Reconstruction of Hemiglossectomy Defects With the Lateral Arm Fasciocutaneous Flap

Abstract

Introduction:

Partial tongue reconstruction requires a thin pliable flap to restore volume and mobility. The lateral arm flap is well suited to this as it is a thin fasciocutaneous flap that has consistent vascular anatomy, reliable perfusion, short harvest time and low donor site morbidity. The authors report our experience with use of this flap for reconstruction of hemi-glossectomy defects.

Methods:

This is a retrospective cohort of patients who underwent reconstruction of hemi-glossectomy and floor of mouth defects with a lateral arm flap, at Aga Khan University Hospital, Karachi (Pakistan) from November 2016 to January 2020. Flaps were harvested from the nondominant upper extremity. Data were collected for patient demographics, size of defect, size of flap, recipient vessels, postoperative complications and functional outcome.

Results:

Over a 3-year period, 8 hemi-tongue and extended hemi-tongue, and floor of mouth reconstructions were performed with a lateral arm fasciocutaneous flap. A standard lateral arm flap was harvested in 3 patients and an extended lateral arm flap in 5 patients. Mean flap size was 65.75 cm² (48 – 76 cm²). The recipient artery in all cases was the superior thyroid artery. The recipient veins were the common facial vein in 1 patient, the internal jugular in 3 patients, the external jugular in 1 patient and both external and internal jugular in 3 patients. Donor sites were closed primarily. There were no total or partial flap losses. All patients were able to resume an oral diet (unrestricted in 1, soft in 4, pureed in 2 patients). Postoperative speech was intelligible to patients’ family in 4 patients and to strangers in 3 patients. One patient succumbed to progressive disease in the early postoperative period. Orocutaneous fistulas developed in 3 patients, all of which healed with non-operative management.

Conclusions:

The lateral arm fasciocutaneous flap is well suited for reconstruction of hemiglossectomy and floor of mouth defect. It has the advantages of straightforward harvest, thin and pliable soft tissue, and low donor site morbidity.

Tongue reconstruction is one of the most challenging under-takings in head and neck reconstruction. The tongue has a critical role in speech, feeding and airway protection. The goal is to create a tongue that is mobile, sensate and of optimal volume.¹ For partial glossectomy defects, the reminder of the tongue musculature retains reasonable function for tongue movements. Surgery thus aims to restore tongue mobility by creating a neo-tongue that is untethered and not overly bulky.

Soft tissue flaps for partial glossectomy defect reconstruction need to be thin and pliable; too bulky of a flap will weight down the tongue and a stiff flap will not be able to recreate the sharp contours of the tongue. The radial forearm free flap is considered the workhorse flap for partial tongue reconstruction because of these favorable characteristics.² Its major downside is donor site morbidity, necessitating a skin graft for closure and sacrifice of a major vessel to the hand.³ The anterolateral thigh fasciocutaneous flap has emerged as a viable alternative; however, it is frequently too bulky and stiff for hemi-tongue reconstructions. Another option is the lateral arm flap.^4,5 This is a thin and pliable fasciocutaneous flap, that does not require sacrifice of a major vessel, has a sensory nerve and leaves an aesthetically acceptable donor site that can be closed primarily.

We have used the lateral arm fasciocutaneous flap for reconstruction of composite hemi-glossectomy or extended hemi-glossectomy and floor of mouth defects. This article presents our experience with the use of this flap in tongue reconstruction, along with our surgical technique.

MATERIALS AND METHODS

Patients

This is a retrospective cohort of all patients who underwent hemi-tongue and floor of mouth reconstruction with lateral arm fasciocutaneous free flap from November 2016 to January 2020 at Aga Khan University Hospital in Karachi (Pakistan). All tumors were squamous cell carcinomas, diagnosed preoperatively by biopsy. Tumor staging was performed with computed tomographic scans of the head and neck. Resections were performed by the otolaryngology service. All reconstructive procedures were consecutive operations performed by the senior author (FR). Data were collected from patient charts for demographics, tumor characteristics, defect size, flap size, recipient vessels, and postoperative complications. Patients were evaluated for donor site outcome, speech intelligibility, and type of oral diet.

The study was approved by the institutional review board (IRB# 2020–4956-11067).

Anatomy of the Lateral Arm Flap

The lateral arm flap is based on the posterior radial collateral artery (PRCA), which is a continuation of the profundal brachii artery (Fig. 1A). The profundal brachii is the largest branch of the brachial artery. It enters the posterior compartment of the arm, along with the radial nerve, through the triangular space formed by the humerus, long head of the triceps and the teres major muscle. It travels on the radial groove on the posterior surface of the humerus accompanied by the radial nerve, deep to the lateral head of the triceps. It then gives off the middle collateral artery to the medial head of the triceps and continues as the radial collateral artery. This vessel then divides into the anterior radial collateral artery which accompanies the radial nerve between the brachialis and brachioradialis, and the PRCA that enters the lateral intermuscular septum between the triceps posteriorly and the brachialis and brachioradialis anteriorly. The PRCA gives off multiple septocutaneous perforators which supply the overlying skin. Venae commitantes accompany the arteries. The sensory innervation of the lateral arm is by the lower lateral cutaneous nerve of the arm. This nerve arises from the radial nerve, perforates the lateral head of the triceps just posterior to the deltoid insertion, and goes through the deep fascia to become subcutaneous. The posterior cutaneous nerve of the forearm travels with the pedicle in the intermuscular septum. It is usually divided and taken with the flap, although it can sometimes be separated from the pedicle and preserved.

FIGURE 1.

(A) Anatomy of the lateral arm flap. (B) Flap design. The proximal part of the flap is used for floor of mouth and the distal part for tongue reconstruction. (C) Harvested lateral arm flap. The lateral cutaneous nerve of the arm has been dissected. (D) Flap shaping. The distal part of the flap is rolled onto itself to create the dorsal and ventral tongue. The proximal part is rotated 90 to 180 degrees and is used to reconstruct the floor of the mouth. (E) Flap inset. Parachuting sutures placed in the base of the tongue. Drill holes made in the mandible for suturing the lateral edge of the floor of mouth component of the flap. (F) Completed flap inset for an extended hemiglossectomy defect.

Surgical Technique

A percutaneous endoscopic gastrostomy tube is placed at the beginning of the case. Tumor extirpation is performed by the otorhinolaryngology service via a submandibular and lip-splitting incisions. On occasion, mandibulotomy is performed for better access to the base of the tongue. Depending on the extent of the tumor, resection may consist of half of the tongue (hemi-glossectomy) or 3/4th of the tongue (extended hemi-glossectomy, the adjacent floor of mouth and neck dissection. A tracheostomy is performed for airway protection.

The lateral arm flap is elevated in the supine position. The nondominant extremity is used. The elbow is flexed 90 degrees with the forearm and hand on the chest and the arm and elbow on the arm rest. The central axis of the flap is marked by drawing a line from the deltoid insertion to the lateral epicondyle. This denotes the location of the lateral intermuscular septum. The tongue and floor of mouth defects are analyzed and measured. These measurements are transposed on the lateral arm and centered on the flap axis. The proximal part of the flap is used for the floor of mouth and the distal part for the tongue defect (Fig. 1B).

The posterior incision is made first. Subfacial dissection is carried out to the lateral intermuscular septum, where the septocutaneous vessels are identified. The PRCA is identified at the base of the septum. Muscular branches going to the triceps are ligated. The PRCA is then traced back proximally between the brachialis and the long head of triceps and more proximally between the deltoid and the long head of the triceps up to the spiral groove. The pedicle length is then measured; an approximately 7 cm long pedicle is typically needed for anastomosis to neck vessels. If the pedicle is short, one of 2 strategies can be employed:

1. The flap design can be simply shifted down onto the forearm, also known as the extended lateral arm flap (ELAF), or

2. the pedicle can be dissected more proximally in the spiral grove of the humerus to gain extra length.

However, care must be taken to not cause injury to the radial nerve as it also lies in the spiral groove. The anterior incision is then made. Subfacial flap elevation is performed up to the intermuscular septum. A cuff of brachioradialis muscle is taken with the flap as it is attached to the septum. The lateral cutaneous nerve of the arm is preserved and dissected to its origin from the radial nerve, if sensory reinnervation is planned. The flap is then elevated distal to proximal. The distal end of the pedicle is ligated. The pedicle is released by cutting the periosteum of the humerus and elevating the septum off the bone. Pedicle is elevated to the spiral groove (or further if longer pedicle length is needed). Flap perfusion is then assessed (Fig. 1C).

Recipient vessels are prepared in the neck. Flap inset is performed before microvascular anastomosis as the flap swells after perfusion which makes inset difficult. Inset also sets the length of the pedicle so that any redundancy may be shortened and adjusted. The proximal part of the flap is used for the floor of the mouth and the distal part for the tongue tip (Fig. 1D). The inset is begun at the posterior most aspect of the defect, that is, the base of the tongue. Parachuting sutures are placed at the base of the tongue and left long. They are then passed through the flap and tightened down. The proximal most part of the flap is rotated 90 degrees and used for floor of mouth reconstruction. If the entire floor of mouth has been resected, drill holes are made in the mandible and the flap inset to the bone (Fig. 1E). The distal part of the flap is folded and inset to the create the dorsal and ventral oral tongue (Fig. 1F). Vessels are painted with ink on their superior surface for orientation, and passed medial to the mandible to the floor of the mouth. Microvascular anastomosis is then performed. Our recipient inflow vessel of choice is the superior thyroid artery as it has a good size match with the PRCA. The internal jugular system is the recipient flap drainage system of choice. Venous anastomoses of choice, in order, are end to end to common facial vein, end to side to internal jugular vein and end to end to external jugular vein. If the external jugular vein is of sufficient length to mobilize medially, a second venous anastomosis is performed. The neck is washed out thoroughly and a suction drain is placed away from the vessels.

A suture is tied to the tip of the flap and taped to the cheek to prevent the flap from falling posteriorly. Two folded sheets are placed on either side of the head to prevent the patient from turning the head. Flap monitoring is performed by checking capillary refill and pin prick. Aspirin is administered for a week. Patients are typically discharged on the 5th postoperative day.

RESULTS

A total of 8 patients underwent reconstruction with the lateral arm flap over a 3-year period. Patient data are summarized in Supplementary Digital Content, Table 1, http://links.lww.com/SCS/C550. The mean age was 46.5 years (32 – 68 years). There were 6 males and 2 females. The mean body mass index was 21.34 kg/m² (17.51 – 24.77 kg/m²). The pathology was squamous cell carcinoma in all patients. Tumor extirpation consisted of hemiglossectomy (1/2 of the tongue) and floor of mouth resection in 5 patients and extended hemi-glossectomy (3/4th of the tongue) and floor of mouth resection in 3 patients. A standard lateral arm flap was harvested in 3 patients and an ELAF in 5 patients. The average size of the flap was 65.75 cm² (48– 76 cm²). The size of the flap for the tongue component ranged from 7 to 10 cm (length) by 4 to 6 cm (width) and that for the floor of mouth component ranged from 4 to 6 cm (length) by 3 to 4 cm (width). The recipient artery was the superior thyroid in all cases. The recipient veins were the common facial vein in 1 patient, the internal jugular in 3 patients, the external jugular in 1 patient and both external and internal jugular in 3 patients. All donor sites were closed primarily.

Patient outcomes are summarized in Supplementary Digital Content, Table 2, http://links.lww.com/SCS/C550. All flaps healed without any total or partial flap losses. Orocutaneous fistulas developed in 3 patients. These fistulas manifested as neck erythema and drainage. They occurred due to break down of suture line anteriorly at junction of flap, mandible and floor of the mouth. Conservative management was undertaken, consisting of saline gargles, mouth wash and oral restriction to only water intake. This resulted in healing of the fistulas. All donor sites healed without any complications. All patients received adjuvant radiation.

Follow up duration was a mean of 16 months (4–41 months). One patient succumbed to recurrent disease 4 months after surgery and thus functional outcome could not be clearly assessed. All patients reported that they were happy with the appearance of the donor site. All patients had anesthesia in the lateral forearm in the territory of the posterior antebrachial cutaneous nerve. However, they all reported that this was not bothersome. Tongue movements were assessed by asking patients to protrude the tongue as far anteriorly as possible; 2 patients were able to protrude tongue up to the lips, 3 were able to protrude up to the incisors, and 3 were able to move the tongue but not touch the incisors. Speech was found to be intelligible to patients’ family in 4 patients and to strangers in 3 patients. All patients resumed an oral diet, with 1 patient resuming an unrestricted diet with solid foods, 4 taking soft foods and 2 patients taking a pureed diet. Gastrostomy tubes were removed at an average of 136 days (101–155 days).

DISCUSSION

Tongue resections can be classified as partial, hemi, extended hemi, and total glossectomies.⁶ Tumor ablation may include the adjacent floor of mouth, mandible/vestibule, pharynx, and palate. The movements of the tongue are extremely complex, controlled by 4 intrinsic and 4 extrinsic muscles, arranged in an intricate three-dimensional arrangement.⁷ These complicated movements cannot be replicated with current reconstructive techniques. The goal of reconstruction is to restore the lost volume of the tongue while maintaining its remaining mobility. In a hemiglossectomy, there is usually sufficient muscle left behind to move the tongue. Reconstruction is performed with a skin flap that is thin and pliable so that excessive volume of tissue does not block the airway, obstruct movement of food bolus or impair articulation. Furthermore, it is critical that the tongue not be weighed down or tethered so as to maintain its finger function, which is important for feeding, speech and oral hygiene. The workhorse flap for hemi-tongue reconstruction is the radial forearm flap. It is thin and pliable, has consistent anatomy, reliable perfusion, and a long vascular pedicle, making it an excellent choice for tongue reconstruction. Its major drawback is the unfavorable donor site. Flap harvest necessitates sacrifice of a major inflow vessel to the hand. An incomplete deep palmar arch is present is approximately 5% of individuals, precluding the use of this flap.⁸ The donor site typically cannot be closed primarily and requires a skin graft. Skin grafting over tendon then mandates immobilization in a splint, which can result in hand stiffness. If the paratenon is inadvertently injured during harvest or if immobilization is suboptimal, delayed healing of the skin graft can occur. A recent meta-analysis looking at donor site problems in radial forearm free flaps revealed an overall complications rate of 7.9%.⁹ These complications included delayed healing requiring regrafting, contractures, decrease in grip strength, lymphedema, and injury to radial sensory nerve. The anterolateral thigh flap is also a popular choice for tongue reconstruction. Its primary advantage is a very favorable donor site. A recent meta-analysis looking at tongue reconstruction with radial forearm flap and anterolateral thigh flap showed equivalent swallowing capacity and speech intelligibility but worse donor site numbness and aesthetic outcomes with radial forearm free flap.¹⁰ However, the anterolateral thigh flap may be too bulky for hemiglossectomy reconstructions, except in very thin patients. Thinning of the flap up to 4 mm has been described.^2,11 However, in an obese patient, a thinned anterolateral thigh flap is not very pliable which makes flap inset and creation of tongue contours very challenging.

The lateral arm flap was described by Song et al in 1982 and further elaborated by Kotsaros.^4,5 It is an intermediate thickness fasciocutaneous flap with sensory innervation. The triceps muscle and tendon, and humerus bone can be incorporated in the flap, although this is rarely done since there are better options for chimeric bone and muscle flaps.¹² The soft tissue thickness, flap size, and pedicle length required for hemi-tongue reconstruction makes the lateral arm flap a good option for partial tongue reconstructions. Faria et al performed 210 head and neck reconstructions with the lateral arm flap, out of which 53 were tongue reconstructions.¹³ They reported a flap elevation time of 25 to 40 minutes and a flap viability rate of 96.8%. Although all patients had paresthesia of the forearm, the most common patient complains were hyper-trophic scar and dog ear of the donor site. Thankappan et al¹⁴ performed partial glossectomy reconstruction in 48 patients with the lateral arm free flap. Flap survival rate was 93.8%. The most common donor site problem was a wide scar, seen in 75% patients. Functional outcomes were measured in 37 patients. Speech was reported to be normal or near normal in all patients. Diet was unrestricted in 75% and soft in 25% patients, whereas tongue movements were normal in 57% (able to protrude tongue out of mouth), mildly restricted in 30% (protrusion to lip) and severely restricted in 13% (protrusion to teeth). Restrictions of diet and tongue movement were associated with administration of adjuvant therapy. Reinert performed intraoral reconstruction in 25 patients with a lateral arm flap, out of which 10 underwent partial tongue reconstruction.¹⁵ He reported 1 flap loss and “no significant complications” at the donor sites, all of which were closed primarily. Oh et al¹⁶ utilized a bilobed design for reconstruction of the hemi-tongue defect, with one lobe and floor of mouth defect with the other lobe. All 13 of their patients healed satisfactorily and returned to normal diet. The donor sites were closed primarily and pedicle lengths were found to be adequate. In our series, we found a direct relation of speech outcome with tongue mobility; the more mobile the tongue, better the quality of speech (Supplementary Digital Content, Table 2, http://links.lww.com/SCS/C550). We did not find a direct relation of final diet with the extent of resection or tongue mobility. The swallowing mechanism is complex and involves the coordinated function of several oropharyngeal structures. Injury to any of these structures, from resection or radiation, can impair the swallowing mechanism.

A major advantage of the lateral arm flap is the favorable donor site. The flap does not require sacrifice of a major vessel and the donor site can be closed primarily for flaps less than approximately 6 to 7 cm.^17,18 Although larger flaps can be harvested, the donor site may not be able to be closed primarily thus losing a main benefit of this flap. Furthermore, radial nerve compression syndrome has been reported if too tight a closure is attempted.^12,19 Flap elevation is straightforward and quick. The anatomy is fairly consistent. Anatomic variations that have been described include duplication of the PRCA and intramuscular course of the posterior cutaneous nerve of the forearm through the triceps.¹³ The biggest drawback of the lateral arm flap is the short pedicle and the proximity of the pedicle to the radial nerve. The usual pedicle length is 7 to 9 cm, with an arterial diameter of 1.2 to 1.3 mm and 2 venae commitantes of 2.5 mm diameter.^12,20,21 This pedicle length is usually sufficient for tongue reconstruction. Pedicle length can be increased by extending the vascular dissection proximally up to the profundal brachii artery, which is also a larger size vessel. The average diameter of the profundal brachii artery just distal to its takeoff from the brachial artery is 2.45 mm.²² However, great care must be taken during dissection as the radial nerve runs with the profundal brachii vessels in the spiral groove. Pedicle length can also be increased by a more distal flap design beyond the lateral epicondyle and on the forearm.²³ This ELAF can have a pedicle length up to 12 cm. Extension of flap design also allows harvest of a larger flap. Another advantage of the ELAF is that the soft tissue of the forearm is thinner due to less sub-cutaneous fat. This decrease in subcutaneous fat from the arm to the forearm can be used to tailor tongue reconstructions; the thicker arm tissue can be used for the tongue base and the thinner forearm tissue for tongue tip.²⁴ Yang et al²⁵ used the ELAF for partial tongue reconstruction and reported good speech, feeding and tongue mobility outcomes. The lateral arm flap typically necessitates sacrifice of the posterior cutaneous nerve of the arm which leaves an area of anesthesia on the proximal lateral forearm. Song et al, in series of 16 patients who underwent lateral arm flap for oral cavity reconstruction, found anesthesia of the lateral elbow and forearm in all patients.²⁰ However, they reported that the numbness faded over time and was “rarely noticed” by patients. Fogdestam et al¹⁸ reported that the posterior cutaneous nerve of the forearm could be preserved in more than half of flap harvests. Graham et al²⁶ performed 123 lateral arm flaps for reconstruction of extremity wounds and reported lateral epicondylar pain in almost 20% of patients. One major disadvantage of this flap in head and neck reconstruction is that simultaneous harvest of the flap at the time of tumor ablation, although possible, may be difficult due to the proximity of the arm to the head and neck area. Our study has several limitations. This is a small cohort of patients without a comparison group. Long term functional evaluation could not be performed on everyone as some patients succumbed to their disease and others did not return for long term follow up. It also has selection bias, which is inherent in nonrandomized studies. The ideal candidate for the lateral arm flap is a patient with normal body mass index and some soft tissue laxity of the arm. Obese patients have thick arms that are not suitable for partial tongue reconstruction with lateral arm flap. We have also found the superior thyroid artery to be a good size match for the PRCA. Facial and lingual arteries, because of their larger diameter, may make microvascular anastomosis more difficult. We did not perform sensory neurotization in our patients. Although it has been shown that the sensory recovery in innervated radial forearm and anterolateral thigh flap tongue reconstructions is superior to non-innervated flaps, there is paucity of such data for lateral arm flap tongue reconstruction.^25,27 Further research needs to be done to evaluate sensory recovery of the neo-tongue in innervated lateral arm flaps and if sensory recovery has an impact on overall function.

CONCLUSIONS

The lateral arm flap is a good option for partial tongue and floor of mouth reconstruction in the appropriate candidate. Its advantages are consistent vascular anatomy, straightforward harvest, obviation of major vessel sacrifice and ability to close the donor site primarily. Its disadvantages are smaller pedicle size, proximity of pedicle to the radial nerve, numbness of the forearm and difficult to harvest at the same time as tumor resection in a two-team approach. It is a useful flap for reconstructive surgeons to have in their reconstructive toolbox.