Qu, Rui, MD^*,†; Wang, Jianxue, MD^*; Li, Jinyi, MD^*; Dong, Zhiyong, MD^*; Yang, Jingge, MD^*; Liu, Daosheng, MD^†; Wang, Cunchuan, PhD^*

Various endoscopic thyroidectomy techniques have been introduced to reduce or avoid neck scarring caused by open thyroidectomy. Of these, several have become popular, including the axillary approach,¹ areola approach,² and subclavian approach.³ However, even these procedures leave a visible scar on the body surface. Accordingly, some reports have described the use of an oral cavity approach^4,5 for female patients who desire minimal or no dermal scars. Despite the esthetic advantage, however, several difficulties of the oral cavity approach have been reported,^6,7 including (1) serious wounds to the oral cavity and potential damage to the sublingual gland, salivary gland, and deep vessels; and (2) a reduced ability to compress the incision and inhibit bleeding. To address these issues, we improved this approach and subsequently were the first to report^8,9 endoscopic thyroidectomy via the oral-vestibular approach (ETOVA), with excellent outcomes in terms of efficacy, safety, and esthetic factors.

When learning any new procedure, a surgeon will encounter a learning curve and should be able to pass through the early stage smoothly. To date, however, no reports have addressed the learning curve for ETOVA. Therefore, this study aimed to evaluate and potentially improve the learning curve for this procedure.

SUBJECTS AND METHODS

Subjects

This study enrolled 101 consecutive patients in the Department of General Surgery, First Affiliated Hospital of Jinan University (Guangzhou, China) from July 2012 to August 2017. The inclusion criteria were as follows: (1) a maximal mass diameter of ≤5.0 cm; (2) a lack of preoperative suspicion or evidence of malignancy; (3) no history of thyroiditis, neck surgery, or irradiation; and (4) the tumor should be ≤2 cm, if it was diagnosed with differentiated thyroid cancer by intraoperative frozen section. The exclusion criteria were as follows: (1) a maximal mass diameter of >5 cm; (2) a suspected cancer with lateral lymph node metastasis; (3) serious comorbid oral infective diseases (eg, periodontitis); and (4) third-degree goiter (the goiter can be seen and touched on the neck, and exceeding the outside of sternocleidomastoid muscle). The therapeutic principles were in accordance with the 2015 American Thyroid Association Guidelines.¹⁰ All patients freely opted to undergo ETOVA after receiving a detailed explanation and gave informed consent before surgery. However, the local ethics committee waived the requirement for ethical approval for this study, given the retrospective design.

The standard clinical characteristics of all cases are shown in Table 1. All procedures completed by surgeon A (C.W.) and surgeon B (J.L.). The cases completed by each surgeon were classified according to whether the surgery was performed during the primary or advanced stage of training, as described in detail in the Results section. Operative parameters such as the operation time, estimated blood loss, complications, and postoperative hospital stay duration were analyzed.

TABLE 1

Surgical Procedures

Although under general anesthesia, the patient was placed in a supine position with the neck stretched and the shoulder lifted. After conventional disinfection procedures, the whole oral cavity was sterilized with 3 chlorhexidine treatments. With the surgeon and assistant positioned above the patient’s head, an expansion solution comprising 1 mg of adrenaline in 500 mL of saline was injected into the middle of the oral vestibule toward the anterior neck (Fig. 1). A 10-mm incision was made in the oral-vestibular mucosa, and a dissection stick was used to open a channel from the area of the lower lip down to the anterior neck. After inserting a 10-mm extended trocar into this channel for endoscopy, CO₂ was injected at a constant pressure of 6 mm Hg. Subsequently, two 5-mm incisions for the operating trocars were made on either side of the oral vestibule (Fig. 2). Under direct endoscopic vision, the loose subcutaneous tissue was separated using an ultrasonic scalpel to create the working space. The scope (upper to lower border) ranged from the larynx to the suprasternal notch, and with the sternocleidomastoid muscles forming bilateral borders. After the white line of the neck was separated using an ultrasonic scalpel, the strap muscle was extended from the thyroid cartilage level to the submandibular level. A scalpel was used to pull the infrahyoid muscles to the respective bilateral sternocleidomastoid muscles, and these muscles were directed toward the skin with sutures to maintain the working space.

FIGURE 1

FIGURE 2

The thyroid isthmus was identified, dissected, and resected using an ultrasonic scalpel (Fig. 3). In addition, the tissue clinging to one thyroid lobe, superior thyroid vessels, and middle thyroid veins were dissected and divided. The gland was then elevated and dissected toward the lower pole of the thyroid (Fig. 4), while exercising caution near Berry’s ligament. Subsequently, the inferior thyroid vessels were sealed by ultrasonic scalpel. Depending on amplified and direct visualization of endoscopy, the recurrent laryngeal nerve, parathyroid glands, and trachea were identified and protected (Fig. 5). Then, the resected tissues were removed via the observation incision for pathologic analysis (Fig. 6). Finally, the wounds were closed with sutures, the chin was compressed with bandages for 24 hours, and a 5-mm drainage tube was reserved through the neck surface for 24 to 72 hours if necessary (especially in the early stage). As ETOVA is considered a type II wound, levofloxacin was administered for 2 to 3 days postoperatively to prevent infection of the wound or operative area. In addition, all patients were recommended to receive the oral calcium or calcitriol supplementation for 1 week to 1 month.

FIGURE 3

FIGURE 4

FIGURE 5

FIGURE 6

Statistical Analysis

Continuous and categorical variables are presented as means (±SD) and frequencies, respectively. A t test or χ² test was used to compare the differences between the 2 training phases. A P<0.05 was considered statistically significant. All of the data were analyzed using SPSS version 17.0 software (SPSS Inc., Chicago, IL).

RESULTS

ETOVA was performed successfully in 101 consecutive patients. Figure 7 presents trends in the operation times of the 2 surgeons over time. After 19 or 20 cases (primary stage), both surgeons achieved a reduced operation time and steady state (advanced stage). The operative results are shown in Table 2. Both surgeons had exhibited significant reductions in the mean operative times during the advanced stage (primary vs. advanced; A: 113.5±20.1 vs. 78.9±16.4 min; B: 132.8±24.5 vs. 101.0±12.9 min; P<0.05). Similarly, the drainage tube reservation rates decreased significantly in the advanced stage for both surgeons (A: 60.0% vs. 11.1%; B: 55.0% vs. 12.0%; P<0.05). However, no significant differences between the stages were observed in terms of the operative blood loss (A: 24.0±10.6 vs. 20.2±6.0 mL; B: 19.5±8.9 vs. 17.9±4.8 mL; P>0.05) or postoperative hospital stay length (A: 7.4±1.5 vs. 7.5±1.9 d; B: 6.8±2.2 vs. 7.3±1.3 d; P>0.05).

FIGURE 7

TABLE 2

The results of complications are listed in Table 3. None of the cases required conversion to open surgery. One patient treated by each surgeon experienced transient hoarseness (total: 2.0%; A: 1/56 cases, 1.8%; B: 1/45 patients, 2.2%). Transient hypoparathyroidism (serum level of parathyroid hormone<12 pg/mL) was observed in 8 of the 101 patients (7.9%), including 4 of 56 cases (7.1%) treated by surgeon A and 4 of 45 patients (8.9%) treated by surgeon B. There were 5 patients (5.0%) who needed parathyroid autograft, including 1 of 56 cases (1.8%) in surgeon A and 4 of 45 patients (8.9%) in surgeon B. The rate of postoperative skin ecchymosis was 3%, and included 2 of the 56 cases (3.6%) treated by surgeon A and 1 of the 45 patients (2.2%) treated by surgeon B. Postoperative subcutaneous fluid was observed in 2 patients (2.0%) treated by surgeon A. No patients experienced recurrent laryngeal injury, permanent hypoparathyroidism, facial numbness/paresis, maxillofacial/wound infection, or other serious complications (eg, tracheal injury, esophagus injury, teeth injury, fever, bleeding). There were no patients requiring reoperation or readmission after the operation.

TABLE 3

DISCUSSION

According to some previously published articles on transoral vestibular endoscopic thyroid surgery,^8,9,11,12 ETOVA is associated with good safety and feasibility, despite some challenges, and allows the complete avoidance of visible cutaneous scars.^8,11 However, the oral-vestibular approach is limited with respect to the ability to create an artificial workspace. Furthermore, the operating and observation ports are very close together, thus causing increased interference and operating difficulties. Therefore, the learning curve and steps associated with ETOVA are distinct from those of other thyroidectomy approaches.

The learning curves vary for different endoscopic thyroidectomy procedures. For areola approach, Cao et al¹³ reported that the primary stage comprised 25 cases, whereas Liu et al¹⁴ reported that at least 60 cases were required to achieve a shorter operative time and low complication rate. In our opinion, different institutes and surgical teams may experience different learning curves even for the same procedure. In the present study, the 56 and 45 patients treated by surgeons A and B, respectively, were divided according to stage. From the learning curve determined in our study (Fig. 7), both surgeons reached advanced levels of procedural skills, proficiency, and stability after 19 or 20 cases. ETOVA was initially reported by surgeons with sufficient endoscopic thyroidectomy experience^2,15 and a detailed understanding of the anatomy of the neck. Accordingly, no significant differences between stages were observed in terms of the intraoperative blood loss and postoperative hospital stay duration. During the primary stage, both surgeons gained a perceptual awareness of the procedural steps while refining their skills, leading to significant reductions in operation times and a steady state after a small number of cases (~20). In addition, the rate of drainage tube reservation decreased greatly during the advanced stage.

We have identified several challenges that should be noted. (1) The separation of the mucous membrane in oral vestibule and loose tissue in the neck represented a critical difficulty. Accordingly, sufficient cases are needed to ensure that the surgeon’s manipulation and workspace-building skills improve. We recommend that a training surgeon initially imitate and practice under the watchful guidance and feedback of a supervisor. (2) Insertion of the trocars around the mandibular chin also represented a considerable difficulty and required an appropriate understanding of the operating steps. (3) The subcutaneous dissection area during this procedure was reduced in comparison with that of the areola or breast approach. However, operative bleeding and injury were minimized with increased experience, and drainage tubes were rarely needed during the advanced stage. (4) A training surgeon should initially adapt to the change in endoscopic view and overcome the challenging operating restrictions. As these restrictions are onerous, we recommend that trainees have surgical experience with endoscopy. (5) As the operating and endoscopy ports are located in close proximity, the endoscope holder should cooperate expertly with the surgeon by adjusting the position and angle of the endoscope to maintain a good video image and prevent fogging. (6) During the primary stage of training, this technique should only be used to target benign thyroid disease or malignant tumors <3 cm in size. With increased experience and the evolution of surgical skills (eg, other learning curves),^16,17 the indications for ETOVA can be expanded to include PMTC, total thyroidectomy, and central lymph node dissection. In this study, we found that ETOVA provided a better endoscopic view and approach for the steps of central lymph node dissection.

In this study, we observed a low rate of complications. Transient hoarseness and recurrent laryngeal nerve injury occurred at respective rates of 2.0% and 0%, which were similar to those reported for conventional open surgery.^18,19 The transient and permanent hypoparathyroidism rates of 7.9% and 0%, respectively, were also comparable with those of open thyroidectomy^20,21 (0% to 11% and 0% to 5.7%, respectively). The skin ecchymosis (n=3) and subcutaneous fluid (n=2) accumulation occurred in the primary stage, but none of these cases occurred in the advanced stage. The reason was that inexperience skill mainly attributed to excessive damage during workspace building. In our opinion, most complications can be controlled and prevented as experience accumulates (especially in the advanced stage). However, transient hoarseness and transient hypoparathyroidism still occurred in the primary and advanced stages. Both of these complications are the most common problems in thyroid surgery. It was worth mentioning that no other serious complications occurred. Furthermore, in comparison with the oral cavity approach, ETOVA allowed us to more easily treat or compress incisions to avoid bleeding and wound infection. Therefore, we consider ETOVA to be safe and effective.

We note that this study of the learning curves of 2 surgeons had some limitations. First, the patients enrolled in our study were almost exclusively female (92.1%), young (mean age: 28.7±9.1 y), and nonoverweight (body mass index, 20.8±1.9 kg/m²). Therefore, future studies should include larger, more diverse patient populations (eg, male, overweight/obese, and/or older patients). Second, more surgeons (especially those with little endoscopic experience) and a larger number of cases are needed to further assess the learning curve. Third, beyond the operation time, additional parameters should be evaluated to more accurately assess a surgeon’s proficiency. Finally, a randomized and prospective study should be conducted to assess the effects of a standardized training program on the learning curves of training surgeons.

CONCLUSIONS

On the basis of our results, a training surgeon with sufficient endoscopic surgical experience could be expected to achieve proficiency with ETOVA and a stable state after a learning curve of ∼20 cases. We further note that for selected patients, the postoperative outcomes and complication rates were comparable with those reported for the conventional open approach. Finally, more research is needed to determine the accurate assessment criteria for the learning curve and to establish a standardized training program.

REFERENCES