Potential of imipenem as single-agent empiric antibiotic therapy of febrile neutropenic patients...

Potential of Imipenem as Single-Agent Empiric Antibiotic Therapy of Febrile Neutropenic Patients with Cancer

JAMES C. WADE, M.D. HAROLD C. STANDIFORD, M.D. GEORGE L. DRUSANO, M.D. DAVID E. JOHNSON, Ph.D. MARCIA R. MOODY, Ph.D. CARLOS I. BUSTAMANTE, M.D. JAI H. JOSHI, M.D. CARLOS deJONGH, M.D. STEPHEN C. SCHIMPFF, M.D. Baltimore, Maryland

From the University of Maryland Cancer Center, and Division of Infectious Diseases, Department of Medicine, University of Maryland School of Medi- cine, and Baltimore Veterans Administration Medi- cal Center, Baltimore, Maryland. Supported in part by grants (1-R01-CA370460-0 and I-P50- CA32107-01) from the National Cancer Institute. Requests for reprints should be addressed to Dr. James C. Wade, University of Maryland Cancer Center, 22 South Greene Street, Baltimore, Mary- land 21201.

Infection remains a major cause of morbidity and mortality for the patient with cancer wh o experiences episodes of severe granulocytopenia. The search continues for new antimicrobial agents with improved efficacy and lower incidence of toxicity. Imipenem is a new carbapenem antibiotic which possesses a broad antibacterial spectrum with excellent activity against Pseudomonas aeruginosa and the other commonly recovered enteric gram-negative bacilli that infect the granulocytopenic patient with cancer. The combination of imipenem plus an aminoglycoside has shown in vitro synergy against P. aeruginosa and Staphylococcus aureus whereas the combination of imipenem plus piperacillin or the extended spectrum cephalosporins have frequently shown antagonism when tested against P. aeruginosa and Serratia marcescens. The use of a P. aeruginosa-infected neutropenic rat model has provided an in vivo system to evaluate the activity of new antibiotics or antibiotic combinations. Monotherapy with imipenem is as effective in this model as any of the currently available synergistic antibiotic combinations. This degree of activity has not been found with other broad-spectrum antibiotics when used alone. Imipenem provides serum bactericidal activity well above a 1:8 dilution for the four most commonly isolated pathogens: P. aeruginosa, Escherichia coil, Klebsiella species, and S. aureus. In addition, imipenem's post-antibiotic effect against P. aeruginosa may be pertinent. Imipenem is a unique antibiotic, with properties that make it well suited for study as monotherapy for fever and suspected infection in granulocytopenic patients with cancer. A prospective randomized, double-blind study comparing imipenem with a control regimen of piperacillin plus amikacin as empiric antibiotic therapy of febrile granulocytopenic patients with cancer is currently underway at the University of Maryland Cancer Center.

Granulocytopenia has a profound effect upon the incidence of infection in patients with cancer [1]. The incidence of infection increases as the granulocyte count decreases to a level of less than 500//~1, with a steep rise in the incidence of severe infection and bacteremia at levels below 100//~1. Because of this predisposition, infection remains a major cause of morbidity and the leading cause of death in patients who undergo marrow transplantation or receive intensive therapy with myelosuppressive agents.

The presence of granulocytopenia is responsible for muted and frequently absent signs of infection (i.e., pain, heat, redness, and swelling) [1]. Fever, however, is invariably present and often the only early sign of

62 May 31, 1985 The American Journal of Medicine Volume 78 (suppl 5A)

IMIPENEM/CILASTATIN SYMPOSIUM--WADE ET AL

infection. Studies relating episodes of fever with subsequent documentation of infection among patients with granulocytopenia have repeatedly shown that approximately 20 percent of these episodes have an associated bacteremia, 20 percent are due to non-bacteremic micro- biologically documented infections, 20 percent to clinically documented infections, and the remaining 40 percent of febrile dpisodes represent possible or doubtful infections [2]. Therefore, at least 60 percent of all new febrile episodes (defined as a temperature greater than or equal to 38°C) occurring among granulocytopenic patients with cancer will be found to be of infectious origin and will re- quire antibiotic therapy.

Infections that occur in patients who are granulocytopenic originate primarily along the alimentary tract (i.e., pharynx, esophagus, anorectum), the respiratory tree (i.e., lungs, sinus), and the skin [3]. These are the sites where chemotherapy has the greatest effect upon muco- sal barriers and ciliary function or where iatrogenic procedures are most likely to disrupt the integument. Responsi- ble pathogens have historically been few. Staphylococcus aureus, the major pathogen prior to 1960, is, for most centers, still the most common pathogenic gram-positive organism [1]. However, a number of centers have reported increasing occurrence of coagulase-negative staphylococci (i.e., Staphylococcus epidermidis) as significant pathogens [4-6]. Escherichia coli, Klebsiella species, and Pseudomonas aeruginosa are the most common gram- negative pathogens, with 60 percent of all initial infections caused by enteric gram-negative bacilli. Fungi rarely cause initial infections but are frequent pathogens among patients who have received broad-spectrum, systemic, antibacterial antibiotics during periods of profound (less than 100//~1), prolonged granulocytopenia. Infecting organisms are frequently found to be colonizing at, or near, the site of infection, and greater than 50 percent of these pathogens will be acquired by the patient following his admission to the hospital.

The prevalence of granulocytopenia-associated infectious complications has required the development of improved preventive and therapeutic strategies. Classic prophylactic approaches have been based on attempts to suppress pathogenic endogenous flora, reduce new organism acquisition, limit invasive procedures, and improve host defenses. Treatment modalities have centered around the use of empiric broad-spectrum antibiotics. The appropriate choice of empiric antibiotics for the febrile granulocytopenic patient requires that the following fac- tors be considered: (1) the antibiotic spectrum must be broad and should include the most commonly infecting gram-positive and gram-negative bacteria since the specific organism can seldom be determined in advance on clinical grounds, and (2) antibiotics should be given intravenously, in high dose, and should be bactericidal in action.

During the past 15 years at the University of Maryland Cancer Center, we have taken the approach of using antibiotic combinations for the empiric therapy of the febrile, granulocytopenic patient. This approach has been advo- cated for two reasons. First, until recently no single drug was capable of "covering" the spectrum of organisms which frequently cause infection in the granulocytopenic patient. Thus, a combination of antibiotics was necessary simply to ensure adequate antibacterial spectrum. Equally important was the concern that if an organism was resistant to one antibiotic, there was still the potential for the second antibiotic to have activity against the infecting pathogen. Second, for a small subset of patients with profound and persistent granulocytopenia and a bacteremia due to gram-negative bacilli, the synergistic activity of an antimicrobial combination appeared important [7,8]. Pa- tients whose granulocyte count increases during therapy appear to do well if they are treated with an appropriate single agent. Yet, the outcome for those patients who remain profoundly granulocytopenic (less than 100//1.1) during therapy is enhanced if they receive two antibiotics to which the infecting organism is susceptible. These response rates can be further improved if the in vitro antibiotic combination shows synergistic killing against the pathogenic organism. Therefore, antibiotic combinations, preferably synergistic in activity, have been found to broaden the antibacterial coverage and improve clinical response rates.

The combination of an aminoglycoside (i.e., gentamicin, tobramycin, amikacin, or netilmicin) plus either a broad- spectrum anti-pseudomonal penicillin (i.e., carbenicillin, ticarcillin, piperacillin, mezlocillin, or azlocillin) or an extended-spectrum cephalosporin (i.e., moxalactam, cefotaxime, or cefoperazone) are currently the most frequently used empiric antibiotic combinations. The development of the newer broad-spectrum penicillins and highly active cephalosporin-like drugs has resurrected interest in their use as monotherapy for infection in the patient with granu- Iocytopenia. A number of such single-agent studies have been completed or are ongoing. Pickard and colleagues [9] have suggested that monotherapy with moxalactam is equivalent to a standard combination approach. However, the activity of this antibiotic against P. aeruginosa, a common pathogen for this host, is marginal. Pizzo et al [10] compared ceftazidime with the antibiotic combination of gentamicin, carbenicillin, and cephalothin. They con- cluded that ceftazidime was equivalent to the antibiotic combination. However, in neither of these studies has there been a sufficient number of patients at high-risk (i.e., profound and persistent granulocytopenia, and gram-negative bacteremias) for adequate comparison to be performed.

We have recently initiated a prospective, randomized, double-blind controlled trial which compares a standard combination (piperacillin plus amikacin) to imipenem/ci-

June 7, 1985 The American Journal of Medicine Volume 78 (suppl 6A) 63

IMIPENEM/ClLASTATIN SYMPOSIUM--WADE ET AL

105cf~ S oureus 42 s ] r~ns Concentration (ug /ml ) °

~39 78 ~ $J3 625 125 25 50 I ~ ~00 400 g~ ~OO IO$cfu

K pneum~ 4911tolns

ConcentratJoff ( ug /mO o ~39 78 I ~ 3~5 625 125 ~ 50 IO0 200 400 800 ~gOO

ILt .j ,5 o ~ <_005 OI 02 O4 .08 .16 ~2 ~ 12. ~S~ ~ IO 2o~s~b

IM IPENEM Concen t ra t i on (~g /m l l

~o~ f . Concen t ra t i on ( u g / m l ) ~

oos 2 ~ 6 32

I M I P E N E I V Concen t ra t ion ( ug / r o l l

10 2 cfu Concentration ( ug /m l )a P. l e rugJno~ . IOI s l r6 i~s ~39 78 ~56 3r3 625 125 25 5o IOO 2oo 400 aoo _>800

ioo

~o

70

so

"6 50

40

30

~ ao

. ~,i, -~"

_~ oos a l a2 o4

[ M ] P E N E M Concentrat ion ( ug /m l ) IM IPENEM Coneent ra t i on (ug /m l )

Figure 1. The cumulative percentage of inhibition of clinical isolates of Pseudomonas aeruginosa, Klebsiella pneumonia, Escherichia coli, and Staphylococcus aureus by imipenem and five broad-spectrum penicillins.

lastatin. We have taken this step with some reservations because of the issues raised earlier, but believe at the present time that imipenem/cilastatin has the greatest potential among the agents currently available, or currently being investigated, to be effective as monotherapy of the febrile granulocytopenic patient. A series of studies that have encouraged us to proceed with this randomized controlled trial are reviewed herein.

IN VITRO SPECTRUM AND ACTIVITY

Imipenem is a new carbapenem antibiotic which is stable in the presence of beta-lactamase and is a potent inhibitor of various plasmid- and chromosomal-mediated beta-lactamases. In vitro testing has shown imipenem to possess a broad antibacterial spectrum with excellent activity against P. aeruginosa and other commonly recovered enteric gram-negative bacilli that infect the granulocytopenic patient. Clinical isolates recovered from patients with cancer treated at the University of Maryland Cancer Center were tested against the new acylampicillins (i.e., piperacillin, mezlocillin, azlocillin), the carboxypenicillins (i.e., ticarcillin, carbenicillin), and imipenem. The relative activities

of.these antibiotics against the four most commonly en- countered pathogenic organisms among patients who are granulocytopenic are shown in Figure 1 as cumulative percentage of inhibition. One hundred and one strains of P. aeruginosa were tested. The MIC5o and MICgo values for imipenem were 1.28 and 2.56 /~g/ml, respectively (Table I). The MIC5o and MICgo values for E. coli, Klebsi- ella pneumoniae, and Serratia marcescens isolates ranged from 0.08 to 0.16 ~g/ml and from 0.08 to 0.32 /~g/ml, respectively. In an attempt to compare relative sus- ceptibilities of bacteremic gram-negative isolates with antibiotics with different peak serum levels and sensitivity breakpoints, we devised a grading system on the basis of the susceptibility breakpoints of National Committee for Clinical Laboratory Standards. Minimal inhibitory concentrations for each isolate are converted to a value which represents the number of dilutions above (positive) or below (negative) the Standard breakpoint. The breakpoints (zero) were 8/~g/ml for imipenem, 64/~g/ml for piperacillin and azlocillin, and 16/~g/ml for ceftazidime, moxalactam, and amikacin. These drugs were then tested against bacteremic isolates of P. aeruginosa which had

64 June 7, 1985 The American Journal of Medicine Volume 78 (suppl 6A)

iMIPENEM/CILASTATIN SYMPOSIUM--WADE ET AL

been recovered from patients with cancer hospitalized at the University of Maryland Cancer Center during the past 14 years and had been saved and stored in our Pseudo- monas culture collection. The mean minimal inhibitory concentrations for imipenem were consistently two to three dilutions below the susceptibility breakpoint (Figure 2). Imipenem was at least one dilution more active against P. aeruginosa isolates than either amikacin, ceftazidime, or moxalactam. Imipenem demonstrated activity similar to that of azlocillin, but was approximately one dilution less active than piperacillin.

Imipenem has also shown activity against a wide range of gram-positive aerobes (Table I) and many anaerobes. Imipenem provides substantial activity against S. aureus and S. epidermidis, an increasingly frequent pathogen isolated from patients with cancer. Although Streptococ- cus faecalis is a rare pathogen among patients with cancer who are granulocytopenic, the activity of imipenem against it is nevertheless noteworthy.

Imipenem is an inhibitor of cell wall biosynthesis and is bactericidal in its action. Most investigators have found equivalence between the minimal inhibitory and minimal bactericidal concentrations when tested against inocu- lums ranging from 103 to 107 colony-forming units/ml.

Therefore, in general, imipenem is more active against many of the enteric gram-negative bacilli, S. aureus and S. epidermidis, than the carboxy and acylampicillin penicillins, with equivalent anti-pseudomonal activity. In con- trast to the new cephalosporins, imipenem is more active against the gram-positive bacteria and P. aeruginosa, with equivalent activity against the Enterobacteriaceae.

Synergistic interaction of two drugs may expand their spectrums of activity and eliminate or prevent the emer-

TABLE I Antibacterial Activity of Imipenem

MICso MICgo Organisms* Humber Tested (#g/mi) (pg/mi)

Gram-negative Escherichia coil 50 0.08 0.08 Klebsiella pneumoniae 49 0.08 0.16 Pseudomonas aeruginosa 101 1.28 2.56 Serratia marcescens 50 0.16 0.32

Gram-positive Staphylococcus epidermidis

(coagulase-negative) 50 0.08 1.28 Staphylococcus aureus 42 0.02 0.02 Streptococcus faecalis 29 1.28 2.56

Inoculum = 105 colony-forming units/ml. *Organisms recovered from single patients at the University of Mary- land Cancer Center.

gence of resistance. Evaluation of imipenem in combination with other antibiotics has been limited but it appears that by bactericidal kinetics the combination of imipenem and an aminoglycoside will provide a beneficial interaction against P. aeruginosa and S. aureus. Checkerboard tech- niques have shown little synergy when imipenem has been combined with an aminoglycoside and tested against P. aeruginosa. This latter observation may, in large part, reflect the inherent anti-pseudomonal activity of imipenem which may make synergy by the checkerboard technique more difficult to demonstrate.

Recently, pairs of beta-lactam antibiotics have been combined and used as therapy of infection among patients with cancer who are granulocytopenic. These como binations have been utilized to take advantage of the observed synergy against some gram-negative bacilli and eliminating the need for the more toxic aminoglycoside

Figure 2. Relative susceptibility of bacteremic isolates of Pseudomonas aeruginosa against imipenem and five broad- spectrum antibiotics.

2

| , AMIKACIN I~,

I¢ MOXALACTAM ; 0 . . . . . . . . m

m CEFTAZ!DIM E ~ -1

Z IAI

AZLOCILUN ~ -2

K

PIPERACILUN ~ -3

[MIPENEM UI~ O -4

. , ~ ........... j ~ f ~ ' ~ .... ~ > ~ -- ~ - -~//Z- .................

/

/ 5 /

TIME PERIOD 83-84

June 7, 1985 The American Journal of Medicine Volume 78 (supp! 6A) 66


TABLE II Mortality of Neutropenic Rats at 72 Hours after Challenge with Pseudomonas aeruginosa

Challenge Dose (dead/challenged)

Therapy 4 LDso 13 LDso 250 LDso

Saline solution 17/20* 33/40 20/20 Moxalactam 19/25* 13/20* 17/20 Amikacin 1/25* 15/40* 18/20 Imipenem -- 0/20* 8/20* Moxalactam plus amikacin 0/20 1/20 10/20" Imipenem plus amikacin - - - - 1/20" *p <0.05.

antibiotics. However, imipenem and piperacillin in vitro are frequently antagonistic against strains of P. aeruginosa [11]. The combination of imipenem plus one of the third- generation cephalosporins--cefoperazone or cefotaxime, for example--has also demonstrated antagonism against some strains of P. aeruginosa and S. marcescens [12]. The observed antagonistic effect with these imipenem- containing antibiotic combinations produced.a four-fold increase in the minimal inhibit()ry concentration of piperacillin or the third-generation cephalosporin but did not increase the minimal inhibitory concentration of imipenem. Induction of beta-lactamase by imipenem was noted in each of these studies and was presumed to be the mech- anism of the observed unidirectional antagonism. The ability of imipenem to induce beta-lactamases may create potential therapeutic problems, and at present would appear to limit its ability to be combined with other beta-lactam antibiotics.

IN VlVO STUDIES

Studies comparing imipenem with the newer acylampicillins (i.e., piperacillin, mezlocillin, azlocillin) or new cepha- Iosporins (i.e., moxalactam, cefoperazone, ceftazidime) in discriminative animal models of infection are few. How- ever, at the University of Maryland Cancer Center we have utilized a neutropenic rat model to test new antibacterial agents alone and in combination against bacterial isolates recovered from blood culture specimens of granulocytopenic patients treated at our center. In brief, this model employs female Sprague-Dawley rats who are con- ditioned in our laboratory. Stool surveillance is performed to guarantee freedom from colonization with the organism to which they are to be challenged. Cyclophosphamide is administered intraperitoneaUy providing a reproducible period of profound neutropenia. Challenge organisms are administered intraperitoneally tothe granulocytopenic rats at varying multiples of the predetermined LDso. Rats are evaluated daily for survival with final mortality recorded 70 hours after bacterial challenge. Challenge organisms are chosen from previously isolated bacteremic strains from

patients treated at the University of Maryland Cancer Center. These strains are characterized according to their in vitro sensitivities to the study antibiotic and, if antibiotic combinations are to be utilized, the in vitro effect of the combination is also determined by a checkerboard technique. Antimicrobial agents tested are given in doses and schedules that provide serum concentrations in rats which closely mimic those achieved in humans.

Imipenem.h.as been employed in this model against a challenge strain of P. aeruginosa [13]. Imipenem was compared with moxalactam and amikacin as single-agent therapy, the single agents were compared with combinations of imipenem or moxalactam plus amikacin, and fi- nally the antibiotic combinations were compared. The P. aeruginosa strain chosen for challenge was highly pathogenic for neutropenic rats, showed in vitro susceptibility to the three antimicrobial agents studied, and in vitro synergy was observed for the aminoglycoside, beta-lactam combinations.

At the challenge dose of 4 LD5o (4 x 107 colony-forming units/ml) moxalactam failed to offer significant protection from death, whereas amikacin did protect (Table II). At 13 LDso (1 x 108 colony-forming units/ml) survival rates were significantly greater for imipenem4reated rats than for rats treated with either moxalactam or amikacin as a single agent. At this challenge dose, the survival rates for rats treated with imipenem as monotherapy were similar to the survival rates for rats treated with the combination of amikacin plus moxalactam. A challenge dose of 250 LD5o (2 x 109 colony-forming units/ml) led to similar survival rates for single-agent imipenem and the combination of amikacin plus rnoxalactam; however, the combination of imipenem plus amikacin provided a significantly better survival rate (Figure 3). The isolation of imipenem- resistant strains of P. aeruginosa during therapy was noted with the single-agent therapy, but occurred no more frequently than among mice treated with the combination of imipenem plus amikacin.

Single-agent imipenem is the only antibiotic we have tested in this neutropenic rat model that has provided survival rates against P. aeruginosa challenge which are similar to those obtained with presently available synergistic antibiotic combinations. Monotherapy with piperacillin, ticarcillin, moxalactam, ceftazidime, and amikacin, when used as treatment of P. aeruginosa bacteremia in this rat model, has been found to be significantly inferior to currently available synergistic antibiotic combinations. These data suggest the possibility that a successful outcome will be obtained for P. aeruginosa infections in granulocytopenic hosts when imipenem/cilastatin is used as single- agent therapy.

PHARMACOKIN ETICS/BACTERICIDAL ACTIVITY

We have administered imipenem/cilastatin as a 1 g dose to volunteers to determine the pharmacokinetics of this



I00-

ANTIBIOTIC

n- O 80 > > n," Or) 60

I--- Z W U 4 0 n,." b.I 0.

20

ADMINISTRATION HOURS POST CHALLENGE 2 8 14 22 30 38 46 54 62 I I I I I I I i I

~ ~ IMIPENEM 30mg/kg

AMIKACIN 20mg/kg

\\ "o MOXALACTAM 150rng/kg / ~, PLUS

" ~ AMIKAClN 20mg/kg

\ SALINE AMIKAClN 2Omg/kg

! I I I I I I I I I I I 1 I I

I0 20 30 40 50 60 70

HOURS POST BACTERIAL CHALLENGE

Figure 3. Antibiotic therapy in neutropenic rats challenged with 250 LDso (2.2 × 109) of Pseudomonas aeruginosa strain number 228.

agent and to establish the bactericidal activity in volunteer serums at one and six hours following antibiotic administration against a stock battery of commonly infecting organisms recovered from our patient population. Imipenem/ cilastatin resembles other beta-lactam antibiotics in that it distributes essentially in the extracellular fluid volume and has a terminal elimination half-life of approximately one hour. Following a 1 g dose given every six hours, mean peak serum levels were 52/~g/ml, one hour post-infusion mean levels were 20/~g/ml, and mean trough levels were 2/~g/ml. The protein binding of imipenem is between 15 and 20 percent. These serum levels will provide free drug concentrations above the MIC9o for the entire six-hour dosing interval for the pathogens (with the exception of P. aeruginosa) most frequently causing bacteremia in granu- Iocytopenic patients with cancer (Table III).

Most beta-lactam antibiotics do not possess a post-antibiotic effect and, consequently, once the antibiotic concentration falls below the minimal inhibitory concentration rapid re-growth of gram-negative bacilli occurs. Aminogly- cosides have a post-antibiotic effect and we have demonstrated that imipenem also has a post-antibiotic effect against P. aeruginosa (Figure 4) [14]. Compared with the control, imipenem or ceftazidime produces rapid bacterial killing. Yet, when a polyvalent beta-lactamase, which de-

stroys either ceftazidime or imipenem, was added to the cultures, P. aeruginosa began to immediately re-grow in the culture samples containing ceftazidime. However, in the imipenem-containing specimens there was a two-hour lag before log phase growth began. If we extrapolate this information to the pharmacokinetic-free drug data, it is possible to predict that the effective time above the MIC9o for P. aeruginosa would be 5.7 hours (3.7 hours plus an additional two hours due to the post-antibiotic effect).

TABLE III Duration of Free Drug Concentration of Imipenem in Excess of the MICgo for Pathogens Frequently Causing Bacteremia in Neutropenic Patients with Cancer

Pseudomonas Klebsiella Escherichia Staphylococcus aeruginosa pneumoniae coli aureus

Number of organisms 101 49 50 42

MIC9o (/~g/ml) 2.56 0.16 0.08 0.02 Duration

above MIC9o* 3.7 >6 >6 >6

*Hours the free drug level exceeds the minimal inhibitory concentration of frequently recovered bacteremic organisms.


IMIPENEM/CILASTATIN SYMPOSIUM- -WADE ET AL

$ Conlrol e.w--e CeflozldlmQ o ~

6

5

4

3

2

CilfloxMImi ~, ~tlaloclomal I / • ~ ~ , 4 ]mlp|nom t bllolo¢lomoll ..~d"

r , . . . l ' ' e , ~

~,~, O~ j j O ~"

&-- ~ I o

~ ""J

I I I I I I L I

I 2 3 4 5 6 7 e TIME(hours)

Figure 4. Post-antibiotic effect of imipenem against Pseu- domonas aeruginosa,

Whether or not this protracted antibacterial activity occurs in the clinical setting remains to be clarified.

Serum bactericidal activity was assessed among healthy volunteers who were administered imipenem (1 g) or moxalactam (2 g) intravenously in a cross-over manner one week apart [15]. Serum bactericidal activity was determined for each antibiotic at one and five and a half hours after administration against 24 strains of P. aeruginosa and six strains each of E. coli, K. pneumoniae, and S. aureus. These isolates were recovered from patients at the University of Maryland Cancer Center (Table IV). The serums from volunteers receiving imipenem had significantly higher bactericidal titers against P. aeruginosa and S. aureus than did the serums of the same volunteers when given moxalactam. Moxalactam produced higher bactericidal titers against the Enterobacteriaceae, but imipenem titers were excellent against these bacteria and

much greater than 1:8 at one hour. Data obtained from a separate trial evaluating the bactericidal activity of the combination of ticarcillin plus amikacin showed its activity against P. aeruginosa to be similar to that of imipenem (Table IV) [16], The activity of imipenem against S. aureus was superior to the antibiotic combination but the combination was more active against E. coil and K. pneumoniae strains, although once again the imipenem bactericidal titers were excellent for these organisms.

CLINICAL EVALUATION AND THERAPY

Imipenem is a unique antibiotic, and its properties suggest that it should be carefully evaluated in patients at high risk, such as those patients with cancer who are granulocyto- penis. The broad antibacterial spectrum is appropriate for such patients, and its pharmacokinetic parameters, post- antibiotic effect, and high degree of serum bactericidal activity against those organisms frequently recovered from granulocytopenic hosts with serious infection make it a promising new beta-lactam antibiotic. These characteristics, as well as the high level of activity against P. aeruginosa used to challenge neutropenic rats--activity which is equivalent to the presently available synergistic bread- spectrum antibiotic combinations--make imipenem an antibiotic worthy of study as monotherapy for fever and suspected infection in granulocytopenic patients with cancer. At present there are no published clinical data using imipenem as monotherapy in such patients. However, Winston et al [17] have studied imipenem as therapy of P. aeruginosa infections and other serious bacterial infections in an open uncontrolled trial. The underlying disease of these patients was varied but 15 of 35 (43 percent) were believed to have life-threatening or ultimately fatal disease, and five of 35 had been receiving immunosup- pressive therapy. An overall favorable clinical response of 89 percent was obtained with imipenem therapy; seven of eight bacteremias were improved or cured. Nausea, with or without emesis, was the only major antibiotic-related side effect, which occurred in one fifth of the patients studied and for the most part was ameliorated by increasing the infusion duration. A Corynebacterium septicemia in a

TABLE IV Comparative Geometric Mean Serum Bactericidal Activity from Volunteers

Imipenem* Moxalastam*

Number One Five and a Half One Five and a Half Organism Tested Hour Hours Hour Hours

Ticarcillin t Plus Amikacin*

Number One Six Tested Hour Hours

Pseudomonas aeruginosa 24 13 3 5 2 31 12 2 Escherichia coil 6 69 6 256 219 7 126 8 Klebsiella pneumoniae 6 46 6 256 149 7 86 8 Staphylococcus aureus 6 224 21 7 2 7 24 3 *Reciprocal mean geometric bactericidal titer. tBactericidal titers from [16].


IMIPENEM:CILASTATIN SYMPOSIUM--WADE ET AL

neutropenic patient undergoing bone marrow transplantation was the only superinfection associated with imipenem therapy. Yet, imipenem-resistant organisms emerged during therapy in nine of 35 patients. The resistant organisms were P. aeruginosa is six cases, Pseudomonas malto- philia in two, and Enterobacter cloacae in one. Three of these nine isolates were of clinical significance, two were associated with treatment failure, and one with infection relapse. All three clinically significant isolates were P. aeruginosa.

The lack of prospective, controlled clinical data regarding imipenem's role in immunocompromised hosts and patients who are granulocytopenic have prompted us at the University of Maryland Cancer Center to undertake such an investigation. In a prospective, randomized double-blind study, imipenem at a dose of 1 g intravenously every six hours is being compared with the control regimen of amikacin (24 mg/kg/per day) plus piperacillin (300 mg/kg/per day) (Figure 5). University of Maryland Cancer Center patients who are granutocytopenic (less than 1,000 granulocytes//~l) and febrile (38°C), with or without a documented site of infection will be eligible for study. Pa- tients will undergo a thorough examination, appropriate cultures will be performed, and patients will then be randomly assigned to receive one of the two treatment regimens. Efficacy, toxicity, and a prospective cost analysis will be examined in both study arms. An unblinded ob- server will review clinical and surveillance microbiology data and the patient's clinical course on a daily basis to provide maximal patient protection.

A number of questions are being posed by this study, but the most important is whether monotherapy with imipenem is equivalent to combination therapy with piperacillin plus amikacin. The sample size needed to answer that question has been calculated on the basis of the following assumptions: alpha error of 5 percent; a beta error of 5 percent or power of 95 percent; an expected control group response rate of 70 percent; and a response rate difference of less than 20 percent to be clinically insignificant. A sample size of 240 patients is necessary to meet these criteria. Assuming that 53 percent of the patients entered into the trial will have a documented infection and be eval- uable for efficacy, a total sample size of 450 is needed. Dependent variables such as toxicity (i.e., nephrotoxicity, ototoxicity, diarrhea) and efficacy (i.e., death due to infection and time to cure) can be examined with this projected sample size. It is anticipated that results from this trial will be available in early 1987.

The significance of the observation of Winston et al [17] regarding the emergence of resistance of gram-negative bacilli among patients treated with single-agent imipenem is unclear at this time, but certainly of potential concern for patients who are profoundly granulocytopenic and who will receive imipenem as monotherapy. Our own animal data (neutropenic rat model) suggested that the develop-

UMCC HIh'TOKY

CANCEff EXAMINATION

GRAN U LOCYTOPENM CULTURES

(< IO00/UL) FEVER (>~o)

Z

Figure 5. A double-bfind, prospective, randomized trial comparing imipenem as monotherapy with piperacillin plus amikacin.

ment of resistance among strains of P. aeruginosa was no more likely to occur with single-agent imipenem therapy than when a synergistic antibiotic combination was employed. All patients treated in our clinical trial will be seri- ally monitored on the basis of culture samples of the nose, gingiva, and rectum to be able to accurately monitor the development of resistance and screen those patients who are at high risk of treatment failure because of colonization with antibiotic resistant flora.

PREVENTION

The prevention of an initial infection or a further (subsequent) infection is important. Many of the currently used prophylactic methods are utilized on the basis of the concept of "colonization resistance." The observations leading to the concept of colonization resistance were made in mouse studies performed by van der Waaij and colleagues [18]. Normally, to induce lasting coionization of the normal mouse alimentary tract, composed of an endogenous flora of aerobes and anaerobes, required an oral inoculum of 107 organisms. Colonization with an ex- ogenous organism of the alimentary tract of a germ-free mouse will occur with as few as 10 to 100 orally administered organisms. Yet, if the alimentary tract flora of the mouse is composed entirely of anaerobes, colonization resistance remains essentially intact suggesting that the alimentary tract anaerobic flora is important in maintaining a barrier against colonization with new species.

Colonization appears to be a prerequisite for the development of infection in many patients who are granulocytopenic. Prophylactic therapy with trimethoprim/sulfameth- oxazole, which suppresses the endogenous, potentially pathogenic, colonizing aerobic gram-negative bacilli, but which preserves the anaerobes, may decrease the incidence of infection for patients at high risk (i.e., patients with acute leukemia undergoing induction or reinduction therapy). However, when fever develops and the need for broad-spectrum antibiotics occurs, suppression of the alimentary tract anaerobes by therapeutic antibiotics frequently follows. Reduction in the alimentary tract anaerobes potentially could be responsible for the increase in



TABLE V Effect of Antibiotic Therapy on Anaerobes Isolated from Stool Specimens

Imipenem Piperacillin Plus Amikacin

Total anaerobic count Number of patients 3

Before therapy 1.5 X 108 During or after therapy* • 2.2 X 107

Qualitative anaerobes t

4 4.7 X 107 3.1 X 10 2

Number of patients 3 2 Bacteroides No change J, 106 Clostridia No change $103 Bifidobacteria No change J, 106 Peptococcus No change J, 105 Veillonella ,~ 10 2 --

Eubacteria - - - -

Actinomyces $103 - -

Fusobacteria No change --

*Stool culture performed during (at least six days) or at completion of therapy. tLog reduction ( $ ) calculated by comparing anaerobe quantitation in paired stool specimens.

subsequent infections observed among those patients at high risk. The EORTC showed many years ago that the risk of further infection was dependent upon the duration of granulocytopenia, but that even for those patients whose granulocyte count returned to normal, the duration of antibiotic therapy was a critical, independent factor in the development of further infections [2]. Antibiotic therapy duration of less than five days was associated with only a 2 percent occurrence of further infection whereas the frequency increased to 25 percent for patients receiving antibiotic therapy for more than 15 days. Thus, it appeared that broad-spectrum antibiotic therapy given for a prolonged period of time increased the risk of later infection. The antibiotics utilized in that trial were the various

two-drug combinations of carbenicillin, cephalothin, and gentamicin. Carbenicillin has reasonably good activity against anaerobic bacteria including some strains of Bac- teroides. This may be pertinent because suppression of anaerobic flora, the dominant flora of the intestinal tract, may allow for overgrowth of organisms which otherwise exist in lesser numbers or for colonization with newly acquired organisms. Thus, the specific antibiotics utilized as therapy may be critical in the development of further infection. We have recently shown a significant difference in the occurrence of further infections when ceftazidime plus tobramycin (105 patients treated) was compared with ceftazidime plus piperacillin (99 patients treated) [19]. Cefta- zidime plus tobramycin, an antibiotic combination known to leave the alimentary tract anaerobic flora intact had a lower 0c(~urrence of further infections than did the alimentary tract anaerobe suppressive regimen of ceftazidime plus piperacillin.

The importance of preservation or suppression of alimentary tract anaerobes remains to be answered. Imipe- nero has characteristics which would be beneficial for such an investigation. Imipenem has a broad-spectrum of activity against the anaerobic flora known to populate the alimentary tract. Yet, less than 1 percent of imipenem is excreted through the biliary tree and recent data from the University of Maryland Cancer Center and other institu- tions have demonstrated that systemically administered imipenem has no quantitative or qualitative effect upon the alimentary tract anaerobes (Table V) [20]. This lack of activity against the alimentary tract anaerobic flora by imipenem is in contradistinction to the profound generalized suppression seen with the antibiotic combination of amikacin plus piperacillin.

Preliminary studies have led to the initiation of a study to answer the question regarding the importance of ali-

PROPHYLAXIS ELIGIBILITY I

TRIMETHOPRIM~SULFA MErHOXAZOLE 4-

AMPHOTERICIN-B ACUTE LEUKEMIA

NEWLY DIAGNOSED RELAPSED

NOT INFBCTE D/APEB RILE INTRODUCTION CHEMOTHERAPY

ANAEROBE PRESERVATION

~ 1MIP~NEM

ANAEROBE SUPPRESSION

EVALUATE L ALIMENTARY CANAL ANAEROBES 2. ACQUISITION 3. SUBSEQUENT INFECTIONS

Figure 6. Study design to test the importance of alimentary tract anaerobe preservation on future colonization and subsequent infection.



mentary tract anaerobe preservation (Figure 6). Patients with acute leukemia or chronic myelogenous leukemia in blast crisis who will undergo induction or reinduction chemotherapy and at the initiation of study are afebrile, not infected, and not receiving therapeutic systemic antibiotics, will be given prophylactic trimethoprim/sulfameth- oxazole plus oral amphotericin-B. At the time of development of fever they will be randomly assigned to receive therapy with an anaerobe-preserving regimen (imipenem/ cilastatin) or an anaerobe-suppressing regimen (amikacin plus piperacillin). The acquisition of new organisms and the development of subsequent infections will be corre- lated with the preservation or suppression of alimentary tract anaerobes. It is anticipated that at least three years of patient recruitment will be required before information from this study will be forthcoming.

COMMENTS

Infection remains a major cause of morbidity and mortality for the patient who experiences periods of profound and persistent granulocytopenia. The need for the development of new more efficacious and less toxic therapy for the treatment of such infections remains important. Imipe- nem is a new beta-lactam antibiotic with a broad-spectrum of activity against the commonly isolated enteric gram- negative bacilli including P. aeruginosa and S. aureus. Imipenem in vitro appears to have activity against coagulase-negative staphylococci, a group of pathogens which

are frequently being recovered. Imipenem produces a high level of bactericidal titer activity in the serums of healthy volunteers as determined by testing against the commonly infecting organisms. The laboratory observation of a post-antibiotic effect against P. aeruginosa may also be pertinent. Imipenem remains the only single agent to provide the same degree of protection as the currently available synergistic antibiotic combinations against P. aeruginosa infections induced in a neutropenic rat model. These characteristics make it a logical choice of study as single-agent therapy of fever and suspected infection in granulocytopenic patients with cancer. Imipenem's lack of effect on the alimentary tract anaerobic flora will be ex- ploited to answer the question regarding the importance of preservation of alimentary tract anaerobes and the development of further infections. Questions relating to the frequency with which imipenem-resistant organisms will develop, or the toxicity of such single antibiotic therapy when compared with combination therapy remain to be answered.

The approach to new antibiotic testing as outlined by the group from the University of Maryland Cancer Center and Division of Infectious Diseases of the University of Maryland School of Medicine and Baltimore Veterans Medical Center which includes in vitro, in vivo, human volunteer studies, and well-designed clinical trials will provide the answers necessary to determine the role of new antibiotics such as imipenem in the treatment of infection in patients with granulocytopenia.

REFERENCES

1. Schimpff SC: Therapy of infection in patients with granulocytopenia. Med Clin North Am 1977; 61 ! 1101-1118.

2. EORTC International Antimicrobial Therapy Project Group: Three antibiotic regimens in the treatment of infection in febrile granulocytopenic patients with cancer. J Infect Dis 1978; 137: 14-29.

3. Levine AS, Schimpff SC, Graw RG Jr, Young RC: Hematologic malignancies and other marrow failure states: progress in the management of complicating infections. Semin Hematol 1974; 11: 141-202.

4. Wade JC, Schimpff SC, Newman KA, Wiernik PH: Staphylococ- cus epidermidis: an increasingly but frequently unrecognized cause of infection in granulocytopenic patients. Ann Intern Med 1982; 97: 503-508.

5. Murphy MT, Campbell B J, Marsh J, et ai: Characteristics of coagulase-negative staphylococcus isolated from marrow trans- plant patients. Am Soc Microb 1980; 139: 297.

6. Winston DJ, Dudnick DV, Chapin M, Ho WG, Gale RP, Martin W J: Coagulase-negative staphylococcal bacteremia in patients receiving irnmunosuppressive therapy. Arch Intern Med 1983; 143: 32-36.

7. Love LJ, Schimpff SC, Schiffer CA, Wiernik PH: Improved prog- nosis for granulocytopenic patients with gram-negative bacte- ramia. Am J Med 1980; 68: 1-10.

8. Standiford H, Joshi J, Wade J, et ah Correlation of serum bactericidal activity and response in profoundly granulocytopenic cancer patients with gram-negative bacteremia (abstr 390).

In: Proceedings of the Twenty-fourth Interscience Conference on Antimicrobial Agents and Chemotherapy, Washington, 1984; 159.

9. Pickard W, Durack D, Gallis H: A randomized trial of moxalactam versus tobramycin plus ticarcillin (T + T) in 50 febrile neutropeic patients (abstr 5). In: Proceedings of the Twenty- second Interscience Conference on Antimicrobial Agents and Chemotherapy, Miami Beach, Florida, 1982; 66.

10. Pizzo P, Thaler M, Hiemenz J, et al: Monotherapy versus combination antibiotics for the initial empiric management of febrile (F + ) granulocytopenic (G + ) cancer patients (abstr 380). In: Proceedings of the Twenty-fourth Interscience Conference on Antimicrobial Agents and Chemotherapy, Washington, 1984; 158.

11. Bertram MA, Young LS: Imipenem antagonism of the in vitro activity of piperacillin against Pseudomonas aeruginosa. An- timicrob Agents Chemother 1984; 26: 272-274.

12. Miller MA, Finan M, Yousuf M: in vitro antagonism by N-formirni- doyl thienamycin and cefoxitin of second- and third-generation cephalosporins in Aeromonas hydrophilia and Serratia marcescens. J Antimicrob Chemother 1983; 11:311-318.

13. Johnson DE, Calla FM, Snyder MJ, Warren JW, Schimpff SC: Imipenem therapy of Pseudornonas aeruginosa bacteraemia in neutropenic rats. J Antimicrob Chemother 1983; 12: 89-96.

14. Bustamante CI, Drusano G, Tatem B, Standiford H: Post-antibiotic effect of imipenem against Pseudomonas aeruginosa. Antimicrob Agents Chemother 1984; 26: 678-682.



15. Standiford H, Drusano G, Bustamante C, Rivera G, Tatem B: Thienamycin versus moxalactam: comparative serum pharmacokinetics and bactericidal activity in normal volunteers. In: Proceedings of the Thirteenth International Congress of Chemotherapy, Vienna, Austria, 1983; 95/51.

16. Standiford H, Drusano G, Fitzpatrick B, Tatem B, Schimpff S: Bactericidal activity of ceftazidime in serum compared with that of ticarcillin combined with amikacin. Antimicrob Agents Chemother 1984; 26: 339-342.

17. WinSton DJ, McGrattan MA, Busuttil RW: Imipenem therapy of Pseudomonas aeruginosa and other serious bacterial infections. Antimicrob Agents Chemother 1984; 26: 673-677.

18.' van der Waaij B, Berghuis-de Vries JM, Lekkerkerk-van der

Wees JEC: Colonization resistance of the digestive tract in conventional and antibiotic-treated mice. J Hyg (Lond) 1971; 69:405-411.

19. Joshi J, Ruxer R, Newman K, et ah Double Beta-lactam versus an aminoglycoside +Beta-lactam combination as empiric therapy for granulocytopenic cancer patients (abstr 383). In: Proceedings of the Twenty-fourth Interscience Conference on Antimicrobial Agents and Chemotherapy, Washington, 1984; 158.

20. Joshi JH, Wade JC: Changes in the faecal anaerobic flora by antibiotic administration (abstr 109). In: Proceedings of the Third International Symposium on Infections in the Immuno- compromised Host, York University, Toronto, Canada, 1984.


Potential of imipenem as single-agent empiric antibiotic therapy of febrile neutropenic patients...

Documents

Transcript of Potential of imipenem as single-agent empiric antibiotic therapy of febrile neutropenic patients...