- Email: [email protected]
Prospective, randomized comparison of sequential intravenous followed by oral ciprofloxacin with intravenous ceftazidime in the treatment of serious infections
Prospective, Randomized Comparison of Sequential Intravenous followed by Oral Ciprofloxacin with Intravenous Ceftazidime the Treatment of Serious Infections
in
JAMESE. PEACOCK,JR.,M.D., P. SAMUELPEGRAM,M.D., STEPHENF. WEBER,M.D., PETER A. LEONE, M.D.
Winston-Salem, North Caroiina
In a prospective, comparative trial, 47 hospitalized patients with serious infections that required parenteral antibiotic therapy were randomly assigned to receive either ciprofloxacin (200 mg every 12 hours intravenously followed by 500 mg every 12 hours orally at a time dependent on the patients’ clinical and bacteriologic responses) or ceftazidime (2 g every eight to 12 hours intravenously). All evaluable subjects (39 patients) had documented infections, 23 percent of which were associated with bacteremia. The mean/median duration of intravenous antibiotic use for ciprofloxacin was 7.37ifive days and for ceftazidime 9.95keven days; 63 percent of the ciprofloxacin patients received an additional 1’7 days of oral therapy with ciprofloxacin, whereas intravenous therapy with ceftazidime was followed by an average of 12 days of an oral regimen in 55 percent of patients. Overall response rates for patients receiving ciprofloxacin and ceftazidime were 76 percent (16 of 21) and 82 percent (18 of 221, respectively. Four out of five bacteremias in each group were successfully treated. Overall, 69 percent of the pathogens were gram-negative aerobes, and 47 percent of the infections involved the urinary tract. Failure of therapy was most often associated with pneumonia (two of five failures with ciprofloxacin and three of four failures with ceftazidime). Adverse effects occurred in approximately 20 percent of patients in each group and were mild and reversible. Superinfections occurred in five of 19 (26 percent) ciprofloxacin recipients and seven of 20 (35 percent) ceftazidime recipients. All fungal superinfections involved the genitourinary tract and occurred most often in association with chronic indwelling catheters. Enterococcal superinfections occurred in both groups (a bacteremic urinary tract infection in a ceftazidime patient and osteomyelitis in a ciprofloxacin patient). Clostridium difticile-associated diarrhea was documented in a ceftazidime recipient. The mean duration of hospitalization following the onset of antibiotic treatment was 10.45 days in the ciprofloxacin group and 12.95
From the Department of Medicine, Wake Forest University Medical Center, WinstonSalem, North Carolina. Requests for reprints should be addressed to Dr. James E. Peacock, Jr.. Section on Infectious Diseases, Depariment of Medicine, Wake Forest University Medical Center, Winston-Salem, North Carolina 27103.
November
days in the ceftazidime group. Sequential intravenous/oral ciprofloxacin was as safe and effective as intravenous ceftazidime in the treatment of infections due to susceptible gram-positive and gram-negative organisms.
ver the past decade, numerous extended-spectrum beta-la&am antibiotics have been introduced into clinical practice [l]. Foremost among these have been the so-called “third generation” cephalosporins with their enhanced spectrum of activity against most gram-negative aerobic bacteria [21. A major differentiating feature among these extended-spectrum cephalosporins has been their activity against Pseudomonas ael-uyinosa [a]. To date, only cefoperazone and ceftazidime have exhibited clinically useful antipseudomonal activity [2], a property that has encouraged their usage, especially in the nosocomial setting or among immunocompromised patients, situations in which P. aeruginosa is a frequent cause of infection. At our institution, ceftazidime has emerged as the “third-generation” cephalosporin of choice, primarily because of its superior activity against P. aeruginosa. Drawbacks to its usage have included the necessity for dosing every eight hours and the lack of an oral preparation with a similar spectrum of activity. This deficiency precludes completion of therapy as an outpatient and early discharge of patients. Ciprofloxacin, a quinolone carboxylic acid derivative, also has excellent in vitro activity against most aerobic gram-negative pathogens, including Serratia marcescens, Enterobacter sp., and P. aemcginosa [351. In contrast to ceftazidime, ciprofloxacin is very active against staphylococci, including methicillinresistant strains [3,5,6], but ceftazidime has superior activity versus streptococci that are not highly susceptible to ciprofloxacin [5,61. Neither ciprofloxacin nor ceftazidime is useful clinically in the therapy of systemic anaerobic or enterococcal infections [2,4-71. Unlike ceftazidime, ciprofloxacin is available in both parenteral and oral formulations, the latter being well absorbed from the gastrointestinal tract and yielding bactericidal serum levels higher than the minimal inhibitory concentrations of most targeted pathogens [71. A potential advantage of ciprofloxacin over ceftazidime may therefore be the early conversion from parenteral to oral therapy in serious infections [8]. This use of sequential intravenous/oral therapy is likely to result in fewer hospital days and thus lower health care costs for individual patients [g-11]. 0
30, 1989
The American Journal of Medicine
Volume 87 (suppl 5A]
5A-185s
SYMPOSIUM
ON CIPROFLOXACIN
/PEACOCK
ET AL
The present study was designed to compare the effectiveness and safety of sequential intravenous/oral ciprofloxacin with intravenously administered ceftazidime in the therapy of serious infections caused by gram-negative or gram-positive organisms susceptible to both drugs. In addition, the potential economic consequences of sequential intravenous/oral therapy were explored. PATIENTS AND METHODS The study was a prospective, controlled, randomized, non-blinded trial conducted at a tertiary care teaching hospital. The study protocol was reviewed and approved by the institutional human studies review board prior to its inception, and all patients provided written informed consent for study participation prior to entry. Subjects enrolled into the trial were hospitalized patients 18 years of age or older who required parenteral antibiotic therapy for severe bacterial infections due to gram-positive or gram-negative aerobes susceptible to the study drugs. Types of infections qualifying for enrollment included, but were not limited to, bacteremia, respiratory tract infection, skin or skin-structure infection, bone or joint infection, and intra-abdominal infection. Patients with urinary tract infections were enrolled only if the infection was considered to be complicated, or severe and recalcitrant to other antibiotics; caused by P. aeruginosa or other organisms resistant to conventional antibiotics; or presumed to be accompanied by bacteremia. Patients with multiple infection sites were eligible for enrollment if each site could be evaluated separately. Prior therapy with other antimicrobials did not preclude entry if a causative organism susceptible to both study drugs could be isolated from the infection site before initiating therapy. Patients previously enrolled in the study were not eligible for re-entry even if infection developed at a site different from that initially treated. Potential study candidates were excluded from entry if they were allergic to a study drug or related compound (including penicillin), were pregnant or nursing, had mild infections not requiring parenteral therapy, exhibited severe impairment of renal function (i.e., serum creatinine greater than 2.0 mg/dl or creatinine clearance less than 50 ml/minute/l.73 m’), or were not expected to survive 30 days post-treatment because of severe underlying disease (exclusive of infection status). Probable need for concomitant therapy with an antimicrobial agent with a spectrum of activity similar to that of the study drugs also precluded participation, but the use of clindamycin or metronidazole for suspected anaerobic infection, or the addition of nafcillin or vancomycin to ceftazidime treatment for suspected staphylococcal infection, was allowed. In addition, patients with the following selected infections or conditions were not eligible for enrollment: meningitis, endocarditis, acute pulmonary exacerbations of cystic fibrosis, infections in immunocompromised patients (i.e., granulocyte counts less than l,OOO/mm”), gross lower or upper urinary tract obstruction that was not relieved during therapy, severe neurogenic bladder disease, chronic indwelling urinary catheter, and pelvic infections. Patients who met enrollment criteria and provided written informed consent were assigned to one of the two drug regimens in accordance with a computer5A-186s
November
30, 1989
The American Journal of Medicine
generated randomization code provided by the study sponsor. Patients assigned to therapy with ciprofloxatin were routinely given 200 mg intravenously every 12 hours for two to five days, followed by 500 mg orally every 12 hours for the duration of therapy. For patients with severe infections caused by organisms only moderately susceptible to ciprofloxacin and resistant to most other conventional antibiotics, the intravenous dosage of ciprofloxacin could be increased to 300 mg every 12 hours and the oral dose to 750 mg every 12 hours. The conversion from intravenous to oral therapy was dependent on the severity of the infection, as well as the patients’ clinical and bacteriologic responses; the duration of therapy was individualized for each patient by one of the study investigators. Patients assigned to treatment with ceftazidime received 2 g intravenously every eight to 12 hours, with the individual dose determined by the nature of the infection under therapy. Intravenous therapy was continued for a minimum of four days but usually for a period of seven to 14 days. In patients who responded promptly to ceftazidime, intravenous treatment could be discontinued (after the minimum of four days of therapy) and an appropriate oral antimicrobial agent (except ciprofloxacin) substituted. The actual duration of intravenous therapy with ceftazidime, the choice of orally administered antimicrobials, and the duration of oral therapy were at the discretion of the investigator. For those patients receiving oral therapy after intravenous ceftazidime treatment, clinical and bacteriologic evaluation was performed at the conclusion of both intravenous and oral regimens. For patients developing renal insufficiency after enrollment in the study, published guidelines were utilized to adjust the dosages of both ciprofloxacin and ceftazidime. A complete clinical, laboratory, and bacteriologic examination was performed for each patient upon study entry. The clinical examination consisted of a complete medical history and physical examination by one of the investigators or his designate. Laboratory studies, including a complete blood cell count with white blood cell differential and platelet count, erythrocyte sedimentation rate measurement, serum chemistry tests, and complete urinalysis, were performed prior to initiating therapy and then every three to five days thereafter. In addition, a pregnancy test was performed on all women of childbearing potential. Additional laboratory or diagnostic tests or procedures were performed whenever indicated. Culture specimens were obtained from all potential infection sites within the 48 hours prior to the onset of therapy and thereafter as often as indicated based upon the clinical course of the disease process. However, in patients with infection of the urinary tract, culture specimens were routinely repeated on Day 3 or 4 of treatment. All isolated pathogens were tested for susceptibility to the study drugs using both the modified Kirby-Bauer procedure with 5-pg ciprofloxacin and 30-pg ceftazidime disks, and a standard twofold tube dilution technique using Mueller-Hinton media and an inoculum of approximately lo5 colony-forming units/ ml. Ciprofloxacin susceptibility was defined as a zone of inhibition of 21 mm or more and minimal inhibitory concentrations of 1 pgiml or less. Moderate susceptibility to ciprofloxacin was defined as a zone of inhibi-
Volume 87 (suppl 5A)
SYMPOSIUM
tion of 16 to 20 mm and minimal inhibitory concentrations of greater than 1 to 2 pgiml. At the completion of therapy, the physical examination and laboratory studies were repeated, and culture specimens were obtained from previously infected sites if material was still available. For patients with urinary tract infections, culture specimens of urine were obtained five to nine days and four weeks after therapy. Chest radiographic examinations were routinely obtained post-therapy in all patients with pneumonia. For study purposes, infections were classified as documented if patients exhibited typical clinical symptoms or signs appropriate for the site under consideration (e.g., purulent sputum production and infiltrates on radiographic examination of the chest in a patient with presumed pneumonia). Culture specimens of the site may or may not have been positive for a definite pathogen. Bacteriologically documented infections were those clinically documented infections that also had positive culture specimens for a definite pathogen isolated from the infection site. All patients with documented infections were considered to be evaluable for efficacy. Assessment of response to therapy was based upon clinical, radiologic, and bacteriologic parameters. A satisfactory clinical response was considered to have occurred if patients exhibited resolution of or improvement in symptoms and signs of infection (e.g., defervescence of fever, clearing of pus or other signs of inflammation, resolution of infiltrates on chest radiographs, and so on). Bacteriologic cure was defined as elimination of causative organisms from cultures of infection sites during treatment. In addition, for patients with urinary tract infections, a negative urine culture specimen at five to nine days of therapy was also required. Treatment failures encompassed those patients who did not exhibit either clinical response or bacteriologic cure or both. Superinfections were defined as new infections occurring during the treatment period with organisms different from those pathogens isolated on pre-therapy culture specimens. Superinfections may have been at either the original infection site or at a different site. During therapy, patients were monitored on a daily basis by one of the study investigators. In the event of a poor clinical response to the assigned study medication, the patient withdrew from the study and alternative therapy as deemed appropriate was substituted. Likewise, for perceived severe adverse drug reactions, study therapy was discontinued, and the reaction managed in the appropriate manner. The development of serious superinfection while receiving therapy mandated termination of study participation. All patients exposed to study drugs for any duration were evaluated for drug safety, whereas only those receiving therapy for four or more days were evaluable for efficacy (unless there was clear-cut evidence of therapeutic failure or resolution of infection).
/ PEACOCK ET AL
TABLE I Clinical Characteristics
I
TreatmentGroup
Characteristic
Ciprofloxacin
Total number of cases Number of evaluable cases Documented infections* Number of nonevaluable cases No documented Infection Protocol violation Mean years of age (range] Sex (male/female] Significant underlying disease Neopiastic Nonneoplasbc Concomitant clrndamycin
Ceftazidime
23
24
:4
:i
ii
i
5; (26-83) 1118 19 7 12
5: (22-89) 8112
18 4 14
6
7
LThere were two patients in each group with two different sites of InfectIon.
1 TABLE II Documented infections Treatment Group Ciprofloxacin Total number Clrnrcallydocumented Clinicallyand bacteriologically documented Bacteremia Primary Secondarv Site ’ Urinary tract Lung Boneifornt Skin/soft tissue Gallbladder Nosocomial Multrple infection
Ceftazidime
21
22
1:
1:
:
i
2
5 1; 151 3
:
2
L
*Numbers in parentheses, number of accompanying bacteremias
tween the ciprofloxacin and ceftazidime groups and included four patients without documented infections and four patients whose study participation was terminated because of protocol violations. The study population evaluable for determination of efficacy consisted of 20 women and 19 men with a mean age of 56.5 years (range, 22 to 89 years). Thirty-seven of these 39 patients had significant underlying illnesses, 30 percent (11 of 37) of which were neoplastic. Types of documented infections in the 39 evaluable study patients are shown in Table II. Two patients in each group had two sites of infection (all microbiologitally documented with different pathogens) resulting in 21 infections in 19 evaluable ciprofloxacin patients and 22 infections in 20 ceftazidime recipients. Overall, 20 of 43 infections (47 percent) involved the urinary tract and 11 infections (26 percent) involved the lower respiratory tract. Twenty-three percent (10 of 43) of all infections were associated with bacteremia; all five of the bacteremias in the eeftazidime patients were associated with urinary tract infections, but three of the five bacteremias in the ciprofloxacin groups were primary and unaccompanied by other sites of infection. The majority of infections were community acquired (34 of 43; 79 percent). Seventy-four percent of
RESULTS Of the 47 patients with serious infections who were enrolled in the trial, 39 patients (83 percent) were evaluable for determination of efficacy (Table I). The eight nonevaluable cases (included in the toxicity but not the efficacy analysis) were evenly distributed beNovember
ON CIPROFLOXACIN
30,
1989
The Amerrcan Journal of Medicine
Volume 87 (suppl 5A)
5A-187s
SYMPOSIUM
ON CIPROFLOXACIN
/ PEACOCK ET AL
TABLE III Bacteriologically
Documented Infections Treatment Group Ciprofloxacin
Ceftazidime
15
17
Total number of infections Organism Gram-oositive aerobes 124%overall)
Stabhylococcus au&s Staphylococcus epidermidis Streptococcuspneumoniae Enterococcussp. Alpha-hemolyticStreptococcus GrF$gahve
; : 1
fill
aerobes (69%overall) 3/u 3 (11 2
Enterobacter sp. K/ebs/e//asp. F aerugir7osa F1mirabjk Citrobacter freundh Providencia siuarfii
; 151
1
l(1)
Anaerobes (7%overall) Polymicrobial infections? (19% overall)
! 6
i
lumber in parentheses, number of accompanying bacteremias. wo or more isolates.
TABLE IV Favorable Clinical Response Rates Treatment Group Ciprofloxacin
Ceftazidime
Patients
Percent
Patients
Overall
16121
:i
18122
82
Clinicallyand Clinrcally documented bacteriologically documented Bacteremia Primary Secondarv Bv site -, Urinary tract
13115 316
87
16117 315
i!
Bone/joint Lung Skin/soft tissue Gallbladder By microblology Gram-positive aerobes Gramnegative aerobes Anaerobes
415 213 212
!
Percent
80
100
415 010 415
9110
zi
9110
90
213 214 0
67
313 417 212
l/l
100
0
314 11113 0
;z
416 14116
313
8:
1:;
100 67
Ii
all the infections were bacteriologically documented, and 26 percent were only clinically proved. Gram-negative aerobes accounted for 69 percent of all bacteriologically documented infections; E. coLi was the causative pathogen in 12 infections and Enterobacter sp. in five (Table III). Six of 10 bacteremias were due to E. coli, all of which originated from the urinary tract. P. aeruginosa and anaerobes were isolated from only three patients each. Overall, 19 percent of the infections (two of 21 infections in the ciprofloxacin group and six of 22 infections in the ceftazidime group) were polymicrobial, involving two or more isolates. Response to therapy, differentiated by treatment group, is summarized in Table IV. Overall, 76 percent of patients receiving ciprofloxacin and 82 percent of those treated with ceftazidime exhibited satisfactory clinical responses. Patients with microbiologically doc5A-188s
November
30, 1989
The American Journal of Medicine
umented infections had favorable responses in 91 percent of the cases, but only 55 percent of the patients with clinically documented infections had favorable responses. The poorest response rates in both groups occurred in patients with infections of the lower respiratory tract, with 50 percent of ciprofloxacin-treated and 57 percent of ceftazidime-treated patients demonstrating a satisfactory response. Of the 10 episodes of bacteremia, four of five in each group were successfully managed. A ciprofloxacin recipient died and had both alpha-hemolytic streptococci and Trichoderma sp. isolated from the bloodstream in addition to a polymicrobial (Proteus mirabilis and Enterococcus sp.) urinary tract infection. The single patient dying with bacteremia receiving ceftazidime developed E. coli sepsis secondary to pyelonephritis and died seven days after the onset of therapy. Overall response rates for infections due to grampositive aerobes were 75 percent (three of four patients) in the ciprofloxacin group compared to 67 percent (four of six patients) of ceftazidime recipients (Table IV). Both ciprofloxacin and ceftazidime proved to be excellent therapy for gram-negative aerobic infections with response rates of 85 and 88 percent, respectively. Documented anaerobic infections occurred only in ceftazidime recipients, with three of three patients responding to therapy. Five of 21 infections (24 percent) treated with ciprofloxacin were considered treatment failures. Two of these occurred in patients with underlying neoplastic disease complicated by clinical pneumonia (i.e., no pathogen isolated). The other three treatment failures included an osteomyelitis (no pathogen isolated), a polymicrobial urinary tract infection (P. mirabilis and Enterococcus sp.), and polymicrobial bacteremia (alpha-hemolytic streptococci and Trichoderma sp.); the latter two infections were documented in a single patient who was the only patient with ciprofloxacin treatment failure to die. Four of 22 (18 percent) infections treated with ceftazidime did not respond. Three of these failures were associated with patients with pneumonias, two of whom died. The other ceftazidime failure occurred in a patient with bacteremic E. coli pyelonephritis who also died. Superinfections occurred in five of 19 (26 percent) ciprofloxacin recipients and seven of 20 (35 percent) ceftazidime recipients (Table V). Enterococcal superinfections appeared in both groups (a bacteremic urinary tract infection in a ceftazidime patient and osteomyelitis in a ciprofloxacin patient). A patient with an initial E. coli urinary tract infection developed a KLebsieLLa pneumoniae urinary tract infection five days after treatment with ceftazidime. Clostridium diffitile-associated diarrhea was documented in a ceftazidime recipient. All fungal superinfections involved the genitourinary tract and occurred most often in association with chronic indwelling urinary catheters. The lengths of antibiotic therapy and hospitalization are compared in Table VI. The mean duration of hospitalization following the start of antibiotic treatment was 10.42 days in the ciprofloxacin group and 12.95 days in the ceftazidime group. The mean duration of intravenous antibiotic use for ciprofloxacin was 7.37 days and for ceftazidime 9.95 days; the median duration of intravenous therapy was five days for ciprofloxacin and seven days for ceftazidime. Sixty-three percent of the ciprofloxacin patients received an addi-
Volume 87 (suppl 5A)
SYMPOSIUM
tional 1’7 days of oral treatment with ciprofloxacin, whereas intravenously administered ceftazidime was followed by an average of 12 days of an oral regimen in 55 percent of patients. Oral antibiotics following discontinuation of ceftazidime included trimethoprimi sulfamethoxazole (36 percent), amoxicillin/clavulanic acid (36 percent), and others (amoxicillin, cephalexin, and clindamycin) . Adverse effects occurred in approximately 20 percent of patients in each group, as shown in Table VII. Only one side effect was associated with discontinuation of the study drug (ciprofloxacin due to shortness of breath), and the remainder of adverse effects were -mild and reversible. COMMENTS New antibiotics with microbiologic, pharmacologic, and potential clinical advantages over older agents are increasingly being introduced. Some of these advantages are extended spectrum of activity (including many highly resistant pathogens), prolonged half-life, oral preparation, simplification of treatment (particularly if monotherapy could replace the use of drug combinations), and reduction in the cost of treating infections [1.,2,10,11]. The two classes of antimicrobial agents receiving the most attention are beta-lactams and quinolones. For example, ceftazidime is a relatively new cephalosporin with superior gram-negative aerobic activity including P. aemginosa that has been used successfully as empiric monotherapy in febrile, granulocytopenic patients [12]. Ciprofloxacin, a new quinolone with a broad spectrum of activity that includes staphylococci and P. aemginosa, is effective as a twice-daily, oral drug for a number of serious infections [ll]; it is also available in a parenteral formulation that is currently undergoing extensive clinical testing [B]. In this study, intravenous ciprofloxacin and ceftazidime were evaluated as initial therapy for a variety of infections in a population of patients with serious underlying disease. The majority of the infections were community acquired (79 percent) and caused by gramnegative aerobes (69 percent); approximately one half of the infections involved the urinary tract (all complicated) and one fourth involved the lower respiratory tract. Twenty-three percent of all documented infections were associated with bacteremia, 80 percent due to gram-negative aerobes. Both antibiotics were comparable in efficacy with overall response rates of 76 percent for ciprofloxacin and 82 percent for ceftazidime. The use of either agent was especially successful in the treatment of microbiologically documented infections (91 percent), urinary tract infections (90 percent), infections caused by gram-negative aerobes (86 percent), and bacteremias (80 percent). Although the of patients studied in the trial . was small, _ number _ ._ I these data provide support that parenteral ciprofloxatin is as effective as ceftazidime in the initial therapy of serious infections. Despite the small number of patients with documented lower respiratory tract infections, the failure rate of both eiprofloxacin (50 percent) and ceftazidime (43 percent) was noteworthy. The majority of the pneumonias were only clinically documented without a pathogen being isolated; historically, old and new antibiotics in this setting have performed poorly [13]. Although the protocol was designed to evaluate seNovember
ON CIPROFLOXACIN
I PEACOCK ET AL
) TABLE V Superinfections Treatment Group Ciprofloxacin
Ceftazidime 20
19
Total number of cases Superinfections BacterIaI Enterococcus sp. UTI with bacteremia Osteomvelitis
5 (26)*
l(35)
0
1
Fi
0
1
K. pneumdnfae UT1
C dific//e 0
Diarrhea Fungal UTI Indwelling catheter Vaginrtrs I
UTI = urrnary tract rnfection. *Numbers rn parentheses, percent.
TABLE VI Length of Therapy/Hospitalization Treatment Group Ciprofloxacin Mean duratron hosprtalizationfrom start of antibiotics (days) tkat;: IV (days) Median Percentage receiving PO Mean duration PO (days)
Ceftazidime
10.42
12.95
7.31
9.95 7
6: 17
:;
IV = intravenous treatment; PO = oral treatment
1 TABLE VII 1 Adverse Effects Treatment Group Ciprofloxacin
Ceftazidime
5123 (22)*
5124 (21)
1
2 i
Overall Adverse effect Gastrorntestinal Diarrhea Elevatedliver enzymes Dizziness Shortness of breatht Phlebitrs Pruritus Eosrnophrlra
i i
:
1
0
i
i
*Numbers in parentheses, percent. tRequired discontrnuation of antibiobc treatment
. . quential intravenous/oral ciprofloxacin, the ceftazidime recipients were also followed for any subsequent oral antibiotic therapy. The mean duration of intravenous ciprofloxacin therapy was 7.37 days, and the median duration was five days. The mean and median durations of intravenous ceftazidime use were 9.95 and seven days, respectively. Sixty-three percent of the ciprofloxacin patients received an average of 17 additional days of oral treatment with ciprofloxacin; several of these patients were infected with pathogens (e.g., P. aeruginosa) at sites for which no other oral 30, 1989
The American Journal of Medicine
Volume 87 (suppl 5A)
5A-189s
SYMPOSIUM
ON CIPROFLOXACIN/PEACOCK
ETAL
agent would have been appropriate. Fifty-five percent of the ceftazidime patients also switched to an oral regimen. A number of different oral agents were employed in the ceftazidime group and were given for an average of 12 days. After the start of antibiotic therapy, the mean duration of hospitalization for the ciprofloxacin patients was 10.42 days and for the ceftazidime patients 12.95 days. There was no statistical difference between the two study drugs, in part because of the sample size and the longer use of oral ciprofloxatin. Superinfections were identified in over one third of all study patients. Two thirds of these were associated with fungal pathogens, all of which involved the genitourinary tract. Enterococcal superinfections occurred in both groups. Recent reports suggest that the enterococcus can be an important cause of serious illness and mortality in hospitalized patients [14,151. These studies also suggest that the increasing incidence of nosocomial enterococcal infections might be correlated with accelerated use of broad-spectrum antibiotics, especially those without good enterococcal activity, such as the cephalosporins and most recently the quinolones [6]. As an example of this phenomenon, a ceftazidime recipient in our study developed a bacteremit enterococcal urinary tract infection eight days after begin@ng therapy. In addition, a patient receiving ciprofloxacin for a cellulitis developed enterococcal osteomyelitis at the infection site after 33 days of treatment. Although concern has been raised about the potential for the emergence of resistance with the use of newer antibiotics (more so with the betalactams than with the quinolones), this phenomenon was not identified in our study [4]. C. d$‘iicile-associated diarrhea, which occurred in one of our patients receiving ceftazidime, can be anticipated with the use of all antibiotics including vancomycin [16,171. Adverse effects occurred in approximately 20 percent of patients treated with either ciprofloxacin or ceftazidime. Excluding one patient who experienced subjective shortness of breath necessitating the discontiimation of ciprofloxacin, all side effects were mild and reversible. Sequential intravenous followed by oral therapy with ciprofloxacin was effective in our study as ther-
5A-190s
November
30, 1989
The American Journal of Medicine
apy for seriously ill patients with severe infections requiring hospitalization. The oral bioavailability combined with an excellent spectrum of activity should lead to increased use of sequential intravenous/oral ciprofloxacin if other studies and clinical experience further support our data. The twice-daily dosing of both intravenous (with attendant decreased pharmacy, administration, and nursing time) and oral (with increased patient compliance) preparations is an attractive advantage. The use of ciprofloxacin has the potential to significantly reduce the cost of treating certain infections by switching from intravenous to oral treatment and hopefully decreasing the duration of hospitalization. REFERENCES 1. Donowitz GR, Mandell GL: Drug therapy: beta-lactam antibiotics. N Engl J Med 1988; 318: 419-426, 490-500. 2. Neu HC: The new beta-iactamase-stable cephalosporlns. Ann Intern Med 1982; 97: 408-419. 3. Wolfson JS, Hooper DC: The fluroquinolones: structures, mechanism of action and resistance, and spectra of activity in viiro. Antimicrob Agents Chemother 1985; 28: 5% 586. 4. Sanders CC: Clprofloxacin: in vitro activity, mechanism of actlon, and resistance. Rev Infect Dis 1988; 10: 516-527. 5. Sanders CC, Sanders WE Jr, Goering RV: Overvlew of preclinical studies with ciprofloxacln. Am J Med 1987; 82 (suppl 4A): 2-11. 6. Barry AL, Jones RN: In vitro activity of ciprofloxacin against gram-positive COCCI.Am J Med 1987; 82 (suppl 4A): 27-32. 7, Neu HC: Ciprofloxacin: an overview and prospective appraisal. Am J Med 1987; 82 (suppl 4A): 395-404. 8. Scully BE, Neu HC: Treatment of serious infections with Intravenous ciprofloxacin. Am J Med 1987; 82 (suppl 4A): 369-375. 9. Barriere SL: Economic Impact of oral ciprofloxacin: a pharmacist’s perspective. Am J Med 1987; 82 (suppl 4A): 387-390. 10. Qulntilianl R, Cooper SW, Briceland LL, Nightingale CH: Economic impact of streamlinIng antibiotic administration. Am J Med 1987; 82 (suppl 4A): 391-394. 11. Sanders WE Jr: Efficacy, safety, and potential economic benefits of oral ciprofloxacin in the treatment of Infections. Rev Infect Dis 1988; IO: 528-543. 12. Pizzo PA. Hathorn JW. Hiemenz J: A randomized trial cornoaring ceftazidime alone with combination antlbiotic therapy in cancer patients with fever and nktropenla. N Engl J Med 1986; 315: 552-558. 13. PIZZOPA, Commers J, Gress J: Approaching the controversies in antibacterial management of cancer patients. Am J Med 1984; 76: 436-449. 14. Moellering RC Jr: Infections due to group D streptococci. Rev Infect Dis 1981; 6: l-17. 15. Moellering RC Jr: The enteromccus: high level resistance to gentamicin and production of beta-lactamase. Clin Mlcrobiol Newsletter 1988: 10: 129-132. 16. Analsse EJ, Fainstein V, Bodey GP, et a/: Randomized trial of beta-lactam regimens in febrile neutropenic cancer patients. Am J Med 1988; 84: 581-589. 17. Gerding DN: Disease associated with Clostridium &k//e infection. Ann intern Med 1989; 110: 255-257.
Volume 87 (suppl 5A)
in
JAMESE. PEACOCK,JR.,M.D., P. SAMUELPEGRAM,M.D., STEPHENF. WEBER,M.D., PETER A. LEONE, M.D.
Winston-Salem, North Caroiina
In a prospective, comparative trial, 47 hospitalized patients with serious infections that required parenteral antibiotic therapy were randomly assigned to receive either ciprofloxacin (200 mg every 12 hours intravenously followed by 500 mg every 12 hours orally at a time dependent on the patients’ clinical and bacteriologic responses) or ceftazidime (2 g every eight to 12 hours intravenously). All evaluable subjects (39 patients) had documented infections, 23 percent of which were associated with bacteremia. The mean/median duration of intravenous antibiotic use for ciprofloxacin was 7.37ifive days and for ceftazidime 9.95keven days; 63 percent of the ciprofloxacin patients received an additional 1’7 days of oral therapy with ciprofloxacin, whereas intravenous therapy with ceftazidime was followed by an average of 12 days of an oral regimen in 55 percent of patients. Overall response rates for patients receiving ciprofloxacin and ceftazidime were 76 percent (16 of 21) and 82 percent (18 of 221, respectively. Four out of five bacteremias in each group were successfully treated. Overall, 69 percent of the pathogens were gram-negative aerobes, and 47 percent of the infections involved the urinary tract. Failure of therapy was most often associated with pneumonia (two of five failures with ciprofloxacin and three of four failures with ceftazidime). Adverse effects occurred in approximately 20 percent of patients in each group and were mild and reversible. Superinfections occurred in five of 19 (26 percent) ciprofloxacin recipients and seven of 20 (35 percent) ceftazidime recipients. All fungal superinfections involved the genitourinary tract and occurred most often in association with chronic indwelling catheters. Enterococcal superinfections occurred in both groups (a bacteremic urinary tract infection in a ceftazidime patient and osteomyelitis in a ciprofloxacin patient). Clostridium difticile-associated diarrhea was documented in a ceftazidime recipient. The mean duration of hospitalization following the onset of antibiotic treatment was 10.45 days in the ciprofloxacin group and 12.95
From the Department of Medicine, Wake Forest University Medical Center, WinstonSalem, North Carolina. Requests for reprints should be addressed to Dr. James E. Peacock, Jr.. Section on Infectious Diseases, Depariment of Medicine, Wake Forest University Medical Center, Winston-Salem, North Carolina 27103.
November
days in the ceftazidime group. Sequential intravenous/oral ciprofloxacin was as safe and effective as intravenous ceftazidime in the treatment of infections due to susceptible gram-positive and gram-negative organisms.
ver the past decade, numerous extended-spectrum beta-la&am antibiotics have been introduced into clinical practice [l]. Foremost among these have been the so-called “third generation” cephalosporins with their enhanced spectrum of activity against most gram-negative aerobic bacteria [21. A major differentiating feature among these extended-spectrum cephalosporins has been their activity against Pseudomonas ael-uyinosa [a]. To date, only cefoperazone and ceftazidime have exhibited clinically useful antipseudomonal activity [2], a property that has encouraged their usage, especially in the nosocomial setting or among immunocompromised patients, situations in which P. aeruginosa is a frequent cause of infection. At our institution, ceftazidime has emerged as the “third-generation” cephalosporin of choice, primarily because of its superior activity against P. aeruginosa. Drawbacks to its usage have included the necessity for dosing every eight hours and the lack of an oral preparation with a similar spectrum of activity. This deficiency precludes completion of therapy as an outpatient and early discharge of patients. Ciprofloxacin, a quinolone carboxylic acid derivative, also has excellent in vitro activity against most aerobic gram-negative pathogens, including Serratia marcescens, Enterobacter sp., and P. aemcginosa [351. In contrast to ceftazidime, ciprofloxacin is very active against staphylococci, including methicillinresistant strains [3,5,6], but ceftazidime has superior activity versus streptococci that are not highly susceptible to ciprofloxacin [5,61. Neither ciprofloxacin nor ceftazidime is useful clinically in the therapy of systemic anaerobic or enterococcal infections [2,4-71. Unlike ceftazidime, ciprofloxacin is available in both parenteral and oral formulations, the latter being well absorbed from the gastrointestinal tract and yielding bactericidal serum levels higher than the minimal inhibitory concentrations of most targeted pathogens [71. A potential advantage of ciprofloxacin over ceftazidime may therefore be the early conversion from parenteral to oral therapy in serious infections [8]. This use of sequential intravenous/oral therapy is likely to result in fewer hospital days and thus lower health care costs for individual patients [g-11]. 0
30, 1989
The American Journal of Medicine
Volume 87 (suppl 5A]
5A-185s
SYMPOSIUM
ON CIPROFLOXACIN
/PEACOCK
ET AL
The present study was designed to compare the effectiveness and safety of sequential intravenous/oral ciprofloxacin with intravenously administered ceftazidime in the therapy of serious infections caused by gram-negative or gram-positive organisms susceptible to both drugs. In addition, the potential economic consequences of sequential intravenous/oral therapy were explored. PATIENTS AND METHODS The study was a prospective, controlled, randomized, non-blinded trial conducted at a tertiary care teaching hospital. The study protocol was reviewed and approved by the institutional human studies review board prior to its inception, and all patients provided written informed consent for study participation prior to entry. Subjects enrolled into the trial were hospitalized patients 18 years of age or older who required parenteral antibiotic therapy for severe bacterial infections due to gram-positive or gram-negative aerobes susceptible to the study drugs. Types of infections qualifying for enrollment included, but were not limited to, bacteremia, respiratory tract infection, skin or skin-structure infection, bone or joint infection, and intra-abdominal infection. Patients with urinary tract infections were enrolled only if the infection was considered to be complicated, or severe and recalcitrant to other antibiotics; caused by P. aeruginosa or other organisms resistant to conventional antibiotics; or presumed to be accompanied by bacteremia. Patients with multiple infection sites were eligible for enrollment if each site could be evaluated separately. Prior therapy with other antimicrobials did not preclude entry if a causative organism susceptible to both study drugs could be isolated from the infection site before initiating therapy. Patients previously enrolled in the study were not eligible for re-entry even if infection developed at a site different from that initially treated. Potential study candidates were excluded from entry if they were allergic to a study drug or related compound (including penicillin), were pregnant or nursing, had mild infections not requiring parenteral therapy, exhibited severe impairment of renal function (i.e., serum creatinine greater than 2.0 mg/dl or creatinine clearance less than 50 ml/minute/l.73 m’), or were not expected to survive 30 days post-treatment because of severe underlying disease (exclusive of infection status). Probable need for concomitant therapy with an antimicrobial agent with a spectrum of activity similar to that of the study drugs also precluded participation, but the use of clindamycin or metronidazole for suspected anaerobic infection, or the addition of nafcillin or vancomycin to ceftazidime treatment for suspected staphylococcal infection, was allowed. In addition, patients with the following selected infections or conditions were not eligible for enrollment: meningitis, endocarditis, acute pulmonary exacerbations of cystic fibrosis, infections in immunocompromised patients (i.e., granulocyte counts less than l,OOO/mm”), gross lower or upper urinary tract obstruction that was not relieved during therapy, severe neurogenic bladder disease, chronic indwelling urinary catheter, and pelvic infections. Patients who met enrollment criteria and provided written informed consent were assigned to one of the two drug regimens in accordance with a computer5A-186s
November
30, 1989
The American Journal of Medicine
generated randomization code provided by the study sponsor. Patients assigned to therapy with ciprofloxatin were routinely given 200 mg intravenously every 12 hours for two to five days, followed by 500 mg orally every 12 hours for the duration of therapy. For patients with severe infections caused by organisms only moderately susceptible to ciprofloxacin and resistant to most other conventional antibiotics, the intravenous dosage of ciprofloxacin could be increased to 300 mg every 12 hours and the oral dose to 750 mg every 12 hours. The conversion from intravenous to oral therapy was dependent on the severity of the infection, as well as the patients’ clinical and bacteriologic responses; the duration of therapy was individualized for each patient by one of the study investigators. Patients assigned to treatment with ceftazidime received 2 g intravenously every eight to 12 hours, with the individual dose determined by the nature of the infection under therapy. Intravenous therapy was continued for a minimum of four days but usually for a period of seven to 14 days. In patients who responded promptly to ceftazidime, intravenous treatment could be discontinued (after the minimum of four days of therapy) and an appropriate oral antimicrobial agent (except ciprofloxacin) substituted. The actual duration of intravenous therapy with ceftazidime, the choice of orally administered antimicrobials, and the duration of oral therapy were at the discretion of the investigator. For those patients receiving oral therapy after intravenous ceftazidime treatment, clinical and bacteriologic evaluation was performed at the conclusion of both intravenous and oral regimens. For patients developing renal insufficiency after enrollment in the study, published guidelines were utilized to adjust the dosages of both ciprofloxacin and ceftazidime. A complete clinical, laboratory, and bacteriologic examination was performed for each patient upon study entry. The clinical examination consisted of a complete medical history and physical examination by one of the investigators or his designate. Laboratory studies, including a complete blood cell count with white blood cell differential and platelet count, erythrocyte sedimentation rate measurement, serum chemistry tests, and complete urinalysis, were performed prior to initiating therapy and then every three to five days thereafter. In addition, a pregnancy test was performed on all women of childbearing potential. Additional laboratory or diagnostic tests or procedures were performed whenever indicated. Culture specimens were obtained from all potential infection sites within the 48 hours prior to the onset of therapy and thereafter as often as indicated based upon the clinical course of the disease process. However, in patients with infection of the urinary tract, culture specimens were routinely repeated on Day 3 or 4 of treatment. All isolated pathogens were tested for susceptibility to the study drugs using both the modified Kirby-Bauer procedure with 5-pg ciprofloxacin and 30-pg ceftazidime disks, and a standard twofold tube dilution technique using Mueller-Hinton media and an inoculum of approximately lo5 colony-forming units/ ml. Ciprofloxacin susceptibility was defined as a zone of inhibition of 21 mm or more and minimal inhibitory concentrations of 1 pgiml or less. Moderate susceptibility to ciprofloxacin was defined as a zone of inhibi-
Volume 87 (suppl 5A)
SYMPOSIUM
tion of 16 to 20 mm and minimal inhibitory concentrations of greater than 1 to 2 pgiml. At the completion of therapy, the physical examination and laboratory studies were repeated, and culture specimens were obtained from previously infected sites if material was still available. For patients with urinary tract infections, culture specimens of urine were obtained five to nine days and four weeks after therapy. Chest radiographic examinations were routinely obtained post-therapy in all patients with pneumonia. For study purposes, infections were classified as documented if patients exhibited typical clinical symptoms or signs appropriate for the site under consideration (e.g., purulent sputum production and infiltrates on radiographic examination of the chest in a patient with presumed pneumonia). Culture specimens of the site may or may not have been positive for a definite pathogen. Bacteriologically documented infections were those clinically documented infections that also had positive culture specimens for a definite pathogen isolated from the infection site. All patients with documented infections were considered to be evaluable for efficacy. Assessment of response to therapy was based upon clinical, radiologic, and bacteriologic parameters. A satisfactory clinical response was considered to have occurred if patients exhibited resolution of or improvement in symptoms and signs of infection (e.g., defervescence of fever, clearing of pus or other signs of inflammation, resolution of infiltrates on chest radiographs, and so on). Bacteriologic cure was defined as elimination of causative organisms from cultures of infection sites during treatment. In addition, for patients with urinary tract infections, a negative urine culture specimen at five to nine days of therapy was also required. Treatment failures encompassed those patients who did not exhibit either clinical response or bacteriologic cure or both. Superinfections were defined as new infections occurring during the treatment period with organisms different from those pathogens isolated on pre-therapy culture specimens. Superinfections may have been at either the original infection site or at a different site. During therapy, patients were monitored on a daily basis by one of the study investigators. In the event of a poor clinical response to the assigned study medication, the patient withdrew from the study and alternative therapy as deemed appropriate was substituted. Likewise, for perceived severe adverse drug reactions, study therapy was discontinued, and the reaction managed in the appropriate manner. The development of serious superinfection while receiving therapy mandated termination of study participation. All patients exposed to study drugs for any duration were evaluated for drug safety, whereas only those receiving therapy for four or more days were evaluable for efficacy (unless there was clear-cut evidence of therapeutic failure or resolution of infection).
/ PEACOCK ET AL
TABLE I Clinical Characteristics
I
TreatmentGroup
Characteristic
Ciprofloxacin
Total number of cases Number of evaluable cases Documented infections* Number of nonevaluable cases No documented Infection Protocol violation Mean years of age (range] Sex (male/female] Significant underlying disease Neopiastic Nonneoplasbc Concomitant clrndamycin
Ceftazidime
23
24
:4
:i
ii
i
5; (26-83) 1118 19 7 12
5: (22-89) 8112
18 4 14
6
7
LThere were two patients in each group with two different sites of InfectIon.
1 TABLE II Documented infections Treatment Group Ciprofloxacin Total number Clrnrcallydocumented Clinicallyand bacteriologically documented Bacteremia Primary Secondarv Site ’ Urinary tract Lung Boneifornt Skin/soft tissue Gallbladder Nosocomial Multrple infection
Ceftazidime
21
22
1:
1:
:
i
2
5 1; 151 3
:
2
L
*Numbers in parentheses, number of accompanying bacteremias
tween the ciprofloxacin and ceftazidime groups and included four patients without documented infections and four patients whose study participation was terminated because of protocol violations. The study population evaluable for determination of efficacy consisted of 20 women and 19 men with a mean age of 56.5 years (range, 22 to 89 years). Thirty-seven of these 39 patients had significant underlying illnesses, 30 percent (11 of 37) of which were neoplastic. Types of documented infections in the 39 evaluable study patients are shown in Table II. Two patients in each group had two sites of infection (all microbiologitally documented with different pathogens) resulting in 21 infections in 19 evaluable ciprofloxacin patients and 22 infections in 20 ceftazidime recipients. Overall, 20 of 43 infections (47 percent) involved the urinary tract and 11 infections (26 percent) involved the lower respiratory tract. Twenty-three percent (10 of 43) of all infections were associated with bacteremia; all five of the bacteremias in the eeftazidime patients were associated with urinary tract infections, but three of the five bacteremias in the ciprofloxacin groups were primary and unaccompanied by other sites of infection. The majority of infections were community acquired (34 of 43; 79 percent). Seventy-four percent of
RESULTS Of the 47 patients with serious infections who were enrolled in the trial, 39 patients (83 percent) were evaluable for determination of efficacy (Table I). The eight nonevaluable cases (included in the toxicity but not the efficacy analysis) were evenly distributed beNovember
ON CIPROFLOXACIN
30,
1989
The Amerrcan Journal of Medicine
Volume 87 (suppl 5A)
5A-187s
SYMPOSIUM
ON CIPROFLOXACIN
/ PEACOCK ET AL
TABLE III Bacteriologically
Documented Infections Treatment Group Ciprofloxacin
Ceftazidime
15
17
Total number of infections Organism Gram-oositive aerobes 124%overall)
Stabhylococcus au&s Staphylococcus epidermidis Streptococcuspneumoniae Enterococcussp. Alpha-hemolyticStreptococcus GrF$gahve
; : 1
fill
aerobes (69%overall) 3/u 3 (11 2
Enterobacter sp. K/ebs/e//asp. F aerugir7osa F1mirabjk Citrobacter freundh Providencia siuarfii
; 151
1
l(1)
Anaerobes (7%overall) Polymicrobial infections? (19% overall)
! 6
i
lumber in parentheses, number of accompanying bacteremias. wo or more isolates.
TABLE IV Favorable Clinical Response Rates Treatment Group Ciprofloxacin
Ceftazidime
Patients
Percent
Patients
Overall
16121
:i
18122
82
Clinicallyand Clinrcally documented bacteriologically documented Bacteremia Primary Secondarv Bv site -, Urinary tract
13115 316
87
16117 315
i!
Bone/joint Lung Skin/soft tissue Gallbladder By microblology Gram-positive aerobes Gramnegative aerobes Anaerobes
415 213 212
!
Percent
80
100
415 010 415
9110
zi
9110
90
213 214 0
67
313 417 212
l/l
100
0
314 11113 0
;z
416 14116
313
8:
1:;
100 67
Ii
all the infections were bacteriologically documented, and 26 percent were only clinically proved. Gram-negative aerobes accounted for 69 percent of all bacteriologically documented infections; E. coLi was the causative pathogen in 12 infections and Enterobacter sp. in five (Table III). Six of 10 bacteremias were due to E. coli, all of which originated from the urinary tract. P. aeruginosa and anaerobes were isolated from only three patients each. Overall, 19 percent of the infections (two of 21 infections in the ciprofloxacin group and six of 22 infections in the ceftazidime group) were polymicrobial, involving two or more isolates. Response to therapy, differentiated by treatment group, is summarized in Table IV. Overall, 76 percent of patients receiving ciprofloxacin and 82 percent of those treated with ceftazidime exhibited satisfactory clinical responses. Patients with microbiologically doc5A-188s
November
30, 1989
The American Journal of Medicine
umented infections had favorable responses in 91 percent of the cases, but only 55 percent of the patients with clinically documented infections had favorable responses. The poorest response rates in both groups occurred in patients with infections of the lower respiratory tract, with 50 percent of ciprofloxacin-treated and 57 percent of ceftazidime-treated patients demonstrating a satisfactory response. Of the 10 episodes of bacteremia, four of five in each group were successfully managed. A ciprofloxacin recipient died and had both alpha-hemolytic streptococci and Trichoderma sp. isolated from the bloodstream in addition to a polymicrobial (Proteus mirabilis and Enterococcus sp.) urinary tract infection. The single patient dying with bacteremia receiving ceftazidime developed E. coli sepsis secondary to pyelonephritis and died seven days after the onset of therapy. Overall response rates for infections due to grampositive aerobes were 75 percent (three of four patients) in the ciprofloxacin group compared to 67 percent (four of six patients) of ceftazidime recipients (Table IV). Both ciprofloxacin and ceftazidime proved to be excellent therapy for gram-negative aerobic infections with response rates of 85 and 88 percent, respectively. Documented anaerobic infections occurred only in ceftazidime recipients, with three of three patients responding to therapy. Five of 21 infections (24 percent) treated with ciprofloxacin were considered treatment failures. Two of these occurred in patients with underlying neoplastic disease complicated by clinical pneumonia (i.e., no pathogen isolated). The other three treatment failures included an osteomyelitis (no pathogen isolated), a polymicrobial urinary tract infection (P. mirabilis and Enterococcus sp.), and polymicrobial bacteremia (alpha-hemolytic streptococci and Trichoderma sp.); the latter two infections were documented in a single patient who was the only patient with ciprofloxacin treatment failure to die. Four of 22 (18 percent) infections treated with ceftazidime did not respond. Three of these failures were associated with patients with pneumonias, two of whom died. The other ceftazidime failure occurred in a patient with bacteremic E. coli pyelonephritis who also died. Superinfections occurred in five of 19 (26 percent) ciprofloxacin recipients and seven of 20 (35 percent) ceftazidime recipients (Table V). Enterococcal superinfections appeared in both groups (a bacteremic urinary tract infection in a ceftazidime patient and osteomyelitis in a ciprofloxacin patient). A patient with an initial E. coli urinary tract infection developed a KLebsieLLa pneumoniae urinary tract infection five days after treatment with ceftazidime. Clostridium diffitile-associated diarrhea was documented in a ceftazidime recipient. All fungal superinfections involved the genitourinary tract and occurred most often in association with chronic indwelling urinary catheters. The lengths of antibiotic therapy and hospitalization are compared in Table VI. The mean duration of hospitalization following the start of antibiotic treatment was 10.42 days in the ciprofloxacin group and 12.95 days in the ceftazidime group. The mean duration of intravenous antibiotic use for ciprofloxacin was 7.37 days and for ceftazidime 9.95 days; the median duration of intravenous therapy was five days for ciprofloxacin and seven days for ceftazidime. Sixty-three percent of the ciprofloxacin patients received an addi-
Volume 87 (suppl 5A)
SYMPOSIUM
tional 1’7 days of oral treatment with ciprofloxacin, whereas intravenously administered ceftazidime was followed by an average of 12 days of an oral regimen in 55 percent of patients. Oral antibiotics following discontinuation of ceftazidime included trimethoprimi sulfamethoxazole (36 percent), amoxicillin/clavulanic acid (36 percent), and others (amoxicillin, cephalexin, and clindamycin) . Adverse effects occurred in approximately 20 percent of patients in each group, as shown in Table VII. Only one side effect was associated with discontinuation of the study drug (ciprofloxacin due to shortness of breath), and the remainder of adverse effects were -mild and reversible. COMMENTS New antibiotics with microbiologic, pharmacologic, and potential clinical advantages over older agents are increasingly being introduced. Some of these advantages are extended spectrum of activity (including many highly resistant pathogens), prolonged half-life, oral preparation, simplification of treatment (particularly if monotherapy could replace the use of drug combinations), and reduction in the cost of treating infections [1.,2,10,11]. The two classes of antimicrobial agents receiving the most attention are beta-lactams and quinolones. For example, ceftazidime is a relatively new cephalosporin with superior gram-negative aerobic activity including P. aemginosa that has been used successfully as empiric monotherapy in febrile, granulocytopenic patients [12]. Ciprofloxacin, a new quinolone with a broad spectrum of activity that includes staphylococci and P. aemginosa, is effective as a twice-daily, oral drug for a number of serious infections [ll]; it is also available in a parenteral formulation that is currently undergoing extensive clinical testing [B]. In this study, intravenous ciprofloxacin and ceftazidime were evaluated as initial therapy for a variety of infections in a population of patients with serious underlying disease. The majority of the infections were community acquired (79 percent) and caused by gramnegative aerobes (69 percent); approximately one half of the infections involved the urinary tract (all complicated) and one fourth involved the lower respiratory tract. Twenty-three percent of all documented infections were associated with bacteremia, 80 percent due to gram-negative aerobes. Both antibiotics were comparable in efficacy with overall response rates of 76 percent for ciprofloxacin and 82 percent for ceftazidime. The use of either agent was especially successful in the treatment of microbiologically documented infections (91 percent), urinary tract infections (90 percent), infections caused by gram-negative aerobes (86 percent), and bacteremias (80 percent). Although the of patients studied in the trial . was small, _ number _ ._ I these data provide support that parenteral ciprofloxatin is as effective as ceftazidime in the initial therapy of serious infections. Despite the small number of patients with documented lower respiratory tract infections, the failure rate of both eiprofloxacin (50 percent) and ceftazidime (43 percent) was noteworthy. The majority of the pneumonias were only clinically documented without a pathogen being isolated; historically, old and new antibiotics in this setting have performed poorly [13]. Although the protocol was designed to evaluate seNovember
ON CIPROFLOXACIN
I PEACOCK ET AL
) TABLE V Superinfections Treatment Group Ciprofloxacin
Ceftazidime 20
19
Total number of cases Superinfections BacterIaI Enterococcus sp. UTI with bacteremia Osteomvelitis
5 (26)*
l(35)
0
1
Fi
0
1
K. pneumdnfae UT1
C dific//e 0
Diarrhea Fungal UTI Indwelling catheter Vaginrtrs I
UTI = urrnary tract rnfection. *Numbers rn parentheses, percent.
TABLE VI Length of Therapy/Hospitalization Treatment Group Ciprofloxacin Mean duratron hosprtalizationfrom start of antibiotics (days) tkat;: IV (days) Median Percentage receiving PO Mean duration PO (days)
Ceftazidime
10.42
12.95
7.31
9.95 7
6: 17
:;
IV = intravenous treatment; PO = oral treatment
1 TABLE VII 1 Adverse Effects Treatment Group Ciprofloxacin
Ceftazidime
5123 (22)*
5124 (21)
1
2 i
Overall Adverse effect Gastrorntestinal Diarrhea Elevatedliver enzymes Dizziness Shortness of breatht Phlebitrs Pruritus Eosrnophrlra
i i
:
1
0
i
i
*Numbers in parentheses, percent. tRequired discontrnuation of antibiobc treatment
. . quential intravenous/oral ciprofloxacin, the ceftazidime recipients were also followed for any subsequent oral antibiotic therapy. The mean duration of intravenous ciprofloxacin therapy was 7.37 days, and the median duration was five days. The mean and median durations of intravenous ceftazidime use were 9.95 and seven days, respectively. Sixty-three percent of the ciprofloxacin patients received an average of 17 additional days of oral treatment with ciprofloxacin; several of these patients were infected with pathogens (e.g., P. aeruginosa) at sites for which no other oral 30, 1989
The American Journal of Medicine
Volume 87 (suppl 5A)
5A-189s
SYMPOSIUM
ON CIPROFLOXACIN/PEACOCK
ETAL
agent would have been appropriate. Fifty-five percent of the ceftazidime patients also switched to an oral regimen. A number of different oral agents were employed in the ceftazidime group and were given for an average of 12 days. After the start of antibiotic therapy, the mean duration of hospitalization for the ciprofloxacin patients was 10.42 days and for the ceftazidime patients 12.95 days. There was no statistical difference between the two study drugs, in part because of the sample size and the longer use of oral ciprofloxatin. Superinfections were identified in over one third of all study patients. Two thirds of these were associated with fungal pathogens, all of which involved the genitourinary tract. Enterococcal superinfections occurred in both groups. Recent reports suggest that the enterococcus can be an important cause of serious illness and mortality in hospitalized patients [14,151. These studies also suggest that the increasing incidence of nosocomial enterococcal infections might be correlated with accelerated use of broad-spectrum antibiotics, especially those without good enterococcal activity, such as the cephalosporins and most recently the quinolones [6]. As an example of this phenomenon, a ceftazidime recipient in our study developed a bacteremit enterococcal urinary tract infection eight days after begin@ng therapy. In addition, a patient receiving ciprofloxacin for a cellulitis developed enterococcal osteomyelitis at the infection site after 33 days of treatment. Although concern has been raised about the potential for the emergence of resistance with the use of newer antibiotics (more so with the betalactams than with the quinolones), this phenomenon was not identified in our study [4]. C. d$‘iicile-associated diarrhea, which occurred in one of our patients receiving ceftazidime, can be anticipated with the use of all antibiotics including vancomycin [16,171. Adverse effects occurred in approximately 20 percent of patients treated with either ciprofloxacin or ceftazidime. Excluding one patient who experienced subjective shortness of breath necessitating the discontiimation of ciprofloxacin, all side effects were mild and reversible. Sequential intravenous followed by oral therapy with ciprofloxacin was effective in our study as ther-
5A-190s
November
30, 1989
The American Journal of Medicine
apy for seriously ill patients with severe infections requiring hospitalization. The oral bioavailability combined with an excellent spectrum of activity should lead to increased use of sequential intravenous/oral ciprofloxacin if other studies and clinical experience further support our data. The twice-daily dosing of both intravenous (with attendant decreased pharmacy, administration, and nursing time) and oral (with increased patient compliance) preparations is an attractive advantage. The use of ciprofloxacin has the potential to significantly reduce the cost of treating certain infections by switching from intravenous to oral treatment and hopefully decreasing the duration of hospitalization. REFERENCES 1. Donowitz GR, Mandell GL: Drug therapy: beta-lactam antibiotics. N Engl J Med 1988; 318: 419-426, 490-500. 2. Neu HC: The new beta-iactamase-stable cephalosporlns. Ann Intern Med 1982; 97: 408-419. 3. Wolfson JS, Hooper DC: The fluroquinolones: structures, mechanism of action and resistance, and spectra of activity in viiro. Antimicrob Agents Chemother 1985; 28: 5% 586. 4. Sanders CC: Clprofloxacin: in vitro activity, mechanism of actlon, and resistance. Rev Infect Dis 1988; 10: 516-527. 5. Sanders CC, Sanders WE Jr, Goering RV: Overvlew of preclinical studies with ciprofloxacln. Am J Med 1987; 82 (suppl 4A): 2-11. 6. Barry AL, Jones RN: In vitro activity of ciprofloxacin against gram-positive COCCI.Am J Med 1987; 82 (suppl 4A): 27-32. 7, Neu HC: Ciprofloxacin: an overview and prospective appraisal. Am J Med 1987; 82 (suppl 4A): 395-404. 8. Scully BE, Neu HC: Treatment of serious infections with Intravenous ciprofloxacin. Am J Med 1987; 82 (suppl 4A): 369-375. 9. Barriere SL: Economic Impact of oral ciprofloxacin: a pharmacist’s perspective. Am J Med 1987; 82 (suppl 4A): 387-390. 10. Qulntilianl R, Cooper SW, Briceland LL, Nightingale CH: Economic impact of streamlinIng antibiotic administration. Am J Med 1987; 82 (suppl 4A): 391-394. 11. Sanders WE Jr: Efficacy, safety, and potential economic benefits of oral ciprofloxacin in the treatment of Infections. Rev Infect Dis 1988; IO: 528-543. 12. Pizzo PA. Hathorn JW. Hiemenz J: A randomized trial cornoaring ceftazidime alone with combination antlbiotic therapy in cancer patients with fever and nktropenla. N Engl J Med 1986; 315: 552-558. 13. PIZZOPA, Commers J, Gress J: Approaching the controversies in antibacterial management of cancer patients. Am J Med 1984; 76: 436-449. 14. Moellering RC Jr: Infections due to group D streptococci. Rev Infect Dis 1981; 6: l-17. 15. Moellering RC Jr: The enteromccus: high level resistance to gentamicin and production of beta-lactamase. Clin Mlcrobiol Newsletter 1988: 10: 129-132. 16. Analsse EJ, Fainstein V, Bodey GP, et a/: Randomized trial of beta-lactam regimens in febrile neutropenic cancer patients. Am J Med 1988; 84: 581-589. 17. Gerding DN: Disease associated with Clostridium &k//e infection. Ann intern Med 1989; 110: 255-257.
Volume 87 (suppl 5A)