Hypothesis and objectives of the trial

The hypothesis of the study is that in IOM of long bone fracture, a short antibiotic treatment is equally effective as a long antibiotic treatment in patients selected for debridement and implant retention. Antibiotic duration can be selected according to the time of diagnosis of IOM as it takes into account the biofilm formation and the state of fracture consolidation:

1. In patients with early infections (those that occur in the first 2 weeks after implantation of the osteosynthesis material) in whom early debridement is performed, it is possible to shorten the duration of antibiotic treatment to 8 weeks (vs. 12 weeks);
2. In patients with delayed infections (those that occur between 3 and 10 weeks after implantation of the osteosynthesis material) in whom it is not possible to remove the implant due to instability/lack of fracture healing, it is possible, after surgical debridement, to shorten the duration of antibiotic treatment to 12 weeks (vs. maintaining antibiotic treatment until fracture healing or implant removal).
3. In patients with late infections (those that occur more than 10 weeks after implantation of the osteosynthesis material) healing is not possible until the implant is removed, but when removal is not possible, antibiotic therapy can be maintained until fracture healing or implant removal.

The main objective of this trial is to evaluate whether, after performing surgical debridement in patients with IOM due to a large bone fracture, a short antibiotic treatment is as effective as a longer treatment. Secondary objectives are: a) to evaluate the efficacy and safety of different antimicrobials used in IOM, b) to evaluate the development of antimicrobial resistance during antibiotic treatment, c) to evaluate the need for new surgeries during follow-up, d) to provide information about the functional prognosis and quality of life of the patient according to each of the treatment strategies, e) to evaluate the consumption of health resources with each type of strategy, and f) to evaluate the different reconstruction strategies (bone and soft tissue) carried out in order to recover lost functionality (degree of mobility and autonomy).

Trial design, sites and study period

DURATIOM is designed as a prospective, multicenter, non-inferiority, open labelled, pragmatic randomized clinical trial comparing different durations of antibiotic therapy in IOM after long bone fractures treated with debridement and material retention. A 36-month recruitment period is planned. The study is coordinated from Hospital Universitario Virgen Macarena and sponsored by Fundación Pública Andaluza para la gestión de la Investigación en Sevilla (FISEVI). Twenty-six public Spanish hospitals where trauma surgery is performed will participate (Table 1). Each hospital has a research team that must include at least one trauma surgeon and a specialist in infectious diseases or internal medicine. Currently, the management of osteoarticular infections is very similar in all the participating centers, since the researchers work together in study groups and consensus documents.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Spanish hospitals participating in DURATIOM.

https://doi.org/10.1371/journal.pone.0286094.t001

The trial obtained competitive, public funding from the Spanish Ministry of Health (Instituto de Salud Carlos III, reference ICI21/00014, www.isciii.es). Funding for the trial was communicated on November 2021, available for study expenses in January 2022. The funder will not have a role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The study is supported by the Spanish Clinical Research Network (SCReN) from the Spanish Ministry of Health (Instituto de Salud Carlos III, PT20/ 00123) with the internal study code 21.009. The study is registered at ClinicalTrials.gov (identifier, NCT05294796) on Jan 26^th 2022 and at the European Union Drug Regulating Authorities Clinical Trials (EUDRACT) (identifier, 2021-003914-38) on Jul 16^th 2021 (S1 File).

The study protocol adheres to the SPIRIT 2013 guidelines for clinical trials and the Statement Consolidated Standards of Reporting Trials (CONSORT) for pragmatic designs (S2 File). Trial authorization by the Spanish Agency for Medicines (AEMPS) was obtained on Jul 27th 2022 (version 2.0 dated Jun 8th 2022) and any modifications to the protocol will be reported to the clinical research ethics committee, AEMPS, and Health Products and ClinicalTrials.gov, and will be stated in the final manuscript. We plan to grant public access to the full protocol, participant level data, and the statistical code if required.

Study population

The study target population is formed by patients older than 14 years, with IOM implanted after a long bone fracture diagnosed during the study inclusion period, and who are treated by DAIR.

IOM is considered when at least one microbiological and one clinical criterion are met:

1. Microbiological criterion: isolation of the same microorganism (same species and antibiogram) in ≥2 tissue samples taken intraoperatively (≥1 in the presence of virulent microorganisms such as Staphylococcus aureus) or from the combination of cultures taken preoperatively by deep aspirate plus intraoperative cultures.
2. Clinical criteria:
   • Presence of clinical syndrome comprising any of the following: localized pain, localized redness or edema, temperature >38.0°C, or discharging wound.
   • Presence of a sinus tract communicating with the osteosynthesis material.
   • Presence of pus around the implant.
   • Presence of histopathological signs of acute inflammation in the peri-implant tissues taken at the time of debridement.

Patient selection and enrolment

Candidate patients will be detected by daily review of admissions to the Traumatology Units both from the Emergency Departments and from outpatients, as well as in the daily ward rounds of the Traumatology wards and outpatient clinics. Screening procedures to select the accurate candidate for the study will involve the Traumatology team, the nursing team, and the Infectious Diseases/Internal Medicine team. Patients will be considered for inclusion after review by the principal investigators of the team.

Informed consent will be obtained from each participant by good clinical practice (GCP) trained research staff after assessing the patient information sheet. Participation in the study is voluntary and patients can give up their willingness to participate in the study at any time with no changes in the clinical attendance after the decision. Otherwise, the sponsor decision of subject withdrawal from the study are the following; if a major protocol violation occurred, in case of clinical failure, or according to safety criteria.

Randomization

Patients will be randomized once inclusion and exclusion criteria are checked, and informed consent is signed; randomization must be performed no later than 72 h after knowing the microorganism causing the IOM and the susceptibility results, preferably from samples obtained in DAIR.

Patients signing the informed consent form will be centrally randomized using an automated online system with a 1:1 randomization. Randomization will be stratified according to whether they are early IOM (those that occur in the first 2 weeks from the implant surgery) or delayed IOM (those that occur between 3 and 10 weeks from the implant surgery), and using blocks of 20 patients. The randomization list will be obtained using Epidat 4.0 software. The automatic randomization system is integrated in the electronic case report form (e-CRF) of the study. One written copy of the randomization list will be stored at the Clinical Trials Unit of the Hospital Universitario Virgen del Rocío (UEC-HUVR) in the event that technical issues avoid the automatic system working, and the assignation of study intervention will be obtained by phone call.

Interventions and study treatment

Patients will be classified according to the age of the osteosynthesis material implantation, i.e., according to the time from implant surgery to the diagnosis of IOM, as it takes into account the biofilm formation and the state of fracture consolidation [2]:

1. Early IOM: those that occur and are diagnosed in the first 2 weeks from the implant surgery.
2. Delayed IOM: those that occur and are diagnosed between 3–10 weeks from the implant surgery.

Patients with early or delayed infections who meet the inclusion criteria and were managed with DAIR will be allocated to one of the following arms (S3 File):

• Experimental arm: short-term antibiotic treatment; 8 weeks in early IOM or 12 weeks in delayed IOM.
• Control arm: long-term antibiotic treatment; 12 weeks in early IOM or antibiotic until fracture healing or implant removal in delayed IOM.

The limit to consider an IOM as early or delayed is given by the moment in which the DAIR is performed. In other words, an IOM is considered early when 14 days or less have elapsed between the date of the implant surgery and the DAIR, and delayed when 3 to 10 weeks have elapsed between the date of the implant surgery and the DAIR.

Antibiotics administered after DAIR will be carried out empirically according to the protocols of each hospital. The duration of intravenous treatment will be at the discretion of the investigator, but it is advised that in the absence of sepsis or post-surgical complications it should not exceed 7 days. Antibiotics administered before DAIR will be permitted in both groups but will not be governed by the trial protocol. Once intravenous treatment is considered to be discontinued, the antibiotic treatment administered orally will be agreed upon by the Traumatology-Infectious Diseases/Internal Medicine-Microbiology team. Given the absence of controlled studies carried out specifically on IOM, it will be performed according to the Spanish prosthetic joint infection guidelines [15]. Adjunctive oral agents such as rifampin will be permitted at the infection specialist’s discretion, reflecting usual practice. In the switch to the oral route, antibiotics with good bioavailability should be used [15], and if no options are available outpatient parenteral antimicrobial therapy may be used. In case of intercurrent infection unrelated to IOM, up to 7 days of concomitant antibiotic treatment is allowed.

Rescue medication is not foreseen. Since the antibiotics to be used are those of standard practice, if the antibiotic initially administered is not effective or an adverse effect occurs, it can be changed according to clinical guidelines or standard practice, without this being considered a failure or violation of the protocol. The alternative may be intravenous treatment on an outpatient or inpatient regimen. If there is no alternative and the antibiotic treatment must be withdrawn, the patient will be dropped from the study, but will continue to be followed according to the protocol and will be included in the intention-to-treat analysis. Changes in treatment, the reason, and adverse reactions will be recorded in the CRF.

Adherence to oral antibiotic treatment will be measured by registering the medication given with expiration day and official batch number in each delivery of drugs.

Minimizing bias

In order to inform about the generalizability of the results, all patients diagnosed of IOM in the participating sites will be evaluated and the reasons for not being included will be collected as “screening failures”. The study is conceived as open-labelled due to the logistic and budget limitation for masking a great group of drugs needed for this study. Blinding will not be used; to avoid bias in the assessment of endpoints, a blinded investigator from the coordinating hospital (HUVM) will also evaluate the outcomes based on data collected.

To control selection bias, it has been developed: a) a randomization procedure that will be carried out when the patient is evaluated by means of an automatic system, so investigator is blinded to the assignment of treatment arm until the procedure is completed; b) stratification according to infection type (early/delayed); c) central randomization distributed in 1:1 ratio using blocks into participating sites.

The potential effect of residual confounders will be controlled by performing a multivariate logistic regression analysis; because the type of surgical treatment and the promptness with which it is performed may vary across sites, the sites will be classified into those with high and low cures rates as a potential confounder. Antibiotic therapy may also vary, such as the tendency to prolonged intravenous treatment. In order to standardize practice and variables collection, investigating teams will be trained at each center.

Outcome variables

Endpoints will be identified by prospective follow-up of patients for at least 12 months after antibiotic therapy withdrawal.

• Primary outcome

The primary outcome will be the composited variable "cure” including:

1. clinical cure in the test of cure (TOC) (see definition below);
2. radiological healing;
3. definitive soft tissue coverage at TOC.

The TOC will be performed 12 months after the end of antibiotic therapy prescribed to treat IOM following DAIR.

Clinical cure is defined by the presence of any of the following:

1. The absence of clinical signs and symptoms of infection (persistence of symptoms of infection, relapse of infection after a period without symptoms, or superinfection by a different microorganism) without antibiotic therapy and with CRP <10 mg/L (unless another cause justifies a higher CRP value);
2. No need for chronic suppressive antibiotic therapy to "control" of the infection.

Persistence of infection is defined as no disappearance of clinical signs and symptoms of IOM, produced by the same micro-organism that produced the initial infection.

Relapse of infection is defined as the reappearance of signs or symptoms of IOM after a period without symptoms, by the same micro-organism that produced the original infection.

Superinfection is defined as a re-infection produced with a different micro-organism than the one that caused the initial infection.

Radiological healing is defined as the presence of radiological signs of fracture consolidation (plain radiographic or computerized tomography) of the infected bone. Non-union or absence of consolidation of a fractured bone is defined when it has not completely healed within 9 months after osteosynthesis surgery or when it has not shown progression towards fracture callus formation in 3 consecutive months on serial radiographs [16]. Given the subjectivity that can sometimes occur in the assessment of bone healing criteria, a blinded investigator will evaluate the concordance of clinical and radiological healing criteria performed by the local investigator. The REBORNE bone healing score will be used to perform a detailed evaluation of long bone non-union consolidation in radiographs and CTs if it is necessary [17].

Clinical failure occurs when any of the following is present:

1. Absence of clinical and microbiological cure of the IOM.
2. Absence of radiological healing.
3. Exposure of bone and/or osteosynthesis material.
4. Death related to the infection.

If the patient dies from a cause unrelated to the infection, it will be considered lost to follow-up or censored.

The main outcome variable will also be evaluated by subgroups:

1. Type of infection: early and delayed.
2. Type of fracture: open (Gustilo and Anderson classification) or closed.
3. Type of osteosynthesis material: endomedullary nail or needle, plates or screws.
4. According to the etiology of the infection: infections by different microorganisms, infections by multidrug-resistant microorganisms.
5. According to type of patient: comorbidities, elderly, renal insufficiency, gender.

• Secondary outcomes

1. a) Clinical endpoints: efficacy of each group of antibiotics, secondary infections, infection recurrence rate (relapses and reinfections), need for new surgeries (debridement, material removal, covering, amputation), reconstructive surgery strategies (bone and soft tissue), functional status (defined as restoration of limb function prior to fracture) and state of health.

The assessment of the patient’s functional status is complex given the number of fractures and bones involved. To facilitate it, the following classification of functional status (normal o reduced with respect to the situation prior to fracture) has been chosen to be performed at TOC:

• For lower limbs mobility functional status will be classified as walks unaided, with 1 cane, with 2 canes, with walker, "home alone ambulation" and no ambulation.
• For upper limbs mobility functional status will be used the Quick-DASH questionnaire [18].

To assess generic health outcomes from the patient’s perspective we will used the SF-12 questionnaire or Barthel in people over 70 years of age [19, 20].

1. b) Microbiological endpoints: Resistance development to antibiotic used during therapy and C. difficile infection during therapy and at 30-day.
2. c) Adverse events (AE) and complications: Adverse events and severity, including death (that is, all-cause mortality).
3. d) Consumption of health resources: Consumption of health resources will be evaluated for each type of strategy: days of antibiotic treatment, length of hospital stay, readmissions, number and type of surgeries performed.

Visits and follow-up of participants

• All visits and procedures to be performed at each one are specified in Fig 1.
• All patients will be assessed:
• The day of recruitment/randomization (day 0, visit 1).
• The day of switching from intravenous to oral treatment (day 7±3, visit 2) or at discharge if the patient goes home with oral antibiotic.
• At 1 month (day 28±7, visit 3), for safety of treatment.
• At 8 weeks (±7 days) (visit 4) for withdrawal of antibiotic treatment in the experimental arm of early infections.
• At 12 weeks (±7 days) (visit 5) for withdrawal of antibiotic treatment in the experimental arm of delayed infections and in the control arm of early infections.
• At 6 months (±14 days) (visit 6), for evaluation of patients who completed treatment and for assessed the control arm of delayed infection who are still on antibiotic treatment.
• At 12 months (±1 month) (visit 7), for evaluation only of patients in control arm with delayed infections.
• At 1 year after completion of antibiotic treatment (±1 month) (TOC, visit 8), all patients.
• If any event occurs between visits (relapses, new hospitalization or surgery, adverse effects, antibiotic changes), additional visits will be performed and collected.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Visits and follow-up of participants of DURATIOM trial.

The total follow-up time will be one year after completing the antibiotic treatment prescribed after randomization. Patients with delayed infections randomized to the long-treatment regimen (until fracture union) who had undergone 6 months or more of antibiotic treatment will be evaluated 18 months after the start of treatment (or randomization).

https://doi.org/10.1371/journal.pone.0286094.g001

Microbiological studies

Preoperative wound aspirate and intraoperative tissues specimen will be sent to microbiology laboratory. For intraoperative will be required at least three biopsies from bone- or peri-implant tissue. Tissue specimens will be cultured for 7–10 days at 35°C on aerobic and anaerobic blood agar as well as in thioglycolate broth. Samples may be incubated for up to 14 days in agreement with the clinician and the microbiologist in case of suspected infection by fastidious micro-organisms or in patients with previous anti-microbial treatment. In case of any orthopaedic devices is removal (e.g. a loose screw from a plate), it will be additionally sonicated and sonication fluid will be cultured [21]. Standard microbiological techniques will be used to identify microorganisms recovered and to determine antimicrobial susceptibility. In case of diarrhea during the antibiotic treatment and up to one month after its completion, stool culture and detection of C. difficile toxin in stool shall be performed.

Radiological and laboratory studies

Blood tests and radiological studies will be performed at the initial visit and subsequently as specified in the visit schedule (Fig 1). Laboratory tests will be performed at each center in accordance with the usual clinical practice standards. The following tests will be performed:

• Blood count including leukocyte, neutrophil, haemoglobin and platelet counts.
• Blood biochemistry including sodium, potassium, creatinine, C-reactive protein, AST (GOT), ALT (GPT) and total bilirubin (if altered also direct bilirubin).

Sample size

To our knowledge, no previous trials on IOM treatment had been performed; a meta-analysis of observational studies found a clinical cure of IOM when DAIR was performed (regardless antimicrobial therapy duration) of 86–100% in early IOM, and 82–89% in delayed IOM. Although our primary outcome is a composite outcome including clinical cure, radiological healing and soft tissue coverage which might reduce the cure rate, but the fact that in this randomized trial we will only include patients in whom a debridement is performed, we estimated that 85% of patients in both arms will reach that primary outcome. With a 10% non-inferiority margin, a 5% level of significance (α value) set at 5%, 80% power, and 15% loss rate, we would need to recruit 182 patients per arm (364 in total). Subgroups analyses for the primary outcome as defined above will be exploratory, as demonstration of non-inferiority in each of them is unfeasible.

Interim analysis will be conducted when 75 patients have been included to assess the frequency of the event rate and inclusion rate to adjust the intended size of the study population.

Statistical analysis

The populations for analysis will be:

• Intention-to-treat (ITT) population: all appropriately-included and randomized patients.
• Per protocol population (PP): all patients who received the complete antibiotic treatment regimen according to the inclusion group.
• Clinically evaluable (CE) population: all patients with evaluation of success in the cure test at 12 months after completion of treatment, or who have clinical failure before TOC.

The primary analysis will be the absolute comparison with one-sided 95% confidence interval of the proportion of patients reaching the composited primary outcome variable "cure" in the ITT population. Secondary analysis will include subgroups analyses for the primary outcome using the same procedure, and the comparison of the composite variable components (clinical cure, radiological healing and soft tissue coverage) and secondary variables in the ITT, PP and CE populations, after excluding incorrectly assigned patients. For continuous secondary outcomes, parametric and nonparametric tests will be used as appropriate.

Finally, a multivariate analysis will be performed in order to control the potential residual effect of variables other than duration of antibiotic therapy on the primary outcome by logistic regression. Potential confounders will be identified by bivariate analysis of their associations with the primary outcome, considering a p value <0.15, and will be entered into the multivariate model to check if the effect of the duration of therapy is significantly modified. Expected potential confounders include age, Charlson, delay in DAIR, type and location of fracture, type of osteosynthesis material, microorganism, and antibiotic group. The effect of study sites will be explored by classifying the sites as those with higher and lower cure rates, and including it as a potential confounder. A formal statistical planning has been prepared before the initiation of the study, with all the details, variables, and statistical test to be used during the study (S4 File).

Safety and adverse event reporting

Safety for all antibiotic used will be carefully followed according to Regulation (EU) No. 536/2014 and Royal Decree 1090/2015. For this objective a clear definition of adverse events, serious adverse events and serious and unexpected adverse events will be included in the final protocol for the study; the accomplishment of serious criteria will be followed by additional information to be completed and sent to the Pharmacovigilance department of UICEC-HUVR which will be acting in representation of the sponsor for the pharmacovigilance activities that includes the evaluation of casualty and the expedited notification to regulatory Authorities depending on the final classification of the event. Cases of relapse of infection that meets seriousness criteria and cases of readmission for relapse of infection has been approved for a waiver for expedited reporting. Yearly reports on safety will be issued in a data safety updated report to be send to Ethics committees and AEMPS.

The antibiotics to be used are standard practice in osteoarticular infections, but due to alerts from the Food and Drug Administration (FEA) and the AEMPS, especially with quinolones and even for the specific considerations of pharmacovigilance activities within the performance of a clinical trial, all the Summary of Characteristic Products will be used as a reference information, taken specially consideration for regulatory agencies safety alerts.

Data and safety monitoring

An external data safety monitoring board formed by 3 expert members not participating as investigators in the project will independently evaluate the data when the first 75 patients have been enrolled and monitored to ensure that there are no safety or efficacy reasons to stop the trial. The safety committee will make recommendations after the interim analyses. A charter with the formal activities to be performed will be included in the sponsor documents. A report elaborated with data on recruitment, adverse events notified to the pharmacovigilance center for the study, and data of primary and secondary outcomes will be provided to the committee before the meetings. The Data and Safety Monitoring Board can recommend that the current study continue without modification, continue with specified modifications, discontinue the study or halt or modify the study until more information is available.

Quality assessment

Clinical Trials Unit of the Hospital Universitario Virgen del Rocío (UEC-HUVR) and the study team will coordinate the activities in collaboration with the ISCIII Clinical Research Network (PT20) for authorization of the study, sites activation and training and for the performance of monitoring activities for the entire duration of the study. A monitoring planning will be approved for the follow-up of correct data collected according to approved protocol and based in the source data available for each patient. Local investigators will need to have accredited training in Good Clinical Practice or Good Clinical Laboratory Practice, as appropriate; the local teams will be adequately trained before the study initiation in a formal site initiation visit performed by the UICEC-HUVR-PT team, on which all the procedures for the study, timelines, eCRF functioning and Pharmacovigilance tasks will be documented. On site monitoring will be performed by trained monitors. A specific monitoring plan will be developed with the indication of the number of on-site visits to be performed during the study and with the percentage of source document verification to be performed in total, the time for completing the data on the eCRF and even for the monitoring report to be issued.

The coordination team of the study will establish a quality system of follow-up, with the continuous surveillance of events happening during the study to be discussed for resolution in their formal meetings. Standard operating procedures for activities as monitoring or pharmacovigilance will be used during the complete study.

To ensure the participation and follow-up of the centers during the development of the trial, a weekly reminder will be sent with the cases included and the screening failures, and a new monthly letter with a summary of the progress of the trial.

Ethical issues and dissemination

Trial approval by the Provincial Ethic Committee of Seville was obtained on Jul 7th 2022 (version 2.0 dated Jun 8th 2022). An approved informed consent (version 2.0, dated Jun 8^th 2022) form must be signed before any study specific procedures is performed. As patients with 14–18 years old can participate, a specific information sheet adapted for minors is included, approved with the same version and date. The study will be carried out under the ethical principles regarding human experimentation developed for the medical community by the World Medical Association (WMA), declaration of Helsinki and following all the applicable legislation regarding the realization of clinical trials in the European Union (EU No 536/2014) and Spanish Royal Decree 1090/2015. The study will not start until the approval of the AEMPS and the Ethic committees give their authorization. The results of the study will be submitted for publication to a scientific journal following the CONSORT recommendations.

Strengths and limitations of this study

The design of this randomized study has limitations, including the open design, the heterogeneity of patients, type and fracture location, the antibiotic treatment, the clinical practice between centers, and the subjectivity in the assessment of healing criteria and functional status. The sample size will not allow to demonstrate the non-inferiority of shorter treatment in subgroups of patients. Other risks include reaching the sample size and loss to follow-up due to the long evaluation time.

Some strengths are its pragmatic design which we hope will allow the appropriate representation of patients with the target infections, the multicenter participation, and previous experience in trials in which the antibiotic used is that of standard practice.