The Transaxillary,Totally Endoscopic Approach for Head and Neck Endocrine Surgery in Children

Introduction

Since the development of advanced endoscopic surgical techniques and instruments, surgeons around the world have persisted in their attempts to expand the application of minimally invasive approaches to nearly every accessible body cavity. The head and neck represents an anatomic region that, while not a natural body cavity, is well suited to minimally invasive procedures performed after creating an artificial space. Most of the focus on head and neck minimal access surgery has been on endocrine lesions involving the thyroid and parathyroid glands.

Over the past 5 years, we have developed a totally endoscopic minimal access approach to head and neck lesions and reviewed our experience in detail at this time to assess the efficacy, safety, and learning curve for this approach in children.

Materials and Methods

Between June 2005 and October 2009, 31 consecutive children with head and neck endocrine problems presented to our service and were subjected to the transaxillary, totally endoscopic (TATE) approach. While we have reported on our initial experience before, ¹ some of the details of the procedure have changed. Therefore, we will detail herein our current technique.

In each case, patients were referred to us by our pediatric endocrinologists who evaluated and treated the patients with medication when indicated before recommending surgery as an option. None of our patients had previously undergone radioablation.

In our discussions with patients and their families, we routinely discuss all the surgical approaches, including all the potential benefits and risks of each operative option. All of the patients referred to us came because the endocrinologists touted the TATE approach as their procedure of choice, primarily because it leaves no visible scar.

In the operating room, all patients are placed supine on the table and are administered a general anesthetic using endotracheal intubation. None of the children in our series underwent recurrent laryngeal nerve monitoring during the procedure.

A rolled towel or sheet is placed behind the neck and the neck is slightly extended and displaced forward in a sniffing position so that the thyroid is prominent in the field and there is a relatively level plane from the axilla to the thyroid gland.

The arms extended anterior at the shoulder and the elbows are bent at right angles. The arms then are suspended from an ether screen using a soft gauze wrap (see Fig. 1). This position shortens the distance from the axilla to the thyroid and makes the surgery easier. It also allows more downward freedom of the endoscopic instruments during the performance of the procedure.

OPEN IN VIEWER

FIG. 1.

Five-year-old boy in position for transaxillary, totally endoscopic thyroidectomy.

After the skin is prepped and draped, three incisions are made in the axilla of the symptomatic side or in both axilla in the case of a total thyroidectomy. All the incisions are made in skin creases just posterior to the anterior axillary line. The largest incision is in the center of the axilla and is extended posterior for ∼1 cm or is made sufficiently long enough to accommodate the surgeon's index finger for the initial blunt dissection. Two smaller incisions, each ∼5 mm, are made in the skin for the introduction of 5-mm instrument trocars after the initial finger dissection creates a space that is big enough to accommodate all the instruments and to begin the dissection.

To create this space, the surgeon inserts the index finger of the dominant hand, and by advancing it and sweeping it from side to side creates an initial working space just anterior to the pectoralis fascia. This blunt dissection is carried medially and cephalad as far as possible using blunt finger dissection alone until it is sufficiently big enough to accommodate trocar placement and instruments for dissection.

A 0.5-cm sleeve made from a rubber catheter is placed over the end of the two 5-mm cannulae and is used for trocar fixation to the skin. The two 5-mm trocars are then introduced through their lateral incisions and fixed to the skin with 3-0 silk sutures so that they will not become dislodged during the dissection.

A purse-string suture is placed in the skin around the larger central incision and used to occlude the space around the 10-mm cannula and secure the cannula in place so that it cannot become dislodged from the incision and so that the incision is gas tight.

Under 10-torr CO₂ insufflation, a 10-mm telescope is inserted through the central cannula, a dissector is inserted through one of the 5 mm cannula, and an “L” hook electrocautery or the Harmonic Scalpel is inserted through the other so that dissection can proceed and so that the space between the subcutaneous fascia and platysma muscle just anterior to the pectoralis fascia can be enlarged to the level of the clavicle cephalad and the sternocleidomastoid muscle medially. At that juncture, dissection is extended cephalad until the sternomastoid muscle is clearly displayed.

Depending on the size and location of the lesion, dissection then continues either underneath or between the two heads of the sternomastoid muscle, medially toward the strap muscles. The junction of the sternomastoid and omohyoid muscles serves as a useful landmark as the omohyoid lies over the anatomic junction between the upper and lower two-thirds of the thyroid gland.

Dissection continues deeper, dividing the strap muscles as required until the thyroid gland itself is observed. At that point, the thyroid is further exposed until the upper and lower poles and their vessels can be easily seen and there is sufficient room to divide them safely and well away from the thyroid gland and any potentially important structures. We carefully search for parathyroid glands and remain fully aware of the position of the recurrent laryngeal nerve throughout the case.

For large glands, it is easier to approach the thyroid isthmus first as it lies immediately anterior to the trachea. We divide the isthmus in the midline with the Harmonic Scalpel to make it easier to continue with our dissection of the individual thyroid lobes.

After the polar vessels are divided, attention is turned to the hilar structures. The vessels are carefully dissected from the hilum and displaced posteriorly, taking care to identify the recurrent laryngeal nerve and parathyroid glands during the dissection so as not to injure them. We prefer to leave a small bit of thyroid tissue adherent to the nerve when further dissection might risk injury to the nerve. Once the hilar structures are free and allowed to fall posterior, the intact lobe is then dissected from the trachea using the Harmonic Scalpel until it is completely free and the tracheal rings are completely freed of any thyroid tissue.

The freed lobe is placed out of the line of sight so that the bed of dissection can be inspected to make certain that there is no residual hemorrhage and that the bulk of the complete lobe was removed.

We then remove one of the dissecting instruments from either of the 5-mm ports and replace it with a 5-mm camera. A grasper is placed through the 10-mm port, and under direct vision, the intact lobe is grasped and extracted through the incision. Rarely, for a large gland, the incision needs to be enlarged slightly to deliver the specimen.

Once the intact lobe is extracted, a Petite Wound drain (Axiom) is placed in the bed of the thyroid and brought out through a stab wound between two of the axillary incisions.

After the drain is sutured in place to the skin, the CO₂ is evacuated and the skin wounds are closed. Surgical glue is used to dress the wounds. Typically, we remove the drain the morning after surgery.

For cases of total thyroidectomy, two surgeons work simultaneously toward the midline, one from each axilla performing the procedure described above. Rarely do the two surgeons interfere with each other's dissection. On the contrary, both are aware of what the other is doing throughout the case; one surgeon may pause the dissection from time to time to allow the other to make some progress.

In cases of hyperparathyroidism, we get at least two preoperative studies documenting the location of the adenoma. We find that a computed tomography scan and a Sestamebe scan help us best to locate the adenoma.

At the time of surgery, we prepare to go directly for the localized adenoma from the side of the lesion. We draw preoperative parathyroid hormone levels before we make our incision and use rapid parathormone levels postoperatively to confirm that we have removed the abnormal gland.

Dissection is more direct, toward the site of the localized adenoma, and thus does not require the extensive dissection we usually perform for total thyroidectomy.

In every case, we have been able to localize the adenoma and confirm successful removal of the gland by observing a precipitous drop in the parathormone level toward normal.

For each case, we recorded the diagnosis, age and gender of the patient, the weight of the specimen, the duration of the surgery, and any complications.

Data

We evaluated 31 children whose mean age was 12.7 years. Girls predominated our series (5:1). The removed thyroid glands ranged in size from 10 g (in a small 5-year-old) to 63 g (in one 16-year-old). The mean size of the thyroid was 29.0 g.

No cervical incisions were required to facilitate completion of the operation and there were no conversions to open surgery.

Three patients in our early experience underwent a robot-assisted procedure (previously reported). ¹

Two patients suffered from hyperparathyroidism and had adenomas removed using the TATE approach as described above.

Of the remaining 29 children, 2 girls had an adenoma of the right thyroid lobe removed uneventfully and the reminder all had Graves disease, which was treated successfully with a total thyroidectomy.

Complications in this series included transient hypocalcemia in 3 patients all of whom were also on steroids for unrelated medical problems. These patients remained in the hospital 1 or 2 days extra to stabilize their serum calcium levels. In each case, the hypocalcemia resolved within a week to 10 days, after which no additional calcium supplementation was necessary.

We observed a transient neuropraxia in 3 patients who had very large glands. We believe that this resulted from traction on the recurrent laryngeal nerve as we mobilized the large gland for excision.

There was 1 patient who developed a postoperative subcutaneous hematoma on one side of her neck from a drain that appeared to malfunction. This stabilized by the next morning and resolved uneventfully.

One 5-year-old girl with severe asthma developed a spontaneous apical pneumothorax that presented as subcutaneous gas in the neck 6 hours after the surgery. We performed bronchoscopy to confirm that there was no laryngeal injury and then inserted a thoracostomy tube. The pneumothorax resolved with this maneuver and she recovered uneventfully.

One teenager with an extremely large gland had a laryngeal tear noted after dissection of the gland from the larynx. This was repaired with sutures endoscopically and the patient recovered uneventfully.

The average operative time for the first 10 patients was 288 minutes and for the subsequent patients was 155 minutes (range, 92–210 minutes).

All patients in the series resolved the symptoms that prompted their operation and none required any additional procedures other than those mentioned that were done for complications.

Discussion

Since minimally invasive surgical approaches became mainstream, beginning in the late 1980s, surgeons began using the available instrumentation and adapted the techniques to procedures beyond laparoscopy and thoracoscopy.

The first attempts at endoscopically assisted thyroidectomy were reported in 1997. ² The approach was to make a small cervical incision through which endoscopic instruments were used to remove the thyroid.

Similar approaches have been described for the removal of abnormal parathyroid glands in conjunction with radioactive localization of the adenomas. ³

Many different approaches to thyroidectomy have appeared in the literature using incisions in the anterior chest wall, circumareolar incisions,^4,5 and transaxillary incisions to gain access to the neck.^6,7 Some of these approaches promoted the use of flexible endoscopes to improve observation.

The main advantage of each of the approaches was to eliminate the highly visible collar incision in the neck.

More recently, reports of a transaxillary approach using robotics have been promoted. ⁸ This requires retraction through a much larger incision than we use in our technique and it does not make use of insufflation. Early reports seem satisfactory, but in our opinion, this approach is less applicable to children and young teens.

The advantage the approach described herein is that it does not need special instruments beyond those ordinarily used in laparoscopic surgery. While we initially thought that the robot might be useful, we abandoned its use for these cases as it seemed simpler and perfectly adequate to perform the procedure without the additional time and expense.

As has been noted for other types of endoscopic surgery, there is a learning curve of about 10 cases, after which the surgeon has a clearer concept of the anatomic relationships and can perform the procedure more quickly.^9,10

With rare exceptions, the sort of complications we observed are similar to those observed with open endocrine surgery in the neck. In this early experience, we have not encountered any permanent complications of nerve injury or hypoparathyroidism as one often sees with similar series of cases performed using open techniques.

We attribute the low complication rate to better observation of the structures with the magnification provided by using the telescope.

Conclusions

Consistent with reports of similar approaches, ¹¹ the TATE approach appears to be as safe and effective as open surgery for cervical endocrinopathies in children. Only transient, minor complications were observed. Operative time significantly decreases with experience, making the length of surgery comparable to its open counterpart. The number of cases it takes to become accustomed to the anatomic relationships and to safely and efficiently remove the pathology appears to be consistent with other reports and is about 10 cases.