MAJOR ARTICLE

Randomized, Double-Blind, Multicenter Trial Comparing Clinaoxacin with Imipenem as Empirical Monotherapy for Febrile Granulocytopenic Patients
Drew J. Winston,1 Hillard M. Lazarus,2 Roy A. Beveridge,3 James W. Hathorn,4 Rasim Gucalp,5 Reuben Ramphal,6 Anthony W. Chow,7 Winston G. Ho,8 Ruth Horn,9 Ronald Feld,10 Thomas J. Louie,11 Mary C. Territo,1 Jeffrey L. Blumer,2 and Kenneth J. Tack12
1 Division of Hematology-Oncology, Department of Medicine, University of California Los Angeles Medical Center, Los Angeles; 2Case Western University Hospital, Cleveland; 3Fairfax Hospital, Falls Church, Virginia; 4Duke University Medical Center, Durham, North Carolina; 5Monteore Medical Center, Bronx, New York; 6University of Florida Medical Center, Gainesville, Florida; 7Vancouver Hospital, Vancouver, British Columbia; 8St. Josephs Hospital, Orange, California; 9Division of Infectious Diseases, Department of Medicine, Royal Victoria Hospital, Montreal, Quebec; 10 Princess Margaret Hospital, Toronto, Ontario; 11Calgary General Hospital, Calgary, Alberta; and 12Parke-Davis Pharmaceutical Research, Ann Arbor, Michigan

In a double-blind, multicenter trial, 541 febrile granulocytopenic patients were randomized to receive either intravenous (iv) clinaoxacin (200 mg every 12 h) or iv imipenem (500 mg every 6 h) as empirical monotherapy. More baseline pathogens were susceptible to clinaoxacin (259 [99%] of 262 organisms) than to imipenem (253 [95%] of 265; P p .03). Initial favorable clinical response rates for clinaoxacin (88 [32%] of 272 patients) and imipenem (89 [33%] of 269) were similar. After addition of other antimicrobial agents, overall response rates were 259 (95%) of 272 for clinaoxacin and 251 (93%) of 269 for imipenem. During the study, only 13 clinaoxacin (5%) and 18 imipenem (7%) recipients died. Both drugs were generally well tolerated. Drugrelated skin rash occurred more often with clinaoxacin (11% vs. 6%; P p .07 ), whereas nausea (2% vs. 5%; P p .16), Clostridium-difcileassociated diarrhea (3% vs. 8%; P p .02 ), and seizures (0% vs. 2%; P p .06) occurred more often with imipenem. These results suggest that clinaoxacin and imipenem have similar efcacy as empirical monotherapy in febrile granulocytopenic patients. Prompt administration of empirical antimicrobial agents is standard therapy for fever in patients with granulocytopenia. Several options are currently available [1]. Combinations of an aminoglycoside plus an antipseudomonal b-lactam drug have been used most frequently. Recently, several studies have shown that monotherapy with ceftazidime, imipenem, cefepime, or meropenem can be as effective as combination therapy [212]. Because of the increasing incidence of grampositive infections in febrile granulocytopenic patients, a third option is to use vancomycin in combination with ceftazidime [1, 13, 14]. Although each of these regimens is effective in the majority of patients, all have certain limitations. Many granulocytopenic patients receive other nephrotoxic drugs (amphotericin B, cyclosporine, and certain chemotherapeutic agents), which makes concomitant use of the aminoglycosides potentially hazardous. Ceftazidime lacks signicant activity for coagulase-negative

Received 8 March 2000; revised 14 June 2000; electronically published 30 January 2001. Financial support: Parke-Davis Pharmaceutical Research. Informed consent approved by the institutional review board at each study center was obtained from each patient or responsible relative. Reprints or correspondence: Dr. Drew J. Winston, Room 42-121 CHS, Dept. of Medicine, UCLA Medical Center, 10833 Le Conte Ave., Los Angeles, CA 90095. Clinical Infectious Diseases 2001; 32:381390 2001 by the Infectious Diseases Society of America. All rights reserved. 1058-4838/2001/3203-0006$03.00

Clinaoxacin vs. Imipenem in Granulocytopenia CID 2001:32 (1 February) 381

staphylococci, Staphylococcus aureus, and other gram-positive organisms that now commonly cause infection in granulocytopenic patients [8, 12, 15]. Monotherapy with ceftazidime can also be associated with the emergence of b-lactamresistant gram-negative bacillary organisms [16]. Gastrointestinal toxicity, seizures, and superinfections with Stenotrophomonas maltophilia may complicate imipenem therapy [58, 17]. The use of cefepime, with its better coverage of gram-positive organisms, or meropenem, with its excellent safety prole, may overcome some of the shortcomings of treatment with ceftazidime or imipenem. However, these newer agents are costly and have not been shown in controlled clinical trials to have any signicant advantages over other b-lactam drugs [912, 18]. Because of cross-allergies, neither the cephalosporins nor the carbapenems can be used comfortably in patients with histories of severe hypersensitivity reactions to b-lactam drugs [19]. Finally, the association of vancomycin-resistant enterococcal infections with the excessive use of vancomycin has made the inclusion of vancomycin as part of any empirical antimicrobial regimen less desirable [20]. Clinaoxacin is a new broad-spectrum uoroquinolone with enhanced activity for a wide range of gram-positive, gramnegative, and anaerobic organisms [21]. Clinaoxacin retains the activity of the older uoroquinolones against Pseudomonas aeruginosa and other gram-negative bacilli but is much more active against gram-positive bacteria and anaerobic organisms [22, 23]. In vitro, clinaoxacin is especially active against the Enterobacteriaceae, P. aeruginosa, staphylococci, and streptococci, as well as many organisms resistant to ciprooxacin and the b-lactam drugs. The role of the uoroquinolones in the empirical treatment of febrile granulocytopenic patients has not been rmly established. Monotherapy with ciprooxacin or ooxacin has produced both favorable and unfavorable results and is generally not recommended [1, 24, 25]. Similarly, studies of quinolone drugs as part of combination therapy have mostly involved small numbers of patients and provided inconclusive results [1, 24, 26]. For these reasons, we performed a randomized, double-blind multicenter study to compare the efcacy and safety of clinaoxacin with imipenem as monotherapy for febrile granulocytopenic patients. Imipenem was chosen as the comparative agent because some previous trials [6, 27] had found imipenem alone to be more effective than ceftazidime alone in febrile granulocytopenic patients and many of the study sites participating in this trial had concerns about the presence of b-lactamase-producing gram-negative bacillary organisms resistant to ceftazidime [16].

PATIENTS AND METHODS Patients. Hospitalized patients were eligible for enrollment in the study if they satised the following criteria: 18 years
382 CID 2001:32 (1 February) Winston et al.

old; hematologic malignancy or solid tumor; granulocyte count 500/mm3 or 1000/mm3 and expected to fall below 500/ mm3 within 48 h of starting treatment with study drug; 2 measurements of elevated temperature 138C or 1 temperature measurement of 138.3C unrelated to the administration of blood products or drugs; no allergy to b-lactam drugs or quinolones; and no history of seizures. Patients were excluded from the study if any of the following criteria applied: severe hepatic disease (dened as an increase in the level of total serum bilirubin 13 times normal or in the level of serum aspartate aminotransferase or serum alanine aminotransferase 15 times normal); requirement for dialysis; infection with HIV; treatment with a quinolone within 7 days of randomization; or previous enrollment in the study. Women were required to have a negative test for pregnancy. Antibiotic regimens. Eligible patients at each study site were randomly assigned in a double-blind fashion to receive either clinaoxacin or imipenem. Randomization was done in a 1:1 ratio and stratied by type of underlying malignancy (hematologic or solid tumor) and by whether the patient had received a bone marrow or stem cell transplant. Clinaoxacin (Parke-Davis) was given iv at a dosage of 200 mg every 12 h. The imipenem (Primaxin; Merck) dosage was 500 mg iv every 6 h. Patients assigned to receive clinaoxacin every 12 h also received an iv placebo (100 mL of 5% dextrose in water) every 12 h so that an iv infusion was given every 6 h. Each dose of study medication was infused over an interval of 60 min. The doses of clinaoxacin and imipenem were adjusted for patients with impaired renal function by a pharmacist not blinded to the patients study drug. The dosage of clinaoxacin was decreased to 100 mg iv every 12 h for patients with creatinine clearance of 1050 mL/min and to 100 mg iv every 24 h for patients with creatinine clearance of !10 mL/min. The dosage of imipenem was decreased to 500 mg iv every 12 h for patients with creatinine clearance of !40 mL/min. Laboratory studies. Before the start of therapy, specimens of blood and from other suspected sites of infection were obtained for culture. Patients with bacteremia or a documented site of infection associated with a positive result of culture had additional cultures performed during and after therapy, unless the site of infection was not accessible for repeated sampling. All organisms isolated from clinical specimens were identied by standard criteria in the clinical microbiology laboratory at each study site and conrmed by a reference laboratory (SciCor) [28]. Similarly, the antimicrobial susceptibility of bacterial isolates to clinaoxacin, imipenem, and other antimicrobial agents was tested by the disk diffusion and broth microdilution methods at both the study site and the reference laboratory [29]. Susceptibility break points were as follows: clinaoxacin, MIC 1 mg/mL or a zone of inhibition 21 mm (5 mg disk); imipenem, MIC 4 mg/mL or zone 16 mm (10 mg disk).

Complete blood cell counts, prothrombin and partial thromboplastin times, determinations of serum electrolyte, creatinine, and blood urea nitrogen levels, urinalyses, and liver function tests (total bilirubin, aspartate aminotransferase, alanine aminotransferase, and alkaline phosphatase levels) were performed before, during, and after therapy to assess possible treatmentrelated side effects. Patients were also examined during and after therapy for clinical symptoms and signs of side effects related to the study antibiotics. All side effects, adverse events, or abnormal laboratory values that occurred during the study were recorded and classied blindly by investigators as unrelated, possibly related, or denitely related to the study drug. Evaluation of response. Initial febrile episodes were classied as caused by one of the following: microbiologically documented infection (both site of infection and organism identied), clinically documented infection (site of infection identied but no organism isolated), or possible infection (infection suspected but no site or organism identied). The response to therapy was classied as improvement (disappearance of fever, overall clinical improvement, and eradication of any infecting organism without modication of antibiotic therapy) or failure (persistence of fever or of the infecting organism requiring modication of antibiotic therapy or any death during treatment with the study drug). Criteria for modifying initial treatment with the study drug included persistence or worsening of clinical symptoms and signs of infection, resistance of the baseline bacterial pathogen to the study drug, persistence of the initial pathogen despite therapy, or emergence of a new pathogen or focus of infection. Superinfection was dened as a new microbiologically documented infection that occurred either during therapy or within 7 days of the discontinuation of antibiotic therapy. All classications of infection and responses to therapy were done by investigators blinded to the patients study therapy. Statistical analysis. Data summaries and analyses were performed by use of SAS version 6.09 [30]. All statistical tests were done as 2-tailed tests. The Cochran-Mantel-Haenszel test was used to compare differences in proportions. Times to specic events were compared by using Kaplan-Meier analysis and the log rank test [31]. Treatment interactions by center and type of malignancy were determined by using the Breslow-Day test [32]. All patients were included in both the efcacy and safety analyses (intent-to-treat analysis).

RESULTS Patient characteristics. A total of 541 patients from 15 medical centers in the United States and Canada were entered into the trial from November 1994 through August 1996. There were 272 patients in the clinaoxacin group and 269 patients in the imipenem group. Patients in the 2 treatment groups were of similar age, sex, underlying disease, degree and duration of

granulocytopenia, and use of hematopoietic growth factors (table 1). A total of 149 (55%) of 272 patients in the clinaoxacin group and 153 (57%) of 269 patients in the imipenem group had received a bone marrow or stem cell transplant. Most patients in both groups had initial granulocyte counts of !100 cells/mm3. Persistent granulocytopenia (!500 granulocytes/ mm3 throughout the course of therapy with study drug) was associated with 25% of the treatment courses. About 40% of the febrile episodes in each treatment group were caused by either a microbiologically documented infection (bacteremia or localized bacterial infection) or a clinically documented infection. The median duration of therapy with study drug was 10 days in each treatment group. Susceptibility of pretherapy isolates to antibiotics. Baseline organisms from both treatment groups were tested for susceptibility to both clinaoxacin and imipenem. Among pretherapy gram-negative organisms available for susceptibility testing, 146 (100%) of 146 were susceptible to clinaoxacin and 145 (98%) of 148 were susceptible to imipenem. Among pretherapy gram-positive organisms available for susceptibility testing, 113 (97%) of 116 were susceptible to clinaoxacin and 108 (92%) of 117 were susceptible to imipenem. Thus, 259 (99%) of 262 baseline organisms tested for susceptibility were susceptible to clinaoxacin, whereas 253 (95%) of 265 baseline organisms were susceptible to imipenem (P p .03 ). The 3 baseline organisms not susceptible to clinaoxacin were 2 isolates of coagulase-negative staphylococci and 1 isolate of Enterococcus faecalis. The 12 baseline organisms not susceptible to imipenem were 8 isolates of coagulase-negative staphylococci, 1 isolate of Corynebacterium species, and 3 isolates of S. maltophilia. Clinical response to therapy. Table 2 summarizes the clinical responses to therapy. The overall favorable clinical response rate of the clinaoxacin group was 32% (88 of 272 patients), which was not signicantly different from the overall response rate (89 [33%] of 269 patients) of the imipenem group. The response rate of patients with bacteremia was higher for those treated with imipenem (21 [27%] of 79) than for those treated with clinaoxacin (13 [18%] of 74; P p .24). The response rates for localized bacterial infections, clinically documented infections, and possible infections were similar for each treatment group. The median time to defervescence was 4 days for both treatment groups. The most common reason for failure of initial monotherapy with clinaoxacin or imipenem was the addition of other antimicrobial agents (table 2). A total of 171 (63%) of 272 patients in the clinaoxacin group and 162 (60%) of 269 patients in the imipenem group were given additional antimicrobial agents for improvement. The most common drugs used for modication of initial therapy were vancomycin (90 [33%] of 272 clinaoxacin recipients vs. 79 [29%] of 269 imipenem recipients), aminoglycosides (53 [19%] of 272 clinaoxacin recipients vs. 47 [17%] of 269 imipenem recipients), and cephalosporins

Clinaoxacin vs. Imipenem in Granulocytopenia CID 2001:32 (1 February) 383

Table 1. Characteristics of patients in study of clinaoxacin versus imipenem as empirical monotherapy for febrile granulocytopenic patients.
Clinaoxacin (n p 272) 47 (1986) 126/146 92 (34) 31 (11) 94 (35) 55 (20) 0 181 (67) 76 (28) 15 (5) 5.0 (131) 9.3 (143) 37 (14) 70 (26) 221 (82) 74 (27) 27 (10) 4 (2) 167 (61) 10 (351) Imipenem (n p 269) 47 (1876) 135/134 82 (30) 31 (12) 100 (37) 53 (20) 3 (1) 174 (65) 87 (32) 8 (3) 4.8 (119) 9.3 (136) 27 (10) 67 (25) 208 (77) 79 (29) 26 (10) 7 (3) 157 (58) 10 (242)

Characteristic Age, mean y (range) Sex, no. male/no. female Underlying disease Hematologic malignancy Solid tumor Bone marrow or stem cell transplant Hematologic malignancy Solid tumor Unknown Pretherapy granulocyte count, cells/mm3
!100

100500
1500

Duration of granulocytopenia, mean d (range)

!100 cells/mm3 !500 cells/mm3

Persistent granulocytopenia, cells/mm3

!100 !500

Treatment with hematopoietic growth factor Etiology of initial fever Bacteremia Localized bacterial infection Clinically documented infection Possible infection Duration of therapy with study drug, median d (range)
NOTE.

Data are no. (%) of patients unless otherwise indicated.

(27 [10%] of 272 clinaoxacin recipients vs. 23 [9%] of 269 imipenem recipients). Overall, 259 (95%) of the 272 patients in the clinaoxacin group and 251 (93%) of the 269 patients in the imipenem group improved either while being treated with study drug alone or after modication of therapy to include other antimicrobial agents. Only 13 treatment failures (5%) in the clinaoxacin group and 18 (7%) in the imipenem group were caused by death of the patient during the study. Only 1 clinaoxacin recipient and 1 imipenem recipient died from a baseline infection. The clinaoxacin recipient was treated with study drug for fever related to possible but undocumented infection and died of multiorgan failure and a clinical syndrome consistent with sepsis. The imipenem recipient died of septic shock related to baseline bacteremia caused by Klebsiella pneumoniae, viridans group streptococci, and coagulase-negative staphylococci. Other causes of death in the clinaoxacin group were clinical sepsis with negative cultures (1 patient), fungal superinfection (4 patients), toxicity from chemotherapy (4), underlying ma384 CID 2001:32 (1 February) Winston et al.

lignancy (2), and hemorrhage (1). Other causes of death in the imipenem group were fungal superinfection (4), toxicity from chemotherapy (4), underlying malignancy (6), hemorrhage (2), and graft-versus-host disease (1). Response by bacterial pathogen. Table 3 compares the response to therapy according to bacterial pathogen. Among 286 baseline pathogens, 159 (56%) were gram-negative organisms and 127 (44%) were gram-positive bacteria. Escherichia coli and Klebsiella species were the most common gram-negative baseline pathogens, whereas coagulase-negative staphylococci and viridans group streptococci were the most common gram-positive baseline pathogens. Eighteen (23%) of 79 gram-negative infections in the clinaoxacin group and 20 (25%) of 80 gram-negative infections in the imipenem group improved during treatment with study drug alone. Most other gram-negative infections in each treatment group improved after modication with additional antimicrobial agents. However, more gram-negative infections in the imipenem group (15%) were associated with death during

Table 2. Clinical responses to therapy among febrile granulocytopenic patients treated with clinaoxacin or imipenem.
Response, type of infection Improvement without modication Bacteremia Localized bacterial Clinically documented Possible infection Failure Modication required for improvement Death
NOTE. (%).
a

Clinaoxacin (np 272) 88 (32) 13/74 (18) 11/27 (41) 1/4 (25) 63/167 (38)

Imipenem (n p 269) 89 (33) 21/79 (27) 11/26 (42) 2/7 (29) 55/157 (35)
a

171 (63) 13 (5)

162 (60) 18 (7)

Data are no. of patients (%) or no. improved/total with infection

P p .24 vs. clinaoxacin patients with bacteremia.

the study than were gram-negative infections in the clinaoxacin group (5%; P p .06; table 3). Similarly, although most gram-positive infections in each treatment group responded to either the study drug alone (20% for clinaoxacin vs. 33% for imipenem) or subsequent modied therapy with other agents (80% for clinaoxacin vs. 54% for imipenem), more grampositive infections were associated with death in the imipenem group (13%) than in the clinaoxacin group (none) during the study (P p .003; table 3). Overall, 20 documented bacterial infections in the imipenem group compared with only 4 in the clinaoxacin group were associated with death during the study (P p .001; table 3). Nevertheless, except for 1 case of bacteremia in an imipenem recipient caused by K. pneumoniae, viridans group streptococci, and coagulase-negative staphylococci, none of these gram-negative or gram-positive infections were considered by investigators (blinded to the study drug) to be the primary cause of death. No baseline organism susceptible to clinaoxacin or imipenem before treatment developed resistance to that study drug during therapy. Response related to duration of granulocytopenia. The response of patients was also evaluated in relation to the duration of granulocytopenia (granulocyte count 500 cells/ mm3) during the study. Among patients with granulocytopenia 7 days duration after starting the study drug, the response rates were 53 (34%) of 157 patients in the clinaoxacin group and 57 (33%) of 172 patients in the imipenem group. Similarly, among patients with granulocytopenia 17 days duration after starting the study drug, the response rates were 35 (30%) of 115 patients in the clinaoxacin group and 32 (33%) of 97 patients in the imipenem group. None of these differences was signicant. Superinfections. The incidence of superinfections was 11% in clinaoxacin recipients and 10% in imipenem recipients (table 4). Except for more superinfections caused by coagulase-

negative staphylococci in patients treated with imipenem (8 [3%] of 269 vs. 3 [1%] of 272; P p .14), the types of bacterial and fungal organisms causing superinfections were similar for each study drug. Candida and Aspergillus species were the most common fungal organisms causing superinfection. Three (19%) of 16 bacterial organisms causing superinfection in the clinaoxacin group and 7 (47%) of 15 bacterial organisms causing superinfection in the imipenem group were resistant to clinaoxacin and imipenem, respectively (P p .13). Antibiotic-related adverse events. A total of 85 (32%) of 270 patients in the clinaoxacin group and 69 (26%) of 265 in the imipenem group experienced adverse events possibly or denitely related to the study drug (P p .18 ). The most common and signicant adverse events associated with each study drug are shown in table 5. Skin rashes were more common in patients treated with clinaoxacin. On the other hand, nausea and diarrhea associated with Clostridium difcile occurred more often in patients receiving imipenem. Seizures related to study drug occurred only in patients treated with imipenem. Drugrelated nephrotoxicity or liver dysfunction was very uncommon with either study drug. One patient treated with clinaoxacin experienced a moderate photosensitivity reaction involving the face on day 3 of therapy, which resolved after the clinaoxacin was discontinued. Drug-related hypoglycemia (serum glucose level, 4467 mg/dL) occurred in only 4 clinaoxacin recipients and was symptomatic (diaphoresis and shakiness) in 2. Drugrelated adverse events caused discontinuation of study drug in 13 clinaoxacin recipients (5%) and in 21 imipenem recipients (8%). Skin rash (7 patients), pruritus (1), nausea (1), vomiting (1), altered mental status (1), hypotension (1), and a photosensitivity reaction (1) were the drug-related adverse events leading to discontinuation of clinaoxacin. Skin rash (11 patients), nausea (4), diarrhea (1), seizures (3), fever (1), and altered mental status (1) were the drug-related adverse events leading to discontinuation of imipenem.

DISCUSSION Recently, several randomized, controlled trials have demonstrated that monotherapy with newer, extended-spectrum bactericidal cephalosporins (ceftazidime, cefepime) or carbapenems (imipenem, meropenem) is an effective alternative to traditional combination therapy in febrile granulocytopenic patients [26, 10, 12]. Potential advantages of monotherapy are less toxicity, decreased costs, and simplicity of administration. Consequently, at many oncology centers, single-agent therapy with ceftazidime or imipenem is now standard for suspected bacterial infection in the granulocytopenic patient. Use of the uoroquinolones as monotherapy for the febrile granulocytopenic patient has been controversial. Initial trials in small numbers of febrile granulocytopenic patients with bac-

Clinaoxacin vs. Imipenem in Granulocytopenia CID 2001:32 (1 February) 385

Table 3.

Response to therapy with clinaoxacin or imipenem by bacterial pathogen.

Clinaoxacin therapy Failure caused by Imipenem therapy Failure caused by Death Improvement Modication of therapy Death

Pathogen Gram-negative

Improvement

Modication of therapy

Escherichia coli Klebsiella species Enterobacter species Serratia species Pseudomonas aeruginosa
Othera Total Gram-positive C-N staphylococci

3/29 3/17 2/7 1/2 1/4 8/20 18/79 (23) 5/25 1/9 1/12 3/5 1/4 1/6 1/3 13/64 (20) 31/143 (22)

25/29 12/17 5/7 1/2 2/4 12/20 57/79 (72) 20/25 8/9 11/12 2/5 3/4 5/6 2/3 51/64 (80) 108/143 (75)

1/29 2/17 0/7 0/2 1/4 0/20 4/79 (5) 0/25 0/9 0/12 0/5 0/4 0/6 0/3 0/64 4/143 (3)

8/34 4/22 2/6 1/2 0/1 5/15 20/80 (25) 6/19 2/5 5/19 5/9 1/6 0/1 2/4 21/63 (33) 41/143 (29)

20/34 14/22 3/6 1/2 1/1 9/15 48/80 (60) 10/19 2/5 11/19 4/9 4/6 1/1 2/4 34/63 (54) 82/143 (57)

6/34 4/22 1/6 0/2 0/1 1/15 12/80 (15)b 3/19 1/5 3/19 0/9 1/6 0/1 0/4 8/63 (13)
d

Staphylococcus aureus
Viridans group streptococci

Enterococcus species
Other streptococci

Corynebacterium species
Other
c

Total Grand total

NOTE.
a

20/143 (14)e

Data are no. of patients who improved or experienced treatment failure/total (%). C-N, coagulase-negative.

Clinaoxacin group: Citrobacter species (3), Moraxella catarrhalis (3), Fusobacterium species (3), Proteus mirabilis, Pseudomonas uorescens, Pseudomonas putida, Stenotrophomonas maltophilia, Capnocytophaga species, Aeromonas sobria, Campylobacter minuti, Neisseria elongata, aerobic gram-negative bacilli (3). Imipenem group: Proteus species (3), Acinetobacter species (2), M. catarrhalis (2), Aeromonas species (2), Citrobacter species, Campylobacter jejuni, aerobic gram-negative bacilli (4). b P p .06 vs. clinaoxacin. c Clinaoxacin group: Bacillus species (2), aerobic gram-positive cocci. Imipenem group: Bacillus species, Micrococcus species, Lactobacillus species, aerobic gram-positive cocci. d P p .003 vs. clinaoxacin. e P p .001 vs. clinaoxacin.

teremias, pneumonias, and soft-tissue infections suggested that single-agent therapy with iv ciprooxacin is effective [24, 33, 34]. However, a subsequent prospective, randomized comparison of iv ciprooxacin monotherapy with a combination regimen of piperacillin plus amikacin done by the European Organization for Research and Treatment of Cancer (EORTC) found that ciprooxacin was less effective as empirical therapy for febrile granulocytopenic patients [35]. This study was discontinued prematurely because patients treated with ciprooxacin alone had a signicantly lower response rate (31 [65%] of 48 patients) than did patients treated with piperacillin plus amikacin (48 [91%] of 53 patients; P p .002). Patients with gram-positive bacteremia had a particularly poor outcome. Similar trials evaluating oral ooxacin alone versus combinations of parenteral antibiotics in febrile granulocytopenic patients also provided inconclusive results [36, 37]. Consequently, current guidelines established by the Infectious Diseases Society of America and other experts for use of antimicrobial agents
386 CID 2001:32 (1 February) Winston et al.

in granulocytopenic patients have not recommended quinolone monotherapy [1, 25]. Clinaoxacin is much more active relative to ciprooxacin, other older uoroquinolones, and b-lactam drugs against grampositive bacteria, while maintaining excellent activity for P. aeruginosa and other gram-negative bacillary pathogens [2123]. Clinaoxacin alone was found to be as effective as imipenem alone for empirical therapy, in this present doubleblind trial, which is the largest trial of quinolone monotherapy in febrile granulocytopenic patients ever reported. A total of 99% of all gram-positive and gram-negative baseline pathogens were susceptible to clinaoxacin, and there were no substantial differences in the overall clinical responses to treatment with clinaoxacin (32%) or imipenem (33%). As in other multicenter trials of empirical antibiotic therapy among febrile granulocytopenic patients [12, 18, 26], most patients in this trial had modication of their monotherapy to achieve a successful outcome. Nonetheless, only 1 patient of each treatment group

Table 4. Superinfections associated with therapy with clinaoxacin or imipenem in febrile granulocytopenic patients.
Clinaoxacin (n p 272) Imipenem (n p 269)
a

Type of infection Gram-positive bacteremia Coagulase-negative staphylococci

3 1 0 2 1 (1) 2 1 1 1 1 1 0 1 (1) 1 (1) 5 5 6 1 1 1 0 30 (11) 35

8 (3) 0 1 0 4 (2) 0 0 0 0 0 0 2 (2) 0 0 2 5 4 0 0 0 1

Staphylococcus aureus Stomatococcus species

Viridans group streptococci

Enterococcus species Corynebacterium species Bacillus species Lactobacillus species Clostridium ramosum
Gram-negative bacteremia

Escherichia coli Pseudomonas aeruginosa Stenotrophomonas maltophilia Fusobacterium nucleatum Leptorichia buccalis
Fungal infections Localized candidiasis Systemic candidiasis Invasive aspergillosis Localized Fusarium infection Disseminated Fusarium infection Disseminated Cryptococcus infection Disseminated Trichosporon infection Total no. of patients with superinfection (%) Total no. of superinfections

26 (10) 27

NOTE. Data in parentheses indicate isolates resistant to study drug (clinaoxacin, MIC 12 mg/mL; imipenem, MIC 18 mg/mL).
a

P p .14 vs. clinaoxacin.

died from their baseline infection, and overall mortality during the study was low in both groups (13 [5%] of 272 with clinaoxacin, 18 [7%] of 269 with imipenem). The initial response rates observed for both groups in this study (88 [32%] of 272 with clinaoxacin, 89 [33%] of 269 with imipenem) were relatively low. However, in accordance with recent guidelines from the US Food and Drug Administration and previously proposed by the Immunocompromised Host Society and the Infectious Diseases Society of America, any modication of the study antibiotic regimen was considered a treatment failure [3840]. In many cases, additional antibiotics were given despite negative results of follow-up cultures and the absence of any clinical deterioration. Using similar criteria for failure, other recent trials evaluating ceftazidime, imipenem, meropenem, or ciprooxacin either as monotherapy or as part of a combination regimen have also observed rates

of response as low as 21%54% among patients whose therapy was not modied [7, 11, 12, 18, 26]. Furthermore, cefepime recently became the rst antibiotic approved by the US Food and Drug Administration for monotherapy for febrile granulocytopenic patients on the basis of pooled response rates without modication of therapy of 34% and pooled response rates with modication of therapy of 93% [41]. These response rates with cefepime are similar to the response rates seen with clinaoxacin and imipenem in this study. Finally, in contrast to most other studies, our study was double-blinded, which prevented primary physicians from knowing which study drug was being given to a patient. This blinding could have contributed to more frequent modications of therapy by physicians who were uncomfortable about the antibiotic that was being used to treat seriously ill patients who did not show immediate improvement. Expanded use of the uoroquinolones worldwide has been associated with a gradual emergence of resistance among P. aeruginosa, E. coli, and other organisms in certain countries [42]. Such resistance needs to be considered when evaluating the appropriateness of empirical monotherapy with clinaoxacin or other newer quinolones at a particular institution. There is also concern that single-agent therapy with a uoroquinolone for febrile granulocytopenic patients could foster the development and spread of resistant organisms. Indeed, when uoroquinolones have been used for prophylaxis of infection in granulocytopenic patients, there has been a reduction in gram-negative bacillary infections at the expense of an increased risk for gram-positive bacterial infections caused by quinoloneresistant organisms, especially viridans group streptococci [24, 43]. Occasional emergence of quinolone-resistant gram-negative bacilli has also been observed in patients given quinolone prophylaxis [44]. In this study, in which clinaoxacin was used for empirical therapy rather than prophylaxis of infection, 99% of all baseline pathogens were sensitive to clinaoxacin and none developed resistance to clinaoxacin during treatment. Furthermore, only 3 (1.1%) of 272 patients had superinfections caused by bacterial organisms resistant to clinaoxacin. Thus, reserving use of a broad-spectrum uoroquinolone such as clinaoxacin for treatment rather than prevention of infection may be associated with less emergence of quinolone-resistant organisms. Of note, excluding patients from this study who previously took a quinolone for prophylaxis may have also contributed to the low incidence of quinolone resistance and a greater number of baseline gram-negative infections relative to gram-positive infections (table 3). Both clinaoxacin and imipenem were generally well tolerated in this trial. Recently, reports of severe liver injury with trovaoxacin and of cardiac arrhythmias with grepaoxacin have raised concerns about the safety of newer uoroquinolones [45, 46]. A third uoroquinolone, temaoxacin, was re-

Clinaoxacin vs. Imipenem in Granulocytopenia CID 2001:32 (1 February) 387

moved from the market in the early 1990s because of the unusual occurrence of an immune hemolytic anemia syndrome [47]. We were unable to note any hepatic, cardiac, or hematologic toxicity related to clinaoxacin in this study. A thorough review of liver-related adverse events, cardiac arrhythmias, and hematologic abnormalities in all investigational clinical trials involving clinaoxacin also found a similar occurrence of these adverse events among patients treated with clinaoxacin and those receiving comparative drugs [48]. The higher incidence of gastrointestinal adverse events (nausea, C. difcileassociated diarrhea) and seizures among patients receiving imipenem in this study has also been reported in other trials that evaluated imipenem in febrile granulocytopenic patients [58, 17]. Hypoglycemic reactions and photosensitivity, which occurred in 4 and 1 clinaoxacin recipients, respectively, in this study (table 5), appear to happen more frequently among patients treated with clinaoxacin than among patients treated with other classes of antimicrobial agents. Clinaoxacin and other quinolones can stimulate insulin release and thereby induce hypoglycemia [49]. Quinolones also absorb light in the UV spectrum, which can cause degradation of the drug and the release of toxic free radicals and oxygen products [50]. The overall incidences of drug-related hypoglycemia and phototoxicity in 11500 patients treated with clinaoxacin during investigational clinical trials were 2.5% and 2.1%, respectively [48]. Phototoxicity has been noted more often in patients treated with oral clinaoxacin (16% incidence) than in patients receiving iv clinaoxacin (1% incidence). The very low incidence of phototoxicity observed with iv clinaoxacin in this study was likely related to its exclusive use in seriously ill hospitalized patients with limited sunlight exposure. The overall mortality rate during this study was similar for the clinaoxacin and imipenem groups. However, more microbiologically documented baseline infections in the imipenem group (20 patients) than in the clinaoxacin group (4 patients) were associated with death during the study (table 3). Only 1 of these baseline infections (a case of polymicrobial bacteremia in the imipenem group) was considered by investigators blinded to the study drug to be the primary cause of death. Thus, although the lower mortality rate among patients with microbiologically documented baseline infections treated with clinaoxacin is noteworthy, other noninfectious factors may have been responsible for the different mortality rates of the 2 treatment groups. Another potential use for quinolone monotherapy is the outpatient treatment of low-risk patients with fever and granulocytopenia. Low-risk patients with fever and granulocytopenia have been characterized as granulocytopenic patients with stable cancer who develop fever as outpatients, have no serious concurrent underlying condition, and are expected to have a duration of granulocytopenia 7 days [51]. In this situation,
388 CID 2001:32 (1 February) Winston et al.

Table 5. Common and signicant adverse events associated with study drugs.
Clinaoxacin (n p 270) 30 (11) 6 (2) 6 (2) 9 (3) 7 (3) 4 (2) 2 (1) 0 1 (0.4) 4 (2) Imipenem (n p 265) 17 (6) 12 (5) 4 (2) 22 (8) 2 (1) 1 (0.4) 1 (0.4) 4 (2) 0 0

Adverse event Skin rash Nausea Vomiting Diarrheaa Hypokalemia Increased serum creatinine Increased liver function tests Seizures Photosensitivity Hypoglycemia
NOTE.
a

P
.07 .16 .75 .02 .18 .37 1.00 .06 1.00 .12

Data are no. (%) of patients.

Associated with Clostridium difcile infection.

oral therapy with a combination of ciprooxacin plus amoxicillin-clavulanate or ooxacin alone has been reported to be as effective as iv therapy [36, 37, 52, 53]. Most patients in our study, which included many patients who had undergone transplantation or who had advanced malignancy, were high-risk patients. The broader antibacterial spectrum of clinaoxacin relative to ciprooxacin and other quinolones, however, could facilitate the outpatient management of low-risk febrile granulocytopenic patients. Levooxacin, an approved quinolone with an antibacterial spectrum similar to clinaoxacin, is currently being evaluated in randomized controlled trials as monotherapy for both inpatients and low-risk outpatients with fever and granulocytopenia [54]. In summary, the results of this large, randomized, doubleblind trial suggest that monotherapy with clinaoxacin is a suitable option for empirical treatment of febrile granulocytopenic patients. Patients who are intolerant of b-lactam drugs because of hypersensitivity or bacterial resistance or of the aminoglycosides due to toxicity may especially benet from clinaoxacin monotherapy. PARTICIPATING CENTERS AND INVESTIGATORS Principal Investigator(s) (Institution): Drs. Drew J. Winston and Mary C. Territo (UCLA Medical Center); Drs. Hillard M. Lazarus and Jeffrey L. Blumer (Case Western Reserve University Hospital, Cleveland); Dr. Roy A. Beveridge (Fairfax Hospital, Falls Church, VA); Dr. James W. Hathorn (Duke University Medical Center, Durham, NC); Dr. Rasim Gucalp (Monteore Medical Center, Bronx, NY); Dr. Reuben Ramphal (University of Florida Medical Center, Gainesville, FL); Drs. Anthony W. Chow and H. Grant Stiver (Vancouver Hospital, Vancouver, British Columbia, Canada); Dr. Winston G. Ho (St. Josephs

Hospital, Orange, CA); Drs. Ruth Horn and Hugh G. Robson (Royal Victoria Hospital, Montreal); Drs. Ronald Feld and Alison McGeer (Princess Margaret Hospital, Toronto); Drs. Thomas J. Louie, David Megran, and Ronald Read (Calgary General Hospital, Calgary, Alberta, Canada); Dr. Nigar Kirmani (Loyola University Medical Center, Maywood, IL); Dr. Julio Alberto Ramirez (Veterans Hospital, Louisville, KY); Dr. Eric J. Bow (University of Manitoba Hospital, Winnipeg, Canada); Dr. James S. Tan (Akron City Hospital, Akron, OH). Study Coordinator (Institution): Ann Gogesch, RN (UCLA Medical Center); Eloise Lemon, RN, Sandy Fiala, RN (Case Western Reserve University Hospital, Cleveland); Kristin Friedman, RN (Monteore Medical Center, Bronx, NY); Kenna Sleigh, RN, MSN (Vancouver Hospital, Vancouver, British Columbia, Canada); Charlene Barker, RN (Royal Victoria Hospital, Montreal); Trish Muir, RN (Calgary General Hospital, Calgary, Alberta, Canada).

12.

13.

14.

15.

16.

17.

Acknowledgment

18.

We thank Katharine Fry for manuscript preparation.

19.

References
1. Hughes WT, Armstrong D, Bodey GP, et al. 1997 guidelines for the use of antimicrobial agents in neutropenic patients. Clin Infect Dis 1997; 25:55173. 2. Pizzo PA, Hathorn JW, Hiemenz J, et al. A randomized trial comparing ceftazidime alone with combination antibiotic therapy in cancer patients with fever and neutropenia. N Engl J Med 1986; 315:5528. 3. De Pauw BE, Deresinski SC, Feld R, et al. Ceftazidime compared with piperacillin and tobramycin for the empiric treatment of fever in neutropenic patients with cancer: a multicenter randomized trial. Ann Intern Med 1994; 120:83444. 4. Norrby SR, Vandercam B, Louie T, et al. Imipenem/cilastatin versus amikacin plus piperacillin in the treatment of infections in neutropenic patients: a prospective, randomized multi-clinic study. Scand J Infect Dis Suppl 1987; 52:6578. 5. Winston DJ, Ho WG, Bruckner DA, et al. b-Lactam antibiotic therapy in febrile granulocytopenic patients: a randomized trial comparing cefoperazone plus piperacillin, ceftazidime plus piperacillin, and imipenem alone. Ann Intern Med 1991; 115:84959. 6. Rolston KVI, Berkey P, Bodey GP, et al. A comparison of imipenem to ceftazidime with or without amikacin as empiric therapy in febrile neutropenic patients. Arch Intern Med 1992; 152:28391. 7. Freifeld AG, Walsh T, Marshall D, et al. Monotherapy for fever and neutropenia in cancer patients: a randomized comparison of ceftazidime versus imipenem. J Clin Oncol 1995; 13:16576. 8. Winston DJ, Bartoni J, Bruckner DA, et al. Randomized comparison of sulbactam/cefoperazone with imipenem as empirical monotherapy for febrile granulocytopenic patients. Clin Infect Dis 1998; 26:57683. 9. Biron P, Fuhrmann C, Cure H, et al. Cefepime versus imipenemcilastatin as empirical monotherapy in 400 febrile patients with short duration neutropenia. J Antimicrob Chemother 1998; 42:5118. 10. Yamamura D, Gucalp R, Carlisle P, et al. Open randomized study of cefepime versus piperacillin-gentamicin for treatment of febrile neutropenic cancer patients. Antimicrob Agents Chemother 1997; 41: 17048. 11. Meropenem Study Group of Leuven, London, and Nijmegen. Equivalent efcacies of meropenem and ceftazidime as empirical mono-

20.

21.

22.

23.

24.

25. 26.

27.

28.

29.

30. 31.

therapy of febrile granulocytopenic patients. J Antimicrob Chemother 1995; 36:185200. Cometta A, Calandra T, Gaya H, et al. Monotherapy with meropenem versus combination therapy with ceftazidime plus amikacin as empiric therapy for fever in granulocytopenic patients with cancer. Antimicrob Agents Chemother 1996; 40:110815. Ramphal R, Bolger M, Oblon AJ, et al. Vancomycin is not an essential component of the initial empiric treatment regimen for febrile neutropenic patients receiving ceftazidime: a randomized prospective study. Antimicrob Agents Chemother 1992; 36:10627. Viscoli C, Moroni C, Boni L, et al. Ceftazidime plus amikacin versus ceftazidime plus vancomycin as empiric therapy in febrile neutropenic children with cancer. Rev Infect Dis 1991; 13:397404. Cometta A, Zinner S, De Bock R, et al. Piperacillin-tazobactam plus amikacin versus ceftazidime plus amikacin as empiric therapy for fever in granulocytopenic patients with cancer. Antimicrob Agents Chemother 1995; 39:44552. Johnson MP, Ramphal R. b-Lactamresistant Enterobacter bacteremia in febrile neutropenic patients receiving monotherapy. J Infect Dis 1990; 162:9813. Bodey G, Abi-Said D, Rolston K, et al. Imipenem or cefoperazonesulbactam combined with vancomycin for therapy of presumed or proven infection in neutropenic cancer patients. Eur J Clin Microbiol Infect Dis 1996; 15:62534. Cordonnier C, Herbrecht R, Pico JL, et al. Cefepime/amikacin versus ceftazidime/amikacin as empirical therapy for febrile episodes in neutropenic patients: a comparative study. Clin Infect Dis 1997; 24:4151. Weiss ME, Adkinson NF. b-Lactam allergy. In: Mandell GL, Bennett JE, Dolin R, eds. Principles and practice of infectious diseases. 4th ed. New York: Churchill Livingstone, 1995:2728. Carmeli Y, Samore MH, Huskins C. The association between antecedent vancomycin treatment and hospital-acquired vancomycin-resistant enterococci: a meta-analysis. Arch Intern Med 1999; 159:24618. Wise R, Ashby JP, Andrews JM. In vitro activity of PD 127391, an enhanced-spectrum quinolone. Antimicrob Agents Chemother 1988; 32:12516. Cohen MA, Huband MD, Gage JW, et al. In vitro activity of clinaoxacin, trovaoxacin, and ciprooxacin. J Antimicrob Chemother 1997; 40:20511. Ednie LM, Jacobs MR, Appelbaum PC. Comparative activities of clinaoxacin against gram-positive and gram-negative bacteria. Antimicrob Agents Chemother 1998; 42:126973. Winston DJ. Use of quinolone antimicrobial agents in immunocompromised patients. In: Hooper DC, Wolfson JS, eds. Quinolone antimicrobial agents. 2d ed. Washington, DC: American Society for Microbiology, 1993:43588. Pizzo PA. Management of fever in patients with cancer and treatmentinduced neutropenia. N Engl J Med 1993; 328:132332. Peacock JE, Wade JC, Lazarus HM, et al. Ciprooxacin/piperacillin vs. tobramycin/piperacillin as empiric therapy for fever in neutropenic cancer patients, a randomized double-blind trial. In: Program and abstracts of the 37th Interscience Conference on Antimicrobial Agents and Chemotherapy (Toronto). Washington, DC: American Society for Microbiology, 1997:373. Liang R, Yung R, Chiu E, et al. Ceftazidime versus imipenem-cilastatin as initial monotherapy for febrile neutropenic patients. Antimicrob Agents Chemother 1990; 34:133641. Murray PR, Baron EJ, Pfaller MA, et al., eds. Manual of clinical microbiology. 6th ed. Washington, DC: American Society for Microbiology, 1995. Performance standards for antimicrobial susceptibility testing, fth international supplement. Wayne, PA: National Committee for Clinical Laboratory Standards, 1994; document M100-55. SAS users guide: statistics. Version 6 edition. Cary, NC: SAS Institute, 1990. Harris EK, Albert A, eds. Survivorship analysis for clinical studies. New York: Marcel Dekker, 1991:512, 5373, 7996.

Clinaoxacin vs. Imipenem in Granulocytopenia CID 2001:32 (1 February) 389

32. Breslow NE, Day NE. Statistical methods in cancer research. Vol. 1: the analysis of case-control studies. Lyon, France: International Agency for Research on Cancer, 1980. 33. Rolston KVI, Haron E, Cunningham C, et al. Intravenous ciprooxacin for infections in cancer patients. Am J Med 1989; 87(Suppl 5A): 261S5S. 34. Johnson PRE, Liu Yin JA, Tooth JA. A randomized trial of high-dose ciprooxacin versus azlocillin and netilmicin in the empirical therapy of febrile neutropenic patients. J Antimicrob Chemother 1992; 30: 20314. 35. Meunier F, Zinner SH, Gaya H, et al. Prospective randomized evaluation of ciprooxacin versus piperacillin plus amikacin for empiric antibiotic therapy of febrile granulocytopenic cancer patients with lymphomas and solid tumors. Antimicrob Agents Chemother 1991; 35: 8738. 36. Malik IA, Abbas Z, Karim M. Randomized comparison of oral ooxacin alone with combination of parenteral antibiotics in neutropenic febrile patients. Lancet 1992; 339:10926. 37. Malik IA, Khan WA, Karim M, et al. Feasibility of outpatient management of fever in cancer patients with low-risk neutropenia: results of a prospective randomized trial. Am J Med 1995; 98:22431. 38. Guidance for industry. Empiric therapy of febrile neutropeniadeveloping antimicrobial drugs for treatment. Draft document. Rockville, MD: US Food and Drug Administration, 1998. 39. Immunocompromised Host Society. Design, analysis, and reporting of clinical trials on the empirical antibiotic management of the neutropenic patient. J Infect Dis 1990; 161:397401. 40. Hughes WT, Pizzo PA, Wade JC, et al. Evaluation of new anti-infective drugs for the treatment of febrile episodes in neutropenic patients. Clin Infect Dis 1992; 15(Suppl 1):S20615. 41. Product information: Cefepime [package insert]. Princeton, NJ: Bristol-Myers Squibb, 1998. 42. Hooper DC. New uses for new and old quinolones and the challenge of resistance. Clin Infect Dis 2000; 30:24354.

43. Bochud PY, Calandra T, Francioli P. Bacteremia due to viridans streptococci in neutropenic patients: a review. Am J Med 1994; 97:25664. 44. Kern WV, Andriof E, Oethinger M, et al. Emergence of uoroquinolone-resistant Escherichia coli at a cancer center. Antimicrob Agents Chemother 1994; 38:6817. 45. Lumpkin MM. Trovan (trovaoxacin/alatrooxain mesylate). Public health advisory. Rockville, MD: US Food and Drug Administration, 9 June 1999. 46. Kent RS. Raxar (grepaoxacin) product withdrawal [letter]. Zebulon, NC: GlaxoWellcome, 28 October 1999. 47. Blum MD, Graham DL, McCloskey CA. Temaoxacin syndrome: review of 95 cases. Clin Infect Dis 1994; 18:94650. 48. Tack KJ. Clinaoxacin new drug application studies. Ann Arbor, MI: Parke-Davis, 1999. 49. Maeda N, Tamagawa T, Niki I, et al. Increase in insulin release from rat pancreatic islets by quinolone antibiotics. Br J Pharmacol 1996; 117:3726. 50. Gondorelli G, De Guidi G, Giuffrida S, et al. Membrane and DNA damage photosensitized by uoroquinolone antimicrobials agents: a comparative screening. Medecine Biologie Environnement 1996; 24: 103110. 51. Talcott JA, Finberg R, Mayer RJ, et al. The medical course of cancer patients with fever and neutropenia: clinical identication of a lowrisk subgroup at presentation. Arch Intern Med 1988; 148:25618. 52. Freifeld A, Marchigiani D, Walsh T, et al. A double-blind comparison of empirical oral and intravenous antibiotic therapy for low-risk febrile patients with neutropenia during cancer chemotherapy. N Engl J Med 1999; 341:30511. 53. Kern WV, Cometta A, De Bock R, et al. Oral versus intravenous empirical antimicrobial therapy for fever in patients with granulocytopenia who are receiving cancer chemotherapy. N Engl J Med 1999; 341:3128. 54. Davis R, Bryson HM. Levooxacin. A review of its antibacterial activity, pharmacokinetics, and therapeutic efcacy. Drugs 1994; 47:677700.

390 CID 2001:32 (1 February) Winston et al.