Gram-Positive Bacterial Infections with Teicoplanin

Transcription

1 ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Dec. 1990, p /90/ $02.00/0 Copyright 1990, American Society for Microbiology Vol. 34, No. 12 Treatment of Bone, Joint, and Vascular-Access-Associated Gram-Positive Bacterial Infections with Teicoplanin RICHARD N. GREENBERG Division of Infectious Diseases, University of Kentucky, Lexington, Kentucky Received 9 February 1990/Accepted 26 September 1990 Teicoplanin, a glycopeptide antibiotic, was evaluated for safety and efficacy in the treatment of vascularaccess-associated bacteremias and of bone and joint infections due to susceptible gram-positive organisms. Of 35 patients enrolled, 26 had osteomyelitis, 8 had vascular-access-associated bacteremias, and 1 had a joint infection. A total of 38 gram-positive isolates were identified: 23 Staphylococcus aureus and 6 coagulasenegative staphylococcus and 9 streptococcus isolates. After at least 6 months of follow-up, 17 patients were evaluable for efficacy: 10 of 14 (71%) with osteomyelitis and 3 of 3 with vascular-access-associated bacteremias had full resolution of their infections. Inadequate debridement, the presence of metal, and inadequate dosing were likely causes of two failures and two relapses in patients with osteomyelitis. For all but two organisms, teicoplanin MICs were <2,ug/ml. Patients who responded had median peak and trough serum bactericidal levels at serum dilutions of 1:64 and 1:16; trough levels of teicoplanin in serum were >30 pg/ml. Patients did not respond as expected to daily doses of 4 mg/kg of body weight, which consequently were increased to :15 mg/kg. Audiology testing of 20 patients found 2 with a mild loss of high-frequency hearing; 1 patient complained of tinnitus. Patients tolerated peak levels in serum as high as 127,tg/ml and trough levels of 49,Lg/ml. However, 5 of 18 patients (28%) whose daily dose was -12 mg/kg developed drug fever and rash and had teicoplanin discontinued. Further study of the antibiotic at such higher doses is needed. Teicoplanin, a glycopeptide antibiotic, has potential for use as an alternative to vancomycin in the treatment of gram-positive bacterial infections, especially those caused by methicilhin-resistant pathogens. Similar to vancomycin in structure, mechanism of action, and in vitro activity (5, 7, 11, 15, 17), teicoplanin is, however, highly protein bound (>90%) and has a much longer half-life in serum than vancomycin (>50 h) (7a, 16). In addition, teicoplanin may have less vestibular toxicity and ototoxicity than vancomycin and has the advantage that it can be given intramuscularly as well as intravenously (8, 9, 14). This study was performed to evaluate the safety and efficacy of once-daily teicoplanin in the treatment of bone and joint infections and of vascular-access-related bacteremia caused by gram-positive bacteria. The study also generated data on dose levels of teicoplanin necessary for the successful treatment of these infections. Serum bactericidal activity and levels of teicoplanin in serum were documented during the study. MATERIALS AND METHODS Patients. Patients, all enrolled from August 1986 through June 1989, were hospitalized at the St. Louis University Medical Center and had either vascular-access-associated bacteremia or bone or joint infection due to a gram-positive pathogen. Vascular-access-associated bacteremias were defined as bacteremias that were considered the result of a contaminated vascular-access device and required removal of the device for treatment. That is, positive blood cultures of the same pathogen were obtained from blood sampled through a peripheral site (not through an infected site) and from a contaminated vascular-access device. The study was a phase two, open, nonrandomized study of infections that could be treated with a glycopeptide antibiotic. After initial administration in the hospital, teicoplanin therapy was designed to continue on an outpatient basis. Patients were excluded if they had a history of hypersensitivity to vancomycin other than "red neck" syndrome or if they were pregnant or nursing, had a serum creatinine level of >2.5 mg/100 ml, had liver function test results that were more than four times the hospital laboratory normal, were unlikely to survive for 48 h after enrollment, or had a history of hearing loss or abnormal vestibular function. Except for metronidazole and aztreonam, patients could not receive other antimicrobial agents for more than 72 h after enrollment. There were three exceptions: patient 1 received amikacin for 6 days for a mixed infection of bone due to Pseudomonas aeruginosa and Staphylococcus simulans. Patient 2 inadvertently took cephradine for 10 days and ampicillin for 11 days while being treated for a Staphylococcus aureus vertebral osteomyelitis. Patient 6 received ciprofloxacin for 30 days; this patient had a bone infection due to enterococci, Klebsiella sp., Proteus mirabilis, and S. aureus, as well as a streptococcal infection resistant to ciprofloxacin. Culture procedures. All specimens (blood, joint fluid, bone biopsy, and aspirate) were obtained via sterile sites. Material for bone cultures was obtained either by bone biopsy through a sterile site or by a needle aspirate taken through uninfected skin cleansed with a povidone-iodine solution. For patients recruited at the beginning of the study, teicoplanin was administered as an intravenous infusion of 6 mg/kg of body weight every 12 h for three doses and then 6 mg/kg every 24 h. However, dosing regimens evolved during the study on the basis of clinical and microbiologic observations. Thus, for patients enrolled by the end of the recruitment period, the dose administered was 15 mg/kg every 12 h for at least the first three doses and then 15 mg/kg once every 24 h. Patient monitoring. Prior to treatment, patients were monitored by physical examinations, appropriate X rays or scans, appropriate cultures for aerobes and anaerobes (usually from sterile sites), blood cultures, complete blood counts, prothrombin times, partial thromboplastin times, platelet counts, erythrocyte sedimentation rates, direct 2392

2 VOL. 34, 1990 Coomb's tests, liver function tests, renal function tests, electrolyte tests, uric acid tests, urine analysis, and (when possible) audiometry in a booth. Patients were observed daily while in the hospital. Laboratory tests were performed on study days 3 and 10 and weekly thereafter. Patients were seen at least every 4 weeks as outpatients and were contacted weekly by phone. Laboratory tests were performed weekly on outpatients. Audiometry studies were performed by the audiometry technician at the hospital at the beginning and end of teicoplanin treatment. All patients gave their informed consent for participation in the study, which received approval from the Institutional Review Board of St. Louis University. Documentation of satisfactory response to therapy was based on complete resolution of clinical findings and negative cultures (when practical) at the completion of therapy. When patients did not show a complete resolution, further cultures were prepared to identify the pathogen. Treatment regimens of such patients were classified as failures in the event of incomplete clinical resolution or if the original pathogen persisted. Serum studies. An agar-disk diffusion susceptibility bioassay with Bacillus subtillis (ATCC 6633) was used to determine peak and trough levels of teicoplanin in serum samples drawn at dosing periods 3 and 7 and weekly thereafter (4). Serum bactericidal levels were assayed by a microtiter method that used serum from the specimens drawn for serum drug level determinations (13). Heat-treated human serum in Mueller-Hinton II broth (1:1 [vol/vol]) was tested in serial dilutions with an inoculum of the patient's pathogen (3 x 105 CFU/ml); the endpoint was the maximum dilution causing at least 99.9% killing. Microbiology. Aerobic and anaerobic cultures were processed by standard methods. Gram-positive pathogens were tested by disk diffusion and in most cases by microtiter assay for susceptibility to oxacillin, vancomycin, and teicoplanin (1, 10). Microtiter dilutions were made with Mueller-Hinton II broth, and teicoplanin was dissolved in 0.2 M phosphate buffer, ph 7.4. MBC determinations were not carried out. RESULTS Study patients. Thirty-five patients with gram-positive bacterial infections received teicoplanin. Twenty-six had osteomyelitis, one of whom (patient 15) also had an infected knee joint. Eight had infections related to intravenous catheters, and one (patient 27) had a knee joint infection. There were 27 males (mean age, 56 years, ranging from 24 to 80 years) and 8 females (mean age, 61 years, ranging from 32 to 82 years). Other antibiotics given during treatment with teicoplanin included aminoglycosides (3 patients), metronidazole (9 patients), aztreonam (12 patients), ampicillin (1 patient), cephradine (1 patient), and ciprofloxacin (1 patient) (Table 1). Osteomyelitis and joint infection. The essential role of surgical treatment of osteomyelitis cannot be overemphasized. All patients were evaluated either clinically or with flow studies to assess adequate vascular flow to infected bone. Debridement or vascular repair or both were performed when indicated. Cultures of remaining bone tissue from these study patients grew the pathogen. No antibioticimpregnated beads were used in the study. Of the 27 patients with osteomyelitis or joint infection, 5 (patients 21 to 25) could not be studied for efficacy: 4 had negative cultures or the isolate could not be tested, and 1 patient was transferred to another hospital after receiving TREATMENT WITH TEICOPLANIN 2393 only two doses of teicoplanin. A further seven patients clearly improved but did not complete therapy and must be considered unevaluable. Five of these seven (patients 16 to 20) had therapy discontinued because of an adverse reaction to teicoplanin between days 5 and 12 of treatment, one was sent home on oral therapy (patient 26), and one (patient 27) was switched to oxacillin at the patient's request. In 2 of the remaining 15 patients (patients 14 and 15), the outcome was failure, while in 13 (86.6%) (patients 1 to 13), infections were resolved completely at the end of therapy. After a follow-up period of -6 months, 1 (patient 11) was lost to follow-up and 2 (patients 12 and 13) had relapsed, bringing to 4 the number of failures from 14 patients evaluable after 6 months. Therefore, the persistence of successful resolution was 71% at that time. Among the resolved cases was that of patient 2, who was treated for 219 days according to the recommendation of the orthopedic surgery service; their recommendation was followed because the duration of treatment of vertebral osteomyelitis is controversial and the resolution of infection can be slow. The two treatment failures (patients 14 and 15) illustrate the difficulties of treating deep-seated staphylococcal infections. One failure (patient 14) had vertebral body (L4) osteomyelitis. She quickly improved once therapy began and was sent home on once-daily teicoplanin therapy through a central intravenous catheter. She relapsed and developed a large psoas abscess while receiving the drug for 76 days. A careful review of her X-ray and computerized tomography studies identified a nidus of infection that should have been debrided before her discharge home. In tests of her serum bactericidal level, effective dilutions remained between 1:16 (trough) and 1:64 (peak) throughout therapy with teicoplanin; this may explain why she did well for so long while incubating a major collection of pus. The MIC for her pathogen increased from 1 to 4,ug/ml as treatment continued. The other patient for whom teicoplanin treatment failed (patient 15) was a diabetic female with staphylococcal infection of bone, joint, and skin. Staphylococcal organisms persisted in joint fluid after 13 days of treatment with teicoplanin: cultures of bone and joint fluid continued to grow the pathogen, and signs of infection remained. This patient ultimately required a long duration of treatment with parenteral antibiotics to resolve her infection. This failure may have been due to inadequate dosing of teicoplanin, which was at the lower levels used in the initial part of this study (Table 1). Vascular-access-associated bacteremias. Eight patients with vascular-access-associated bacteremias received teicoplanin. Five were given three infusions of 6 mg/kg every 12 h and then one infusion of 6 mg/kg every 24 h. After the three initial loading doses of 6 mg/kg, one patient received only maintenance doses of 1.5 mg/kg daily. Another patient received maintenance doses ranging from 2 to 4 mg/kg daily after the three 6-mg/kg loading infusions. A further patient was given three loading doses of 12 mg/kg at 12-h intervals followed by daily maintenance doses of 15 mg/kg every 24 h. Five patients could not be evaluated for efficacy: two had negative cultures and three received teicoplanin for fewer than 3 days. The other three patients responded to teicoplanin therapy. The blood cultures grew Staphylococcus epidermidis (patient 28), multiple pathogens (including Staphylococcus haemolyticus, S. epidermidis, and Streptococcus sanguis [patient 29]), and S. aureus (patient 30). There were no relapses (Table 1). Microbiology. Twenty-three positive cultures of S. aureus

3 2394 GREENBERG ANTIMICROB. AGENTS CHEMOTHER. TABLE 1. Patients, pathogens, and results Patient Infection Organism Doseb Serum teicoplanin level no.a site (MIC [tg/mlj) (mg/kg) x days (Lg/ml), associated serum Commentsd bactericidal level" 1 Sternum S. simulans 6 b.i.d. x 2, 26 (peak); (0.125), P. 6 q.d. x 40 9 (trough) aeruginosa S. aureus (1.0) 2 Vertebral body (L1, L2) 3 Foot 4 Foot 5 Foot 6 Foot 7 Foot 8 Hand 9 Foot 10 Foot 11 Cranium 12 Sternum 13 Tibia 14 L2-L5 vertebral bodies 15 Foot, knee 16 Tibia 17 Femur 18 Hand 19 Femur 20 Foot S. epidermidis (1.0) Enterococci S. aureus (0.5) S. aureus (0.125), Proteus mirabilis, enterococci, group A streptococci, Klebsiella sp. S. aureus (0.125) S. aureus (0.25) S. aureus S. aureus S. aureus (0.25) S. aureus (0.25) S. aureus (0.125) S. aureus (1.0-i' 4.0) S. aureus (1.0) S. aureus (0.5) S. aureus (1.0) S. aureus (0.5) S. aureus (0.25), group B streptococci S. aureus (0.125), group B streptococci 6 q.d. x 6, 10 q.d. x 2, 15 q.d. x 2, 16 q.d. x 32, 8 q.d. x b.i.d. x 1, 12 q.d. x b.i.d. x 5, 10 q.d. x q.d. x 13, 10 q.d. x 28 8 b.i.d. x 5, 8 q.d. x q.d. x 1, 22 q.d. x 1, 15 q.d. x b.i.d. x 4, 12 q.d. x 2, 16 q.d. x 19, 24 q.d. x 1, 30q.d. x 1, 16 q.d. x b.i.d. x 3, 15 q.d. x b.i.d. x 5, 15 q.d. x 38 6 q.d. x 10, 9 q.d. x 18 7 t.i.d. x 3, 10 b.i.d. x 27, 12 q.d. x 16 6 q.d. x 5, 8 q.d. x q.d. x 9, 6 q.d. x 64 6 q.d. x 7, 6 b.i.d. x 2, 12 q.d. x 4 4 b.i.d. x 1, 10 b.i.d. x 2, 11 q.d. x 1, 16 b.i.d. x 4, 16 q.d. x 3 12 q.d. x 2 12 b.i.d. x 5, 12 q.d. x 7 15 b.i.d. x 5, 15 q.d. x 5 12 b.i.d. x 5, 15 q.d. x (peak), 1:64; 70 (trough), 1:16 97 (peak); 25 (trough) 71 (peak), 1:4; 27 (trough), 1:2 64 (peak), 1:32; 30 (trough), 1:16 79 (peak), 1:64; 32 (trough), 1: (peak), 1:64; 36 (trough), 1: (peak); 46 (trough) 88 (peak), 1:32; 42 (trough), 1:4 88 (peak), 1:128; 46 (trough), 1:32 47 (peak); 14 (trough) 46 (peak); 7 (trough) 84 (peak), 1:64; 43 (trough), 1:16 40 (peak, 1:4; 9 (trough) 70 (peak) 59 (peak); 21 (trough) 107 (peak), 1:4; 41 (trough), 1:2 65 (peak), 1:2; 41 (trough), <1:2 80 (peak); 34 (trough) Amikacin for first 5 days; AZM; MET; no relapse at 1 yr Oral ampicillin and oral cephradine days of study; no relapse after 1 yr; mild high-frequency hearing loss (10 to 30 db [both ears], 3-8 khz) No relapse after 6 mo No relapse at 6-mo follow-up No relapse after 6 mo AZM; MET; ciprofloxacin for 30 days AZM; MET No relapse after 6 mo AZM; MET; loose stools MET; AZM Lost to follow-up; AZM Relapsed after 1 mo; needed more debridement; tinnitus Metal appliance; relapsed Required further debridement Gentamicin for first 2 days; AZM; MET Drug fever; 10-dB hearing loss at 4 khz Drug fever Drug fever Drug fever; AZM Drug fever Continued

4 VOL. 34, 1990 TREATMENT WITH TEICOPLANIN 2395 TABLE 1-Continued [Ong/ml) (mg/kg) x days (,ug/ml), associated serum Commentsd Patient Infection Organism Doseb Serum teicoplanin level no.'a site (MIC bactericidal level' 21 Hip No growth 6 q.d. x 7 26 (peak); AZM; loose stools 5 (trough) 22 Knee S. aureus (0.25) 6 q.d. x 2 Transferred elsewhere 23 Foot No growth 15 b.i.d. x 4, 15 q.d. x 2 24 Foot No growth 12 q.d. x 2, 29 (peak); AZM; MET 5 q.d. x 8 17 (trough) 25 Femur S. epidermidis 15 b.i.d. x 6, 73 (peak); Isolate lost and not 15 q.d. x 7 36 (trough) tested 26 Foot S. aureus, group 12 b.i.d. x 4, 82 (peak); Sent home; lost to B streptococci 12 q.d. x 2 32 (trough) follow-up 27 Knee S. aureus (0.25) 12 b.i.d. x 6, 61 (peak), 1:4; Changed to oxacillin at 15 q.d. x (trough), <1:2 patient's request 28 Blood, vascular- S. epidermidis 6 b.i.d. x 2, 26 (peak); AZM; prolonged course access (2.0) 2 q.d. x 15, 18 (trough) due to possible associated 4 q.d. x 19 endocarditis 29 Blood, vascular- S. haemolyticus 6 b.i.d. x 1 30 (peak); Tobramycin for first 3 access (8.0), Strepto- 6 q.d. x (trough) days associated coccus sanguis (0.125), S. epidermidis 30 Blood, vascular- S. aureus (0.5) 6 b.i.d. x 1, 72 (peak); access 12 b.i.d. x 4, 41 (trough) associated 15 q.d. x 4 a Patients 1 to 27 had bone and joint infections: the infections of patients 1 to 10 were resolved without relapse for at least 6 months; patient 11 had a resolved infection but was lost to follow-up; patients 12 to 15 had failure or relapse; patients 16 to 20 could not be evaluated because of drug fever or rash; patients 21 to 27 could not be evaluated because of other reasons. Patients 28 to 30 had vascular-associated bacteremias and were cured. b b.i.d., Twice a day; q.d., once a day; t.i.d., three times a day. Boldface doses are specifically associated with the peak and trough teicoplanin levels. c Not all pathogens were tested for MICs or serum bactericidal levels; all pathogens were susceptible by disk assay. d AZM, Received aztreonam; MET, received metronidazole. were recovered (Table 2). In addition, there were four positive cultures of S. epidermidis, one of S. simulans, one of S. haemolyticus, and nine of streptococci (group B, five cultures; enterococci, two cultures; S. sanguis, one culture; and group A, one culture). Four gram-negative isolates were recovered from bone from the study patients: P. aeruginosa, two isolates; P. mirabilis, one isolate; and Klebsiella sp., one isolate. These organisms were not tested with teicoplanin and were presumed to be resistant. Patients with positive cultures of gram-negative isolates received concurrent therapy with appropriate antibiotics (Table 1). All gram-positive Pathogen (n) TABLE 2. Susceptibility of pathogens MIC range (p.g/ml) Oxacillin Teicoplanin Vancomycin Oxacillin-susceptible S. aureus" (15) Oxacillin-resistant > S. aureus (8) S. haemolyticus (1) > Other coagulase > negative staphylococcib (5) Streptococcic (9) NDd ND ND 14 (MIC of a Includes one pathogen isolated during therapy from patient teicoplanin increased from 1 to 4 b plg/ml). S. simulans (n = 1) and S. epidermidis (n = 4) Group B (n = 5), enterococci (n = 2), S. sanguiis (n = 1), and group A (n = 1); all were susceptible to all three drugs by disk testing, except the enterococci, which were susceptible only to teicoplanin and vancomycin. d ND, Not done. isolates were susceptible to teicoplanin, and for all but two staphylococcal isolates, MICs were c2,ug/ml or 30-,ug disk zone sizes were.16 mm. After 2 months of teicoplanin treatment for an S. aureus isolate from patient 14, the MIC increased from 1 to 4,ug/ml and disk zone size decreased from 16 to 14 mm. For the single S. haemolyticus isolate, the MIC was 8,ug/ml and the disk zone size was 15 mm. Teicoplanin dosing and levels in serum. Dose levels were increased during the study because several patients with joint or bone infections failed to show clinical improvement at 6 mg/kg. Striking improvement was evident once doses of.12 mg/kg were administered. Eventually the protocol was amended so that everyone received a standard loading dose of 15 mg/kg every 12 h for 3 days and then 15 mg/kg every 24 h. Accumulation of teicoplanin was not apparent in those patients with a long duration of treatment, including one patient who received the drug for 219 days. Peak and trough levels in serum were measured for all patients at the times of their third and seventh doses and weekly thereafter. Levels in serum increased with dosing, as the following figures show. The mean peak and trough levels (± standard error) were as follows: for 6 mg/kg, 25 ± 2 p.g/ml (n = 16) and 13 ± 2,ug/ml (n = 18); for 12 mg/kg, 79 ± 5,ug/ml (n = 16) and jig/ml (n = 15); for 15 mg/kg, 85 ± 3,ug/ml (n = 19) and 41 ± 1 jig/ml (n = 14). Serum bactericidal activity. Peak and trough serum bactericidal activities were measured whenever possible. In six patients in whom bone or joint infections were resolved, the median peak serum bactericidal level was at a dilution of 1:64 (range, 1:128 to 1:8) and the trough was at 1:16 (range, 1:32 to <1:2). In a patient in whom treatment failed (patient 14), the peak level was at a dilution of 1:64 and the trough

5 2396 GREENBERG was at 1:16; this patient sequestered pus for over 2 months before showing signs of an abscess, and the treatment actually failed because of inadequate debridement. Audiology. Audiology testing was performed whenever possible. Both pre- and posttreatment evaluations were completed for 20 patients, including the patient who was treated for 219 days. Only two patients showed any hearing loss; one had only a 10-dB loss in the right ear at 4 khz, and one had losses of 20 db (left ear) and 30 db (right ear) at 4 khz and 15 db (left ear) and 25 db (right ear) at 8 khz after 219 days. One patient complained of tinnitus, which resolved when the dose of teicoplanin was reduced. Adverse reactions. Two types of adverse events required cessation of teicoplanin treatment. One vascular-access patient experienced acute bronchospasm during the first infusion. The problem resolved with discontinuation of the antibiotic. Five patients developed a macular rash associated with fever and with itching and nausea; these symptoms required at least 4 days to resolve once teicoplanin was discontinued. The rash and associated events occurred between days 5 to 12 of therapy and were seen in 5 of 18 patients (28%) receiving doses of 12 mg/kg or higher. Two patients experienced loose stools (Table 1), which resolved after withdrawal of teicoplanin. DISCUSSION This study confirms that the efficacy, dosing, safety, and serum bactericidal activity of teicoplanin merit special attention. The results suggest that administration of this antibiotic can achieve high serum bactericidal activity and can eradicate susceptible gram-positive bacteria, including S. aureus. Thus, all three evaluable patients with device-related bacteremias had their infections resolved. Of 27 patients with bone and joint infections, 14 could be fully evaluated after follow-up of 6 months or more; 10 of the infections (71%) were successfully resolved. Furthermore, plausible explanations were found for the two failures and two relapses, including underdosing (one patient), inadequate debridement (one patient), or effects of metal devices (two patients). Although two of the three evaluable bacteremias related to catheters did resolve with dosing of 6 mg/kg or less, it cannot be concluded that this is an effective dose for all septic patients, particularly those who do not have a removable source of infection, such as a vascular-access catheter. More-definitive conclusions may be drawn regarding dosing and resolution of joint and bone infections. For example, in an instance in which dosing of 6 mg/kg did not result in a response (i.e., resolution of signs of cellulitis overlying an acute osteomyelitis due to S. aureus), a favorable response was achieved when dosing was increased to yield trough levels in serum of.30,ug/ml (dosing of at least 12 mg/kg). Several other investigators have reported similar results with teicoplanin (2, 6, 8). In time-kiml studies in vitro, we have also found that killing of S. aureus significantly improved with larger concentrations of the drug (7a). The need for higher dosing probably relates to the sequestration of the drug by protein binding (>90%) in serum (7a, 16); the importance of achieving free drug concentrations continuously exceeding the MBCs for gram-positive bacteria has been discussed by others (3, 8, 12). The data in Table 1 allow an estimate of the range of levels of teicoplanin in serum needed to resolve deep-seated infections of bone and joints. Nine of the successfully resolved cases with at least 6 months of follow-up (patients 1 to 10) had trough levels of 25,ug/ml or greater. Furthermore, if ANTIMICROB. AGENTS CHEMOTHER. from that group only cases of osteomyelitis in which no concomitant antibiotics were given, with only one pathogen isolated per patient, are selected, four cases are identified, all of which resolved successfully; their serum trough levels ranged from 25 to 46,g/ml and their peak levels ranged from 64 to 122,ug/ml. On the other hand, levels in serum lower than these were recorded in one case in which treatment failed (patient 15): serum peak and trough levels were 40 and 9,ug/ml, respectively. Together, these data suggest that trough levels of teicoplanin in serum of.30,ug/ml are needed for consistent responses. The safety of the doses of teicoplanin in the range of 6 to 14.4 mg/kg daily has been reported by other investigators (8, 9). In this study, doses as large as 30 mg/kg daily were administered safely to one patient. Another patient was treated for 219 days and for 203 of those days received 8 mg/kg without any adverse effects. Nevertheless, two types of adverse reaction to teicoplanin occurred. One episode of acute bronchospasm was seen during the administration of a first dose. Five instances (28%) of fever and rash were observed in 18 patients who received doses of.12 mg/kg. This reaction developed between days 5 and 12 of therapy and required cessation of teicoplanin treatment. No particular lot of antibiotic was identified with this problem; however, all of the lots used were part of a single batch manufactured for this study. Drug fever of such frequency has not been reported by others who have studied teicoplanin. Tingling and flushing of the face, neck, and thorax (the red neck syndrome) is associated with the infusion of vancomycin; this syndrome was not observed in patients in the present study and was not associated with the observed drug fever and rash. Ototoxicity or renal toxicity presented no major problem; whether such toxicities might occur when aminoglycosides are given concurrently is not known. In summary, when administered at levels yielding trough levels in serum of >30,ug/ml, teicoplanin appears to be effective once-a-day intravenous therapy for deep-seated bone and joint infections caused by susceptible pathogens. Indeed, teicoplanin has the potential to become a useful parenteral antimicrobial agent in outpatient therapy due to its single daily dosing. Further studies are needed to establish effective dosing for other types of deep-seated grampositive bacterial infections and to document safety when the antibiotic is administered both alone and with ototoxic or nephrotoxic agents. ACKNOWLEDGMENTS The research nurses were Kathy M. Wilson, Ann Heston-Durand, Philip Brusca, Karen E. Bowen, and Joy W. Thompson. Donald Kennedy assisted in patient evaluations. The cooperation of the Orthopedic Service, St. Louis University, the microbiology work of David Bowne, Gill Corrigan, Jr., and James Toner, and the secretarial assistance of Kimberly Holt are appreciated. The study was funded by a grant from Merrell-Dow Co. (Cincinnati, Ohio). LITERATURE CITED 1. Bauer, A. W., W. M. M. Kirby, J. C. Sherris, and M. Turck Antibiotic susceptibility testing by a standard disc method. Am. J. Clin. Pathol. 45: Calain, P., K. H. Krause, P. Vaudaux, R. Auckenthaler, D. Lew, F. Waldvogel, and B. Hirshel Early termination of a prospective, randomized trial comparing teicoplanin and flucloxacillin for treating severe staphylococcal infections. J. Infect. Dis. 155: Cantoni, L., A. Wenger, M. P. Glauser, and J. Bile Comparative efficacy of amoxicillin-clavulanate, cloxacillin,

6 VOL. 34, 1990 and vancomycin against methicillin-sensitive and methicillinresistant Staphylococcus aureus endocarditis in rats. J. Infect. Dis. 159: Cavenaghi, L., A. Corti, and G. Cassani Comparison of solid phase enzyme receptor assay (SPERA) and the microbiological assay for teicoplanin. J. Hosp. Infect. 7(Suppl. A): Cynamon, M. H., and P. A. Granato Comparison of the in vitro activities of teicoplanin A2 and vancomycin against staphylococci and enterococci. Antimicrob. Agents Chemother. 21: Glupczynski, Y., H. Lagast, P. Van Der Auwera, J. P. Thys, F. Crokaert, E. Yourassowsky, F. Meunier-Carpentier, J. Klastersky, J. P. Kains, E. Serruys-Schoutens, and J. C. Legrand Clinical evaluation of teicoplanin for therapy of severe infections caused by gram-positive bacteria. Antimicrob. Agents Chemother. 29: Gorzynski, E. A., D. Amsterdam, T. R. Beam, Jr., and C. Rotstein Comparative in vitro activities of teicoplanin, vancomycin, oxacillin, and other antimicrobial agents against bacteremic isolates of gram-positive cocci. Antimicrob. Agents Chemother. 33: a.Greenberg, R. N., and C. A. Benes Time-kill studies with oxacillin, vancomycin, and teicoplanin versus Staphylococcus aureus. J. Infect. Dis. 161: Leport, C., C. Perronne, P. Massip, P. Canton, P. Leclercq, E. Bernard, P. Lutun, J. J. Garaud, and J.-L. Vilde Evaluation of teicoplanin for treatment of endocarditis caused by gram-positive cocci in 20 patients. Antimicrob. Agents Chemother. 33: Martino, P., M. Venditti, A. Micozzi, C. Brandimarte, G. Gentile, C. Santini, and P. Serra Teicoplanin in the TREATMENT WITH TEICOPLANIN 2397 treatment of gram-positive bacterial endocarditis. Antimicrob. Agents Chemother. 33: National Committee for Clinical Laboratory Standards Methods for dilution antimicrobial susceptibility test for bacteria that grow aerobically (tentative standard), 2nd ed., M7-T2. National Committee for Clinical Laboratory Standards, Villanova, Pa. 11. Neu, H. C., and P. Labthavikul In vitro activity of teichomycin compared with those of other antibiotics. Antimicrob. Agents Chemother. 24: Peterson, L. R., J. A. Moody, C. E. Fasching, and D. N. Gerding Influence of protein binding on therapeutic efficacy of cefoperazone. Antimicrob. Agents Chemother. 33: Relier, L. B., and C. W. Stratton Serum dilution test for bactericidal activity. II. Standardization and correlation with antimicrobial assays and susceptibility tests. J. Infect. Dis. 136: Smith, S. R., J. Cheesbrough, R. Spearing, and J. M. Davies Randomized prospective study comparing vancomycin with teicoplanin in the treatment of infections associated with Hickman catheters. Antimicrob. Agents Chemother. 33: Somma, S., L. Gastaldo, and A. Corti Teicoplanin, a new antibiotic from Actinoplanes teichomyceticus nov. sp. Antimicrob. Agents Chemother. 26: Verbist, L., B. Tjandramaga, B. Hendrickx, A. Van Hecken, P. Van Melle, R. Verbesselt, J. Verhaegen, and P. J. De Schepper In vitro activity and human pharmacokinetics of teicoplanin. Antimicrob. Agents Chemother. 26: Williams, A. H., and R. N. Gruneberg Teicoplanin. J. Antimicrob. Chemother. 14: