TREATMENT OF CHANCROID BY CLAVULANIC ACID WITH AMOXYCILLIN IN PATIENTS WITH β-LACTAMASE-POSITIVE HAEMOPHILUS DUCREYI INFECTION

Saturday 4 September TREATMENT OF CHANCROID BY CLAVULANIC ACID WITH AMOXYCILLIN IN PATIENTS WITH &bgr;-LACTAMASE-POSITIVE HAEMOPHILUS DUCREYI INFECTION HERBERT NSANZE MARGARET V. FAST FRANCIS A. PLUMMER LUCE J. D’COSTA IAN W. MACLEAN PETER KARASIRA ALLAN R. RONALD

Department of Medical Microbiology, University of Manitoba, Winnipeg, Manitoba, Canada; Department of Medical Microbiology, University of Nairobi; and Special Treatment Clinic, Nairobi, Kenya Multiresistant strains of Haemophilus ducreyi, the aetiological agent of chancroid, are prevalent in Nairobi, Kenya, where tetracyclines and sulphonamides are no longer very effective in the treatment of chancroid. The following regimens (given three times daily for seven days) were compared in a double-blind randomised

Summary

500 mg, amoxycillin 500 mg and clavulanic acid 125 mg, and amoxycillin 500 mg and clavulanic acid 250 mg. 68 of 100 ulcers were culture-positive for H. ducreyi. All strains of H. ducreyi produced &bgr;-lactamase. At day 7 none of the amoxycillin-treated patients had responded clinically or bacteriologically, whereas all but 2 of 56 patients treated with an amoxycillin/clavulanic-acid regimen had responded clinically and H. ducreyi had been eradicated from their ulcers. The combination of amoxycillin-clavulanic acid appears to be very effective for the treatment of chancroid. The results of this study accord with H. ducreyi as the primary pathogen of chancroid.

trial—amoxycillin

1982

chancroid in many parts of the world. However, they are not always effective and healing time may be delayed.3 Several other agents have been used with varying success, but only the tetracyclines have achieved wide acceptance as treatment for ambulant patients.4 However, treatment failure is common in Kenya and the tetracyclines offer little advantage over sulphonamides except in patients with sulphonamide allergy or sulphonamide-resistant H. ducreyi infection. Despite in vitro susceptibility,5,6 and reports of successful therapy,7 the penicillins have been little used in the therapy of chancroid. Many Nairobi isolates of H. ducreyi produce a (3-lactamase of the TEM-1 type8 mediated by small non-conjugative plasmids.9 Penicillins would, therefore, not be effective in the treatment of chancroid in Kenya if H. ducreyi is the aetiological agent. However, synergistic antibacterial activity of amoxycillin and clavulanic acid, a potent (3-lactamase inhibitor, has been reported for H. ducreyi" and this combination could prove useful for the treatment of chancroid. The present study was undertaken to determine the efficacy of combinations of amoxycillin-clavulanic acid in the therapy of genital ulcer disease in a setting in which many H. ducreyi isolated produce &bgr;-lactamase. We also planned to test the hypothesis that the clinical response of chancroid to treatment would correlate both with bacteriological eradication and with the antimicrobial susceptibility of the -

organism. Materials and Methods

Patients Introduction

CHANCROID, seldom

of the "minor venereal diseases" is in industrialised countries, whereas it is common sexually transmitted disease in the one

recognised

reported to be a tropics. In Kenya most genital ulcers are clinically categorised as chancroid.Although Haemophilus ducreyi has been isolated from some genital ulcers since 1889,’ doubts continue to be expressed about the role of this organism as a primary pathogen. We have isolated H. ducreyi from over 70% of genital ulcers clinically suspected to be chancroid in Kenya. A correlation between the clinical response of genital ulcers to treatment and bacteriological eradication of H. ducreyi by antimicrobial therapy directed against this organism would accord with the role of H. ducreyi as a primary pathogen in chancroid. Sulphonamides are routinely used as first line therapy of

From January, 1981, to June, 1981, 100 men presenting to the special treatment clinic in Nairobi, Kenya, with genital ulcers were enrolled in the study. Informed verbal consent was obtained before we allocated patients to a therapeutic regimen. A standardised history was obtained and patients were examined for evidence of genital ulcers, inguinal adenopathy, urethral discharge, and other relevant physical findings. Patients reporting penicillin allergy or who had received antibiotics in the week before presentation were excluded.

Laboratory Studies Laboratory techniques to identify Haemophilus ducreyi, and Treponema pallidum infection in this clinic population have been described elsewhere.’ Isolation of H. ducreyi was achieved by streaking exudate on the surface of enriched chocolate agar with vancomycin (3 mg/1) and 5% fetal calf serum. This medium was incubated in a candle jar at 35°C for 48 h and the culture was read 8297

510

daily for 7 days. H. ducreyi was identified on the basis of colony morphology, gram stain, and nutritional requirements (X and V factor).11 T. pallidum infections were diagnosed by positive darkfield and/or positive serology-i.e., positive rapid plasma reagin (RPR) in a dilution 1:2 in conjunction with a positive Treponema pallidum haemagglutination (TPHA).12

TABLE II-RESPONSE OF

68

MEN WITH HAEMOPHILUSDUCREYI

POSITIVE GENITAL ULCER DISEASE TO TREATMENT WITH THREE REGIMENS

Therapy and Follow-up The first 45 patients were randomly assigned to one of three 7-day drug regimens: amoxycillin 500 mg three times a day (A), amoxycillin 500 mg plus clavulanic acid 125 mg three times a day (AC375), or amoxycillin 500 mg plus clavulanic acid 250 mg three times a day AC750). The investigators were unaware of the treatment regimen being given to individual patients. The patients were given enough drug for 4 days and asked to return on the 4th day for reexamination and culture for H. ducreyi. At this second visit they were given enough drug to complete the 7 days of therapy. Patients were also asked to return on days 7, 10, 14, and 28 for further evaluation. If on day 7 the ulcer was healed or improved (decrease in size and improvement in symptoms), no further antimicrobial treatment was given. If the ulcers were unchanged or worse (increase in size and persistence of symptoms) treatment was deemed to have failed, the code was broken, and alternative therapy was prescribed (in 7 patients AC375 was prescribed). Serological tests for syphilis were repeated on day 7. On days 10 and 14 the ulcer was re-evaluated and re-cultured if necessary and on day 28 the final serological tests for syphilis were repeated. After 45 patients had been enrolled in the study it became apparent that patients receiving amoxycillin alone failed to improve. This treatment arm was discontinued and the final 55 patients were randomly allocated to treatment with either AC375 or AC750 and followed in an identical manner. All isolates ofH. ducreyi were tested for 0-lactamase production by the chromogenic method.13 Susceptibility of H. isolates to clavulanic acid, and to a combination ducreyi amoxycillin, of amoxycillin and clavulanic acid in a fixed ratio of 2:1 were performed by an agar dilution technique.1O analysis of strains was performed as described elsewhere.9

cephalosporin

Plasmid

Results H. ducreyi was isolated from the ulcers of 68 patients (table I) and all the isolates were /3-lactamase positive. The ulcer in one of these patients was darkfield positive and one further patient had a positive RPR and TPHA. These two patients were treated for syphilis with a long-acting penicillin after completing the study drug and when ulcers had healed. TABLE I-CLINICAL DATA ON

100

MEN WITH GENITAL ULCERS

*Includes 7 patients who initially did not respond clinically or bacteriologically to amoxycilhn alone and were treated with AC375 without randomisation. TABLE III-RESPONSE OF

27

MEN WITH GENITAL ULCER DISEASE OF

UNKNOWN AETIOLOGY TO THE THREE TREATMENT REGIMENS

The response to therapy of patients with H. ducreyi positive ulcers is outlined in table II. Of the first 45 patients 27 had positive cultures for H. ducreyi and 18 were adequately followed. In 9 of these treatment was declared a failure at day 7 because of continuing positive cultures on the 4th day of therapy and no clinical improvement after one week of therapy. When the code was broken, all 9 patients had been randomly allocated to receive amoxycillin alone. Of the 9 patients who were treated with one of the amoxycillin/clavulanic acid regimens, all were culturenegative for H. ducreyi after one week of therapy and had clinically responded. 26 patients with H. ducreyi positive ulcers were randomised to AC375 and an additional 7 amoxycillin failures were treated with AC375. All 28 patients evaluated at the end of the week of therapy were clinically improving or had ulcers that had healed completely. However, 1 of these patients had a positive culture for H. ducreyi despite clinical improvement and this patient was treated with an alternative drug. 30 men with H. ducreyi-positive ulcers were randomised to AC750. All but 1 of the 27 patients evaluated at day 7 was either cured or improving. Treatment was deemed to have failed in this patient; however, therapeutic compliance was

doubtful. The interval from the onset of therapy to complete ulcer healing ranged from 3 to 28 days with a mean time of 12 days in the patients treated with AC375 and 10-4 in patients treated with AC750. 53% of the patients with H. ducreyi-positive ulcers had tender inguinal adenopathy. 6 patients with both ulcers and buboes treated with amoxycillin alone did not improve during the week of therapy. However, none of the buboes progressed to suppuration. On both AC regimens the adenopathy gradually resolved in 3 to 28 days. In most

patients inguinal lymphadenitis responded more slowly to therapy than did the ulcers. 4 patients with H. ducreyi positive ulcers had fluctuant buboes which were aspirated. 1 was fluctuant at the time of initial presentation and the remaining 3 became fluctuant while the patient was on therapy. H. ducreyi was not isolated from these aspirates.

511 In 32 of the 100 men enrolled in the isolated from the ulcers. 5 were

study, H. ducreyi was diagnosed as having were darkfield positive and 3 had positive syphillis (2 All received for appropriate penicillin serology syphillis). therapy and are not considered further here. Table III shows the results of treatment in the remaining 27 men. 2 received amoxycillin and were cured. 12 men were treated with AC375 and 13 with AC750. All the patients who returned for follow-up had resolved or were improving at the end of one week of therapy. The mean interval to complete healing after onset of treatment was 6 - 5 and 7.5 5 days, respectively, in these two groups. It is presumed that many patients’with darkfield negative H. ducreyi negative ulcers may have been infected with H. ducreyi. The sensitivity of the present culture techniques is probably not more than 70%. Further studies are needed to determine the aetiological agents responsible for these "culture negative" ulcers. A total of 50 strains were available for investigation of antibacterial resistance. All produced (3-lactamase and all contained either a 5 -7or a 7 - 0 megadalton plasmid.9 6 patients, 3 on each AC regimen noted nausea but no other adverse effects were reported. not

Discussion

During 1980, studies performed on 25 H. ducreyi strains from Nairobi showed that 19 produced &bgr;-Iactamase. We had intended to compare the efficacy of amoxycillin alone in patients infected with (3-lactamase-positive and (3-lactamasenegative H. ducreyi, thus testing the hypothesis that the antimicrobial susceptibility of H. ducreyi from genital ulcers would correlate closely with the clinical response of the ulcer, with amoxycillin as a therapeutic probe. However, by 1981, all isolates of H. ducreyi in this study produced (3-lactamase and we are unable fully to answer this question. Resistance to the penicillins was mediated by one of two plasmidsl4 with the assistance of a mobilising plasmid. 15 There were no favourable clinical responses to amoxycillin among patients with H. ducreyi positive ulcers and 18 of 20 cultures taken on the 4th or 7th day from the ulcers of patients treated with amoxycillin alone grew H. ducreyi. One would expect that amoxycillin alone would have been effective if patients infected with (3-lactamase-negative H. ducreyi strains had been treated, as clavulanic acid has little in vitro antibacterial action against H. ducreyi" In vitro it facilitates the antibacterial action of amoxycillin, presumably through inhibition offl-lactamase. In this study 53 of 55 patients had a rapid response to treatment with improvement or complete resolution by day 7 and eradication of H. ducreyi from the ulcer in all but one patient. The correlation between the eradication ofH. ducreyi and ulcer healing is evident. Of the 8 patients with positive cultures at the end of one week of therapy, 7 showed no evidence of clinical improvement. On the other hand, 8 of the 10 ulcers that were categorised as clinical failures were also bacteriological failures. These correlations are further evidence for the primary pathogenic role of H. ducreyi in genital ulcer disease. Ulcers from which H. ducreyi could not be cultured responded well to therapy. Although no microbial pathogens can

be

implicated

in these ulcers,

they

share clinical

characteristics with H. ducreyi-positive ulcers. It is possible they may be due to strains of H. ducreyi which to date have eluded detection because of special nutritional requirements. It is also possible that other as yet unidentified microbial agents are responsible for these ulcers.

The efficacy of antimicrobial therapy on the inguinal adenitis of chancroid has been equivocal and buboes have been reported to develop during antimicrobial therapy despite ulcer healing.16 In the present study, all buboes resolved within four weeks with the amoxycillin/clavulanic acid regimen. The pathogenesis of suppurating buboes in chancroid is uncertain and H. ducreyi is rarely cultured from the aspirate. Very few studies have demonstrated the efficacy of clavulanic acid outside the urinary tract. 11,18 H. ducreyi is relatively resistant to clavulanic acid with median minimum inhibitory concentrations (MIC) of 32 mg/1 or more, far in excess of expected serum or tissue levels of 1-2 mg/1. In this study 3-lactamase positive H. ducreyi provide an excellent therapeutic model for the evaluation of clavulanic acid in soft tissue infection. Whereas, amoxycillin alone is unable to eradicate H. ducreyi from ulcers or cure the patient, the addition of clavulanic acid facilitated the antibacterial action of amoxycillin presumably through blockade of &bgr;-lactamase. Clavulanic acid was equally effective in the two doses tested. It is currently marketed as a fixed combination of amoxycillin 250 mg and clavulanic acid 125 mg in each capsule. The present study shows that this combination is effective in a soft tissue infection due to invasive (3-lactamase-producing organisms, and further therapeutic trials with soft tissue infections are warranted. This study was supported by grants from Beecham Pharmaceuticals, U.K., World Health Organization, Canadian international Development Agency, and the Medical Research Council of Canada. M. V. F. was the recipient of the Sidney Israels Fellowship. F. A. P. was the recipient ofa fellowship from the Medical Research Council of Canada.

Correspondence should be addressed to A. R. R., Department of Medical Microbiology, University of Manitoba, Winnipeg, Manitoba, Canada RE3 OW3.

REFERENCES 1. Nsanze

H, Fast M, D’Costa LJ, et al Genital ulcers in Kenya. A clinical and laboratory study of 97 patients Br J Vener Dis 1981; 57: 378-81. 2 Ducreyi A. Experimentelle untersuchungen über den ansteckunsstoff des weichen schnakers and über die bubonen. Montash F Prakt Dermatol 1889; 9: 387-92. 3. Fast MV, Nsanze H, D’Costa LJ, Karasira P, Maclean I, Ronald AR. Antimicrobial therapy of chancroid: An evaluation of five treatment regimens correlated with in vitro sensitivity. Sex Trans Dis (in press). 4. Asin J. Chancroid: A report of 1402 cases. Am J Syph Gonorrhea Vener Dis 1952; 36: 483-87 5. Reymann F. Sensitivity of Hemophilus ducreyi to penicillin, streptomycin, and sulfathiazole. Acta Pathol Microbiol Scand 1949; 26: 309-18. 6 Singer S, Deacon WE. Effects of penicillin G in vitro on Hemophilus ducreyi. Public Health Rep 1956; 71: 1112-14. 7. Hammond GW, Chang JL, Wilt JC, Ronald AR. Antimicrobial susceptibility of Haemophilus ducreyi Antimicrob Agents Chemother 1978; 13: 608-12. 8. Maclean I, Bowden GHW, Albritton WL TEM-type &bgr;-lactamase production in Haemophilus ducreyi Antimicrob Agents Chemother 1980; 17: 897-900. 9. Brunton JL, Maclean I, Ronald AR, Albritton WL. Plasmid-mediated ampicillin resistance in Haemophilus ducreyi. Antimicrob Agents Chemother 1979; 15: 294-99. 10. Girouard YC, MacLean IW, Ronald AR, Albritton WL. Synergistic antibacterial activity of clavulanic acid and amoxicillin against &bgr;-lactamase producing strains of Haemophilus ducreyi. Antimicrob Agents Chemother 1981, 20: 144-45. 11 Hammond GW, Lian GJ, Wilt JC, Albritton WL, Ronald AR. Determination of the hemin requirement of Haemophilus ducreyi: Evaluation of the porphyrin test and media used in the satellite growth test. J Clin Microbiol 1978; 7: 243-46 12. Ursi JP, Van Dyck E, Van Houlte C, Plot P, Colaert J, Diamini M, Meheus A. Syphilis in Swaziland. A serological survey. Br J Vener Dis 1981; 57: 96-99 13 O’Callaghan CH, Morris A, Kirby SM, Shingler SH. Novel method for detection of &bgr;-lactamases by using a chromogenic cephalosporin substrate Antimicrob Agents Chemother 1972, 1: 283. 14. Brunton J, Meier M, Ehrman N, Maclean I, Slaney L, Albritton WL. Molecular epidemiology of beta-lactamase-specifying plasmids of Haemophilus ducreyi. Antimicrob Agents Chemother 1982; 21: 857-63 15. Deneer HG, Slaney L, Maclean IW, Albritton WL Mobilization of nonconjugative antibiotic resistance plasmids in Haemophilus ducreyi. J Bacteriol 1982, 149: 726-32. 16 Marmar JL. The management of resistant chancroid in Vietnam J Urol 1972; 107: 807-08. 17. Leigh DA, Bradnock R, Marriner JM Augmentin therapy in complicated infections due to &bgr;-lactamase producing bacteria. J Antimicrob Chemother 1981; 7: 229-36. 18 Ball P, Watson T, Mehtar S Amoxicillin and clavulanic acid in intra-abdominal and pelvic sepsis. J Antibicrob Chemother 1981; 7: 441-44