This pilot study was designed as a prospective, randomized trial. The Institutional Review Board of Ajou University Medical Center approved this study (No. MED-INT-19-306), and informed consent was obtained from all individual participants included in the study. This trial is registered with CRIS (https://cris.nih.go.kr; No. KCT0006352).

The recruitment period was from February to December 2020, followed by the first postoperative outpatient visit from April 2020 to December 2020. The enrolled patients were randomly assigned to 2 groups in a 1:1 ratio in this concurrent parallel study design as follows: (1) 2 mL of fibrin glue application with 1 closed suction drain or (2) 6 mL of fibrin glue application with 1 closed suction drain. A study coordinator (D.K.Y.) performed stratified randomization using a computerized program based on sex, BMI (<25 and ≥25 kg/m²), and thyroiditis (absent/present). Thyroiditis was defined as a parenchymal disease of the thyroid detected on ultrasonography performed by a radiologist (E.J.H). The treatment allocation was performed by the coordinator (D.K.Y.) after preoperative ultrasonography. Participants were allocated to 8 subgroups based on their characteristics, including sex, BMI, and thyroiditis, after combining the factors considered to have a significant impact on the results. Furthermore, stratified randomization was accomplished by performing simple randomization for participants assigned to subgroups. The surgeon was blinded to the allocation of enrolled patients until the time of fibrin glue application on the operative field in the operating room.

The inclusion criteria in this study were as follows: (1) patients who were scheduled to undergo total thyroidectomy with central compartment neck dissection (CCND), (2) patients older than 20 years, (3) malignancy on fine-needle aspiration cytology or core needle biopsies, (4) normal platelet count of ≥100,000/mL, and (5) availability of informed consent. In contrast, the exclusion criteria were as follows: (1) age of <20 or >65 years, (2) uncontrolled diabetes, (3) on anticoagulation medication, (4) previous thyroid surgery, (5) advanced hepatic or renal disease, (6) cardiovascular disease, (7) on steroid medication, and (8) declination of informed consent.

All surgeries were performed by a single surgeon (J.L., who had specialized in endocrine surgery for 10 years). A 5 to 6-cm transverse incision was made 2 finger-breadths above the sternal notch, and the subplatysmal skin flap was created from the thyroid cartilage to the sternal notch. After dividing the strap muscles in the midline, the thyroid gland was exposed from the strap muscle, and the thyroid was subsequently removed from the thyroid bed after ligating the thyroid vessels with a Harmonic scalpel. Bilateral central compartments, including fibrofatty tissue within an area limited by the thyroid cartilage cranially, sternal notch caudally, and the carotid artery medially, were removed. Bilateral CCND was performed in all enrolled patients.

After total thyroidectomy and CCND, the operative bed was irrigated with 100 mL of normal saline after hemostasis. Subsequently, a coordinator was asked to reveal the allocation concealment, and fibrin glue specially prepared for the clinical trial by an annotated pharmacist was applied to the thyroid bed and the inner side of the sternothyroid muscle, including the thyroid cartilage, trachea, esophagus, carotid artery, and jugular vein, by a surgeon (J.L.). As per the protocol, 2 mL and 6 mL of fibrin glue (Greenplast Q PFS kit, GC Pharma, Yongin, Korea) were applied to patients in the low and high fibrin groups, respectively. Fibrin glue is a liquid form of a mixture of fibrin and thrombin, which is dispensed via a dual syringe set with 2 chambers, a Y-tube, and an application needle. Anti-adhesive agent (Regencol 1 g, Dongsung, Seongnam, Korea) was applied to both the thyroid bed and the subplatysmal space after the application of fibrin glue, and one closed suction drain was placed in all patients. Patients were discharged routinely on postoperative day 4 ± 1 and were routinely followed up between postoperative days 10 and 14 at an outpatient clinic based on the routinely implemented clinical practice guidelines.

The primary outcomes of this study were the drain volume observed during the hospital stay and rate of flow of drain fluid for the first 12 and 24 hours postoperatively, the sample size, and potential harm of future definitive RCTs. The drain volume and check times were assessed by nursing staff who were blinded to the treatment allocation and study design. The drain volumes were measured every 8 hours until discharge. When the amount of drain reduced to <45 mL over a 24-hour period, the drains were removed. In contrast, secondary outcomes were symptomatic seroma formation and the duration of hospital stay. Although patients were routinely discharged within 4 days based on the clinical practice pathway, early or delayed discharge was done according to the clinical features, including drain volumes and postoperative complications such as hypoparathyroidism, pain, hematoma, and infection.

Wound infections were determined to include erythema and/or purulent discharge or aspirate from the incision site and were treated with an oral antibiotic. Symptomatic seromas were assessed on follow-up visits at an outpatient clinic based on the physical examination, patients’ symptoms, and aspiration if needed.

This pilot study initially planned to enroll 40 patients, with a sample size of 20 patients in each group. However, during stratified randomization, 41 patients were included; 21 in the low fibrin group and 20 in the high fibrin group.

Stratified randomization was performed using Microsoft Excel 2016 (Microsoft, Redmond, WA, USA). Continuous and categorical variables were reported as mean ± standard deviation (SD) and proportion, respectively. Two-tailed Student t-test, Mann-Whitney U-test, or chi-square test were used to evaluate the difference in measures between the 2 study groups. Correlation analysis with Pearson correlation and 2-way analysis of variance was used to evaluate the independent factors related to total drain volume of seroma. The analyses were performed using IBM SPSS Statistics ver. 23 (IBM Corp., Armonk, NY, USA). Statistical significance was set at P-value of <0.05, and sample size and power were calculated using PASS 2011 (NCSS, LLC., Kaysville, UT, USA).

Fig. 1 is a flowchart delineating the processes involved in enrollment, randomization, and follow-up in this study. Between February and December 2020, 72 patients were eligible for inclusion. Of these 72 patients, 31 declined to participate or were excluded during stratified randomization. The remaining 41 patients were randomly allocated to each group. No patients were excluded during follow-up and data analyses, and the postoperative follow-up period was from April 2020 to December 2020.

Table 1 summarizes the demographic characteristics of the included patients. The mean age was 48.9 years, and the ratio of males to females was 1:3. The random variables (sex, thyroiditis, and BMI) were not statistically different between the 2 groups. Similarly, there were no significant differences between the 2 groups in terms of tumor/lymph node (LN) stage, numbers of metastatic LNs, number of retrieved LNs, and weight of thyroid (all P > 0.05, respectively).

Table 2 presents the outcomes of the low and high fibrin groups according to intention-to-treat analysis. The primary outcomes, including total drain volumes and flow rate of drain fluid in the first 12 hours after surgery, were significantly different between the 2 groups. Additionally, both the total drain volume and flow rate, during the first 12 hours postoperatively, were significantly lower in the high fibrin group than in the low fibrin group (65.0 mL vs. 47.6 mL, P = 0.008 and 2.7 mL/hr vs. 1.8 mL/hr, P = 0.002, respectively). Although the flow rate of drain fluid during the first 24 hours after surgery was lower in the high fibrin group, no statistical significance was observed (1.6 mL/hr vs. 2.1 mL/hr, P = 0.060).

The calculated sample size of future RCT was 32 patients (16 in the low and high fibrin groups, respectively) with µ₁ of 2.7, µ₂ of 1.8, α of 0.05, and power of 0.8, as revealed by PASS statistics. Additionally, the power for a sample size of 41 patients was 0.91 with µ₁ of 2.7, µ₂ of 1.8, σ of 0.9, and α = 0.05 according to PASS 2011 (µ = mean and σ = SD for both groups).

As for the secondary outcomes, there were only 3 patients (7.3%) in the low fibrin group (Table 3); however, the difference was not statistically significant (P = 0.232). Symptomatic seroma at the outpatient clinic was related to the amount of drain during the first postoperative 12 hours (P = 0.004). In 3 patients, the drainage volume was 11 mL more than those who did not develop seroma (29.5 mL vs. 40.6 mL). Among 3 patients, 1 patient was discharged on postoperative day 1 and the others on postoperative day 2. The volume of aspirate was <5 mL in all 3 patients. There was no patient with hematoma formation or wound infection during hospitalization or follow-up visits. The duration of hospital stay was not statistically different between the 2 groups (P = 0.414).

Table 4 presents the correlation analysis between the total drain volumes and clinicopathological features. Both the duration of hospital stay and the amount of fibrin glue used were statistically correlated factors in the estimation of total drain volume after thyroidectomy (P < 0.05). An interaction between the amount of fibrin glue and the hospital stay could not be demonstrated in this study (F = 0.94, P = 0.4). There was no patient with hematoma, reoperation, or wound infection in this study.