment, whenever technically possible. Minimum inhibitory concentration (MIC) break points were 1 and 4 p~g/ml (Acar et al., 1987) for susceptible and resistant, respectively. Adverse events were categorized, according to the investigator's opinion of their severity, as mild, moderate, or severe. Their relationship to the study drug was assessed by the investigator as nil, unlikely, possible, probable, clinically nonsignificant, or insufficiently documented. A three-scale assessment of clinical response was used: "satisfactory" = clinical improvement or resolution; "unsatisfactory" = deterioration or no change; and "unevaluable." Bacteriologic response was assessed as "satisfactory" = susceptible pathogen eradicated and/or new isolate without clinical symptoms, or clinical cure, making follow-up culture impossible; "unsatisfactory" = persistence of susceptible pathogen with persistence or reappearance of clinical signs, or partial eradication in the event of multiple susceptible pathogens; and "nonevaluable." Statistical Methods We used Student's t-test for continuous variables, the chi-square test for categorical variables (using Fisher's exact test for 2 x 2 tables or the likehl~ood ratio chi-square for larger tables whenever suitable), and Wilcoxon's nonparametric test for ordinal parameters. Overall clinical and bacteriologic response rates were summarized by treatment groups and compared using Mantel-Haenszel statistics with respect to diagnosis. RESULTS A total of 1588 patients from 62 centers and 12 countries entered the study: 793 were randomized to 300mg once-daily group and 795 to 150 mg b.i.d. Fifteen patients (nine in the once-daily group and six in the b.i.d, group) were included, but did not receive any treatment and were lost to follow-up just after inclusion. In the remaining 1573 patients, demographic parameters and base-line data were similar in the two dosage groups (Table 1). In addition, we found no significant difference between the two treatment groups in symptom duration, prior antibiotic therapy, underlying disease, body temperature, blood pressure and pulse, or the findings of physical examination and chest radiography (for lower respiratory tract infection).
113S
Roxithromycin 300 mg Daily vs 150 mg b.i.d.
TABLE 1 Demographic and Disease Characteristics at Presentation Patient Group
Men Outpatients (%) Inpatients (%) Race Caucasian Black Oriental Other Age (years) (m - SD) Weight (kg) (m -+ SD) Height (cm)(m -+ SD) Overall status (%) Asymptomatic Fair Serious Critical History of infection Acute Chronic Recurrent Diagnosis of infection Upper respiratory tract Lower respiratory tract Skin and soft tissue
~ 0 mg Once Daily (n = 784)
150 mg Twice Daily (n = 789)
52.0 60.5 39.4
51.1 59.8 40.2
87.5 2.9 0.1 9.4 46.1 ± 19.5 67.8 -- 13.9 168.1 -+ 9.6
86.9 3 0 10.0 45.8 -+ 19.6 67.6 -+ 13.8 168.1 -+ 9.5
14.8 71.6 13.1 0.5
15.3 68.3 16 2 0
86.0 3.6 10.5
87.8 2.7 9.4
37.5 54.2 8.3
39.5 52.0 8.5
p > 0.05 for all comparisons.
Patients were excluded from the analysis of clinical efficacy (n = 182) either because a resistant (or intermediate) pathogen was identified on pretreatment bacteriology (86 cases) or because of loss to follow-up (45 patients). A further 17 patients were excluded for protocol violations and 14 for diagnostic errors including active tuberculosis. A total of 872 patients were excluded from the analysis of bacteriologic efficacy either because no causative pathogen was identified in the pretreatment bacteriologic specimen (637 cases), susceptibility was not tested on the isolate (130 cases), or no bacteriology was performed at all (126 cases). Treatment duration was 9.8-+ 2.9 days in the oncedaily group and 9.9+2.9 days in the b.i.d, group. During the trial, 754 patients, 377 per group, reported one or more concomitant or intercurrent conditions and received concomitant treatments allowed by the protocol. Statistical analysis of clinical efficacy in 1391 patients showed no significant difference between regimens either in the total population or per diagnosis (see Table 2). Success rates differed by 0.1% between treatment groups (95% conference interval, - 2 . 4 %
to + 2.6%). The Mantel-Haenszel probability, stratified by diagnosis, was 0.90. The clinical efficacy in patients over 65 years of age, w h o accounted for 20.7% of the total population, was separately analyzed a posteriori. Overall clinical efficacy was 88.9% a n d 90.3% with oncedaily versus b.i.d, dosing, respectively (hiS), compared with 95.5% and 94.9% in younger patients. Statistical analysis in 519 patients showed no significant difference between regimens in bacteriologic efficacy either in the total population or per diagnosis (Table 3). Of these 519 patients, clinical cure was achieved in 128, making posttreatment bacteriology impossible. The difference between success rates in the two groups was 1.9% (95% confidence interval, - 1 . 5 % to 5.2%). The Mantel-Haenszel probability, stratified by diagnosis, was 0.345. Eradication rates per pathogen are shown in Table 4. As allowed by the protocol, treatment was started before receiving the bacteriologic results. However, investigators were allowed to continue treatment if the pa'tient was improving clinically. Pretreatment bacteriology identified 69 resistant isolates. This was followed up after treatment in 26 cases: 16 isolates
114S
J.-C. Pech~re
TABLE 2 Clinical Response per Diagnosis in 1391 Patients Treated with Roxithromycin 300 mg Once Daily or 150 mg Twice Daily Satisfactory Clinical Response (%) by GroulY' 300 mg Once Daily (n = 694)
150 mg Twice Daily (n = 697)
Pharyngitis (84) Tonsillitis (139) Sinusitis (262; Otitis (39) Pneumonia (344) Acute bi'ollChitis(222) Acute episode of chronic bronchitis {134) Other respiratory infections (52) Skin and soft tissue infections (115)
95.7 98.4 93.4 92.3 93.3 92.1 97.1
100 98.7 92.9 92.3 89.8 98.1 90.9
95.0
96.9
93.0
93.1
Total patients (1391)
94.1
94.0
Diagnosis (No. of Patients)
"No significantdifferencebetween the two groups. (61.5%) were found to have been eradicated, and clinical cure precluded bacteriology in six cases. An unsatisfactory clinical and/or bacteriologic response was seen in 93 (6.6%) of 1391 patients, 47 (6.7%) in the once-daily group and 46 (6.5%) in the b.i.d, group. Superinfection occurred in 19 patients (six and 13 cases in the orlce-daiiy and b.i.d, groups, respectively), associated in nine cases with persistent clinical infection. The pathogens associated with superinfection were streptococci (including one
Streptococcus pneumoniae), Staphylococcus aureus, Moraxella catarrhalis, Enterobacter aerogenes, and Candida albicans. The incidence of adverse events classified as being "unlikely," "possibly," or "probably" related to roxithromycin therapy is presented in Table 5. Gastrointestinal events were commoner in the once-daily group (p = 0.03), particularly nausea and/or vomiting and abdominal pain. Most gastrointestinal upsets were mild or moderate. Only four events in the oncedaily group and one in the b.i.d, group were assessed as severe by the investigator. Neurologic adverse events, mostly dizziness, vertigo, or headache, were mild or moderate, and thought unlikely to be due to the test drug in 12 of 17 cases. Skin reactions consisted of rash, pruritis, and urticaria, and were possibly due to roxithromycin in eight of 13 cases. Liver function changes consisted of transient enzyme elevation in six of seven cases, and one case of cholecystitis, which was thought unlikely to be due to roxithromycin. Also described were 19 "other" events, of which 18 were thought unlikely to be due to roxithromycin. Ten patients in the once-daily group and seven in the
b.i.d, group were withdrawn due to an adverse event possibly related to the test drug: 12 cases of gastrointestinal upset and five skin reactions. The figures for the events both possibly related to the study drug and defined as severe by the investigators were five in the once-daily group and one in the b.i.d. group. Thirteen subjects (six and seven per group, respectively) developed clinically significant laboratory changes possibly caused by the study drug, including decreases in hemoglobin or platelet count, serum creatinine, SGPT, and/or SGOT. All were relatively mild and transient.
DISCUSSION The ability to offer treatment on once-a-day basis is an appreciable benefit in terms of patient comfort, especially for the ambulatory patients who make up the majority of those likely to be treated with macrolides. Compliance is also likely to be enhanced by once-daily rather than more frequent dosing. The sample size of the present study was designed to be able to detect a 5% difference between the two dosage regimens. The effective power of the study, computed on the basis of the number of evaluable patients, was >99% for the analysis of clinical response and -92% for the analysis of bacteriologic response. The randomization list was adhered to, as evidenced by final results, and no difference was observed between the two groups in base-line variables or treatment duration. No statistically significant difference between the regimens was found in clinical or bacteriologic efficacy, and the overall incidence of sat-
Roxithromycin 300 mg Daily vs 150 mg b.i.d.
TABLE 3
115S
Bacteriologic Efficacy per Diagnosis in 519 Patients Treated with Roxithromycin 300 mg Once Daily or 150 mg b.i.d. Satisfactory Bacteriologic Response (%) by Group~
Diagnosis (No. of Patients)
300 mg Once Daily (n = 257)
150 mg Twice Daily (n = 262)
Pharyngitis (31) Tonsillitis (76) Sinusitis (34) Otitis (19) Pneumonia (113) Acute bronchitis (91) Acute episode of chronic bronchitis (65) Other respiratory infections (15) Skin and soft tissue infections (75)
92.9 89.2 100 100 98.3 95.8 100
88.2 89.7 100 93.3 92.6 100 100
100
100
Total of patients (519)
100
94.2
96.9
95.0
aNo significantdifferencebetween the twe groups.
TABLE 4
Bacteriologic Efficacy per Pathogen in 403 Isolates Percent Eradicated in Groul~ 300 mg Once Daily
150 mg Twice Daily
Other (6)
90.7 98.7 100 94.4 100 100
91.8 97.0 95 91.7 100 100
Total (403)
96.6
94.9
Pathogen (No. of Strain) Streptococcus pyogenes (!03) Streptococcus pneumoniae (146) Staphylococcus spp. (106)b Haemophilus influenzae (30) MoraxeUa catarralis (12)
aNo significantdifferencebetween the two groups. bS. aureus = 105 and S. epidermidis = 1. isfactory responses was high, confirming that roxithromycin is active against the infections selected in this study (Lachat et al., 1986; Gentry, 1987; Champetier de Ribes et al., 1989; Young et al., 1989). This success is due to its antibacterial spectrum, combined with the high concentrations achieved in tissue and body fluids, for example, tonsil, middleear exudate, and bronchial secretions (Periti et al., 1989). In particular, roxithromycin concentrations in bronchoalveolar lavage cells are two and ten times higher than concomitant levels in plasma and epithelial fluid, respectively (Chastre et al., 1987). In individual infections, both regimens produced very similar results in tonsillitis, sinusitis, otitis, and skin and soft tissue infections. The once-daily dose
gave (nonsignificantly) better results in pneumonia (93.3% vs 89.9%) and exacerbations of chronic bronchitis (97.1% vs 90.9%), suggesting that this r e d , h e n could provide improved pharmacokinetic conditions in lung tissue and bronchial secretion. Conversely, the twice-daily regimen was associated with a higher rate of satisfactory clinical response (nonsignificant) in pharyngitis (100% vs 95.7%) and acute bronchitis (98.1% vs 92.1%). In some of the diseases treated in the present study, however, particularly pharyngitis/tonsillitis and acute bronchitis, the efficacy of cure was difficult to evaluate because of the propensity for spontaneous cure and the frequent viral involvement. It was therefore important to consider bacteriologic ef-
1165
J.-C. Pech6re
TABLE 5 Incidence of Reported Adverse Events on Therapy with Roxithromycin 300 mg Once Daily or 150 mg b.i.d. Treatment Group 300 mg Once Daily (n = 784)
150 mg Twice Daily (n = 789)
Gastrointestinal Nausea Vo~mting Abdominal pian Diarrhea/loose stools Flatulence Other Neumlogic Skin reaction Liver Other
54 (7.0%) 19 7 14 11 2 1 10 (1.3%) 5 (0.6%) 4 (0.5%) 7 (0.9%)
35 (4.4%) 12 2 9 10 1 1 7 (0.9%) 7 (0.9%) 3 (0.4%) 12 (1.5%)
0.46 0.57 0.69 0.25
Total of patients with one or more adverse effects
67 (8.5%)
57 (7.2%)
0.33
Adverse Event
0.03a
"Significantdifference,p < 0.05. ficacy. High activity against streptococci was found, as expected from previous reports (Herron, 1987) and low MICs (Pech6re and Auckenthaler, 1987), but the activity of roxithromycin against staphylococci and Haemophilus influenzae, where MICs are higher (Pech6re and Auckenthaler, 1987), deserves spedal mention. Anti-Haemophilus activity by roxithromycin in pneumonia, despite often borderline in vitro potency, agrees with previous fndings (Brfickner et al., 1989) and is probably the result of the drug's favorable pharmacokinetics (Nilsen et al., 1992). The study comprised 61 cases of atypical pneumonia, a critical indication. Clinical cure was achieved in 58 cases. In three patients, including two documented mycoplasma infections, the clinical response was unsatisfactory. All three werc in the twice-daily group. The favorable overall results in atypical pneumonia confirmed those obtained elsewhere (Fernandez-MacLoughli et al., 1988; Breux et al., 1989).
Skin and soft tissue infections consisted mainly of erysipelas and, as expected, most of the isolates were S. aureus and streptococci, evenly distributed in the two dosage groups. The clinical and bacteriologic responses were excellent, with notably a 100% bacteriologic cure rate in the once-daily group. The adverse events potentially due to roxithromycin were generally those expected for this type of drug. Overall, the incidence of side effects was similar in both groups, but with more gastrointestinal events on the once-daily group. According to pharmacokinetic studies, this difference is not due to drug accumulation in tissues on once-daily dosing (Nilsen et al., 1992). On macrolide therapy, gastrointestinal symptoms are due to local irritation, probably by direct action on smooth muscle. Thus, a higher incidence of gastrointestinal symptoms could be expected in the group receiving the higher unit dose (Keller and Follath, 1988).
REFERENCES Acar J, Bergogne-Berezin E, Chabbert Y, Ouzel R, Courtieu A, Courvalin P (1987) Comminqu~s 1987du Comit~ de I'Antlbiogramme de la Soci~t(~Fran(~aise de Microbioiogie. Pathol Biol 35:1101-1004. Barlam T, Neu NC (1984) In-vitro comparison of the activity of RU 28965, a new macrolide, with that of erythromycin against aerobic and anaerobic bacteria. Antimicrob Agents Chemot~r 25:529-531. Breux JP, Fort E, Agius G, Grollier G, Cazenave-Roblot F, Casters M (1989) Empirical therapy for atypical pneu-
monia with roxithromycin. In Abstracts, 16th International Congress of Chemotherapy, Jerusalem, Israel. International Society of Chemotherapy. Briickner O, Trautmann M, Kemmerich B (1989) Roxithromydn compared to a new cephalosporin, cefixime, and dprofloxadn in purulent chest disease and pneumonia. In Abstracts, 16th International Congress of Chemotherapy, Jerusalem, Israel. International Society of Chemotherapy. Champetier de Ribes D, Jockey C, Gaudouen G (1989)
Roxithromycin 300 m g Daily vs 150 ,rig b.i.d.
Etude de refficacit~ et de la tolerance de la roxithromycine dans les infections broncho-pulmonaires. Sem Hop Paris 65:2819-2821. Chang HR Pech~re JC (1987) Effect of roxithromycin on acute toxoplasmosis in mice. Antimicrob Agents Chemother 31:1147-1149. Chastre J, Brun B, Fourtillan JB, Soler P, Basset G, Manuel C, Trouillet JL, Gibert C (1987) Pulmonary disposition of roxithromycin (RU 28965), a new macrolide antibiotic. Antimicrob Agents Chemother 31:1321-1316. Fernandez-MacLoughli GF, Lanoel JL, Staraboulian D, Sarachian B (1988) Roxithromycin in the treatment of atypical pneumonia. Br J Clin Pract 42(Supp155):92-93. Gentry LO (1987) Roxithromycin, a new macrolide antibiotic, in the treatment of infections in the lower respiratory tract: an overview. J Antimicrob Chemother 20(Suppi B):145-152. Herron JM (1987) Roxithromycin in the therapy of Streptococcus pyogenes throat infections. J Antimicrob Chemother 20(Suppl B):139-144. Keller H, Follath F (1988) Meyler's Side-Effects of Drugs, l l t h ed. Ed, MNG Dukes. Elsevier: Amsterdam, pp 543566.
117S
Lachat JM, Ane× JF, Regamey C (1986) Ro×ithromycin, a new macrolide for pulmonary infections. Schweiz Med Wochenschr 116:1739-1741. Nilsen OG, Aamo T, Zahlsen K, Svarva P (1992) Macrolide pharmacokinetics and dose scheduling of roxithromycin. Diagn Microbiol Inf Dis 15(Suppl):71S--76S [this supplement]. Pech~re JC, Auckenthaler R (1987) In-vitro activity of ro×ithromycin against respiratory and skin pathogens. J Antimicrob Chemother 20(Suppl B):l-5. Periti P, Mazzei T, Mini E, Nove|ii A (1989) Clinical pharmacokinetic properties of the macrolide antibiotics. Clin Pharmacokinet 16:261-282. Purl SK, Lassman HB (1987) Roxithromycin: a pharmacokinetic review of a macrolide. J Antimicrob Chemother 20(Suppl B):89-110. Ridgway GL (1987) A review of the in-vitro activity of roxithromycin against genital pathogens. J Antimicrob Chemother 20(Suppl B):7-11. Young RA, Gonzales JP, Sorkin EM (1989) Roxithromycin: a review of its antibacterial activity, pharmacokinetic properties and clinical efficacy. Drugs 37:8-41.
