The submental flap: Be wary

Abstract The submental island flap (SIF) is as an alternative to free flaps in head and neck reconstruction. 10 patients underwent submental flaps. All ten patients suffered failure of SIF as the definitive reconstructive procedure. Despite comparing favorably to free tissue transfer in published reports, our SIF had high failure rate.


| INTRODUCTION
Oral cavity cancer is the sixth most common cancer worldwide, comprising 30% of all H&N cancers. 1 Surgery is considered the gold standard to achieve tumor control. 2 The traditional surgery has been the use of microvascular free flaps, albeit at higher costs, and hospital length of stay than alternatives. Other head and neck resections result in tissue deficits, such as after parotidectomy. The submental island flap (SIF) has emerged as an alternative over the more costly and lengthy free flap, for oral cavity and other head and neck defects.
The first description of the SIF was by Martin and colleagues in 1993. 3 They described it as a reliable alternative to the more traditional free flap procedure and as a means of reconstruction after oncologic procedures of the head and neck. 4 Arising deep to the submandibular gland, the submental artery, a reliable branch of the facial artery, is the main contributor to the SIF. [4][5][6] At the mylohyoid, the artery either continues deep (70%) or superficial (30%) to the anterior belly of the digastric muscle, terminating at the mandibular symphysis. 6,7 Up to 4 cutaneous perforators have been described in other SIF studies 7 ; however, cadaveric studies have only been able to consistently find 1 reliable perforator to supply the skin paddle. 8 Also, the submental vein has been found to be the primary vessel for venous drainage of the SIF. 9 With decreased hospitalizations and shorter operating times, the SIF transformed into a plausible and effective alternative to traditional free flap reconstruction. 5 However, even with its relatively consistent anatomy and large number of positive surgical outcomes, the SIF has had its fair share of complications.
This study seeks to analyze and explain the various post-op complications surrounding submental island flaps. It will estimate the potential impact submental flaps may have relative to the traditional method of using free flap reconstruction, once post-op complications are taken into consideration. It will also serve as a warning to new reconstructive surgeons who consider using a submental flap, as opposed to the more traditional free flap.

| METHODS AND MATERIALS
This retrospective case series with chart review includes 10 consecutive patients that underwent SIF reconstruction following various head and neck procedures by 2 different physicians at a single care facility between November 2016 and April 2018. These surgeons were newly out of fellowship training and embarking on their first attending surgeon employment. Inclusion criteria were adults with a diagnosis of malignancy of the head and neck undergoing surgery with reconstruction using SIF that then went on to fail SIF reconstruction. Demographics and preoperative risks were collected. Data were gathered regarding the type of procedure performed. Postoperative variables and wound dehiscence were recorded.

| RESULTS
Ten total patients underwent submental flaps between 2016 and 2018 (Table 1). Five were female, and 5 were male. Age of patients ranged from 33 to 85, with an average age 60.7 years. Only 2 patients were smokers. Four patients had hypertension, and one had diabetes. Six of the patients had no comorbidities. Nine of the patients had simultaneous neck dissection. None of the patients had prior chemotherapy or radiation. The defects requiring reconstruction were widely varied.
All ten patients suffered failure of the SIF as the definitive reconstruction. Eight of ten patients required a second procedure in the operating room. Three of ten patients received an intraoperative change in reconstruction plan, aborting the SIF during the initial procedure. Patient 1, a 61-year-old woman and tobacco user, had her SIF reconstruction aborted due to pathologic nodes in the submental area which were not present on her preoperative imaging studies (PET and CT). She subsequently underwent a free flap reconstruction the following day. Patient 5, a 62-year-old woman with hypertension, had her SIF aborted because of poor venous flow. Instead, she had an acellular dermal matrix allograft placed. Patient 8, 53-year-old man with no comorbidities, had his SIF reconstruction aborted due to poor blood supply to the island graft. He then underwent a split-thickness skin graft (STSG) the same day.
Six of the 10 patients had initial placement of the SIF and further debridement at a second OR sitting. Three of those had venous congestion, and 3 of those were due to necrosis from poor arterial supply. One patient noted survival of a portion of the SIF for a parotid defect but needed a subsequent cervicofacial rotation flap for closure of the remaining defect.

| DISCUSSION
Although the submental flap is relatively thin, easy-toharvest, and typically well-vascularized, it does have complications. 10   All 10 of our patients had disease-free necks preoperatively, with stage N0 based on preoperative CT and/or PET. Nine of our 10 patients required simultaneous neck dissection. When a neck dissection is needed during a procedure where SIF is planned, the reconstructive surgeon should have a careful discussion with the resecting head and neck surgeon so that the facial artery or vein is not ligated during the neck dissection. In the circumstance that the vein or artery is injured, using that side of the neck for the SIF is not recommended, and the submental flap should be based on the contralateral side.
Three of the patients were noted to have venous congestion, requiring second trip to the OR. The submental vein has been found to be the primary venous drainage of the flap, but in one of our cases, the submental vein was noted to drain into the external jugular system during the bring-back procedure. The external jugular system had been ligated during the initial procedure. Perhaps this could have been avoided with an earlier identification of the anatomy. A different mode of reconstruction could have been undertaken during the initial procedure.
Three of the patients were noted to have necrotic SIF from lack of blood supply. Studies have shown only one reliable perforator of the SIF, which is much smaller than the perforators of the work-horse anterolateral thigh free flaps and radial forearm free flaps. The size difference for vessel handling can be a potential technical challenge.
Our poor SIF results were independent of the defect site. We used SIF for soft tissue defects resulting from composite resection of mandible/tongue/floor of mouth; as well as defects of oral tongue; retromolar trigone; buccal mucosa; floor of mouth; and parotid. Sittitrai and colleagues concluded that the SIF is reliable, is suitable for oral tongue reconstruction, and had a lower complication incidence when compared to the radial forearm free flap. 14 While there is an abundance of support for free flaps and the success is >95% in the literature, there are also as much positive data on the SIF.
Our reconstructive surgeons had a greater than 90% survival rate when performing free flaps and a 0% success rate when using SIF. Thus, technical and training factors were examined. In typical head and neck surgical oncology fellowships, free flaps are a far more common form of reconstruction than SIF. Regenbogen and several others have acknowledged that commonly recommended interventions, like restricting high-complexity operations to experienced surgeons and additional trainings for inexperienced surgeons would lead to an improvement in outcomes. 15 In our hands, free tissue transfers have superior outcomes compared with pedicled flaps due to our experience and knowledge of free flaps and our deficiency of these with SIF.
Studies have shown less cost with pedicled flaps than free flaps. However, 70% of our SIF patients required second trips to the operating room during their initial stay, compared with 5% of our free flap patients. In our hands, patients who underwent SIF did not experience decreased cost, length of stay, and operative time compared with free flaps, as reported in other series. 14, [16][17][18] As Zhou and colleagues reiterate, in regard to intraoperative factors, surgical technique is regarded as the most important component of free flap success. 19,20 In their very own study, Zhou had two surgeons perform the microvascular anastomoses in the free flap reconstructions for his study, with each having been in practice 5 years or more. 21,22 Such experiential and technical rigor has not been analyzed in SIF outcome literature, perhaps because SIF is may be viewed as a simpler procedure. Three of our ten patients received an intraoperative change in reconstruction plan, abandoning the SIF during the initial procedure. One patient was not reconstructed that day and underwent a free flap reconstruction the following day. A second patient had an acellular dermal matrix allograft placed the same day. A third patient underwent a split-thickness skin graft (STSG) the same day. In retrospect, given our team's great success with free flap reconstruction and our dismal success with SIF, we should be prepared for free flap reconstruction in any patient planned for SIF on the same day as their resection. This can likely avoid return to the operating room for a second procedure.

| CONCLUSION
Despite their promising potential and numerous cases with positive outcome results, submental flaps are still open to failure. In our series, surgeon experience may have been more influential in outcomes than patient factors such as age, type of defect, history of radiation, tobacco use, or other comorbidities. In our institution, a free flap is preferable and has lower risk of complications compared with a pedicle flap, and the pedicled flap is the backup secondary option when a free flap is lost.