The efficacy of pivmecillinam in oral step-down treatment in hospitalised patients with E. coli bacteremic urinary tract infection; a single-arm, uncontrolled treatment study

The trial was registered at ClinicalTrials.gov under the identifier: NCT03282006 13/09/2017 and approved by The Regional Committees for Medical Research Ethics South East Norway (2015/2384/REK sør-øst) and the Norwegian Medicines Agency (SLV; reference No 16/06018-09; EudraCT No 2016-000984-18) before initiation. The Clinical Trial Unit at Oslo University Hospital monitored the project. All participants gave written informed consent according to the International Conference on Harmonization of Good Clinical Practice and the Helsinki Declaration.

Patients ≥ 18 years old hospitalised at Vestfold Hospital Trust from September 2017 through March 2020 with E. coli bacteremia due to febrile UTI were eligible for inclusion. Both men and women were recruited. Inclusion was performed during the first 3 days after hospital admission. Inclusion/exclusion criteria are summarized in Table 1.

An overview of the treatment intervention and patient course is summarized in Fig. 1. A variety of blood samples, including blood cultures, and midstream urinary sample or secondarily a catheter sample, for quantitative urine culture, were collected at hospitalisation (day 0) and on discharge from hospital (day 3). Ultrasonic examination or CT-scan of the urinary tract was performed prior to inclusion and chest X-ray within the 1st day. Bacterial isolates from blood and urine were identified by Matrix Assisted Laser Desorption Time of Flight Mass Spectrometry (MALDI-TOF, Bruker Daltonics) and tested for antibiotic susceptibility according to European Committee on Antimicrobial Suseptibility Testing (EUCAST) disc diffusion methodology [20]. Patients infected with fully susceptible isolates (MIC ≤ 1 mg/L) and non-wild-type E. coli isolates with reduced susceptibility to mecillinam (MIC = 2–8 mg/L) were considered candidates for inclusion. Patients with resistant isolates (MIC > 8 mg/L) were not included. Extended-spectrum beta-lactamases (ESBL) producing isolates did not disqualify from study participation. All bacterial isolates from blood cultures were frozen for further research. All participants were scored according to the Charlson Comorbidity Index (CCI) [21]. Self-rated health status was measured using three different methods, Table 2. To compare values in the EQ-5D-3L health state, we used the Danish TTO set of weights [22]. The protocol describes the use of EQ-5D-3L combined with EQ-VAS-thermometer as PROM (Patient Reported Outcome Measure). In designing the CRF (Case Report Form) we added a standardised questionnaire as a third method for quantifying self-rated health status.

The participants were initially given parenteral antibiotics (ampicillin/gentamicin, cefotaxime or piperacillin/tazobactam) according to Norwegian guidelines [23]. Parenteral drug administration was ended on day 3 and the participants were put on oral treatment with investigational drug (PIV) tablets 400 mg (Selexid, LEO-pharma A/S) QID. In case of fever on day 3, the patients were observed in hospital for an additional 24 h without deviation from the study protocol in other respects. The patients received study medication in a pre-filled cassette from the department and the duration of the treatment was 1 week.

Three days after discharge, the patients were contacted by phone by the study personnel (trained nurse or doctor) and queried for adverse events or treatment failure. One week after discharge, the patients delivered blood samples and urine specimen and were interviewed by phone. The test of cure (TOC) clinical assessment was carried out 2 weeks ± 1 day following discharge; i.e. 1 week after treatment cessation. PROMs were performed as outlined above together with blood pressure, pulse and temporal body temperature recordings (Exergen Thermometer TemporalScanner Model TAT-5000), as well as bladder scan for residual urine assessment (BladderScan BVI 3000). A urine specimen was collected for bacteriology. Blood tests were sampled and the medication cassette was compiled and checked regarding compliance.

The patient file was checked 1 month ± 3 days following discharge to document whether the patient was still alive or readmitted to hospital (due to urinary tract infection, Clostridioides difficile Infection (CDI) or other causes). The participant was contacted to clarify whether urinary tract infection or CDI had occurred since TOC.

The primary endpoint was defined as a combination of three elements; afebrility (temporal body temperature < 38 °C), no need for retreatment and improvement in self-reported health status at TOC. In the assessment of treatment success, all three aspects were weighted equally and all three criteria had to be redeemed.

Secondary endpoints included non-significant bacterial growth in urine on TOC, C-reactive protein value < 30 mg/L, treatment-requiring UTI or readmission due to UTI after TOC but less than 1 month after discharge, readmission for other reasons or death less than 1 month after discharge, serious adverse drug events and occurrence and CDI.

Interim analysis was not done, but certain rules for premature study-abortion were established in case of absence of clinical cure on day 10 or need for hospital admission due to SAE or treatment failure for a significant share of patients as described elsewhere (www.​ClinicalTrials.​gov).

Confidence interval (CI) for the observed proportion of success of primary outcome was obtained using binomial exact 95% confidence interval. To analyse potential effect of different variables on probability of success of the treatment, we performed univariate logistic regression and reported odds ratios, 95% CIs and p-values for CRP-values, age, sex, persistent E. coli bacteriuria, IDC (indwelling urinary catheter), presence of non-wild type E. coli, CCI, renal function and diabetes mellitus. Data was analysed using the SPSS Statistics 26 package and Stata.

The trial algorithm, including the number of patients assigned for study treatment, is visualised in Fig. 2. The main reason for exclusion was a different origin of infection, e.g. cholecystitis, cholangitis, diverticulitis, liver abscess or synchronous bacterial infection outside the urinary tract.

Fifty-three patients initiated treatment with the investigational drug. Two patients interrupted the treatment due to adverse events and one refused to continue trial participation without reason. One participant was diagnosed with concomitant viral pneumonia (respiratory syncytial virus) and on discharge referred to a nursing home. After 3 days, the patient was readmitted to the hospital due to fever, dyspnea and clinical deterioration. Broad-spectrum intravenous antibiotics were administered, classifying the participant as treatment failure. The patient was not able to adhere to the following study protocol but was included among the 50 patients evaluated for primary endpoint. Seven participants had malignant comorbidity and one had solid organ transplant (pancreatic allograft). Characteristics of the participants are listed in Table 3.

All of the 476 subjects screened for eligibility had bacteremic growth of E. coli, of whom 23 (4.8%) were infected with mecillinam-resistant strains which corresponds to national susceptibility data [5]. Microbiological characteristics of the participants’ E. coli isolates are summarised in Table 3. Two isolates were expressing ESBL. Forty-five subjects (90%) had growth of E. coli in their urine on admission. The main reason for urine culture failure were pre-sampling antibiotics (three patients). Two patients did not deliver a urine sample.

Of the 50 patients included in the per-protocol evaluation, 44 patients achieved success in the primary composite endpoint (Table 4). None of the patients had fever at TOC (missing data = 1). Four of the patients were readmitted to hospital, of whom two patients were diagnosed with UTI, one with spondylodiscitis and one with calculous cholecystitis. In total five patients needed retreatment. The results from the three different methods measuring patients reported health status, showed considerable degree of intrapersonal divergence and are listed in Additional file 1: Fig. S1. Characteristics of the six patients with treatment failure are summarised in Table 5.

On TOC, 14 out of the 48 patients (29.2%) had significant growth (> 10³ CFU/mL) of E. coli (Table 2). In 12 cases, the isolates exposed the same susceptibility pattern as in the pre-treatment cultivation.

The treatment showed high efficacy with respect to biochemical outcome measures with normalisation of inflammatory markers as total amount of leukocytes, neutrophils, CRP and PCT in 90% of the patients. Five patients had CRP ≥ 30 mg/L on TOC. Among these, four were treatment failures giving a significant lower probability of treatment success on TOC for CRP ≥ 30 mg/L compared to CRP < 30 mg/L (p < 0.001; OR = 0.006, 95% CI 0.00–0.11). The associations between clinical/laboratory parameters vs primary outcome are listed in Additional file 1: Table S2.

At study completion (day 33 ± 3), six patients had received antimicrobial treatment for UTI after TOC. One patient developed CDI 2 weeks after treatment completion following intravenous cephalosporin treatment at readmission due to respiratory infection as described above.

In general, the study drug was well tolerated. The side effects were essentially mild without need for further intervention or treatment. The exception was for two patients who developed intolerable side effects (skin-eruption and nausea/vomiting) which resulted in interruption of the study protocol. An overview of adverse events is displayed in Additional file 1: Table S3. The participants who completed the trial were largely compliant to the study treatment and no patient skipped more than two tablets. Forty-two participants (84%) adhered completely to the treatment regimen.