Background (brief):

Shigellosis is the second leading cause of death due to diarrheal diseases worldwide (>200,000 deaths/year). Though the mortality rate associated with Shigellosis has decreased, the fact that the bacteria have acquired resistance to multiple antibiotics, is a cause for major concern. Oral azithromycin and intravenous ceftriaxone are the recommend first and second line therapies, respectively in Bangladesh. Approximately 20% of Shigella isolates are resistant to azithromycin suggesting that a substantial number of children will require second-line therapy. While resistance to ceftriaxone in shigellosis is low in Bangladesh at 5%, the potential for rapid emergence of antibiotic resistance to this third-generation cephalosporin and ceftriaxone's resource-intensive delivery method, underscore the need for evidence-based alternative antibiotic regimens for multidrug resistant Shigella infections

Hypothesis:

Children treated with tebipenem-pivoxil will have no worse clinical and microbiologic failure rates compared to ceftriaxone.

Primary Aim To determine whether tebipenem-pivoxil is clinically non-inferior to the currently WHO-recommended second line Shigella therapy (ceftriaxone) 3 days after treatment initiation.

Hypothesis: Children randomized to tebipenem-pivoxil experience no more clinical failures than children treated with ceftriaxone 3 days after treatment initiation.

Secondary Aim:

To determine whether tebipenem-pivoxil is clinically non-inferior to the currently WHO-recommended second line Shigella therapy (ceftriaxone) 7 and 30 days after treatment initiation.

To determine whether tebipenem-pivoxil is microbiologically non-inferior to the currently WHO-recommended second line Shigella therapy (ceftriaxone) 7 and 30 days after treatment initiation.

Describe the number of adverse events, between children with shigellosis treated with oral tebipenem-pivoxil or IV ceftriaxone.

Compare the prevalence of ceftriaxone and carbapenam resistance, as well as ESBL-and carbapenemase-producing Escherichia coli, in children treated with tebipenem-pivoxil or ceftriaxone 7 and 30-days after initiation of second-line therapy.

Methods:

Investigators propose a phase IIb randomized controlled trial (RCT) to determine the efficacy and safety of oral tebipenem-pivoxil, compared to IV ceftriaxone, for children with Shigella infections unresponsive to first-line antibiotic therapy. Bangladeshi children aged 24 to 59 months with suspected Shigella infections and no clinical improvement within 48 hours of first-line therapy will be randomized to a 3-day course of oral tebipenem-pivoxil (4 mg/kg 3x daily) or 3-days of IV ceftriaxone (50 mg/kg 1x daily). The children will be evaluated for key clinical, microbiologic, and safety outcomes during the subsequent 30-day period. Additionally, investigators propose a lead in study of 15 patients to confirm the safety profile and pharmacokinetics and efficacy of tebipenem in the study population. During this pharmacokinetic study period investigators will compare 15 children with oral Tebipenem randomizing with 15 children with oral Azithromycin arm. Investigators will also check invitro susceptibility of Tebipenem-pivoxil in 200 shigella isolates prior to the clinical trial in collaboration with Infectious Diseases Division, icddr,b.

Randomization Block randomization (1:1) in random sized blocks of will be used to assign treatment groups at study enrollment by an independent statistician. Treatment allocation (once assigned) will be known to the managing clinician and the participant due to the differing drug delivery mechanisms of the two antibiotics (oral vs. injectable). However, the team conducting the statistical analyses will be blinded to treatment allocation (allocation will appear A and B).

Background of the Project including Preliminary Observations:

Shigellosis is the second leading cause of death due to diarrheal diseases worldwide (>200,000 deaths/year). Among children under 5 years, 60,000 deaths and 74,000,000 cases of diarrhoea were attributed to Shigella in 2016, approximately 20% of which occurred in South Asia. In its severe form, Shigella invades the intestinal tissue resulting in the clinical manifestation of dysentery, blood or mucoid stool. The World Health Organization (WHO) recommends antibiotic therapy for children with Shigella dysentery based on evidence from randomized trials demonstrating clinical and microbiologic benefit of antibiotics for dysentery (>60% of which is caused by Shigella infections). Oral ciprofloxacin and intravenous (IV) or intramuscular (IM) ceftriaxone are the recommend first and second line therapies, respectively. In Bangladesh, based on the prevalence of Shigella isolates with resistance to ciprofloxacin (~70%), the macrolide antibiotic azithromycin is used as a first line therapy with ceftriaxone (second-line) reserved for the most severe cases. Approximately 20% of Shigella isolates are resistant to azithromycin suggesting a substantial number of children will require second-line therapy (Dhaka hospital, data unpublished). While a study conducted at icddrb showed resistance to ceftriaxone in shigellosis is low in Bangladesh at 2-5%, but Dhaka hospital surveillance system identified around 10% ceftriaxone resistance in under 5 children with shigellosis in recent years (from unpublished data), the potential for rapid emergence of antibiotic resistance to this third-generation cephalosporin and ceftriaxone's resource-intensive delivery method, underscore the need for evidence-based alternative antibiotic regimens for multidrug resistant Shigella infections.

In a patency document regarding the dosage of oral tebipenem pivoxil, it was reported that the drug is effective against a number a Gram negative bacteria, including Shigella. To evaluate the susceptibility of Tebipenem -pivoxil among Bangladeshi shigella isolates, investigators will perform in vitro susceptibility testing with previously collected 200 isolates in collaboration with Infectious Diseases Division of icddr,b. Assessing the need for new therapeutic regimes to counter the growing threat of development of antimicrobial resistance and development of multi-drug resistant strains, Investigators propose a phase IIb randomized controlled trial (RCT) to determine the efficacy and safety of oral tebipenem-pivoxil, compared to IV ceftriaxone, for children with Shigella infections unresponsive to first-line antibiotic therapy. Bangladeshi children aged 24 to 59 months with suspected Shigella infections and no clinical improvement within 48 hours of first-line therapy will be randomized to a 3-day course of oral tebipenem-pivoxil (4 mg/kg 3x daily) or 3-days of IV ceftriaxone (50 mg/kg 1x daily). The children will be evaluated for key clinical, microbiologic, and safety outcomes during the subsequent 30-day period. Additionally, investigators will monitor the acquisition of antibiotic resistance, including ESBL- and carbapenemase-producing E. coli, among enrolled children to determine the clinical and public health risk of using carbapenem antibiotics in this context. Furthermore, a lead in study of 15 patients is proposed to confirm the safety profile and pharmacokinetics of tebipenem in the study population. In parallel investigators will enroll another 15 children in Azithromycin arm. At the end of study, investigators will compare the two drugs in terms of safety and efficacy both clinically and microbiologically against shigellosis in children.

Research Design and Methods

Pilot study for efficacy and pharmacokinetics for Tebipenem:

After consent is received, children will be screened in Dhaka Hospital, icddr,b after admission and eligible children with suspected shigella infection (clinical features of fever, mucus and/or blood in stools, tenesmus and RBC and leucocytes >10 per hpf in stool) will be randomized local standard antibiotic azithromycin 10 mg/kg once daily for 5 days or oral tebipenem-pivoxil (4 mg/kg TID x 3 days).

Purposive sampling will be done for pilot study with 15 children in arm. Those who do not respond to treatment within 48 hours or whose symptoms worsen will be switched to IV ceftriaxone. The point of randomization will be considered Day 0; 3, 7, and 30 days after randomization into the trial, children will be assessed clinically by a physical exam. Anthropometric measurements will be taken from children at initiation of first line therapy and at day 3, 7, and 30 of follow-up.

Stool samples will be collected from children at prior to randomization and at day 3, day 7, and day 30 of follow-up. Stool samples will be divided into three equal parts, part 1 used for microbiologic culture and parts 2 and 3 stored at -80°C. For the first stool sample an additional portion of the stool sample will be used for microscopy and fecal leukocyte determination. Microbiologic culture (for isolation of Shigella and E. coli and antibiotic susceptibility testing [AST]) will be performed on fresh stool on day 0, day 3, day 7, and day 30. Quantitative PCR to detect the presence of DNA of Shigella through the identification of the ipaH gene will be performed on day 0, day 3 and day 7 frozen stool samples in a single batch at the end of the study. Culture-results will not be considered as confirmed Shigella infection as previous reports indicate that that approximately 60% of clinically relevant Shigella infections are missed by culture.

3 ml of blood will be collected on day 3 to assess AST, ALT, serum creatinine, and serum carnitine levels by ELISA) and to assess drug exposure (if possible). In addition, 3 ml of blood will be collected from participants at Day 0, Day 7 and Day 30 of follow-up for the isolation of peripheral blood mononuclear cells, which will be subsequently cryopreserved for future immunological studies.To confirm the tebipenem pharmacokinetic and safety profile in the patient population, a lead in study will be performed with the first 15 participants enrolled in the tebipenem-pivoxil arm. The 15 children will have additional 0.5 ml blood samples (one tenth tsf) collected via traditional method at time points distributed over days 0-3 to evaluate tebipenem pharmacokinetics.

As investigators have targeted to evaluate the pharmacokinetics of Tebipenem in shigellosis of 15 children. It will not be possible to have culture report before 48 hours, so investigators planned to evaluate 15 children with Tebipenem arm among those at least 50% children should have confirmed shigella infection in culture.

Assessment of efficacy trial:

The decision will be to progress to the main study if the proportion of clinical successes in the Tebipenem arm is greater than or equal to the number of successes in the Azithromycin arm among shigella confirmed cases.

Objective At the end, investigators might only get a conclusion whether Tebipenem as treatment of Shigella is somewhat better than Azithromycin based on clinical improvement and microbiological results.

Main trial:

Study procedure:

After consent is received, children will be screened and monitored in the study in hospital despite not yet being randomized (Day -2). Children will be immediately started on first-line antibiotic therapy according to the local standard of care (azithromycin 10 mg/kg once daily for 5 days). At or within 48 hours of treatment initiation, the study physician will examine the child and determine whether or not the child is considered to have clinical failure with first-line therapy (presence of fever, clinical deterioration, blood in stool, ≥3 loose/watery stool 48 hours after initiation of therapy). Children who have clinically failed at 48-hours since 1st line therapy initiation will be randomized to one of the two treatment arms: oral tebipenem-pivoxil (4 mg/kg TID x 3 days) or IV ceftriaxone (50 mg/kg QD x 3 days). Children randomized to tebipenem-pivoxil that do not respond to treatment within 48 hours or whose symptoms worsen will be switched to IV ceftriaxone. The point of randomization will be considered Day 0; 3, 7, and 30 days after randomization into the trial, children will be assessed clinically by a physical exam. Anthropometric measurements will be taken from children at initiation of first line therapy, randomization (enrolment in the RCT), and at day 3, 7, and 30 of follow-up.

Stool samples will be collected from children at Day -2 (prior to starting first-line therapy) and among those randomized, prior to administration of the first dose of second-line therapy (day 0), at day 3, day 7, and day 30 of follow-up. In the main clinical trial, we will additionally collect a flocked rectal swab at day-2, day 0 (before randomization), day 3, day 7 and day 30 for storage and eventual biomarker assessment. A series of inflammatory biomarker assays will be conducted on whole stool and rectal swab samples to determine markers of Shigella detection and treatment response upon the availability of funds. Stool samples will be divided into three equal parts, part 1 used for microbiologic culture and parts 2 and 3 stored at -80°C. For the first stool sample (at Day -2) an additional portion of the stool sample will be used for microscopy and fecal leukocyte determination. Microbiologic culture (for isolation of Shigella and E. coli and antibiotic resistance testing [AST]) will be performed on fresh stool on day -2, day 0, day 3, day 7, and day 30. Quantitative PCR to detect the presence of DNA of Shigella through the identification of the ipaH gene, will be performed on day -2, day 0, day 3 and day 7 frozen stool samples in a single batch at the end of the study. Culture-results will not be considered as confirmed Shigella infection as previous reports indicate that that approximately 60% of clinically relevant Shigella infections are missed by culture.

Blood collection at Day 3 to assess AST, ALT, serum creatinine, serum carnitine levels and at Day 0, 7, and 30 to perform any future immunological studies will be done in the pilot phase of the project only, it will not be repeated in the main trial.

Sample Size To the best of our knowledge, no randomized clinical trials have compared treatment options for clinically non-responding children with shigellosis or children with drug resistant (or presumed drug-resistant) Shigella. Therefore, the sample size estimation was derived from the most recent of the three trials of IV/IM ceftriaxone for shigellosis which compared a 3-day course of oral ciprofloxacin to IM ceftriaxone in Israeli children with invasive diarrhea (73 of whom had Shigella) and found 97% microbiologic success and 100% clinical success on day 5. We therefore assumed a clinical and microbiologic success rate of 97%. We chose an absolute 10% non-inferiority margin as the maximum risk difference in clinical success between tebipenem-pivoxil or ceftriaxone that would be clinically acceptable, by consultation with infectious disease and paediatric clinical specialists at the icddr,b.

Assuming an equal cure rate of 97% for both arms, an absolute 10% non-inferiority margin and a 2.5% one-sided alpha level, the study would require 46 confirmed Shigella patients per treatment arm to have 80% power. To achieve 92 children with Shigella-confirmed infection we will need to recruit 124 children in the RCT (approximately 75% of whom will have Shigella infection confirmed by PCR). We anticipate a low (5%) dropout rate because of the short follow-up time period for the primary outcome and the close monitoring of these children, requiring we enroll 132 children in the trial (66 in each arm).

Primary Aim 1: To determine whether tebipenem-pivoxil is clinically non-inferior to the currently WHO-recommended second line Shigella therapy (ceftriaxone) 3 days after treatment initiation. Clinical failure at Day 3 will be defined as presence of fever (axillary temperature ≥38°C), diarrhoea (3 or more abnormally loose or watery stools in the last 24 hours), blood in stool, or abdominal pain/tenderness (defined by localization of pain by a child in response to query of parent/caregiver or an examination during palpation there is any facial expression during compression of any part of abdomen) at Day 3 of follow-up or a death or hospitalization prior to Day 3.

Secondary Aim 1: To determine whether tebipenem-pivoxil is clinically non-inferior to the currently WHO-recommended second line Shigella therapy (ceftriaxone) 7 and 30 days after treatment initiation. Clinical Failure at Day 7 and Day 30 will be defined as presence of fever (axillary temperature ≥38°C), diarrhoea (3 or more abnormally loose or watery stools in the last 24 hours), blood in stool, or abdominal pain/tenderness (defined by localization of pain by a child in response to query of parent/caregiver or an examination during palpation there is any facial expression during compression of any part of abdmen) at Day 7 or 30, respectively of follow-up or a death or hospitalization prior to Day 30. If a child changed treatment at day 3 due to clinical failure, then the day 3 outcome (failure) value will be carried forward to the Day 7 and Day 30 timepoint in primary analyses.

Secondary Aim 2: To determine whether tebipenem-pivoxil is microbiologically non-inferior to the currently WHO-recommended second line Shigella therapy (ceftriaxone) 7 and 30 days after treatment initiation.Microbiological failure will be defined as the presence of Shigella DNA at Ct values of 30 or less enrolment OR Shigella isolated by microbiologic culture, at the follow-up visit. Microbiologic failure will be assessed at Day 7 and Day 30. If a child changed treatment at day 3 due to clinical failure, then the day 3 microbiological outcome will be carried forward to the Day 7 and Day 30 timepoints.

Secondary Aim 3: Describe the number of adverse events, between children with shigellosis treated with oral tebipenem-pivoxil or IV ceftriaxone. Adverse events will be ascertained by caregiver report or identified by the study clinicians during clinical exams in hospital, at scheduled follow-up visits or during unscheduled visits. Severity (grades 1-5) will be defined according to 2014 Division of AIDS Table for Grading the Severity of Adult and Paediatric Adverse Events by the clinical team.

Secondary Aim 4: Compare the prevalence of ceftriaxone and tebipenem-pivoxil resistance, as well as ESBL-and carbapenemase-producing Escherichia coli, in children treated with tebipenem-pivoxil or ceftriaxone 7- and 30-days after initiation of second-line therapy. Resistance cut-offs will be determined using Clinical and Laboratory Standards Institute (CLSI) 2021 Standards. Accordingly, ceftriaxone resistance will be defined as MIC ≥4 as per CLSI guidelines. ESBL will be detected by Vitek 2 system which is an automated bacterial identification and susceptibility testing system. Carbapenemase resistance will be defined as resistance to any of the following: meropenem (MIC ≥4), imipenem (MIC ≥4) or ertapenem (MIC ≥2). MIC cut-offs have not been established for tebipenem-pivoxil in E. coli and therefore will be analysed as a continuous variable.

Pharmacokinetic study:

Dose Administration

The study drug in this study will be tebipenem pivoxil administered orally as the pediatric formulation, Orapenem® (10% fine granules). Proposed dosing will be based on the following weight bands:

3-5.9 kg: 18 mg TID x 3 days (54 mg total daily dose) 6-9.9 kg: 32 mg TID x 3 days (96 mg total daily dose) 10-13.9 kg: 48 mg TID x 3 days (144 mg total daily dose) 14-19.9 kg: 68 mg TID x 3 days (204 mg total daily dose) 20-24.9 kg: 90 mg TID x 3 days (270 mg total daily dose) 25-34.9 kg: 120 mg TID x 3 days (360 mg total daily dose)

Tebipenem pivoxil will be administered in 3 separate doses throughout the day for 3 consecutive days. If clinical research staff members discover that a dose was not administered on time, it should be administered immediately, and the actual time noted. The following/subsequent dose should not be given any sooner than 4 hours after the "late" dose. If the "missed/late" dose is within an hour of the next dose, then the missed dosed will be skipped and the next dose will be given as scheduled (i.e., the subject will have 2 doses rather than 3 in that 24-hour period). If a child vomits within 30 minutes of drug administration, the dose will be repeated.

PK Sampling Schedule Plasma PK Sampling Schedule Below is a proposed sample schedule for PK study that will be done on the first 15 patients. However, the final schedule will be modified to confirm to all blood collection limits by weight, age inclusive of all blood draws within the last 30 days. We plan to have sparse sampling for each participant in the trial. Blood samples will be collected for determination of tebipenem concentrations and PK parameters are shown below. Time of study drug administration, time of nutritional supplement administration, PK blood sampling time, and meal timing will be collected.

Day 1 of tebipenem dosing:

Pre-Dose 1: 0 hours Post-Dose 1: 0.5 hours (±6 minutes) 1 hours (±6 minutes) 4 hours (±24 minutes) Post-dose 2: 1-2 hours post-dose 2 Post-dose 3: 1-2 hours post-dose 3

Day 2 of tebipenem dosing:

Pre-Dose 1: 0 hours Post-Dose 2: 1-2 hours post-dose 2 Post-Dose 3: 1-2 hours post-dose 3

Day 3 of tebipenem dosing:

Pre-Dose 1: 0 hours Post-Dose 1: 0.5 hours (±6 minutes)

1 hours (±6 minutes) 4 hours (±24 minutes) Post-dose 2: 1-2 hours post-dose 2 Post-dose 3: 1-2 hours post-dose 3

Analytical Methods Measurement of tebipenem plasma concentrations will be performed by Q2 Solutions using liquid chromatography-tandem mass spectrometry (LC/MS/MS). The assay methods will be validated for quantification of tebipenem in human plasma. Additional information regarding assay validation will be provided in the Validation Report prepared by Q2 Solutions.

Pharmacokinetic Analysis Methods Analysis Population and Handling of Missing Time Points All subjects receiving study product and having at least 1 measurable drug concentration will be included in the PK population. If there are subjects who do not have sufficient samples for a complete analysis, a modified analysis will be conducted in which some or all PK parameters may not be estimated for those subjects, but available PK concentrations will still be tabulated and graphed. PK sample concentrations identified as laboratory errors will be indicated as such in the report and may be excluded from analyses. Values below the lower limit of quantification (LLOQ) will be referred to as below quantification limit (BQL). BQL values that precede the first PK concentration above the LLOQ will be imputed as 0 for plotting and for all calculations including NCA and summary statistics but will be shown as BQL in listings of concentrations. All other BQL values will be imputed as LLOQ/2 for linear plots of individual concentration-time profiles but will not be included in other graphical displays or analyses. Missing PK sample concentrations will not be imputed for the NCA. A geometric mean of concentrations will be treated as missing for sets of data points containing a BQL value.

Demographic and Baseline Characteristics Sex, race, age, weight, height, and body mass index (BMI) of subjects in the PK population will be listed and summarized by analysis group.

Dosing and Pharmacokinetic Sampling Summary Subject dose administration times will be presented. Special cases that potentially affect the analysis will be discussed.

Protocol deviations related to dosing or PK sampling will be listed and summarized in the PK Report text.

Missed treatment administration Delayed treatment administration Blood not collected Specimen result not obtained Specimen temperature excursion Required specimen collected out of window Any other deviation determined to be related to PK by the PK analyst Plasma drug concentrations will be listed by subject, with BQL, out of sample time window, and PK analyses-excluded samples indicated. The listings will also indicate the nominal and actual time associated with the sample (nominal time is defined as the time in hours since the first dose of the day). Plasma drug concentrations will be summarized by group (age, weight) and timepoint and plotted.

Definition and Estimation of Individual NCA PK Parameters

PK parameters will be initially estimated through a NCA using Phoenix WinNonlin version 8.0 or later (Pharsight Corporation, Cary, NC). Actual post-dose time will be used for the estimation of PK parameters instead of nominal time.

Phoenix WinNonlin NCA will use the following settings to compute parameters from plasma PK data:

Linear Trapezoidal (Linear interpolation) calculation method Uniform weighting Extravascular dose

Cmax Maximum concentration (Cmax) for a given day is defined as the maximum observed drug concentration observed over all PK sample concentrations for that day.

Cmin Minimum concentration (Cmin) for a given day is defined as the minimum observed drug concentration observed over all PK sample concentrations for that day.

Tmax Time of maximum concentration (Tmax) is defined as the time at which the maximum concentration (Cmax) occurs.

AUC AUC is defined as the area under the concentration-time curve. AUC will be estimated using the Linear Trapezoidal calculation method.

AUC0-24 AUC0-24 is defined as AUC from time of the first dose of the day to 24 hours following the first dose of the day (Pre-dose 1 on Day 2). AUC0-24 will be calculated for Day 1.

Descriptive Statistics Summaries of Day 1, Day 2, and Day 3 Plasma PK parameters will be described. All PK parameters will be summarized using descriptive statistics: N, Mean, standard deviation (SD), Min, Max, Median, coefficient of variation as a percent (CV %), and geometric mean (GM).

Population Pharmacokinetic Analysis PopPK analysis will be used to identify additional variables associated with tebipenem exposure. Furthermore, pharmacokinetic/pharmacodynamic (PK/PD) analysis will be performed to determine whether tebipenem plasma exposure is associated with efficacy (microbiological or clinical) and safety (adverse events) outcomes.

Previous tebipenem PopPK models in the literature There is no population pharmacokinetic analysis regarding tebipenem in children with shigellosis. The only published population PK analysis used a one-compartment model to fit the concentration-time data 39. This previous study performed a PopPK analysis using plasma tebipenem concentrations obtained from pediatric patients with otolaryngological infection or bacterial pneumonia (0.5-16 years old; n=217, 395 points) after repeated oral administration of tebipenem pivoxil at a dose of 4 or 6 mg/kg BID. This model will provide a foundation for our PopPK modeling efforts. We expect any impacts on tebipenem concentrations in the shigellosis study population will be evident through changes in the absorption rate constant (ka) and/or bioavailability (F).

Modeling software and settings All population PK modeling will be performed using Phoenix v8.1 NLME or later or NONMEM version 7.2 or later. The First Order Conditional Estimation-Extended Least Square (FOCE-ELS) method will be used for estimation. For models suffering from poor initial estimates, Maximum A Posteriori (MAP) - Naïve Pooling (NP) option may be used to assist the estimation process. Diagnostics (e.g., predicted corrected visual checks) and a bootstrap evaluation will be used to support adequate model fit.

Model development All models will include a mix of multiplicative and additive residual variability. Between subject variability (BSV) for all PK parameters will be modeled in the form: θ_i=tvθ*exp⁡(η_i ),η_i~N(0,ω^2). Based on previously published PopPK models, a 1-compartment model will initially be used. 39. BSV will be included for each PK parameter but may be considered for removal from the model if removal lowers the AIC and does not adversely affect model diagnostics. Inter-occasion variability will be tried on each of the PK parameters. Exponential models will be used to describe the inter-individual variability for the structural PK model parameters. For residual variability, we will consider proportional and combined additive and proportional models. Covariate model development will incorporate demographic data (e.g., age, sex, body weight). The covariate model will be performed using stepwise covariate modeling with a forward inclusion phase followed by a backward elimination phase.

Our final model will contain estimates for the model parameters and associated inter-individual variability. Furthermore, the model will include estimates of covariate effects and residual variability. We will evaluate the final model with both prediction-corrected visual predictive checks and bootstrap evaluation.

PK model application to determine probability of target attainment (PTA) for full trial The final PK model will be used to determine the probability 90% of the participants in the full tebipenem pivoxil trial achieve our target PK/PD parameter. It has been previously reported that carbapenems have concentration independent bactericidal activity and unbound therapeutic serum levels should be above the minimum inhibitory concentration (MIC) for at least 40% of the dosing interval 40,41. Therefore, we will evaluate the probability of target attainment (PTA) for the full efficacy study (n=64 children) with 40%fT>MIC as the PK/PD target. The final PK model will be used to simulate individual tebipenem concentration versus time profiles for the full efficacy study. The simulated concentrations will be used to determine the proportion of pediatric shigellosis patients that achieve free tebipenem serum levels ≥ 0.031 mcg/mL for more than 40% of the 3-day dosing period (28.8 of the 72 hours post treatment initiation). The concentration 0.031 mcg/mL reflects the tebipenem MIC observed with Shigella clinical isolates collected from children in Kenya. If we observe a difference in the tebipenem MIC value for isolates collected from children in Bangladesh, we will revise the target tebipenem concentration for our PK/PD target to reflect this difference.

The Statistical Analysis Plan has been uploaded as a separate document.

Inclusion Criteria:

• Children aged 24-59 months with suspected Shigella infection (clinical features of fever, mucus and/or blood in stools, tenesmus, and RBC and leucocytes >10 per hpf)

Exclusion Criteria:

• Child received study antibiotics (azithromycin, ceftriaxone, and/or tebipenem) for the illness prior to presentation (as confirmed by bottle or prescription)
• Severe acute malnutrition (SAM), defined as weight-for-height z-score less than -3 or mid-upper arm circumference less than 115mm, and/or other signs of infections requiring antibiotics
• Patients with other infectious foci who are potentially unresponsive to treatment with orally administered medication
• Patients in whom the efficacy and safety of the study drug is difficult to determine because of a progressive, complicated, or severe underlying disease believed to critically influence the onset of the infection, its clinical course, and therapeutic efficacy
• Patients with convulsive disorders, such as epilepsy, as an underlying disease
• Patients with a known lipid metabolism disorder or congenital carnitine deficiency
• Patients with severe hepatic or renal dysfunction
• Patients with a history of allergy to β-lactam antibiotics (e.g., carbapenems, penicillin, and cephems)
• Patients who have received other antibiotics for the illness and exhibited improvements
• Patients deemed inappropriate for this study by the attending physician
• Clinically improved after first-line therapy
• Unable to provide a stool sample at enrolment