Safety and Outcomes of the Transoral Endoscopic Thyroidectomy Vestibular Approach

This study compares the safety and outcomes of the transoral endoscopic thyroidectomy vestibular approach with those of open thyroidectomy.

Key Points

Question

Is the transoral endoscopic thyroidectomy vestibular approach safe compared with conventional open thyroidectomy?

Findings

In this study of patients undergoing thyroidectomy in Bangkok, Thailand, the operative time for transoral endoscopic thyroidectomy vestibular approach was longer than that for open thyroidectomy; however, patients who underwent transoral thyroidectomy had less postoperative pain than did patients who underwent open thyroidectomy. Blood loss and complication rates were similar.

Meaning

The transoral endoscopic thyroidectomy vestibular approach can be a safe operation for select patients, with outcomes similar to those of open thyroidectomy.

Abstract

Importance

Natural orifice transluminal endoscopic surgery thyroidectomy is a novel approach to avoid surgical scars.

Objective

To compare the safety and outcomes of the transoral endoscopic thyroidectomy vestibular approach (TOETVA) with those of open thyroidectomy (OT).

Design, Setting, and Participants

This study retrospectively reviewed all TOETVA and OT operations performed from April 1, 2014, through August 31, 2016, at Police General Hospital, Bangkok, Thailand. All patients who underwent TOETVA and patients who underwent OT were included. Exclusion criteria were (1) previous neck surgery, (2) substernal goiter, (3) lymph node or distance metastasis, and (4) suspicious invasion to the adjacent organs. Propensity score matching was conducted to reduce selective bias.

Main Outcomes and Measures

Operative time, blood loss, and complications related to thyroid surgery.

Results

Of the 425 patients who underwent transoral endoscopic thyroidectomy (mean age, 35.3 [12.1] years; age range, 16-81 years; 389 [92.2%] female), 422 successfully were treated with the TOETVA; 3 patients were converted to a conventional operation because of bleeding. Twenty-five patients (5.9%) had transient recurrent laryngeal nerve palsy, and 46 (10.9%) had transient hypoparathyroidism. None had permanent recurrent laryngeal nerve palsy or permanent hypoparathyroidism. Three patients (0.7%) had transient mental nerve injury; all cases resolved by 4 months. One patient developed postoperative hematoma treated by OT. Twenty patients (4.7%) had seroma treated by simple aspiration. Operative time was longer for the TOETVA compared with the OT group (100.8 [39.7] vs 79.4 [32.1] minutes, P = 1.61 × 10⁻¹⁰). The mean (SD) visual analog scale score for pain was lower in the TOETVA group (1.1 [1.2] vs 2.8 [1.2], P = 2.52 × 10⁻³⁸). Estimated mean (SD) blood loss (36.9 [32.4] vs 37.6 [23.1] mL, P = .43) and rate of complications (45 of 216 [20.8%] vs 38 of 216 [17.6%], P = .41) were not significantly different in the TOETVA vs OT group.

Conclusions and Relevance

The TOETVA was performed as safely as OT, requires only conventional laparoscopic instruments, and avoids incisional scars; thus, the approach may be an option for select patients.

Introduction

Many patients who need thyroidectomy are concerned about the scar caused by the incision in the neck. Various techniques of minimally invasive thyroidectomy, including endoscopic and robotic thyroidectomy, have been developed to address this issue. Some are concerned that endoscopic and robotic thyroidectomy are not minimally invasive because of the need for wide flap dissection. Moreover, there are still scars at the incision sites, albeit away from the neck. The techniques of natural orifice transluminal endoscopic surgery as applied to thyroidectomy via the oral cavity access have been developed since 2008, starting in pigs, then cadavers, and finally in humans. Transoral natural orifice transluminal endoscopic thyroidectomy can be performed via the sublingual approach or the oral vestibular approach. The sublingual approach has been less popular because it is associated with more complications. On the other hand, several oral vestibular techniques have been developed and appear to be safer than the sublingual approach.

The transoral endoscopic thyroidectomy vestibular approach (TOETVA) has been successfully refined and applied to the transoral endoscopic parathyroidectomy vestibular approach, with few complications and excellent results. However, debate remains about whether TOETVA is appropriate. Furthermore, only a few published series of TOETVA are available, with none specifically comparing TOETVA with the traditional open thyroidectomy (OT). To our knowledge, the current study is the largest series of TOETVA to date. We also compared the outcomes and complications of TOETVA with those of OT performed by the same surgeons by using propensity score matching to mitigate the potential effect of patient selection.

Methods

Patient Eligibility and Study Design

We retrospectively reviewed all TOETVA and OT operations performed from April 1, 2014, through August 31, 2016, at Police General Hospital, Bangkok, Thailand. All patients who underwent TOETVA and patients who underwent OT were included. Exclusion criteria for this study included (1) previous neck and chin surgery, (2) substernal goiter, (3) clinically evident lateral neck lymph node metastasis, (4) distance metastasis, and (5) suspicious invasion of this metastasis to adjacent organs, such as the esophagus, trachea, or recurrent laryngeal nerve. The study was approved by the Police General Hospital Ethics Committee. Written informed consent was provided by all patients (in Thai language). All data were deidentified.

Either TOETVA or OT was chosen based on the patients’ preferences, size of the tumor, and the nature of the disease. All procedures were performed by 4 of us (A.A., K.K., P.J., T.S.). All patients were evaluated preoperatively with thyroid function tests, fine-needle aspiration cytologic testing, and thyroid ultrasonography. Computed tomography was used for limited indications. No routine dental examination was required.

Surgical Technique

Transoral Endoscopic Thyroidectomy Vestibular Approach

For the TOETVA technique, the patient was placed in a supine position with slight neck extension under nasotracheal intubation. Amoxicillin–clavulanic acid, 1.2 g, was administered 30 minutes before incision. Three laparoscopic ports (a 10- to 15-mm port at midline and two 5-mm ports at the lateral junction between the canine and first premolar teeth) were inserted under the lower lip at the oral vestibular area. A standard 10-mm 30° laparoscope was used, allowing top-down visualization. The working space was created down to the sternal notch with the lateral border at the sternocleidomastoid muscles. The strap muscles were separated in the midline and retracted laterally by a transcutaneous 2/0 silk suture to expose the thyroid and trachea. The thyroid isthmus was divided first. Next, the superior pole was dissected, and the branches of the upper pole vessels were divided using ultrasonic shears (Harmonic Scalpel) on the surface of the thyroid gland. Upper parathyroid and lower parathyroid glands were identified and preserved. The recurrent laryngeal nerve (RLN) was identified at the insertion to the larynx, then followed down and parallel to the trachea inferiorly. A thyroid lobe specimen was placed in a specimen pouch and retrieved through the 10- to 15-mm central incision. For total thyroidectomy, the operation was performed in the same manner on the contralateral side. A No. 10 drain was placed, exiting through a small incision above the clavicle for total thyroidectomy. A pressure dressing was placed around the chin and upper neck for 24 hours. Further details of the operative steps can be watched in the Video.

Video. Operative Steps of the Transoral Endoscopic Thyroidectomy Vestibular Approach.

Download video file ^{(352.7MB, mp4)}

Open Thyroidectomy

For standard OT, the patient was placed in a supine position under orotracheal intubation. Cefazolin, 1 g, was given intravenously 30 minutes before incision. A 5- to 10-cm transverse collar incision was made. Subplatysmal flaps were dissected up to the thyroid cartilage and down to the sternal notch. The strap muscles were separated in the midline to expose the thyroid gland. Lateral dissection of the thyroid gland was performed. The middle thyroid vein was ligated. Superior thyroid vessels were individually ligated close to the thyroid gland. The RLN was identified and dissected near the inferior thyroid artery. Branches of the inferior thyroid vessels were ligated, preserving the lower parathyroid gland. Berry ligament was then dissected, and finally the isthmus was transected. For total thyroidectomy, the operation was performed in the same manner on the contralateral side. A No. 10 drain was placed for total thyroidectomy. A pressure dressing was applied around the anterior part of the neck.

Postoperative Management

All patients received a liquid diet on the day of surgery, followed by a soft diet on the following day. None of the operations were performed as outpatient procedures. For TOETVA, intravenous antibiotics were given for 2 days, followed by a 7-day course of oral antibiotics. For OT, no postoperative antibiotics were administered. Postoperative analgesic medication was the same in the TOETVA and OT groups. All patients received intravenous meperidine in the operative room. In the ward, both groups were routinely prescribed 2 tablets of acetaminophen, 500 mg, for mild to moderate pain and intravenous meperidine, 30 mg, for severe pain. Few patients required additional intravenous analgesia in the ward. All patients were evaluated by laryngoscopy 1 week postoperatively. All patients had follow-up examinations at 1 week, 1 month, 6 months, and 1 year.

Outcome

Medical records were reviewed for patient characteristics and outcomes, including operative time (incision to closure time), estimated intraoperative blood loss, and pain as measured by the visual analog scale (VAS). Mean values and values for the first 3 days postoperatively were reported. Complications were identified. An RLN injury was defined as impaired movement of one or both vocal cords on laryngoscopy. A permanent RLN injury was defined as an injury that did not recover within 6 months. Seroma that required aspiration was recorded. Hypoparathyroidism was defined as a parathyroid hormone level less than 13 pg/mL (to convert to nanograms per liter, multiply by 1) 24 hours postoperatively. Permanent hypoparathyroidism was defined as no recovery within 6 months. Mental nerve injury was characterized as lower lip paresthesia after surgery. Hematoma was defined as obvious postoperative bleeding that caused neck swelling and required a subsequent operation. Infection was defined as postoperative local abscess or high-grade fever with evidence of systemic bacterial spread that required treatment.

Statistical Analysis

Baseline characteristics were compared using the Fisher exact test for categorical variables and the independent, 2-tailed, paired t test for continuous variables. Because Police General Hospital had more TOETVA cases than OT cases, propensity score matching analysis was conducted to reduce bias attributable to potential confounding effects. Six factors that could affect the surgical outcomes were selected: age, sex, diagnosis, tumor size, bilateral involvement, and multicentricity. After propensity score matching was conducted, the 2 groups were compared in terms of characteristics and surgical outcomes. P < .05 was considered to be statistically significant. Statistical analysis was performed with SPSS statistical software, version 20.0 (IBM Inc). All values were expressed as mean (SD) for continuous variables and number (percentage) for categorical variables.

Results

Characteristics and Results Before Matching

Of the 425 patients who underwent TOETVA (mean age, 35.3 [12.1] years; age range, 16-81 years; 389 [92.2%] female), 3 were converted to OT because of excessive bleeding, 2 because of Graves disease (estimated blood loss of 750 and 1000 mL) and 1 because of a goiter larger than 12 cm (estimated blood loss of 800 mL). Therefore, 422 patients successfully underwent TOETVA and were included in the analysis. A total of 245 patients (58.1%) had a single thyroid nodule or cyst and underwent hemithyroidectomy, and 118 patients (27.9%) had multinodular goiter. Thirty-three patients (7.8%) had Graves disease and underwent total thyroidectomy. Twenty-six patients (6.2%) had papillary microcarcinoma and underwent total thyroidectomy with central lymph node dissection. The tumor size was 3.8 (1.5) cm (range, 1-10 cm), with an operative time of 96.6 (36.8) minutes (hemithyroidectomy, 76.8 [20.3] minutes; total thyroidectomy, 124.1 [35.6] minutes). Estimated blood loss was 35.8 (36.8) mL (hemithyroidectomy, 25.7 [28.8] mL; bilateral thyroidectomy, 52.3 [41.5] mL). The overall VAS pain score was 1.3 (1.2), and the VAS pain scores were 2.3 (2.00), 1.1 (1.33), and 0.4 (0.8) for the first 3 days. Twenty-five patients (5.9%) had transient RLN palsy; all cases resolved within 6 months. Forty-six patients (10.9%) had transient hypoparathyroidism; none had permanent hypoparathyroidism. One patient (0.2%) had hematoma and required open cervical incision for decompression 72 hours after the operation. Twenty patients (4.7%) had seroma and were treated by simple aspiration. Three patients (0.7%) had transient mental nerve injury; all cases resolved within 4 months. No patient had wound infection (Table 1). When we performed the analysis by intention to treat and included the 3 patients who were converted to OT, the results remained the same.

Table 1. Demographic Data and Operative Details of the Study Patients^a.

Variable	Finding (N = 422)
Patient age, mean (SD) [range], y	35.3 (12.1) [16-81]
Sex
Male	33 (7.8)
Female	389 (92.2)
Thyroid disease
Single thyroid nodule or cyst	245 (58.1)
Multinodular goiter	118 (27.9)
Graves disease	33 (7.8)
Papillary microcarcinoma	26 (6.2)
TOETVA
Hemithyroidectomy	245 (58.1)
Bilateral thyroidectomy	177 (41.9)
Operative time, mean (SD) [range], min
Total	96.6 (36.8) [45-300]
Hemithyroidectomy	76.8 (20.3) [45-177]
Bilateral thyroidectomy	124.1 (35.6) [45-300]
Blood loss, mean (SD) [range], mL
Total	35.8 (36.8) [6-300]
Hemithyroidectomy	25.7 (28.8) [6-200]
Bilateral thyroidectomy	52.3 (41.5) [7-300]
Tumor size, mean (SD) [range], cm	3.8 (1.5) [1-10]
VAS pain score, mean (SD) [range]
Total	1.3 (1.2) [0-7]
Day 1	2.3 (2.00) [0-7]
Day 2	1.1 (1.3) [0-5]
Day 3	0.4 (0.8) [0-3]
Complications
RLN palsy
Transient	25 (5.9)
Permanent	0
Hypoparathyroidism
Transient	46 (10.9)
Permanent	0
Hematoma	1 (0.2)
Infection	0
Seroma	20 (4.7)
Mental nerve injury (transient)	3 (0.7)
Mediastinal emphysema	0

Abbreviations: RLN, recurrent laryngeal nerve; TOETVA, transoral endoscopic thyroidectomy vestibular approach; VAS, visual analog scale.

^aData are presented as number (percentage) of patients unless otherwise indicated.

Comparison of Surgical Outcomes

Baseline characteristics of the TOETVA vs OT group before propensity score matching are given in Table 2. After matching, 216 pairs of patients were selected for the 2 groups (Table 3). We then compared surgical outcomes between the 2 matched groups, and the results are reported in Table 4.

Table 2. Baseline Characteristics of Patients Before Propensity Score Matching^a.

Characteristic	TOETVA Group (n = 422)	OT Group (n = 216)	P Value
Age, mean (SD), y	35.3 (12.1)	35.3 (12.1)	.48
Sex
Male	33 (7.8)	19 (8.8)	.33
Female	389 (92.2)	197 (91.2)	.38
Extent of surgery
Hemithyroidectomy	245 (58.1)	132 (61.1)	.47
Bilateral thyroidectomy	177 (41.9)	84 (38.9)	.46
Tumor size, mean (range), cm	3.8 (1.5)	4.7 (1.9)	<.001

Abbreviations: OT, open thyroidectomy; TOETVA, transoral endoscopic thyroidectomy vestibular approach.

^aData are presented as number (percentage) of patients unless otherwise indicated.

Table 3. Baseline Characteristics of Patients After Propensity Score Matching^a.

Characteristic	TOETVA Group (n = 216)	OT Group (n = 216)	P Value
Age, mean (SD), y	35.1 (11.9)	35.3 (12.1)	.41
Sex
Male	20 (9.1)	19 (8.8)	.49
Female	199 (90.9)	197 (91.2)	.38
Extent of surgery
Hemithyroidectomy	133 (60.7)	132 (61.1)	.46
Bilateral thyroidectomy	86 (39.3)	84 (38.9)	.38
Tumor size, mean (range), cm	4.1 (1.6)	4.7 (1.9)	.08

Abbreviations: OT, open thyroidectomy; TOETVA, transoral endoscopic thyroidectomy vestibular approach.

^aData are presented as number (percentage) of patients unless otherwise indicated.

Table 4. Comparison of Surgical Outcomes Between the TOETVA and OT Groups After Propensity Score Matching^a.

Variable	TOETVA Group (n = 216)	OT Group (n = 216)	P Value
Operative time, mean (SD) [range], min
Total	100.8 (39.7)	79.4 (32.1)	1.61 × 10−10
Hemithyroidectomy	78.6 (20.8)	64.2 (20.7)	2.06 × 10−8
Bilateral thyroidectomy	135.1 (37.6)	103.3 (23.9)	3.02 × 10−10
Blood loss, mean (SD) [range], mL
Total	36.9 (32.4)	37.6 (23.1)	.43
Hemithyroidectomy	26.8 (26.6)	25.9 (12.6)	.37
Bilateral thyroidectomy	52.6 (31.9)	55.9 (24.2)	.45
VAS pain score, mean (SD) [range]
Total	1.1 (1.2)	2.8 (1.2	2.52 × 10−38
Day 1	2.1 (1.9)	4.9 (1.8)	5.21 × 10−44
Day 2	0.9 (1.3)	2.4 (1.4)	3.87 × 10−26
Day 3	0.4 (0.8)	1.1 (0.9)	2.09 × 10−15
Complication
Total	45 (20.8)	38 (17.6)
RLN palsy			.41
Transient	9 (4.0)	9 (4.0)
Permanent	0	0
Hypocalcemia
Transient	23 (10.6)	20 (9.3)
Permanent	0	0
Hematoma	1 (0.5)	0
Seroma	9 (4.0)	9 (4.0)
Mental nerve injury (transient)	3 (1.4)	0
Infection	0	0

Abbreviations: OT, open thyroidectomy; RLN, recurrent laryngeal nerve; TOETVA, transoral endoscopic thyroidectomy vestibular approach; VAS, visual analog scale.

^aData are presented as number (percentage) of patients unless otherwise indicated.

Discussion

The initial, smaller series of 60 patients found TOETVA to be safe. We present here a larger series of more than 400 patients and additionally compare the outcome of TOETVA with that of OT.

Before matching, the clinical characteristics between the 2 groups were similar except for tumor size. The TOETVA group had smaller tumor sizes compared with the OT group. This difference is most likely attributable to patient selection. The decision about whether to perform TOETVA or OT was based on patient preference, tumor size, and nature of disease (eg, advanced cancer or huge goiter required OT). We used propensity score analysis to reduce confounding factors attributable to patient selection.

Compared with the matched OT group, the TOETVA group had a significantly longer operative time. Most other studies comparing endoscopic thyroidectomy and OT had similar results. Operative time may decrease after a learning curve.

We conclude that TOETVA may be as safe as OT because no significant differences were found in estimated blood loss or complication rates between the 2 groups. There are, however, differences between the techniques. For example, TOETVA allows a magnified view by the endoscope, but the top-down anatomy may not be familiar to the traditional thyroid surgeon (Figure). In OT, the RLN can be identified 2 different ways: initially caudad from the inferior thyroid artery and then traced to the cephalad (bottom up) or initially from the RLN at the insertion point and then traced caudad (top down). For TOETVA, it is easier to dissect the RLN first at its insertion and then follow from top down. Surgeons who are familiar with the top-down, cranial to caudal dissection of RLN may find it easier to adopt the technique. Our rate of temporary RLN injury was 5.9%, which is comparable to other approaches of endoscopic or OT. We had no patients with permanent RLN injury in this study. Some surgeons have reported using intraoperative nerve monitoring to reduce the rate of RLN injury for TOETVA. The rate of hypoparathyroidism in this series was 10.9%, similar to OT and other endoscopic approaches. We had no patient with permanent hypoparathyroidism. The magnified view may allow for easier identification of the parathyroid glands.

Figure. Endoscopic View During the Transoral Endoscopic Thyroidectomy Vestibular Approach Demonstrated a Top-down Anatomy After Right Thyroidectomy.

White arrowhead indicates the right recurrent laryngeal nerve. CA indicates right common carotid artery; CT, cricoid cartilage; LT, left thyroid lobe; RLP, right lower parathyroid gland; RUP, right upper parathyroid gland; and TA, trachea.

Patients in the TOETVA group had less postoperative pain based on the VAS pain score. The incision inside the oral vestibule seems to cause less pain compared with skin incisions. Patients rarely experienced pain from the oral vestibular incision. One advantage of TOETVA is that less flap dissection is required compared with other endoscopic and robotic thyroidectomy techniques, and thus there may be less postoperative pain.

We believe further improvement of TOETVA is likely. Transoral incision is categorized as a clean-contaminated wound and may have higher risk of infection. There was a case of stitch abscess in the series by Wilhelm and Metzig that occurred 3 weeks after transoral endoscopic thyroidectomy; it was treated by incision and drainage. A sublingual incision may have a higher risk of infection than the oral vestibular approach because ours and other series of oral vestibular approach also did not have any case of infection. The use of perioperative antibiotics in our study may not have been optimal, and further study may show antibiotics to be needed for a shorter period. The rate of transient RLN injury may be improved by more meticulous dissection, avoiding thermal injury from hot dissecting blade, and perhaps by using intraoperative nerve monitoring. The rate of transient hypoparathyroidism may also be improved by more meticulous dissection, avoiding thermal injury from hot dissecting blades, and perhaps by using indocyanine green and other aids to identify parathyroid glands. To address the oncologic concern of cancer seeding and local recurrence, we routinely use a specimen pouch to enclose the specimen before extraction. For extracting benign lesions, we used a pair of sponge forceps to grasp and align the thyroid lobe longitudinally and pull on the sponge forceps while also pushing the specimen externally with the other hand. With this technique, the thyroid lobe can be extracted relatively intact. For larger specimens, cutting the thyroid capsule away from the nodule also can help with extraction. Because it is important to extract the tumor intact, patients with thyroid tumors larger than 2 cm are usually not candidates for TOETVA.

The TOETVA technique has a significant learning curve. The surgeons learning this technique should already be proficient in OT and other laparoscopic or endoscopic surgery. In our experience, for endocrine surgeons who routinely perform OT and laparoscopic adrenalectomy, the learning curve of TOETVA is estimated to be 7 to 10 cases. There are concerns about the safety of performing endoscopic thyroidectomy in obese patients. Thai patients have lower body mass indexes (calculated as weight in kilograms divided by height in meters squared) than Western patients. In our series, there were fewer than 20 patients (TOETVA and OT combined) with morbid obesity (body mass index >35). We found that TOETVA can be performed safely in obese patients. The shorter distance from the oral vestibule to the thyroid gland makes it easier in obese patients compared with other endoscopic approaches for thyroidectomy (eg, transaxillary, transareolar). As in other endoscopic operations, we recommend that surgeons become experienced in the technique before performing it in obese patients.

The length of hospital stay for surgery in Asian countries is different from that in the United States. For thyroid surgery, a patient usually stays in the hospital 2 to 3 days after OT or TOETVA. The length of stay depends on the patient’s expectation and culture and the health care system. The cost of stay is relatively inexpensive, and payers do not pressure the patients or hospitals in Thailand for early discharge. Thus, length of stay for patients undergoing OT and TOETVA is not different. Patients from the 3 series from the United States and Italy were discharged after an overnight stay without any adverse outcomes. We believe further study regarding length of stay should be performed to ensure the safety of staying overnight for this procedure.

Limitations

Our study has some limitations. Even though a propensity score was conducted, a single–Asian institute retrospective study with an unbalanced population between both groups does not allow us make conclusions about the true safety of this technique. Furthermore, most patients in this study had benign lesions because of the nature of thyroid disease in our country. We therefore did not have enough data regarding thyroid cancer parameters. Thus, multicenter, multicountry studies, as well as prospective studies with long-term follow-up, are needed to confirm the safety and effectiveness of TOETVA, especially in patients with thyroid cancer. This will be the subject of future studies.

Conclusions

In this large series from a single center, we found that TOETVA can be performed safely while avoiding any skin incision. In addition, TOETVA is associated with longer operating time and less postoperative pain compared with OT and has outcome and complication rates similar to those of OT.