Prophylactic Antibiotic Duration in Head and Neck Cancer Surgery (PROTECT-HN Trial)

Background and Rationale

Surgical site infection (SSI) is a common and serious complication of major head and neck cancer surgery. Operations such as mandibulectomy, pharyngolaryngectomy, and major glossectomy, and procedures that need microvascular free-flap or pedicled flap reconstruction, expose the neck to bacteria from the oral cavity and upper aerodigestive tract for many hours. Several other factors add to the risk: tumor-associated oral dysbiosis, long operative times, flap ischemia, dead space, previous radiotherapy, malnutrition, tobacco and alcohol use, and tracheostomy. As a result, reported SSI rates in these operations range from 20% to 50%. That is much higher than expected for clean-contaminated procedures and closer to the rates seen in infected procedures. SSI is linked to fistula, flap failure, reoperation, longer hospital stays, delays in adjuvant treatment, and higher costs.

Centers for Disease Control and Prevention (CDC) and World Health Organization (WHO) guidelines advise against continuing antibiotic prophylaxis after surgery ends. However, these guidelines do not address the specific risk factors of head and neck reconstruction, and these patients are underrepresented in the trials behind them. Practice varies widely: in one survey of microvascular surgeons, 5 days of prophylaxis was the most common duration, and some authors recommend 7 days for free-flap cases. Retrospective studies disagree about whether longer courses help, and no randomized evidence exists. Data from low- and middle-income countries, where SSI rates are higher, are especially limited.

At the study institution, SSI indicators have been collected prospectively since 2014. SSI rates in major and free-flap head and neck procedures stayed between 33% and 50.9%. This was true even after the prophylactic regimen was changed from clindamicin plus ceftriaxone to ampicillin/sulbactam in 2022, and after a preoperative decolonization protocol was introduced. In a previous prospective cohort of 401 patients, the overall SSI rate was 41.9%. This trial tests whether extending postoperative prophylaxis lowers SSI and length of stay in this high-risk group.

Study Design

Single-center, pragmatic, open-label, randomized, parallel-group superiority trial. The trial is pragmatic because both regimens are already used in clinical practice. Blinding is not feasible because the arms differ in how long the postoperative antibiotic is given.

Randomization

Patients are allocated 1:1 using a maximum tolerated imbalance (MTI) procedure with an imbalance limit of 3. When the limit is reached, the next patient is assigned to the smaller arm. This lowers the predictability of allocation while keeping the arms balanced. Allocation is generated dynamically in REDCap after enrollment, and patients do not cross over between arms.

Standardized Perioperative Care (Both Arms)

Preoperative decolonization for 5 days: 0.12% chlorhexidine oral rinse three times a day, 2% chlorhexidine body wash once a day, and nasal mupirocin three times a day.

Preoperative consultations with nursing, nutrition (with supplementation when needed), and oral medicine.

In the operating room: 2-minute skin antisepsis of the resection and donor sites with 2% chlorhexidine or 10% povidone-iodine, plus oral and pharyngeal cleansing with 0.12% chlorhexidine.

The first antibiotic dose is given at anesthesia induction, 10 to 60 minutes before skin incision, and redosed every 4 hours during surgery.

Data Collection

Data are collected from the electronic medical record with a standardized case report form and entered in REDCap. Variables include:

Demographics and lifestyle: age, sex, smoking (pack-years), alcohol use. Clinical data: comorbidities, American Society of Anesthesiologists Physical Status Classification System, Body Mass Index.

Tumor data: site, histology, UICC-TNM stage, prior radiotherapy or chemotherapy.

Surgical data: main procedure, type of reconstruction, flap ischemia time (from pedicle ligation to completion of the anastomoses), operative time (from skin incision to dressing), salvage surgery, and perioperative blood transfusion.

Cost Analysis

Costs are taken from the institution's patient-level billing database (Base de Atendimento) for the index surgical admission. Direct costs are materials, drugs, and professional fees. Indirect costs are allocated by absorption costing and include idle capacity. The analysis follows the Brazilian Ministry of Health guidelines for economic evaluation, takes the payer perspective, and uses micro-costing. Unit prices come from public procurement references: BEC/SP and the Ministry of Health Health Price Database.

Statistical Analysis

Analyses follow the intention-to-treat principle and are done in R and Python. Categorical variables are compared with the chi-square test and continuous variables with Student's t-test. Multivariable analysis uses Cox proportional hazards models, with the proportional hazards assumption checked by Schoenfeld residuals.

Sample Size The sample size was calculated for the primary outcome (30-day SSI) as a comparison of two independent proportions. The SSI rate in the control group was assumed to be 43%, based on the institutional prospective cohort of patients undergoing major head and neck surgery. An SSI rate of 12% was expected in the intervention group, which is an absolute reduction of 31 percentage points.

With a two-sided significance level of 5% and 80% power, at least 32 evaluable patients per group are needed. Allowing for an estimated 10% loss to follow-up, the sample was set at 40 patients per group, for a total of 80 participants. The calculation was done in G*Power, version 3.1.9.7.

Interim Safety Analysis

A safety analysis will be done once 50% of the planned patients have completed 30 days of follow-up. Adverse events are graded using CTCAE. The two arms will be compared for serious treatment-related complications, including infections other than SSI (such as Clostridioides difficile), drug toxicity, and allergic reactions. If the difference in serious adverse event rates between the arms is greater than 20%, early termination or protocol changes will be considered together with the institutional Infection Prevention and Control Service.