Amoxicillin–clavulanic acid: Clinical Uses and Toxicity

Clavulanic acid has no effect on the activity of amoxicillin against non-beta-lactamase-producing bacteria, which are normally sensitive to amoxicillin (Slocombe et al., 1984). Similarly, ampicillin-susceptible strains would, with rare exceptions, also be susceptible to coamoxiclav. Brumfitt et al. (1983) reported a phenomenon of in vitro ampicillin-susceptible, co-amoxiclav-resistant strains of E. cloacae, C. freundii, and S. marcescens, which was attributed to a greater activity of ampicillin than amoxicillin against these isolates. This is unlikely to be clinically significant as ampicillin would not generally be considered to have activity against these bacteria. 

CLINICAL USES OF THE DRUG

a. Urinary tract infections

Co-amoxiclav is widely used with good cure rates to treat urinary tract infections caused by both beta-lactamase-producing and nonproducing organisms, although in the latter case there is no advantage over amoxicillin alone. Early noncomparative trials of co-amoxiclav generally used 250/125 mg 8-hourly for the treatment of amoxicillin-resistant urinary tract infections and cure rates varied from 64–74% in the elderly or complicated patient to 87–92% for uncomplicated cystitis in young women (Leigh et al., 1981; Ball, 2007). In a small comparative study versus amoxicillin, co-amoxiclav had superior efficacy (absence of bacteriuria within 7 days; 85% vs 25%) for amoxicillin-resistant organisms (Martinelli et al., 1981). In a group of elderly patients, Gallacher et al. (1986) also showed superiority of coamoxiclav over amoxicillin alone (88% vs 43% response, respectively).

Doses of 250/125 mg 8-hourly for 7–10 days were used in larger open-label studies (Iravani and Richard, 1986), with 92–97% early cure rates and 69–97% cure at late follow-up. Co-amoxiclav has efficacy in patients with recurrent urinary tract infections (Brumfitt and Hamilton-Miller, 1984), and for this indication was similar to cephradine in one trial (Brumfitt and Hamilton-Miller, 1990). Comparative studies indicated that co-amoxiclav has similar efficacy to oral cefaclor (Gurwith et al., 1983; Iravani and Richard, 1986) and is similarly as effective as cephalexin for the treatment of bacteriuria during pregnancy (Pedler and Bint, 1985). There is conflicting evidence on the efficacy of co-amoxiclav compared with cotrimoxazole for urinary tract infection. In a study of 52 adults with uncomplicated infection, cure rates were 83% and 100%, respectively (Bailey et al., 1983). However, in a similar-sized study, Fancourt et al. (1984) found marginal superiority of co-amoxiclav, and in a larger study Karachalios (1985) observed higher bacteriologic cure rates with co-amoxiclav (95% vs 83%). In a large open-label study including complicated urinary tract infection, co-amoxiclav was unexpectedly effective against infections caused by S. marcescens, E. cloacae, and C. freundii, which produce beta-lactamases not inhibited by clavulanic acid (Nakazawa et al., 1983). Eradication of bacteria was noted in 63– 89%, and this was attributed to high urinary concentrations of clavulanic acid, which exceeded the MICs for these strains.

b. Respiratory tract infections

While penicillin V is used to treat pharyngitis caused by group A streptococci (GAS) with good clinical success, in many patients eradication of streptococci from the pharynx is not achieved. In contrast, high rates of GAS eradication have been reported with coamoxiclav (Kaplan and Johnson, 1988; Brook, 1989). It has been postulated that beta-lactamases (produced by normal upper respiratory tract flora) inactivate penicillin, leading to persistence of GAS, although this explanation remains controversial and evidence for it is lacking (Gerber et al., 1999). The use of co-amoxiclav is not recommended routinely for GAS pharyngitis and would also be associated (as is amoxicillin) with a high incidence of rash in patients acutely infected with Epstein–Barr virus.

Co-amoxiclav has good activity against the most common bacterial pathogens associated with upper and lower respiratory tract infections (see Table 14.1) including S. pneumoniae, H. influenzae, and M. catarrhalis. These are associated with sinusitis, otitis media, exacerbations of chronic obstructive pulmonary disease (COPD), and community-acquired pneumonia. Co-amoxiclav has the advantage over amoxicillin alone in retaining activity against the 15–20% of H. influenzae and W90% of M. catarrhalis that are beta-lactamase producing (Jacobs et al., 2003).

c. Mixed Gram-negative, Gram-positive, and anaerobic infections

In animal studies, co-amoxiclav is effective for the treatment of mixed B. fragilis and E. coli infection (Beale et al., 1988). Where the i.v. preparation of co-amoxiclav is available, this can be used to treat moderate to severe infections in which mixed aerobic and anaerobic pathogens are likely, including lung abscess, empyema, peritonitis, intra-abdominal collections and postoperative obstetric and gynecologic infections (Mehtar and Ball, 1985; Obwegeser et al., 1989; Schmitt et al., 1989; Walker et al., 1993), and animal bite-associated infections and diabetic foot infections (see later under 7d. Skin and soft-tissue infections). Oral co-amoxiclav can also be used following initial i.v. treatment with either co-amoxiclav or other regimens for the above infections.

d. Skin and soft-tissue infections

In experimental surgical wounds in mice infected with both S. pyogenes and S. aureus, amoxicillin alone did not eliminate S. pyogenes in the presence of beta-lactamase-producing S. aureus, whereas co-amoxiclav eliminated both organisms (Boon and Beale, 1987). Abscesses, cellulitis, and impetigo are most commonly caused by S. pyogenes and S. aureus, and co-amoxiclav has efficacy for these conditions, similar to other agents such as flucloxacillin and cefaclor (Fleisher et al., 1983; Pien, 1983; Dagan and Bar-David, 1989; Gentry, 1992). However, if the pathogen has been identified, narrower-spectrum treatment (e.g. penicillin for S. pyogenes and flucloxacillin for S. aureus) would be preferred. With the emergence of community-acquired MRSA, antimicrobials active against this pathogen should be considered for empiric treatment of more severe skin and soft-tissue infections. For more complicated skin and soft-tissue infections, such as diabetic foot infections and postoperative abdominal and gynaecologic surgical infections in which mixed anaerobic and anaerobic flora is present, co-amoxiclav is an appropriate treatment option, so long as P. aeruginosa is not present (Lipsky et al., 2007).

e. Animal and human bites

The microbiology of infected animal bites is complex, and many infected wounds are polymicrobial, with a median of five isolates per culture (Talan et al., 1999). The most common isolate is Pasteurella spp., with other common isolates including staphylococci, streptococci, Moraxella, Neisseria, and anaerobic organisms (Fusobacterium, Bacteroides, Porphyromonas, and Prevotella spp.) (Talan et al., 1999).

Co-amoxiclav has good activity against these likely pathogens and is an appropriate first choice for infected wounds. Co-amoxiclav is also widely used for prophylaxis after animal bites (Morgan and Palmer, 2007), although there is limited evidence for efficacy.

f. Surgical chemoprophylaxis

In Europe, where the intravenous formulation is marketed, coamoxiclav is used for surgical prophylaxis, particularly in gynecologic and intestinal procedures (van Kasteren et al., 2003). Co-amoxiclav has equal efficacy to gentamicin and metronidazole for prophylaxis in elective colorectal surgery (Hall et al., 1989). Co-amoxiclav was more effective than metronidazole alone as prophylaxis in hysterectomy (Brown et al., 1988), and had similar efficacy to cefazolin for prophylaxis (both as a single dose) in laparoscopic gynecologic surgery (Cormio et al., 2003) and vaginal gynecologic surgery (Cormio et al., 2007).

g. Chemotherapy-related fever and neutropenia

Co-amoxiclav has been used in combination with ciprofloxacin as an oral outpatient regimen for patients with fever and neutropenia following chemotherapy considered at ‘‘low risk’’ (expected duration of neutropenia o10 days, no serious coexisting medical conditions, no hemodynamic instability or abdominal symptoms). In these patients oral ciprofloxacin and co-amoxiclav (500/125 mg 8-hourly) had similar efficacy to conventional intravenous regimens such as ceftazidine or ceftriaxone plus amikacin (Freifeld et al., 1999; Kern et al., 1999).

h. Melioidosis

Burkholderia pseudomallei produces a class A (Bush 2e) beta-lactamase which is inhibited by clavulanic acid (Cheung et al., 2002).

Intravenous co-amoxiclav has been used for treatment of severe melioidosis in Thailand. Mortality was similar to the comparative agent ceftazidime, but more patients in the co-amoxiclav group required crossover to the comparator regimen (Suputtamongkol et al., 1994). Initial uncontrolled studies suggested efficacy of oral co-amoxiclav for the prolonged eradication phase of melioidosis treatment, with resolution of disease in 67% and good tolerability (Suputtamongkol et al., 1991).

i. Mycobacterial infections

There are sporadic reports of the use of co-amoxiclav for drug-resistant tuberculosis (Nadler et al., 1991; Yew et al., 1995). Chambers et al. (1998) observed a quantitative reduction in M. tuberculosis sputum load of 0.32 log10 CFU/ml/day during the first 2 days of co-amoxiclav therapy (compared with 0.6 log10 CFU/ml/day reduction with isoniazid). In contrast, Martiz et al. (2001) found no evidence of early bactericidal activity. It would appear that co-amoxiclav has a limited role in the management of tuberculosis; a recent meta-analysis of the treatment of multidrug-resistant tuberculosis (MDR-TB) suggested the possible addition of co-amoxiclav (or other agents) if there were not five adequate medications in the treatment regimen for MDR-TB (Mukherjee et al., 2004).

j. Nocardia infections

Co-amoxiclav has been reported to be effective in some cases of nocardiosis (attributable to N. asteroides) in renal transplant patients (Arduino et al., 1993). There have, however, been major advances in the taxonomy of Nocardia spp. since that time, including subdivision of N., asteroides, and susceptibility to co-amoxiclav varies (Wallace et al., 1988). Nocardia abscessus (formerly N. asteroides type I drug sensitivity pattern), N. farcinica, and N. brasiliensis are usually susceptible to coamoxiclav, and it is a second-line oral agent (after co-trimoxazole) for the last species (Wallace et al., 1987; Gomez-Flores et al., 2004). Other species, including N. nova complex, N. otitidiscaviarum, N. asteroides (type VI drug sensitivity pattern), and N. cyriacigeorgica, are resistant to co-amoxiclav (Brown-Elliott et al., 2006). Co-amoxiclav may be an option for treatment of selected Nocardia infections, but would need supportive laboratory susceptibility and identification data.

k. Gonorrhoea

Oral co-amoxiclav (single dose of 3 g/125 mg with 1 g probenecid or 3 g/250 mg) is effective for gonorrhoea caused by penicillin-sensitive or beta-lactamase-producing N. gonorrhoeae (Latif et al., 1984; Key et al., 1985; Lawrence and Shanson, 1985). However, resistance is also mediated by chromosomal resistance. These isolates increased in frequency in the USA from 0.5% in 1988 to 5.7% in 1999, but have decreased to a prevalence of 1.2% in 2006 (Centers for Disease Control and Prevention; 2006). Marked regional and country differences exist in resistance, however, and current guidelines recommend ceftriaxone or cefixime as the preferred treatment (Centers for Disease Control and Prevention, 2007).

l. Chancroid

Multiresistant strains of H. ducreyi are now prevalent in many parts of the world and this pathogen has become resistant to the traditionally used drugs such as tetracycline and sulfonamides; and most strains are beta-lactamase positive. In a study from Kenya, where resistance is common, 54 of 56 patients given a 7-day course of co-amoxiclav had a clinical and bacteriologic response (Fast et al., 1982). A subsequent study found that a 3-day course of co-amoxiclav 500/250 mg 8-hourly was effective, but shorter courses or single doses of 3000/350 mg were ineffective (Ndinya-Achola et al., 1986).

m. Bacterial vaginosis

Co-amoxiclav is more effective than amoxicillin for the treatment of bacterial vaginosis, consistent with the role of multiple anaerobic species in this condition (Symonds and Biswas, 1986). However, it is less effective than metronidazole (Van der Meijden et al., 1987). Although it could be used in metronidazole-refractory cases, oral and topical clindamycin is preferred as second-line therapy.

TOXICITY

Co-amoxiclav is relatively well tolerated. Adverse reactions observed in 9700 patients participating in clinical trials with oral co-amoxiclav included diarrhea, 398 (4.1%); nausea, 294 (3%); vomiting, 175 (1.8%); indigestion, 158 (1.6%); rash, 110 (1.1%); urticaria, 9; anaphylaxis, 1; Candida superinfection, 98 (1%); altered liver function tests, 3; and jaundice, 1 (Croydon, 1984).

The majority of the safety data for co-amoxiclav have been obtained with the earlier formulations, mostly given three times daily. The 875/ 125 mg twice-daily regimen appears to have a similar safety profile to the 500/125 mg three times daily dosing (White et al., 2004). A lower incidence of diarrhea with the former dosing (2.9% vs 4.9%; Calver et al., 1997) likely relates to the lower clavulanic acid daily dose. The new ‘‘pharmacokinetically enhanced’’ 2000/125 mg twice-daily formulation appears to have similar adverse events to the 875/125 mg twice-daily preparations (White et al., 2004).

a. Gastrointestinal side-effects

In an analysis of 38,500 patients in clinical trials a similar range of sideeffects occurred, with upper gastrointestinal events (2.5%), nausea (1.4%), and diarrhea (3.4%) the most common (Neu et al., 1993). Diarrhea has been reported more commonly (E10%) by others, and is attributed primarily to the clavulanic acid component (Bush and Johnson, 2000), as diarrhea and other gastrointestinal side-effects such as nausea and vomiting seem to be more common with co-amoxiclav than with amoxicillin alone (Iravani and Richard, 1982; Pien, 1983; Conner, 1985). With the twice-daily dosing schedules and 125-mg individual doses of clavulanic acid (compared with 250-mg doses) gastrointestinal toxicity is less (Crokaert et al., 1982; Lawrence and Shanson, 1985). Administration with food may reduce the severity of gastrointestinal adverse events (Bush and Johnson, 2000). Antibioticassociated diarrhea associated with C. difficile can also occur and ranges from mild diarrhea to fulminant pseudomembranous colitis.

b. Rashes

The incidence of rashes (1.1%) noted by Croydon (1984) was surprisingly low because amoxicillin alone is associated with a higher frequency of rashes. A higher frequency of rashes was noted in one study in which 116 females were given co-amoxiclav – rashes were observed in 4.1% (Iravani and Richard, 1982). Rashes and fever also occurred more frequently in HIV-infected patients given co-amoxiclav (Van der Ven et al., 1994). Other rare dermatologic adverse events include linear IgA bullous disease (Ho et al., 2007) and reactions also observed with ampicillin-class antibiotics including erythema multiforme, exfoliative dermatitis, pruritus, and Stevens–Johnson syndrome.

c. Hepatotoxicity

There are a number of reports of co-amoxiclav-induced cholestatic jaundice (Thomson et al., 1995), including with a fatal outcome (Hebbard et al., 1992; Ersoz et al., 2001). Risk factors include male sex, higher age, and longer treatment duration (Cundiff and Joe, 2007), although cases have also been reported in children, including a vanishing bile duct syndrome (Smith et al., 2005). The overall estimated risk of hepatoxicity is 1:100,000. Hepatotoxicity is a hypersensitivity, usually cholestatic, reaction, which can develop weeks after stopping the drug and recovery over weeks to months generally occurs (Cundiff and Joe, 2007).

d. Other adverse reactions

Other uncommon to rare adverse events include interstitial nephritis, eosinophilia, leucopenia, thrombocytopenia, agitation, convulsions, insomnia and hyperactivity, and anaphylactoid and hypersensitivity reactions (see Chapter 3, Ampicillin, Amoxicillin and Other Ampicillin-Like Penicillins).

e. Risks in pregnancy

Co-amoxiclav is pregnancy category B. There is no evidence of teratogenic effects of co-amoxiclav in animals (Baldwin et al., 1983). Both amoxicillin and clavulanic acid cross the placenta (Matsuda et al., 1992). Czeizel et al. (2001) performed a population-based case– control study of 6935 pregnant women with offspring with congenital abnormalities and 10,238 controls and found no significant differences in co-amoxiclav exposure during pregnancy, including exposure during the first trimester. In another study, 191 women treated with coamoxiclav during the first trimester were prospectively matched with controls and following during pregnancy. Rates of live births, abortions, and major malformations were similar in the two groups (Berkovitch et al., 2004). These studies are reassuring regarding the safety of co-amoxiclav in pregnancy, but of course do not exclude the possibility of infrequent toxicity or malformations.

References

Cundiff J, Joe S (2007). Amoxicillin-clavulanic acid-induced hepatitis. Am . Otolaryngol 28: 28.
Cynamon MH, Palmer GS (1983). In vitro activity of amoxicillin in combination with clavulanic acid against Mycobacterium tuberculosis. Antimicrob Agents Chemother 24: 429.
Czeizel AE, Rockenbauer M, Sorensen HT, Olsen J (2001). Augmentin treatment during pregnancy and the prevalence of congenital abnormalities. A population-based case–control teratologic study. Eur J Obstet Gynecol Reprod Biol 97: 188.
Dagan R, Bar-David Y (1989). Comparison of amoxicillin and clavulanic acid (Augmentin) for the treatment of non-bullous impetigo. Am J Dis Child 143: 916.
Dambro N, FSriedman AD, Alexander ER, Harrison HR (1984). Augmentin therapy for urinary tract infections in children. Postgraduate Medicine, Progress and Perspectives on Beta-Lactamase Inhibition: A Review of Augmentin, p. 263. New York: Custom Communications.
Damrikarnlert L, Jauregui AC, Kzadri M (2000). Efficacy and safety of amoxicillin/clavulanate (Augmentin) twice daily versus three times daily in the treatment of acute otitis media in children. The Augmentin 454 Study Group. J Chemother 12: 79.
Ersoz G, Karazu Z, Yildiz C, Akarca US et al. (2001). Severe toxic hepatitis associated with amoxicillin and clavulanic acid. J Clin Pharm Therap 26: 225.
Evans CM, Purohit S, Colbert JW et al. (1988). amoxicillin–clavulanic acid (Augmentin) antibiotic prophylaxis against wound infections in renal failure patients. J Antimicrob Chemother 22: 363.
Fancourt GJ, Flavell Matts SG, Mitchell CJ (1984). Augmentin (amoxicillin– clavulanic acid) compared with co-trimoxazole in urinary tract infections.
Br Med J 289: 82.
Farmer T, Reading C (1982). Beta-lactamases of Branhamella catarrhalis and their inhibition by clavulanic acid. Antimicrob Agents Chemother 21: 506.
Gerber MA, Tanz RR, Kabat W (1999). Potential mechanisms for failure to eradicate group A streptococci from the pharynx. Pediatrics 104: 911.
Girouard YC, Maclean IW, Ronald AR, Albritton WL (1981). Synergistic antibacterial activity of clavulanic acid and amoxicillin against betalactamase-producing strains of Haemophilus ducreyi. Antimicrob Agents Chemother 20: 144.
Goldstein EJC (1992). Bite wounds and infection. Clin Infect Dis 14: 633.
Goldstein EJC, Citron DM (1988). Comparative activities of cefuroxime, amoxicillin-clavulanic acid, ciprofloxacin, enoxacin, and ofloxacin against aerobic and anaerobic bacteria isolated from bite wounds. Antimicrob Agents Chemother 32: 1143.

You may like