Efficacy and Pharmacokinetic/Pharmacodynamic Parameters of Cefoxitin in Women With Acute Pyelonephritis Without Severity Symptoms Due to Extended-spectrum β-lactamase Producing Escherichia Coli

Overview

Escherichia coli is the primary cause of urinary tract infections and Gram-negative bacteremia worldwide. Since the early years of the 21st century, E.coli has acquired a new mechanism of resistance to antibiotics: extended spectrum β-lactamase (ESBL), type CTX-M. These ESBL inactivate most β-lactams, the preferred class of antibiotics for the treatment of severe E.coli infections. Moreover, the strains that produce these ESBL are often resistant to other classes of antibiotics. Their rapid spread constitutes a major public health concern because of a serious risk of therapeutic impasse. Treatment options in cases of infection with ESBL-producing E.coli are often limited to carbapenems, a class of more recently developed β-lactams. Carbapenems have a very wide spectrum of activity but their effectiveness is threatened by the emergence of strains producing carbapenemases. The development of therapeutic alternatives to treat ESBL-producing E.coli infections is therefore essential. Cephamycins, including cefoxitin, are β-lactams marketed in the seventies but their use was practically abandoned when wide spectrum antibiotics became available. They are distinguished by the presence of an α-methoxy group in position 7 which interferes with the action of the extended-spectrum β-lactamase and renders it ineffective against cephamycins. Cefoxitin is therefore active in vitro against ESBL-producing E.coli and offers the advantage of a narrower antibacterial spectrum, thus reducing the selection pressure and the emergence of resistance. However, the in vivo activity of cefoxitin for the treatment of ESBL-producing E.coli infections has never been measured. Furthermore, available pharmacokinetic and pharmacodynamic (PK/PD) data for cefoxitin are dated and incomplete, which raises many questions concerning the optimal dosage regimen. We have shown in a mouse model of ESBL-producing E. coli CTX-M pyelonephritis that cefoxitin efficacy is comparable to that of carbapenems without selecting resistant mutants. Cefoxitin could thus constitute an alternative to carbapenems for the treatment of pyelonephritis caused by ESBL-producing E.coli.

Clinical Study: This is a prospective, multicentric, non-comparative, proof of concept study evaluating the efficacy of cefoxitin in women with acute ESBL-producing E. coli pyelonephritis without severity symptoms. This antibiotic is already commercially available and its tolerability profile is well known, but it has never been previously evaluated to treat this particular type of infection. The primary objective of this study is to assess the clinical and microbiological response to treatment with cefoxitin in women with acute pyelonephritis without severity symptoms due to ESBL-producing E. coli. The secondary objectives are: In women with pyelonephritis caused by ESBL-producing E.coli: * Gain a better understanding of the current epidemiology of ESBL-producing E. coli associated with febrile UTI * Assess the impact of cefoxitin use on the emergence of resistance in the colonising bacteria of the gastrointestinal tract in women with pyelonephritis treated with cefoxitin and identify the associated mechanisms * Characterise the PK/PD parameters for cefoxitin in order to optimise the dosage strategy for that antibiotic * Assess the tolerance profile of cefoxitin Determination of sample size, feasibility of enrolment No determination of the minimal sample size could be calculated in this proof of concept study but a total enrolment of 40 patients in the 8 participating centers seems adequate to evaluate the efficacy of cefoxitin for treating uncomplicated acute pyelonephritis and to study its impact on the emergence of resistance. Half of enrolled women will have repeated measurement of serum concentration of cefoxitin: 6 blood samples collected within 6 hours following the injection of one of the doses received between the 3rd and the last day of treatment. A security assessment by an independent committee will be conducted on the first five patients enrolled. The precise incidence of ESBL-producing E.coli acute pyelonephritis without severity symptoms in the 8 participating centers is unknown. At BICHAT hospital, the incidence of ESBL-producing E.coli in 2008 was 0.66/100 admissions (57 778 admissions/year). It was 0.46/1000 hospitalisation days at BEAUJON in 2009 with 36.5% from urine samples. The same year, Saint-Louis hospital recorded 18 bacteremias secondary to infection with ESBL-producing E.coli with 6% originating from a UTI. Those last numbers underestimate the actual incidence of ESBL-producing E coli pyelonephritis at that hospital, only taking into account the UTI associated with bacteremias. We estimate that there are approximately 10 cases of ESBL-producing E.coli pyelonephritis in these centers each year. Outline and conduct of the study/investigations performed by the CIC As soon as the antibiogram is available, confirming diagnosis of an ESBL-producing E. coli and after verification of the inclusion criteria, information and obtention of the the written consent, cefoxitin (2 grams intravenous perfusion of 30 minutes every 6 hours) will be given for 10 days. A control visit and urine culture will be performed at day 2, day 7 and at the end of treatment, day 10. A rectal swab will be performed at enrolment before the first dose of antibiotic, at the end of treatment (day 10) and at day 40. The inclusion period will be 2 years and each patient's participation will last 40 days. The PK study will be performed on 20 patients enrolled at one of the 4 CIC at hospital Saint-Louis, BICHAT, HENRI-MONDOR and HEGP. Measurement of serum concentration of cefoxitin will be done on the 3rd day of treatment, or between the 9th and 12th dose of antibiotic. Samples will be taken as follows: t0 (immediately before beginning infusing cefoxitin); t30 min (at the end of infusion); t45 min ; t1h ; t3h ; t6h (before the next dose) for a total of 6 samples. Evaluation criteria for the primary objective Clinical and microbiological response defined at the end of cefoxitin treatment by the presence of the 3 following criteria: (i) afebrile (temperature \> 36° and \< 38°C), (ii) resolution of urinary symptoms present at the time of diagnosis: dysuria, urgency, frequency, cloudy urine, pain on urination, pelvic or lumbar pain (iii) sterile urine culture. Evaluation criteria for the secondary objective (i) Clinical and microbiological response at 48h after beginning treatment with cefoxitin; (ii) Early relapse at day 40; (iii) PK/PD parameters; (iv) side effects of cefoxitin. Statistics, PK/PD modelling, cefoxitin dosing regimen optimisation Qualitative variables will be reported in percentage of the studied subjects and quantitative variables by a median. Comparisons will use non-parametric tests. The tests will be bilateral and use a significance threshold of 5%. The rate of clinical and microbiological response at the end of treatment with cefoxitin will be reported as a percentage with an estimate of the confidence interval (CI) of 95% (using a binomial distribution). PK analysis will be conducted using a population approach and average PK parameters for cefoxitin will be estimated as well as interpatient variability. The PK population analysis will examine different demographic and clinical covariables in order to weigh their influence on PK/PD parameters; the studied covariables will be: age, weight, BMI, blood pressure and creatinine clearance. In order to study the relationship between pharmacokinetics and clinical response at H48 and at the end of treatment as well as at J40±5, the patients' PK parameters will be compared using the Wilcoxon and Mann-Whitney non-parametric tests. ESBL-producing E.coli strains responsible for pyelonephritis in women: typing and identification of subtype ESBL-producing strains will be screened with the synergy method, using a disk containing a 3rd generation cephalosporin, like cefotaxime, and another disk containing a β-lactamase inhibitor, clavulanic acid, according to the recommendations of the competent authorities of antibiogram. The isolated strains will then be analysed further on three main aspects: (i) the phylogenetic group will be determined by multiplex PCR (ii) the strains' sequence-typing will be done using Multiple Sequence Typing (iii) the presence of 46 genes associated with extra-intestinal virulence factors will be identified through multiplex PCR. The determination of ESBL type will also be done by PCR. Detection of cefoxitin resistant strains colonising the gastrointestinal tract of women with pyelonephritis before treatment with cefoxitin ; emergence of resistance under treatment and determination of associated mechanisms of resistance to cefoxitin Screening for enterobacteriaceae with reduced-sensitivity to cefoxitin will be done by plating a rectal swab (or a stool sample) on DRIGALSKI media containing 10 mg/l of cefoxitin. Acquired resistance mechanisms in species normally sensitive to cefoxitin (E. coli, Klebsiella spp.) will be characterised: hyperproduction of Amp C in E. coli, hyperexpression of efflux pumps and changes in purines in E. coli et K. pneumoniae. If comparing strains is necessary, it will be done using ERIC2 PCR as previously described in literature. Ethical aspects The research protocol will be presented to the ethics committee, once consent from the promoter is obtained (with proof of the required insurance and payment of fees). The opinion of the committee will be included in a form delivered by the promoter to the competent authorities before beginning the study. All enrolled patients will be required to sign an informed consent before any investigation or treatment is performed. The consent forms will be kept for 30 years by the CIC. Patients will be informed of the results after completion of the study. Expected public health impact of study The expected benefits of using cefoxitin in women with uncomplicated acute ESBL-producing E.coli pyelonephritis are (i) an clinical and microbiological effectiveness similar to that of carbapenems;(ii) a decrease in the selection pressure on bacteria colonising the gastrointestinal tract, thus contributing to prevent the emergence of carbapenemase-producing Enterobacteriaceae (iii) a decrease in drug-associated cost.