1. Therapy with Combined Antimicrobial Agents
a) additive or supraadditive toxicities
vancomycin and tobramycin - given separately - minimal toxicity
in combination - marked impairment of renal function
bacteriostatic antibiotics (e.g., tetracyclines, erythromycin, chloramphenicol) frequently antagonize the action of a bactericidal drug (e.g., b-lactam antibiotics, vancomycin, aminoglycosides) - because bacteriostatic antibiotics inhibit cell division and protein synthesis, processes that are required for the bactericidal effect of most bactericidal agents
streptomycin + penicillin - synergism
rifampin combinations appear to be an exception to this general rule in vitro - indifferent or antagonistic effect combined with other bactericidal
rifampin combinations -- effective clinically
trimethoprim and sulfamethoxazole synergism - due to sequential inhibition of two steps in the pathway of biosynthesis of tetrafolate
Treatment of Mixed Bacterial Infections
intraabdominal, hepatic, and brain abscesses and some genital tract infections
perforation of colon - contamination and infection with aerobic Enterobacteriaceae, anaerobic and aerobic gram-positive cocci (streptococci), anaerobic bacilli such as Bacteroides fragilis, and anaerobic gram-positive rods such as Clostridium species aminoglycoside for the Enterobacteriaceae
anaerobic microorganisms clindamycin, metronidazole, cefotetan, ceftizoxime, ticarcillin-clavulanate, imipenem-cilastatin
Therapy of Severe Infections in Which a Specific Cause Is Unknown. Combination chemotherapy is used probably most frequently in the empirical therapy of infections in which the causative agent has not been or cannot be identified. In these situations, the goal of treatment is to select antibiotic "coverage" for microorganisms that are most likely involved. Selection of antimicrobials must be based on the physician's clinical judgment, which reflects a knowledge of the signs and symptoms of the various infectious diseases, the microbiology of these diseases, and the antibiotic spectrum of available drugs. The breadth of the antibiotic coverage that is selected for treatment is narrowed by the ability to reduce the list of potential infectious agents, but the severity of illness also must be taken into account. A physician may want to continue broader coverage if the infectious agent is still in doubt and the infection is severe.
Prolonged administration of broad-spectrum or multiple antibiotics may lead to overuse of toxic and expensive drugs. This problem most often arises when the physician fails to obtain adequate cultures prior to the initiation of therapy or fails to discontinue the combination chemotherapy after identifying the microorganism and determining sensitivities. There is an understandable reluctance to change antimicrobial agents when a favorable clinical response has occurred. However, the goal of chemotherapy always should be to use the most selectively active drug that produces the fewest adverse effects.
Enhancement of Antibacterial Activity in the Treatment of Specific Infections. As mentioned above, when two antimicrobial agents are administered together, they may produce a synergistic effect. There are specific clinical indications for the use of combinations of antimicrobial agents, and they are based on documented proof of efficacy (Sande and Scheld, 1980).
Perhaps the best-documented example of the utility of a synergistic combination of antimicrobial agents is in the treatment of endocarditis. In vitro penicillin alone is bacteriostatic against most strains of E. faecalis, whereas a combination of penicillin and streptomycin or gentamicin is bactericidal. Treatment of enterococcal endocarditis with penicillin alone frequently results in relapses, while combination therapy is curative.
Penicillin and streptomycin or gentamicin also are synergistic in vitro against strains of viridans streptococci. In animal models of endocarditis, this combination produces more rapid eradication of bacteria from infected vegetations on heart valves than does penicillin G alone. Wilson and associates (1978) reported a 100% cure rate for patients with this condition who received a short course (2 weeks) of combination chemotherapy, a result identical with that reported for patients treated with penicillin G alone for 4 weeks (Bisno et al., 1989). Synergism in vitro and in experimental models in vivo by a combination of a penicillin and an aminoglycoside has also been demonstrated with Staph. aureus. Selected patients with tricuspid valve endocarditis caused by Staph. aureus can be treated successfully with nafcillin and a low dose of tobramycin or gentamicin administered for a total of 2 weeks, instead of the 4 to 6 weeks traditionally used to treat this disease (Chambers et al., 1988).
Synergistic antibiotic combinations have been recommended in the therapy of infections with Pseudomonas in neutropenic patients. In vitro, antipseudomonal penicillins plus an aminoglycoside are synergistic against most strains of Pseud. aeruginosa. Studies in animals support the superiority of the combination over either drug alone, and clinical studies suggest improved survival with the combination. Despite the fact that the microorganism is sensitive to gentamicin in vitro, administration of gentamicin alone frequently does not cure the infection and may even allow sustained bacteremia. The addition of an antipseudomonal b-lactam antibiotic such as ticarcillin markedly increases the cure rate, a phenomenon that correlates with a more rapid bactericidal effect in vitro. This success may be a reflection of the importance of the use of antibiotics that produce bactericidal effects rapidly when infection occurs in the neutropenic patient (Klastersky and Staquet, 1982).
Sulfonamides combined with trimethoprim are synergistic in vitro and are effective against infections caused by microorganisms that may be resistant to sulfonamides alone. A fixed combination of trimethoprim and sulfamethoxazole is effective treatment of recurrent urinary tract infections, Pneumocystis carinii pneumonia, typhoid fever, shigellosis, and certain infections due to ampicillin- resistant Haemophilus influenzae.
Combinations of an inhibitor of b-lactamase (clavulanate, sulbactam, or tazobactam), which has little or no intrinsic antimicrobial activity, and a b-lactam antibiotic that is susceptible to b-lactamase (amoxicillin, ampicillin, ticarcillin, or piperacillin) allow successful treatment of infections by microorganisms that produce b-lactamase. For example, infections caused by b-lactamase-producing H. influenzae may be treatable with ampicillin-sulbactam or amoxicillin-clavulanate. Thus, the utility of time-tested antibiotics may be restored in infections for which they had become ineffective.
Advances also have been made by combination of synergistic agents in the antimicrobial therapy of fungal infections. A combination of flucytosine and amphotericin B has been shown to be synergistic in vitro and in animal models of infection. In the therapy of cryptococcal meningitis, a combination of flucytosine and a low dose of amphotericin B for 6 weeks was as effective as therapy with a higher dose of amphotericin B for 10 weeks and caused less renal toxicity (Bennett et al., 1979).
Prevention of the Emergence of Resistant Microorganisms. The use of combinations of antimicrobial agents was first proposed as a method to prevent the emergence of resistant mutants during therapy. If spontaneous mutation were the predominant means by which microorganisms acquired resistance to antibiotics, combination chemotherapy would, in theory, be an effective means of prevention. For example, if the frequency of mutation for the acquisition of resistance to one drug is 10-7 and that for a second drug 10-6, the probability of independent mutation to resistance to both drugs in a single cell is the product of the two frequencies, 10-13. This makes the emergence of such mutant resistant strains statistically unlikely. This approach has been most extensively used in the treatment of tuberculosis, where the concomitant use of two or more appropriate agents strikingly reduces the development of drug resistance by the tubercle bacillus.
Disadvantages of Combinations of Antimicrobial Agents. It is important that physicians understand the potential negative aspects of the use of combinations of antimicrobial agents. The most obvious are the risk of toxicity from two or more agents, the selection of microorganisms that are resistant to antibiotics that may not have been necessary, and increased cost to the patient. In addition, as noted above, antagonism of antibacterial effect may result when bacteriostatic and bactericidal agents are given concurrently. The clinical significance of antibiotic antagonism is not fully understood. Although antagonism of one antibiotic by another has been a frequent observation in vitro, well-documented clinical examples of this phenomenon are relatively rare. The most notable of these involves the therapy of pneumococcal meningitis.
In 1951, Lepper and Dowling reported that the fatality rate among patients with pneumococcal meningitis who were treated with penicillin alone was 21%, while patients who received the combination of penicillin and chlortetracycline had a fatality rate of 79%. This conclusion was supported by Mathies and colleagues (1967), who treated children with bacterial meningitis of multiple causes either with ampicillin alone or with the combination of ampicillin, chloramphenicol, and streptomycin. The mortality rate among those treated with ampicillin was 4.3%, while that among those treated with the combination was significantly greater10.5%.
Antagonism between antibiotics is probably relatively unimportant in most infections. If an antagonistic interaction between two antibiotics is to occur, both agents must be active against the infecting microorganism. The addition of a bacteriostatic drug to a bactericidal drug frequently results in only a bacteriostatic effect. In many infections where host defenses are adequate, this is of no consequence. Where host defenses are impaired, as in patients with neutropenia, or with special infections, such as endocarditis and meningitis, achieving a bactericidal effect may be important. In clinical trials the more rapidly bactericidal combinations of antibiotics have been in general more effective than less rapidly bactericidal or purely bacteriostatic drugs in the therapy of gram-negative infections in neutropenic patients.