Journal of Antimicrobial Chemotherapy (1996) 7, 1-18 Phexymethyl penicillin versus co-amoxiclav in the treatment of acute streptococcal pharyngitis, and the role of /Mactamase activity in saliva R. S. Dykhuizen', D. Golder*, T. M. S. Reid* and I. M. Gould' "Infection Unit and ''Department of Medical Microbiology, Aberdeen Royal Infirmary NHS Trust, Forresterhill, Aberdeen AB2 2ZD, UK One hundred and sixty-five consecutive patients (>2 years of age) with acute group A streptococcal (GAS) pharyngitis randomly received co-amoxyclav (79 patients) or phexymethyl penicillin (86 patients). /?-Lactamase activity in saliva was determined for each patient. At follow up after seven days, tonsillar cultures from seven patients (9 6%) in the penicillin V group grew group A streptococcus; three of these patients had tonsillitis clinically. In the co-amoxiclav group these figures were three (.8%) and two respectively (P > 0.05). Within the 12 month follow up period, there were four clinical s (6.1 %) in the penicillin V group and seven (9.%) in the co-amoxiclav group (P > 0.1). /?-Lactamase activity in the saliva was demonstrated in 29 patients (19.2%). Fourteen (74%) of 19 bacteriological s or clinical s had /Mactamase activity, versus 15 (12%) of 129 successfully treated patients (P < 0.001). There is evidence that oral co-amoxiclav is better than oral penicillin V for the first treatment of acute GAS pharyngitis, but bacteriological and clinical are strongly associated with the presence of /?-lactamase activity in commensal flora Introduction Streptococcus pyogenes (group A streptococcus) is one of the most common human pathogens. It causes a wide array of infections, the most frequent of which are acute pharyngitis and impetigo. The burden of this infection is determined by days missed from school or work, physician visits, and costs of antimicrobial treatment. The nsuppurative sequelae of acute rheumatic fever have been the focus of intense clinical interest. The frequent epidemics of severe group A streptococcal (GAS) infections as well as the high incidence of acute rheumatic fever seen in the 19th and early 20th century have been on the decline since the Second World War due to improvements in socio-ecomic conditions and the introduction of antimicrobial agents. Since the early eighties however, there has been a resurgence of group A streptococcal infections and their sequelae affecting all socio-ecomic classes in the United States and Europe (Bis, 1991). Penicillin has been the recommended drug of choice in most cases of GAS pharyngitis for nearly 40 years based on its efficacy in the prevention of acute rheumatic fever. The organism remains uniformly susceptible to penicillin in vitro (Bis, 1991). However, treatment s do occur and speculation has developed about the role of /f-lactamase production by pharyngeal flora rendering penicillin less effective. Brook (1989a) studied 1 005-745/96/0101 + 06 $12.00/0 f; 1996 The British Society for Antimicrobial Chemotherapy
14 R. S. Dykhuizen et al. 4 children with recurrent pharyngitis. Administration of penicillin did t eliminate GAS in six of 20 children treated with a 10 day course. In contrast, all GAS were eliminated following administration of co-amoxiclav. An increase in the number of /2-lactamase producing organisms was ted in the penicillin group, but a reduction of these organisms was seen in the co-amoxiclav group. The rapid emergence of /?-lactamase producing organisms in patients treated with penicillin had been previously reported (Brook & Gober, 1984). In contrast to these findings in children with recurrent GAS pharyngitis, Tanz et al. (1990) could demonstrate therapeutic advantage of co-amoxiclav over penicillin in 89 children who presented with GAS pharyngitis for the first time. In their study bacteriological treatment was unrelated to recovery of /?-lactamase producing organisms at the time of enrolment or after treatment. In the present report, we investigated 151 patients aged 2 49 years, who presented with acute GAS pharyngitis as indicated by a rapid GAS antigen assay. Bacteriological outcome and rate after treatment with co-amoxiclav in comparison with penicillin V, and the relation of the outcome to the presence of /Mactamase in sputum of patients were examined. Materials and methods The study was approved by the Joint Ethics Committee of Grampian Health Board and the University of Aberdeen. In an observer-blind randomised, parallel group, multicentre trial, using a randomisation stratified by age, 165 consecutive patients (>2 years of age) presenting to general practice with GAS pharyngitis, as diagsed by rapid latex streptococcal antigen detection (Strep A, Abbott Laboratories, Maidenhead, U.K.), were entered into the study. Excluded were patients who had received antimicrobial therapy during the preceding days, were hypersensitive to penicillin or co-amoxiclav, had renal or hepatic dysfunction, were pregnant of lactating, or had a history of recurrent pharyngitis (> 1 episode). Patients in whom the presence of GAS was confirmed using the rapid antigen detection kit and who fulfilled the entry criteria, had an ENT and general examination at entry. Throat swabs were transported to the laboratory within 4 h for aerobic and anaerobic culture on horse blood agar with overnight enrichment in Todd-Hewitt broth. GAS strains were serotyped by the Central Public Health Laboratory, Colindale, UK. A sample of saliva was collected in sterile containers for the determination of /Mactamase activity, after centrifugation, by chromogenic cephalosporin (Nitrocefin Unipath, Basingstoke, UK), using the method of O'Callaghan et al. (1972). After obtaining informed consent the patients were randomised into treatment groups stratified by age to receive either penicillin V or co-amoxiclav orally for seven days. Age defined dose regimens were used; 2-5 years: 125 amoxycillin + 1 mg clavulanic acid (paediatric suspension) tds or 125 mg penicillin V (syrup) qds; 6-11 years: 125 mg amoxycillin + 62 mg clavulanic acid (junior suspension) tds, or 250 mg penicillin V (syrup) qds; 12 years and over: 250 mg amoxycillin + 125 mg clavulanic acid (tablets) tds, or 250 mg penicillin V (tablets) qds. The patients were reviewed after seven days when compliance was ascertained by measurement of any unused drug returned. Medical and bacteriological examinations were repeated. Treatment s were defined as patients whose throat swab taken during the follow up visit after seven days still grew group A streptococci. All treatment s
Penicillin versus co-amoxyclav in pharyngitis 15 were withdrawn from further follow up. The other patients were included in a follow up for 12 months to monitor for of clinical tonsillitis. Results were analysed by y} or Fisher's exact test where appropriate. The analyses were carried out on an intention to treat basis. Results Seventy-nine patients were entered into the penicillin V and 86 into the co-amoxiclav treatment group. Entry data for sex, age, and history of pharyngitis in the preceding 12 months were similar for the two patient groups (P > 0.2 for all criteria, Table I). Haemolytic streptococci were isolated from throat swabs in 70 patients of the penicillin V group (89% of total), 68 group A and two group G. Throat swabs of patients in the co-amoxiclav group showed growth of haemolytic streptococci in 74 patients (86% of total), 7 group A and one group B. There were six withdrawals in the penicillin V group; three because of an adverse event, two because of n-compliance, and one patient failed to attend for follow up. In the co-amoxiclav group these figures were eight withdrawals; two adverse events, five n-compliance, and one n-attender (P > 0.1 comparing both treatment groups). The adverse events were gastrointestinal upsets in all cases, except for a skin rash on one occasion in each treatment group. At follow up after 7 days, throat swabs from seven patients (9.6%) in the penicillin V group grew GAS; three of these patients had pharyngitis clinically. In the co-amoxiclav group these figures were three (.8%) and two respectively (P > 0.05). Within the 12 month follow up period, there were four s (6.1%) in the penicillin V group and seven (9.%) in the co-amoxiclav group (P > 0.1). Fifteen (67%) Table I. Entry data of the two treatment groups, stratified by sex, age, previous history of tonsillitis, and /?-lactamase activity in saliva. Originally 165 patients were entered but 14 withdrawals occurred, eight in the co-amoxyclav and 6 in the penicillin V group. Treatment s and s with and without /?-lactamase activity are shown in the table Study entries male female Age range (years) 2-5 years 6-11 years ;> 12 years Previous history of: tonsillitis' /?-Lactamase activity Treatment s Failures with /?- lactamase activity Recurrences < 12 months Recurrences with filactamase activity Co-amoxyclav 78 2 55 2^t9 10 2 45 8 9 7 Penicillin V 7 40 2-49 10 17 46 29 20 7 5 4 Total 151 56 95 2^49 20 40 91 67 29 10 One previous episode only; patients with recurrent tonsillitis were excluded from the study. 8 11 6
16 R. S. Dykhuizen et al. Table II. Serotypes of treatment s and s together with /J-lactamase activity in saliva at entry of the study Sample number 7(C) 9(P) 11 (C) 18 (P) 2 (C) 4 (P) 6 (P) 7 (P) 55 (C) 72 (P) 84 (C) 90 (C) 94 (C) 109 (C) 117 (P) 122 (P) 127 (P) 146 (P) 162 (P) 171 (C) 175 (C) Same serotype T1M1 T1M1 T28OF + T1M1 T28M28 T28M28 T28OF + T6M6 T4M29 Different serotype first: T1OF + second: TOF + first: second: T1M1 first: T1T12M1M12 second: T1M1 first: second: T28M28OF + first: second: T28M28 first: second: T28M28 both n-typable (C), Co-amoxyclav; (P), penicillin V. /Mactamase activity t tested t tested Failure or of the 21 bacteriological s or s had the same serotype of group A streptococcus on repeat isolation. In one case the strain was n-typable on entry and after of treatment (Table II). On entry into the study, /Mactamase activity in the saliva was demonstrated in 29 patients (19.2%); 20 patients in the penicillin V group and 9 patients in the co-amoxiclav group (% 2 = 4.81 with 1 df: P = 0.028). Fourteen (74%), eight penicillin V and six co-amoxiclav, of the 19 bacteriological s or s that were tested for /Mactamase activity were positive, versus 15 (12%) of the 129 successfully treated patients (P < 0.001). Discussion GAS antigen assays are simple to perform, give rapid results and, although they lack sensitivity, a positive test can be used reliably as the basis to initiate immediate antibiotic therapy because they are highly specific for group A streptococci (Gerber 1989; Anhalt et al. 1992). In this study 87.% of the positive antigen assays were confirmed by growth from throat swabs.
Penicillin versus co-amoxyclav in pharyngitis 17 After randomisation, stratified by age, the two treatment groups were well matched, except for the /Mactamase activity in sputum (20 patients in the penicillin V and 9 patients in the co-amoxiclav group; y 2 = 4.81 with 1 df: P = 0.028). Both treatment groups showed good tolerance to the drug used and the number of withdrawals was similar. Penicillin V and co-amoxiclav appeared equally effective at 7 day follow up and they showed similar rates. Despite the relatively short duration of treatment, the efficacy of both drugs was comparable or better than that from previous trials (Pichichero et al., 1987; Milatovic & Knauer, 1989; Tanz et al., 1990). A similar efficacy of penicillin V versus co-amoxiclav was reported before (Tanz et al., 1990), but Brook (1989/)) found a lower rate of bacterial colonisation of the tonsils after co-amoxiclav compared with penicillin V when taking two-monthly tonsillar cultures after treatment. His study was carried out in subjects with recurrent pharyngitis where local /Mactamase production may be present in a high percentage of patients. In our view, the use of clinical and/or bacteriological as an indicator of the efficacy of the antimicrobial agent used is debatable, because reinfection may take place after initial eradication. In the present study repeat swabs were only taken from patients with evidence of clinical. Six out of 21 treatment s or s occurred with different serotypes (Table II). The most likely explanation for the finding of a different serotype during a of clinical pharyngitis (four cases) is reinfection of the pharynx by a different organism and is t an indication of inadequate treatment of the original episode. The finding of a different serotype after seven days of antimicrobial treatment (two cases) could be due to co-existence of two serotypes before treatment or reinfection. /?-Lactamase activity in saliva was demonstrated in 19% of patients at entry into the study. This percentage is lower than that reported in the literature for children and adolescents in Maryland, USA (Brook 1989Z>). This may be due to the different method used in our study to assess /Mactamase production, a lower exposure rate to antimicrobials in our study population, or to the older age group studied. There was a strong association between treatment or and /Mactamase activity (P < 0.001). This finding is in contrast with the findings of Tanz et al. (1990) who found relationship between /Mactamase production and treatment outcome. However, they assessed /Mactamase production in vitro, after culturing throat swabs under anaerobic conditions. In our study, /Mactamase detection was dependent on in-vivo production of /Mactamase in the saliva of patients, which resulted in a lower rate of positivity for /Mactamase detection than encountered by Tanz et al. (1990), but a higher correlation of /Mactamase production with treatment outcome. Despite the strong association between treatment or with /Mactamase activity, co-amoxiclav did t perform better than penicillin V. This could be due to the low prevalence of /Mactamase activity plus the observed bias at entry, which resulted in only 9 patients with /Mactamase activity in saliva receiving co-amoxiclav. Ather explanation could be lack of penetration of co-amoxiclav into infected tissue although this seems unlikely since the drug penetrates well into other respiratory tissues (Gould, Legge & Reid, 1988; Gould et al., 1994). In conclusion, evidence was obtained from this study that oral co-amoxiclav is superior to oral penicillin V for the first treatment of acute GAS pharyngitis. However, an association of treatment with /Mactamase activity in commensal flora was
18 R. S. Dykhuizen et al. encountered in the study, and Brook (1989a) has demonstrated the superiority of co-amoxiclav over penicillin V in children with recurrent tonsillitis. Therefore we would recommend treatment with a /?-lactamase stable agent in the case of proven clinical recurrent of GAS pharyngitis. Ackwledgements Thanks are due to the general practitioners who cooperated in the study and the Central Public Health Laboratory, Colindale, U.K. for typing the isolates. Financial support for the study was provided by SmithKline Beecham UK Ltd, Welwyn Garden City, U.K. The study was partly presented at the 0th Interscience Conference on Antimicrobial Agents and Chemotherapy, Atlanta Georgia, 21-24 October, 1990; Abstract 642. References Anhalt, J. P., Heiter, J. B., Naumovitz, D. W. & Bourbeau, P. P. (1992). Comparison of three methods for detection of group A streptococci in throat swabs. Journal of Clinical Microbiology 0, 215-8. Bis, A. L. (1991). Group A streptococcal infections and acute rheumatic fever. New England Journal of Medicine 25, 78-9. Brook, I. (1989a). Treatment of patients with acute recurrent tonsillitis due to group A /?-haemolytic streptococci: a prospective randomized study comparing penicillin and amoxycillin/clavulanate potassium. Journal of Antimicrobial Chemotherapy 24, 227-. Brook, I. (19896) The concept of indirect pathogenicity by /?-lactamase production, especially in ear, se and throat infection. Journal of Antimicrobial Chemotherapy 24, Suppl. B, 6-72. Brook, I. & Gober, A. E. (1984). Emergence of beta-lactamase producing aerobic and anaerobic bacteria in the oropharynx of children following penicillin chemotherapy. Clinical Pediatrics 2, 8-41. Gerber, M. A (1989). Comparison of throat cultures and rapid strep tests for diagsis of streptococcal pharyngitis. Pediatric Infectious Diseases Journal 8, 820-4. Gould, I. M., Harvey, G., Golder, D., Reid, T. M. S., Watt, S. J., Fnend, J. A R f( al. (1994). Penetration of amoxycillin/clavulanic acid into bronchial mucosa with different dosing regimens. Thorax 49, 999-1001. Gould, I. M., Legge, J. S & Reid, T. M. S. (1988). Amoxycillin/clavulanic acid levels in lower respiratory secretions. Journal of Antimicrobial Chemotherapy 22, 88-9. Milatovic, D. & Knauer, J. (1989). Cefadroxil versus penicillin in the treatment of streptococcal tonsillopharyngitis. European Journal of Clinical Microbiology and Infectious Diseases 8, 282-8. O'Callaghan, C. H., Morris, A., Kirby, S. M. & Shingler, A H. (1972). Novel method for detection of beta-lactamase by using a chromogenic cephalosporin substrate. Antimicrobial Agents and Chemotherapy 1, 28-8. Pichichero, M. E., Disney, F. A., Arovitz, G. H., Talpey, W. B., Green, J L. & Francis, A. B. (1987). Randomized, single-blind evaluation of cephadroxil and phexymethyl penicillin in the treatment of streptococcal pharyngitis. Antimicrobial Agents and Chemotherapy 1, 90-6. Tanz, R. R.. Shulman, S. T., Sroka, P. A., Marubio, S.. Brook, I. & Yogev, R. (1990). Lack of influence of /J-lactamase producing flora on recovery of group A streptococci after treatment of acute pharyngitis. Journal of Pediatrics 117, 859-6 (Received 24 April 1995; returned 6 June 1995; revised July 1995: accepted 16 August 1995)