SPECIAL TESTS OF ANTIMICROBIAL ACTIVITY IN INFECTIOUS DISEASES
Serum Bactericidal Test
Determinations of Minimal Inhibitory and Bactericidal Concentrations
The optimal management of a bacterial infection requires the isolation and identification of a causative pathogen from a body site and demonstration of its susceptibility to antimicrobial agents by standard microbiologic testing. This methodology takes into account achievable serum levels of the antimicrobials and compares them with the amount of an agent determined to be necessary to inhibit the growth of the isolate. Generally, organisms are considered susceptible to antimicrobials when the antimicrobial concentration in serum is two to ten times greater than the amount needed for inhibition of growth. It has not been demonstrated that higher “therapeutic indices” improve clinical outcome. Selection of an antimicrobial is based on degree of activity, safety, cost, host factors, and pharmacokinetic considerations. The microbiologic systems that are commercially available work well for rapidly growing aerobic pathogens but have limitations for those that are more fastidious. In practice, decision making may be complicated by the inability to perform adequate susceptibility testing (e.g., anaerobes and other fastidious organisms), the need to employ an antimicrobial agent not available on the panels of automated systems, or the need to treat infection outside the bloodstream, where the concentration of antimicrobial may be substantially different (higher or lower) from the usual serum levels.
On occasion, more sophisticated antimicrobial testing procedures need to be employed. Those available in many clinical laboratories include measurement of antimicrobial blood levels, serum bactericidal testing (SBT), and the determination of minimum inhibitory concentration (MIC) of an antimicrobial. In general, these studies are not routinely indicated because they are labor-intensive and do not add to the efficacy of treatment; however, knowledge of their indications, benefits, and limitations can help the clinician manage antimicrobials in selected situations.
This test is used when the clinician wants to know the concentration of a specific agent in serum (or rarely other body fluids or tissues). Results are given in micrograms per milliliter (µg/mL) or micrograms per gram (µg/g). The peak level may be compared with the MIC of the organism to help determine “antibiotic efficacy” and toxicity; the trough level is used to guide additional dosing and also to help prevent toxicity. Knowledge of such levels is useful for antimicrobials with therapeutic serum levels close to those associated with toxicity and with pharmacokinetics that may be difficult to predict. In practice, serum levels are most useful for the management of aminoglycosides (especially gentamicin and tobramycin) and perhaps vancomycin. Determination of levels of aminoglycosides and vancomycin is especially useful in patients with rapidly changing renal function. Additionally, numerous other factors, including weight, hematocrit, and body temperature, may confound serum levels of aminoglycosides.
The value of measuring peak levels for aminoglycosides has been recently challenged with the documentation of efficacy and safety of once-daily dosing of aminoglycosides (see Chapter 75). With once-daily dosing, peak levels above 20 µg/mL are generally attained without enhanced toxicity. Thus, measurement of peak levels is rarely indicated. Alternatively, renal and auditory toxicity may occur when trough levels are above 2 µg/mL; these should be measured after the second or third daily dose. Current assay technology allows values to be returned expeditiously. For patients with significantly impaired or rapidly changing renal function, serum levels should be ordered more frequently, so that dosing can be individualized.
Use of serum levels to monitor vancomycin is controversial but still very common. As an example, a large university hospital performed more than 6,500 assays in a year, and estimates are that on a national basis more than 5 million tests are performed annually. This agent is less toxic than aminoglycosides and has a higher therapeutic index. In fact, whether vancomycin is associated with ototoxicity and nephrotoxicity at all is largely unknown! Recent reviews demonstrate a lack of correlation between serum levels of vancomycin and either toxicity or clinical success. With this in mind, most investigators consider that measurement of blood levels is generally not indicated. Situations in which they might be employed include the following: (a) patients with rapidly changing renal function, where use of nomograms is burdensome; (b) those receiving aminoglycosides (in which case potentially toxic synergy may occur); (c) patients receiving supernormal dosages; and (d) those on hemodialysis dialysis with high-flux dialysis membranes.
Determination of levels of antimicrobial in tissues and fluids other than blood is generally employed for investigational use but may occasionally be of clinical benefit. As an example, many antimicrobials are concentrated in urine and may have clinical efficacy therein despite apparent resistance by standard testing (which is based on serum levels). In unusual circumstances, a clinician might wish to document high urine concentrations of a selected antimicrobial and compare that value with the MIC for the organism. Alternatively, for abscesses that cannot be totally extirpated, documentation of significant levels of antimicrobial within the abscess fluid may confirm for the physician that the agent and dosage are correct.
Serum Bactericidal Test
The SBT was developed in 1912 and modified by Schlichter in 1947. In practice, the Schlicter test is an inhibitory rather than a bactericidal test. The SBT attempts to answer the physiologic question of what dilution of a patient’s serum (with accompanying antimicrobials, antibodies, and other inhibitory factors) is capable of killing the organisms causing infection in that person. Results are provided as a dilution of the patient’s serum (e.g., 1:4, 1:16) at which the tested organism is killed. The higher the dilution, the more active the serum. In this test, dilutions of the patient’s serum are prepared with liquid media with or without the addition of pooled human serum and are inoculated with the offending organism. After overnight growth, tubes demonstrating no visible growth (the one with the lowest dilution is the serum inhibitory concentration) are incubated on agar. The SBT result is then defined as the highest serum dilution at which more than 99.9% of the inoculum is killed. In theory, this test should be a useful gauge of antimicrobial treatment because it takes into account the susceptibility of the offending pathogen, the activity of the antimicrobial in conjunction with other inhibitory factors within human serum, and the pharmacokinetics of the product(s). Additionally, it can be performed in patients who simultaneously receive multiple antimicrobial agents. Major problems include lack of interlaboratory standardization, controversy over the importance of peak versus trough SBT results, and what defines optimal results for different infections.
The SBT is most commonly employed to gauge response in infective endocarditis, infections of bone or joint, and infections in neutropenic patients, when bactericidal (as opposed to bacteriostatic) activity may be necessary for a satisfactory clinical outcome. In addition, the SBT has been employed to guide the changeover from parenteral to oral therapy, often in bacterial endocarditis or osteomyelitis. In these situations, it provides the physician with the knowledge that the antimicrobial, in a physiologic setting, is capable of killing the offending pathogen. The use of this test, however, remains controversial, and several authorities feel that its value has not yet been proved.
With regard to infective endocarditis, recent studies suggest that dilutions obtained within 1 hour of antimicrobial administration of more than 1:64 and trough values of more than 1:32 are associated with bacteriologic cures, whereas clinical and microbiologic failures cannot be predicted by SBT determinations. There is no information to support the concept that extremely high dilutions (e.g., >1:1,024) are clinically more effective than lower favorable ones. Similar data are available for osteomyelitis and for selected infections in the neutropenic patient. In my opinion, the SBT should be employed infrequently. It may be indicated for selected cases of infective endocarditis, hematogenous osteomyelitis, and bacteremia in neutropenic patients when infection is caused by unusual pathogens, an unusual drug regimen is employed, or patient response has been unsatisfactory despite apparent adequacy of therapy. In these instances, an adequate SBT result, typically above 1:8 at peak, demonstrates that antimicrobial activity is satisfactory, and adverse clinical or bacteriologic experiences probably do not mean that the antimicrobial employed is ineffective.
Determinations of Minimal Inhibitory and Bactericidal Concentrations
Determinations of MIC and minimum bactericidal concentration (MBC) are laboratory tests that evaluate the minimum concentration of an antimicrobial required to inhibit or kill an organism in vitro. The MIC, when compared with achievable serum levels of an antimicrobial, determines whether the organism is susceptible or resistant. For instance, if an organism has an MIC of 4 µg/mL but achievable serum concentrations of a given agent are only 2 µg/mL, the organism is likely to be resistant. This information is now incorporated into virtually all automated antimicrobial susceptibility systems, and results for any antimicrobial agent versus the offending pathogen are either noted in the standard microbiology report or are available on request from laboratory personnel.
The MBC is determined by incubating on agar those tubes demonstrating no visible growth from the MIC determination. The MBC is an antimicrobial concentration that kills more than 99.9% of the original inoculum. This test is extremely time-consuming for laboratory personnel and is currently of little clinical value. For selected patients in whom knowledge about antimicrobial killing may be of value (e.g., immunocompromised patients with unresponsive, overwhelming bacteremia or with infective endocarditis), the SBT is more physiologic. (R.B.B.)
Cantu TG, Yamanaka-Yuen NA, Lietman PS. Serum vancomycin concentrations: reappraisal of their clinical value. Clin Infect Dis 1994;18:533–543.
After performing a literature review on the subject, the authors concluded that measurement of serum concentrations of vancomycin are generally not indicated. It is not even clear whether the product is associated with ototoxicity or nephrotoxicity when it is employed as monotherapy. For determination of drug levels to be useful, certain criteria must be met: (a) Drug concentrations must be correlated with toxicity/efficacy; (b) there must be variations between patients with regard to pharmacokinetics; (c) efficacy or toxicity of drug must be either difficult to measure or substantially delayed in onset; and (d) a sensitive assay must exist. In the case of vancomycin, the first two criteria are not met.
Fernandez de Gatta MD, et al. Cost effectiveness analysis of serum vancomycin concentration monitoring in patients with hematologic malignancies. Clin Pharmacol Ther 1996;60:332–340.
This was a prospective, randomized investigation of 70 febrile, immunocompromised patients receiving vancomycin. One group received active pharmacologic intervention—monitoring of vancomycin serum concentrations and adjustment of vancomycin dose—while the other arm acted as a control. Patients who received active intervention fared no better regarding clinical response but had significantly less nephrotoxicity and lower hospital costs. However, this finding may not conflict with recommendations against routine testing, as many of these patients might have been receiving other agents that could have caused cumulative/synergistic toxicity.
Follin SL, et al. Falsely elevated serum vancomycin concentrations in hemodialysis patients. Am J Kidney Dis 1996;27:67–74.
Maintenance hemodialysis patients may have significantly overestimated serum vancomycin levels if the tests are performed with the fluorescence polarization immunoassay (FPIA). This widely employed method may result in false determinations of further vancomycin dosing regimens. The authors suggest use of an alternative enzyme multiplied immunoassay technique (EMIT) method.
Jordan GW, Kawachi MM. Analysis of serum bactericidal activity in endocarditis, osteomyelitis, and other bacterial infections. Medicine (Baltimore) 1981;60:49–61.
This retrospective review of a 5-year clinical experience with SBT in the management of a variety of infections concludes that (a) SBT is a useful indicator of antimicrobial efficacy in infective endocarditis; (b) for infections caused by susceptible bacteria, antimicrobial combinations offer no advantage over monotherapy; (c) SBT was useful when second-line antimicrobials were indicated; (d) SBT was useful in adjusting dosages in complex cases; and (e) SBT determinations may be useful in the management of endocarditis caused by “tolerant” staphylococci.
Moellering RC Jr. Editorial: monitoring serum vancomycin levels: climbing the mountain because it is there? Clin Infect Dis 1994;18:544–546.
An editorial response to the Cantu article. Dr. Moellering is in general agreement with the conclusions, but feels that four indications for vancomycin monitoring might exist, as mentioned in the text of this chapter.
Mulhern JG, et al. Trough serum vancomycin levels predict the relapse of gram-positive peritonitis in peritoneal dialysis patients. Am J Kidney Dis 1995;25:611–615.
Thirty-one episodes of gram-positive peritonitis in patients on chronic ambulatory peritoneal dialysis were assessed to discover risks for relapse. All patients received four weekly doses of vancomycin, with the last two adjusted to maintain a trough level greater than 12 m/mL. Relapse was seen only in those who failed to maintain a mean 4-week concentration at this level.
Washington JA II. The role of the microbiology laboratory in the diagnosis and antimicrobial treatment of infective endocarditis. Mayo Clin Proc 1982;57:22–32.
This overview of the SBT and the MBC test describes the techniques and limitations involved in both. It also reviews the history of the use of the SBT in infective endocarditis and describes the differences in several published reports in regard to how the SBT was employed.
Weinstein MP, et al. Multicenter collaborative evaluation of a standardized serum bactericidal test as a prognostic indicator of infective endocarditis. Am J Med 1985;78:262–268.
Employing standardized methodology for SBT, this investigation demonstrated the statistical relevance of peak SBT levels above 1:64 and trough SBT levels above 1:32 as a prediction of bacteriologic cure in cases of infective endocarditis. Trough levels above 1:8 correlated with bacteriologic cure. However, the SBT result failed to predict bacteriologic failure or clinical outcome.
Wolfson JS, Swartz MN. Serum bactericidal activity as a monitor of antibiotic therapy. N Engl J Med 1985;312:968–975.
This excellent overview of the rationale, technique, and advantages and disadvantages of the SBT concludes that far more data are necessary to determine its efficacy in monitoring infections.