Otology & Neurotology 36:879Y885 Ó 2015, Otology & Neurotology, Inc.
The Submental Island Flap for Reconstruction of Temporal Bone Defects *Craig Miller, †‡Jason Charles Hanley, ‡Thomas J. Gernon, ‡Audrey Erman, and †‡§Abraham Jacob *Department of Otolaryngology, The University of Arizona, Tucson; ÞThe University of Arizona Ear Institute, Tucson; þDepartment of SurgeryVDivision of Otolaryngology, The University of Arizona, Tucson; and §The University of Arizona Cancer Center and The University of Arizona Bio5 Institute, Tucson, Arizona, U.S.A.
Objective: Untreated cutaneous malignancies involving the lateral aspect of the cranium often invade the temporal bone, necessitating a resection of this site. The reconstruction of the associated complex defect typically requires a reconstructive flap placement to obliterate the resection cavity and provide an aesthetically pleasing restoration. We performed a retrospective case review of 30 patients undergoing temporal bone resection and reconstruction with a submental island flap (SIF), free flap, or temporalis rotation flap. We sought to evaluate the benefit of the submental island flap over the other reconstructive options in terms of cost benefit, patient aesthetic satisfaction, complications, morbidity, and duration of hospitalization. Setting: Tertiary referral center. Patients: Patients who underwent temporal bone resection requiring reconstruction. Intervention(s): Therapeutic. Main Outcome Measure(s): Main outcome measures included time to functional recovery, patient satisfaction, and hospital stay.
Results: In total, 30 patients were included in this study. Twenty-three patients received a SIF, three underwent a radial forearm free flap, two underwent a temporalis rotation flap, one received a sternocleidomastoid flap, and one received a myocutaneous flap. Average ICU stay after surgery was under 2 days for non-SIF patients. No SIF patients spent time in the ICU nor were there complications reported in this group. Patients who underwent SIF showed a quicker functional recovery, increased satisfaction with appearance of reconstruction, and improved cosmetic results. Conclusions: Submental island flap reconstruction is an appealing option for the reconstruction of temporal bone defects. This technique offers decreased length of ICU stays, increased patient satisfaction, and decreased complication rates compared with other reconstructive techniques. Key Words: Head and neck cancerVMyocutaneous flapVReconstructionVSubmental anatomyVSubmental arteryVSubmental island flapVSubmental nerveVTemporal boneVTemporal bone resection. Otol Neurotol 36:879Y885, 2015.
Cutaneous malignancies involving the lateral cranium sometimes invade the temporal bone, requiring substantial soft tissue and bone resection for extirpating disease and establishing clear margins. The resulting complex defect requires reconstructive techniques that obliterate
dead space within the resection bed and provide a reasonable cosmetic result. In recent years, microsurgical procedures for the reconstruction of craniofacial defects have become popular. Forearm free flaps or other free fasciocutaneous tissue transfers offer thin and pliable tissues with a high degree of versatility in their design (1). However, free tissue transfers from distant sites have a number of disadvantages, including difficulty in matching facial skin cosmetically as well as the potential for failure of the microvascular anastomosis, longer operative times, and increased hospital costs related to ICU stay. Ideally, soft tissue used for reconstruction should be reliable, functionally and cosmetically acceptable, sufficient in size/ volume, cause minimal donor-site morbidity, and closely resemble the recipient site for color, texture, and thickness.
Address correspondence and reprint requests to Abraham Jacob, M.D., The University of Arizona Ear Institute, The University of Arizona Department of Surgery, The University of Arizona Cancer Center, and The University of Arizona Bio5 Institute, Tucson, Arizona, U.S.A.; E-mail: [email protected] Research supported by the Medical Student Research Program, The University of Arizona College of Medicine, Bio5, and The University of Arizona Cancer Center. Funding received from the National Institute on Deafness and Other Communication Disorders [Aj-k08 dc009644- 01a1]. The authors disclose no conflicts of interest.
879
Copyright © 2015 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
880
C. MILLER ET AL.
Locoregional flaps provide an excellent alternative to free flaps for repair of head and neck defects. The submental island flap, first described by Martin and colleagues (2), provides a reliable and cosmetically acceptable option for head and neck reconstruction. This flap, now increasing in popularity, provides more than adequate color and texture match to facial skin, is supple, and has a concealed donor-site incision that can be closed primarily. In older patients with redundant soft tissues in the submental region, this flap can have substantial cosmetic benefit. From a technical standpoint, flap harvest is straightforward, and the submental island flap can be lifted as a fasciocutaneous, myofascial, myocutaneous, or osteocutaneous flap. Additionally, because of increased skin laxity in older patients as well as the large vertical dimension available for safe flap harvest, a single submental artery can sustain a skin territory of up to 10 16 cm; this allows for reconstruction of large defects (3). Invasive skin cancer requiring temporal bone resection as part of a patient’s surgical management is epidemic in the desert Southwest. Therefore, our group has performed an unusually large number of these resections in a relatively short period of time and has used the submental island flap as our preferred reconstructive technique. Here, we describe the harvest and inset of these flaps for ear and periauricular defects by performing a review of patients’ presentations, operative reports, and follow-up. MATERIALS AND METHODS This was a single institution IRB-approved (IRB# HR11037) retrospective analysis of patients undergoing temporal bone resection and subsequent reconstruction between the dates of October 2011 and December 2013. A variety of reconstruction methods were used including submental island flap, radial forearm free flap, temporalis rotation flap, sternocleidomastoid (SCM) flap, and myocutaneous pectoralis major flap. The focus of the current work is the submental island flap, which is described in detail. Patients who underwent temporal bone resection without flap reconstruction were excluded from this study. Preoperative examination and diagnosis, surgical procedure, hospital stay (including duration of ICU stay), cost, follow-up, and outcomes were evaluated. A literature review regarding submental island flaps was also performed.
RESULTS Anatomy The submental artery, a branch of the anterior facial artery, vascularizes the submental island flap. This vessel reliably branches just below the mandible and courses anteriorly in a groove on the superomedial aspect of the submandibular gland, supplying branches to this structure. As it continues anteriorly, it lies on the mylohyoid, feeding it as well as the platysma and the periosteum of the mandible (4). The major outflow to the flap is the submental vein; frequently, however, a secondary venous drainage pattern to the anterior jugular vein will be found. Frequent anatomic variations in the venous pedicle have
been noted (4). At the level of the submandibular gland, the submental artery and vein course together, but just at the anterior aspect of the gland, the submental vein diverges laterally from the artery. Here, it traverses the gland and joins the facial vein at the posterolateral superior aspect of the gland. Once joined, the facial vein drains into the common facial vein and then into the internal jugular vein (IJV). As reported by others (4), the authors have found frequent anatomic variations in the venous pedicle. Therefore, it is essential to map the venous drainage before sacrificing any venous structures. Variations noted for this review include 1) the submental vein coursing from the anterior-medial aspect of the gland caudally over the anterior aspect of the gland and bypassing the facial vein for direct insertion to the IJV and 2) a direct connection of the vein to the external jugular vein through posterior passage across the lateral aspect of the submandibular gland. Many patients needing this flap require a neck dissection; therefore, all venous structures should be preserved unless directly involved with tumor. It is critical to identify and secure the venous pedicle early in the dissection. Surgical Technique Temporal bone resections are performed by a single neurotologist (A.J.) at our institution. Care is taken to preserve perforators, local veins, and the facial nerve when possible, while undertaking an appropriate oncologic resection. Incision planning is critical, as many skin cancers that secondarily involve the temporal bone require resection of all or parts of the pinna along with resection of cheek skin anterior to the ear. The incision must also facilitate parotidectomy or neck dissection when required and allow harvest/inset of the submental flap. A broad pedicle to the pinna is required to prevent vascular compromise. Typically, a large C-shaped incision extending into a neck crease is used. The ear canal is transected when appropriate or if portions of the pinna are involved with tumor, a partial auriculectomy is performed using conchal cuts that leave tumor tissue connected to the underlying temporal bone resection. Other incision options include a lazy-S postauricular incision that curves posteriorly into the temporal hairline or a preauricular incision that curves posteriorly above the pinna. In the latter case, the canal is transected and the pinna is raised as part of a posteriorly based cervical flap. Biopsies are taken to confirm malignancy as well as clear soft tissue margins. A complete, intact-canal mastoidectomy is then performed. All the usual landmarks are identified; the facial nerve is found at the external genu and traced inferiorly. Often, it is decompressed at the stylomastoid foramen to facilitate dissection into the parotid bed. The mastoid tip is removed. Next, an epitympanic dissection is carried to the soft tissues of the temporomandibular joint (TMJ). An extended facial recess approach is then performed. The incus-stapes joint is separated, the incus is removed, the head of malleus resected, and the tensor tympani tendon is cut. The fibrous annulus of the tympanic
Otology & Neurotology, Vol. 36, No. 5, 2015
Copyright © 2015 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
SUBMENTAL ISLAND FLAP FOR TEMPORAL BONE DEFECTS membrane is identified and traced inferiorly and then anteriorly. A chisel is used to crack away ear canal bone deep to the tympanic annulus but lateral to the petrous carotid artery. This leaves the anterior bony ear canal wall tethered only to soft tissues of the TMJ and parotid gland. The wound is then copiously irrigated, mucosa is removed from the middle ear, and the Eustachian tube orifice is packed with bone wax, Surgicel, and muscle. Two head and neck surgeons (A.E. and T.J.G.) remove tumor-involved soft tissues, in continuity with the underlying temporal bone resection specimen. Parotidectomy and neck dissections are performed if needed. Once all ablative portions of the operation are completed, the submental artery island flap is harvested and inset (Figs. 1 and 2). Before harvest, the approximate positions of the facial artery, vein, and the marginal mandibular branch of the facial nerve are identified. The superior skin paddle incision borders the inferior edge of the mandible and the maximal dimensions of the flap can extend from the ipsilateral angle of the mandible to the contralateral angle of the mandible.
881
The superior incision is performed on the undersurface of the mandible but made contiguous to the ipsilateral neck dissection when required. The skin and subcutaneous tissue are elevated from the inferior surface of the mandible up to the attachment of anterior belly of the digastric muscle. Next, the mylohyoid muscle is identified. Care is taken laterally to identify and preserve the marginal mandibular branch of the facial nerve. This may already have been done if a parotidectomy was performed to extirpate disease. The facial vessels are found as they course over the mandible. Next, the inferior skin incision is performed at approximately the level of the hyoid bone. Subplatysmal dissection is performed until the intermediate tendon of the digastric muscle is found. Dissection proceeds laterally until the submandibular gland and lateral border of the mylohyoid muscle are identified. The submandibular gland is removed; this is imperative as the submental artery and vein course over the superior aspect of the submandibular gland often supplying small branches to the gland. These branches should be clipped or ligated but the integrity of the submandibular artery and vein should
FIG. 1. Schematic diagram of the submental island flap and its pedicled vessels before dissection (A). The flap is elevated and rotated to show the orientation of muscle fibers as well as orientation of feeder vessels (B). The flap is inset to the temporal bone defect (C). Rostral aspect of flap is elevated. Otology & Neurotology, Vol. 36, No. 5, 2015
Copyright © 2015 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
882
C. MILLER ET AL.
FIG. 2. Elevation and placement of the submental island flap in a patient. A demonstrates flap elevation after resection of the primary tumor. B demonstrates the temporal bone defect (black arrow). The flap is rotated and placed in the ablative defect with pedicled vessels shown (black arrow; C). Flap is inset, and the wound is closed (D). Two Jackson Pratt drains are placed to prevent fluid accumulation.
not be compromised. It is our preference not to dissect the submental artery and vein to its origin from the facial artery, as these vessels can be small and easily damaged. However, we do perform dissection of the more proximal facial artery and vein as they course under or over the digastric muscle, respectively. At this point, the skin paddle and associated musculature are elevated. The insertion of the ipsilateral anterior belly of the digastric muscle is transected from its attachment to the mandible. One nuance in elevating this flap is that we prefer to harvest the ipsilateral mylohyoid muscle as well. This protects the submental vessels where the anterior belly of the digastric muscle and the mylohyoid muscle encloses them. It also adds bulk to the flap for obliteration of dead space within the temporal bone resection. Inferiorly, the intermediate tendon of the digastric muscle is harvested along with the mylohyoid muscle. Once the ipsilateral muscle attachments have been transected, the flap is elevated from distal to proximal (from the contralateral neck towards the ipsilateral neck). The skin paddle is elevated over the submandibular gland fascia and superficial to the contralateral anterior belly of the digastric muscle. Care should be taken when elevating the skin in the contralateral neck so as not to damage the marginal mandibular branch of the facial nerve. As the elevation proceeds medial to the contralateral anterior belly of the digastric muscle, the dissection plane depends on whether the mylohyoid muscle and ipsilateral anterior belly of the digastric muscle is being included in the flap’s elevation. The submental artery island flap now remains attached only by its vascular pedicle and fascia attach-
ments. The submental artery and vein can be seen, and branches that tether the flap’s mobilization (such as branches to the mylohyoid muscle) are transected. Typically, the distal facial artery and vein are left intact where they course over the mandible; however, they may be ligated if a further arc of rotation is needed. The submental artery island flap is then rotated into the ablative defect. Closure of the donor-site defect requires wide undermining of the lower anterior neck skin. This skin is advanced superiorly to close the defect. The undersurface of the flap should be sewn to the periosteum of the hyoid bone to recreate the cervicomental angle. The neck may need to be flexed to achieve complete closure of the neck skin. Outcomes In total, 17 patients undergoing temporal bone resections between the dates of October 2011 and December 2013 were included in this study. The average age was 73 years, and Table 1 delineates patient specifics such as diagnoses, defect size, flap type, ICU stay, and total cost of hospitalization. Nine patients received a submental island flap(SIF), two underwent a temporalis rotation flap, one underwent a radial forearm free flap, four received a sternocleidomastoid (SCM) flap (one patient required both a SIF and an SCM flap), and one received a myocutaneous pectoralis major flap. Average ICU stay for locoregional flap patients was less than 24 hours; the free flap patient had the longest ICU stay. Overall, six patients did require
Otology & Neurotology, Vol. 36, No. 5, 2015
Copyright © 2015 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
SUBMENTAL ISLAND FLAP FOR TEMPORAL BONE DEFECTS TABLE 1. Defect size (1 = Small, 2 = Medium, Type of Flap Name Gender Age 3 = Large) Free flap
HS
M
65
2
SMI
DB
M
85
3
FS
M
90
3
JG
M
81
Not in op note
CF
F
79
Not in op note
WS
M
75
2
KM
M
70
2
DS
M
56
DL
M
82
Not in op note 2
PC
M
83
GC SP
M M
66 66
AV
F
47
JG
M
81
CT RL LB
M M M
78 75 61
GC
M
65
AA
M
75
Rotational
Not in op note Not in op note Not in op note Not in op note 1 2 2 Not in op note Not in op note
883
Patient demographics
Tumor Size (cm)
Neck Dissection
Parotidectomy?
Diagnosis
ICU Stay Hospital Total (d) Stay (d) Charges
3.1 3.7 3.7 Left selective neck dissection (1Y4) 8.3 6.2 8 Right modified radical neck dissection 4 2.5 1 Supraomohyoid neck dissection
N
SCC of left ear and left parotid gland
4
8
179,035.30
Y
0
6
125,838.30
1
4
104,935.04
5.5 5.1 1.8 Right selective neck dissection (1Y3) 2.5 No
Y
0
8
130,826.32
0
8
114,051.27
1
4
101,439.21
1
3
88,339.90
0
7
99,643.37
4
85,077.15
43
No
Y
SCC right auditory canal and parotid gland SCC of the right ear, right temporal bone, and right parotid SCC of right auricle and right temporal bone BCC of left ear with involvement of the concha SCC of tempora bone (CNVII paralysis) SCC left cheek, auricle, and lateral skull base (CNVII paralysis) SCC of right ear
2.5 2.1 2
Y
SCC of left ear canal
G1
Left modified radical neck dissection No
N
0
5
95,735.82
2 1.4 Unavailable
No No
N N
0 1
1 3
60,570.95 53,887.50
Unavailable
No
N
BCC of left auditory canal Basal cell carcinoma Cholesteatoma right; chronic petrousitis Cholesteatoma
0
0
40,323.05
2
No
Y
SCC
0
9
127,956.97
1 23 Unavailable
No No No
N Y Y
0 0 1
4 6 4
75,912.87 95,299.79 110,044.99
1.4 0.6
No
N
0
1
55,230.80
Unavailable
No
N
BCC of the right ear SCC of left helix SCC of left ear and parotid gland SCC of right auditory canal Osteoradionecrosis of left temporal bone
0
0
46,626.73
2.9 2.4 2.8 Right posterior neck dissection 2.8 2.7 3 Left modified neck dissection
overnight observation in the ICU because of extent of resection, use of multiple drains, medical comorbidities, or poor cardiovascular/respiratory status compromising extubation in the operating room. Average hospital stay was approximately 5 days. There were no flap failures and cosmetic outcomes for the submental island flap were excellent (Fig. 3). Patients were followed after hospitalization by two head and neck microvascular surgeons (A.E./T.J.G.); those who underwent SIF showed a quicker functional recovery and required less outpatient postoperative wound care translating to fewer office visits the first 6 to 8 weeks after surgery. Cost analysis performed using documented facility charges at our institution showed that, on average, patients undergoing submental island flap reconstruction had average cost savings of greater than 40%
Y
N Y Y
compared to the free flap patient. Flap related operative time was less than 50% for all submental island flap elevations/ insets as compared to our free flap patient; however, the primary determinants of cost were length of ICU stay and length of hospitalization. Therefore, we believe that future analysis using larger cohorts of patients and stratifying by presurgical comorbidities could help better define the cost advantages of using the submental island flap.
DISCUSSION This study presents the largest published series of submental island flap reconstructions for patients with temporal bone defects and demonstrates that this flap is Otology & Neurotology, Vol. 36, No. 5, 2015
Copyright © 2015 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
884
C. MILLER ET AL.
FIG. 3. Healed left-sided submental island flap patient 1 year after surgery. Four views demonstrating excellent cosmetic outcome with minimal depression of the pinna.
quick, reliable, and has acceptable cosmetic and functional outcomes. Martin and colleagues first described the flap in 1993. The flap design and technique was investigated on 20 fresh cadavers as well as eight patients undergoing radical neck dissection. With a primary goal of reconstructing orofacial defects, Martin described this flap’s long, reliable pedicle, large cutaneous dimensions, excellent skin color match, and wide arc of rotation. In this investigation, the authors were able to extend the flap to the entire ipsilateral face, with the exception of the superior aspect of the forehead. The original eight patients described utilized this mechanism for reconstruction of both trauma and cancer ablation of the cheek, orbit, auricular region, and temporo-orbitofrontal region. Notably, no patients in this series had defects of the temporal bone nor did the article describe elevation or inset of this flap as an option for these specific defects. However, the authors noted the submental island flap’s overall versatility. More recently, Hayden and colleagues described the use of a variant of the submental flap at their institution between 2002 and 2012. Along with a detailed description of the anatomy and surgical technique, the authors described a variant of the flap, namely the hybrid pedicled flap. This flap is partially pedicled but also utilizes a draining vein close to the recipient site with reanastomosis to the submental or facial vein. In their review, they note successful reanastomosis with the external jugular, retromandibular, common facial, and internal jugular veins. This technique allows for even greater arc or rotation because of the ability to release the tethering by a short outflow pedicle. Their investigation looked at the utility of the flap for a broad range
of facial reconstructions including three cases in the auriculo-temporal region. Unlike microvascular free flaps, the submental island flap is a pedicled flap and decreased flap related operative time by over 50% in our series compared to performing a radial forearm free flap. Furthermore, there was no need for frequent postoperative flap checks and no flap failures during the postoperative period. Also, fewer wound surveillance visits were required immediately after discharge. The submental flap’s skin paddle is large and allows coverage of large parotid-bed defects as well as both partial and total auriculectomy sites. The bulk provided by harvesting the mylohyoid muscle along with this flap allows obliteration of dead space within the largest of temporal bone defects. Our investigation into the cost benefit of utilizing the submental island flap over free flaps found greater than 40% reduction in hospital charges. The free flap patient in this series had the longest ICU stay, providing one explanation for increased costs as validated by studies with published costs for ICU stays (5). The best candidates for this flap are those patients with excess skin laxity along the midline neck and substantial adipose tissue in the submental region. Less than ideal candidates lack neck skin laxity and do not have adequate soft tissue bulk. The latter may result in an inability to close the ablative defect, difficulty with closure of the donor site, and potential dysphagia caused by difficulty in neck extension from soft tissue tethering. Flaps of inadequate bulk can lead to significant depression at the ablative site and poor cosmesis; however, in these patients, abdominal fat grafting to bolster the underside of the flap may be a viable solution. Extremely poor candidates are patients who have undergone previous level
Otology & Neurotology, Vol. 36, No. 5, 2015
Copyright © 2015 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
SUBMENTAL ISLAND FLAP FOR TEMPORAL BONE DEFECTS one neck dissection. In patients with small auricular defects, the submental island flap may be more than what is needed for reconstruction. For these patients with tumors localized to the external auditory canal and ablative sites without large auriculectomy defects, we prefer to use a temporalis muscle rotation flap and abdominal fat placed within the temporal bone defect (deep to the muscle). The posterior aspect of the temporalis muscle is harvested, and these patients recover from surgery without any of the donor-site complications. CONCLUSION Invasive skin cancer requiring temporal bone resection as part of a patient’s surgical management is epidemic in the desert Southwest. As a result, our group has performed an unusually large number of these resections in a relatively short period of time. In appropriately chosen patients with
885
skin and soft tissue laxity in the midline neck along with adequate underlying bulk, the pedicled submental island flap provides a quick, reliable, less costly, and cosmetically acceptable option for reconstruction. REFERENCES 1. Merten SL, Jiang RP, Caminer D. The submental artery island flap for head and neck reconstruction. ANZ J Surg 2002;72:121Y4. 2. Martin D, Pascal JF, Baudet J, et al. The submental island flap: a new donor site. Anatomy and clinical applications as a free or pedicled flap. Plast Reconstr Surg 1993;92:867Y73. 3. Faltaous AA, Yetman RJ. The submental artery flap: an anatomic study. Plast Reconstr Surg 1996;97:56Y60; discussion 61Y2. 4. Hayden RE, Nagel TH, Donald CB. Hybrid submental flaps for reconstruction in the head and neck: part pedicled, part free. Laryngoscope 2014;124:637Y41. 5. Dasta JF, McLaughlin TP, Mody SH, Piech CT. Daily cost of an intensive care unit day: the contribution of mechanical ventilation. Crit Care Med 2005;33:1266Y71.
Otology & Neurotology, Vol. 36, No. 5, 2015
Copyright © 2015 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
The Submental Island Flap for Reconstruction of Temporal Bone Defects *Craig Miller, †‡Jason Charles Hanley, ‡Thomas J. Gernon, ‡Audrey Erman, and †‡§Abraham Jacob *Department of Otolaryngology, The University of Arizona, Tucson; ÞThe University of Arizona Ear Institute, Tucson; þDepartment of SurgeryVDivision of Otolaryngology, The University of Arizona, Tucson; and §The University of Arizona Cancer Center and The University of Arizona Bio5 Institute, Tucson, Arizona, U.S.A.
Objective: Untreated cutaneous malignancies involving the lateral aspect of the cranium often invade the temporal bone, necessitating a resection of this site. The reconstruction of the associated complex defect typically requires a reconstructive flap placement to obliterate the resection cavity and provide an aesthetically pleasing restoration. We performed a retrospective case review of 30 patients undergoing temporal bone resection and reconstruction with a submental island flap (SIF), free flap, or temporalis rotation flap. We sought to evaluate the benefit of the submental island flap over the other reconstructive options in terms of cost benefit, patient aesthetic satisfaction, complications, morbidity, and duration of hospitalization. Setting: Tertiary referral center. Patients: Patients who underwent temporal bone resection requiring reconstruction. Intervention(s): Therapeutic. Main Outcome Measure(s): Main outcome measures included time to functional recovery, patient satisfaction, and hospital stay.
Results: In total, 30 patients were included in this study. Twenty-three patients received a SIF, three underwent a radial forearm free flap, two underwent a temporalis rotation flap, one received a sternocleidomastoid flap, and one received a myocutaneous flap. Average ICU stay after surgery was under 2 days for non-SIF patients. No SIF patients spent time in the ICU nor were there complications reported in this group. Patients who underwent SIF showed a quicker functional recovery, increased satisfaction with appearance of reconstruction, and improved cosmetic results. Conclusions: Submental island flap reconstruction is an appealing option for the reconstruction of temporal bone defects. This technique offers decreased length of ICU stays, increased patient satisfaction, and decreased complication rates compared with other reconstructive techniques. Key Words: Head and neck cancerVMyocutaneous flapVReconstructionVSubmental anatomyVSubmental arteryVSubmental island flapVSubmental nerveVTemporal boneVTemporal bone resection. Otol Neurotol 36:879Y885, 2015.
Cutaneous malignancies involving the lateral cranium sometimes invade the temporal bone, requiring substantial soft tissue and bone resection for extirpating disease and establishing clear margins. The resulting complex defect requires reconstructive techniques that obliterate
dead space within the resection bed and provide a reasonable cosmetic result. In recent years, microsurgical procedures for the reconstruction of craniofacial defects have become popular. Forearm free flaps or other free fasciocutaneous tissue transfers offer thin and pliable tissues with a high degree of versatility in their design (1). However, free tissue transfers from distant sites have a number of disadvantages, including difficulty in matching facial skin cosmetically as well as the potential for failure of the microvascular anastomosis, longer operative times, and increased hospital costs related to ICU stay. Ideally, soft tissue used for reconstruction should be reliable, functionally and cosmetically acceptable, sufficient in size/ volume, cause minimal donor-site morbidity, and closely resemble the recipient site for color, texture, and thickness.
Address correspondence and reprint requests to Abraham Jacob, M.D., The University of Arizona Ear Institute, The University of Arizona Department of Surgery, The University of Arizona Cancer Center, and The University of Arizona Bio5 Institute, Tucson, Arizona, U.S.A.; E-mail: [email protected] Research supported by the Medical Student Research Program, The University of Arizona College of Medicine, Bio5, and The University of Arizona Cancer Center. Funding received from the National Institute on Deafness and Other Communication Disorders [Aj-k08 dc009644- 01a1]. The authors disclose no conflicts of interest.
879
Copyright © 2015 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
880
C. MILLER ET AL.
Locoregional flaps provide an excellent alternative to free flaps for repair of head and neck defects. The submental island flap, first described by Martin and colleagues (2), provides a reliable and cosmetically acceptable option for head and neck reconstruction. This flap, now increasing in popularity, provides more than adequate color and texture match to facial skin, is supple, and has a concealed donor-site incision that can be closed primarily. In older patients with redundant soft tissues in the submental region, this flap can have substantial cosmetic benefit. From a technical standpoint, flap harvest is straightforward, and the submental island flap can be lifted as a fasciocutaneous, myofascial, myocutaneous, or osteocutaneous flap. Additionally, because of increased skin laxity in older patients as well as the large vertical dimension available for safe flap harvest, a single submental artery can sustain a skin territory of up to 10 16 cm; this allows for reconstruction of large defects (3). Invasive skin cancer requiring temporal bone resection as part of a patient’s surgical management is epidemic in the desert Southwest. Therefore, our group has performed an unusually large number of these resections in a relatively short period of time and has used the submental island flap as our preferred reconstructive technique. Here, we describe the harvest and inset of these flaps for ear and periauricular defects by performing a review of patients’ presentations, operative reports, and follow-up. MATERIALS AND METHODS This was a single institution IRB-approved (IRB# HR11037) retrospective analysis of patients undergoing temporal bone resection and subsequent reconstruction between the dates of October 2011 and December 2013. A variety of reconstruction methods were used including submental island flap, radial forearm free flap, temporalis rotation flap, sternocleidomastoid (SCM) flap, and myocutaneous pectoralis major flap. The focus of the current work is the submental island flap, which is described in detail. Patients who underwent temporal bone resection without flap reconstruction were excluded from this study. Preoperative examination and diagnosis, surgical procedure, hospital stay (including duration of ICU stay), cost, follow-up, and outcomes were evaluated. A literature review regarding submental island flaps was also performed.
RESULTS Anatomy The submental artery, a branch of the anterior facial artery, vascularizes the submental island flap. This vessel reliably branches just below the mandible and courses anteriorly in a groove on the superomedial aspect of the submandibular gland, supplying branches to this structure. As it continues anteriorly, it lies on the mylohyoid, feeding it as well as the platysma and the periosteum of the mandible (4). The major outflow to the flap is the submental vein; frequently, however, a secondary venous drainage pattern to the anterior jugular vein will be found. Frequent anatomic variations in the venous pedicle have
been noted (4). At the level of the submandibular gland, the submental artery and vein course together, but just at the anterior aspect of the gland, the submental vein diverges laterally from the artery. Here, it traverses the gland and joins the facial vein at the posterolateral superior aspect of the gland. Once joined, the facial vein drains into the common facial vein and then into the internal jugular vein (IJV). As reported by others (4), the authors have found frequent anatomic variations in the venous pedicle. Therefore, it is essential to map the venous drainage before sacrificing any venous structures. Variations noted for this review include 1) the submental vein coursing from the anterior-medial aspect of the gland caudally over the anterior aspect of the gland and bypassing the facial vein for direct insertion to the IJV and 2) a direct connection of the vein to the external jugular vein through posterior passage across the lateral aspect of the submandibular gland. Many patients needing this flap require a neck dissection; therefore, all venous structures should be preserved unless directly involved with tumor. It is critical to identify and secure the venous pedicle early in the dissection. Surgical Technique Temporal bone resections are performed by a single neurotologist (A.J.) at our institution. Care is taken to preserve perforators, local veins, and the facial nerve when possible, while undertaking an appropriate oncologic resection. Incision planning is critical, as many skin cancers that secondarily involve the temporal bone require resection of all or parts of the pinna along with resection of cheek skin anterior to the ear. The incision must also facilitate parotidectomy or neck dissection when required and allow harvest/inset of the submental flap. A broad pedicle to the pinna is required to prevent vascular compromise. Typically, a large C-shaped incision extending into a neck crease is used. The ear canal is transected when appropriate or if portions of the pinna are involved with tumor, a partial auriculectomy is performed using conchal cuts that leave tumor tissue connected to the underlying temporal bone resection. Other incision options include a lazy-S postauricular incision that curves posteriorly into the temporal hairline or a preauricular incision that curves posteriorly above the pinna. In the latter case, the canal is transected and the pinna is raised as part of a posteriorly based cervical flap. Biopsies are taken to confirm malignancy as well as clear soft tissue margins. A complete, intact-canal mastoidectomy is then performed. All the usual landmarks are identified; the facial nerve is found at the external genu and traced inferiorly. Often, it is decompressed at the stylomastoid foramen to facilitate dissection into the parotid bed. The mastoid tip is removed. Next, an epitympanic dissection is carried to the soft tissues of the temporomandibular joint (TMJ). An extended facial recess approach is then performed. The incus-stapes joint is separated, the incus is removed, the head of malleus resected, and the tensor tympani tendon is cut. The fibrous annulus of the tympanic
Otology & Neurotology, Vol. 36, No. 5, 2015
Copyright © 2015 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
SUBMENTAL ISLAND FLAP FOR TEMPORAL BONE DEFECTS membrane is identified and traced inferiorly and then anteriorly. A chisel is used to crack away ear canal bone deep to the tympanic annulus but lateral to the petrous carotid artery. This leaves the anterior bony ear canal wall tethered only to soft tissues of the TMJ and parotid gland. The wound is then copiously irrigated, mucosa is removed from the middle ear, and the Eustachian tube orifice is packed with bone wax, Surgicel, and muscle. Two head and neck surgeons (A.E. and T.J.G.) remove tumor-involved soft tissues, in continuity with the underlying temporal bone resection specimen. Parotidectomy and neck dissections are performed if needed. Once all ablative portions of the operation are completed, the submental artery island flap is harvested and inset (Figs. 1 and 2). Before harvest, the approximate positions of the facial artery, vein, and the marginal mandibular branch of the facial nerve are identified. The superior skin paddle incision borders the inferior edge of the mandible and the maximal dimensions of the flap can extend from the ipsilateral angle of the mandible to the contralateral angle of the mandible.
881
The superior incision is performed on the undersurface of the mandible but made contiguous to the ipsilateral neck dissection when required. The skin and subcutaneous tissue are elevated from the inferior surface of the mandible up to the attachment of anterior belly of the digastric muscle. Next, the mylohyoid muscle is identified. Care is taken laterally to identify and preserve the marginal mandibular branch of the facial nerve. This may already have been done if a parotidectomy was performed to extirpate disease. The facial vessels are found as they course over the mandible. Next, the inferior skin incision is performed at approximately the level of the hyoid bone. Subplatysmal dissection is performed until the intermediate tendon of the digastric muscle is found. Dissection proceeds laterally until the submandibular gland and lateral border of the mylohyoid muscle are identified. The submandibular gland is removed; this is imperative as the submental artery and vein course over the superior aspect of the submandibular gland often supplying small branches to the gland. These branches should be clipped or ligated but the integrity of the submandibular artery and vein should
FIG. 1. Schematic diagram of the submental island flap and its pedicled vessels before dissection (A). The flap is elevated and rotated to show the orientation of muscle fibers as well as orientation of feeder vessels (B). The flap is inset to the temporal bone defect (C). Rostral aspect of flap is elevated. Otology & Neurotology, Vol. 36, No. 5, 2015
Copyright © 2015 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
882
C. MILLER ET AL.
FIG. 2. Elevation and placement of the submental island flap in a patient. A demonstrates flap elevation after resection of the primary tumor. B demonstrates the temporal bone defect (black arrow). The flap is rotated and placed in the ablative defect with pedicled vessels shown (black arrow; C). Flap is inset, and the wound is closed (D). Two Jackson Pratt drains are placed to prevent fluid accumulation.
not be compromised. It is our preference not to dissect the submental artery and vein to its origin from the facial artery, as these vessels can be small and easily damaged. However, we do perform dissection of the more proximal facial artery and vein as they course under or over the digastric muscle, respectively. At this point, the skin paddle and associated musculature are elevated. The insertion of the ipsilateral anterior belly of the digastric muscle is transected from its attachment to the mandible. One nuance in elevating this flap is that we prefer to harvest the ipsilateral mylohyoid muscle as well. This protects the submental vessels where the anterior belly of the digastric muscle and the mylohyoid muscle encloses them. It also adds bulk to the flap for obliteration of dead space within the temporal bone resection. Inferiorly, the intermediate tendon of the digastric muscle is harvested along with the mylohyoid muscle. Once the ipsilateral muscle attachments have been transected, the flap is elevated from distal to proximal (from the contralateral neck towards the ipsilateral neck). The skin paddle is elevated over the submandibular gland fascia and superficial to the contralateral anterior belly of the digastric muscle. Care should be taken when elevating the skin in the contralateral neck so as not to damage the marginal mandibular branch of the facial nerve. As the elevation proceeds medial to the contralateral anterior belly of the digastric muscle, the dissection plane depends on whether the mylohyoid muscle and ipsilateral anterior belly of the digastric muscle is being included in the flap’s elevation. The submental artery island flap now remains attached only by its vascular pedicle and fascia attach-
ments. The submental artery and vein can be seen, and branches that tether the flap’s mobilization (such as branches to the mylohyoid muscle) are transected. Typically, the distal facial artery and vein are left intact where they course over the mandible; however, they may be ligated if a further arc of rotation is needed. The submental artery island flap is then rotated into the ablative defect. Closure of the donor-site defect requires wide undermining of the lower anterior neck skin. This skin is advanced superiorly to close the defect. The undersurface of the flap should be sewn to the periosteum of the hyoid bone to recreate the cervicomental angle. The neck may need to be flexed to achieve complete closure of the neck skin. Outcomes In total, 17 patients undergoing temporal bone resections between the dates of October 2011 and December 2013 were included in this study. The average age was 73 years, and Table 1 delineates patient specifics such as diagnoses, defect size, flap type, ICU stay, and total cost of hospitalization. Nine patients received a submental island flap(SIF), two underwent a temporalis rotation flap, one underwent a radial forearm free flap, four received a sternocleidomastoid (SCM) flap (one patient required both a SIF and an SCM flap), and one received a myocutaneous pectoralis major flap. Average ICU stay for locoregional flap patients was less than 24 hours; the free flap patient had the longest ICU stay. Overall, six patients did require
Otology & Neurotology, Vol. 36, No. 5, 2015
Copyright © 2015 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
SUBMENTAL ISLAND FLAP FOR TEMPORAL BONE DEFECTS TABLE 1. Defect size (1 = Small, 2 = Medium, Type of Flap Name Gender Age 3 = Large) Free flap
HS
M
65
2
SMI
DB
M
85
3
FS
M
90
3
JG
M
81
Not in op note
CF
F
79
Not in op note
WS
M
75
2
KM
M
70
2
DS
M
56
DL
M
82
Not in op note 2
PC
M
83
GC SP
M M
66 66
AV
F
47
JG
M
81
CT RL LB
M M M
78 75 61
GC
M
65
AA
M
75
Rotational
Not in op note Not in op note Not in op note Not in op note 1 2 2 Not in op note Not in op note
883
Patient demographics
Tumor Size (cm)
Neck Dissection
Parotidectomy?
Diagnosis
ICU Stay Hospital Total (d) Stay (d) Charges
3.1 3.7 3.7 Left selective neck dissection (1Y4) 8.3 6.2 8 Right modified radical neck dissection 4 2.5 1 Supraomohyoid neck dissection
N
SCC of left ear and left parotid gland
4
8
179,035.30
Y
0
6
125,838.30
1
4
104,935.04
5.5 5.1 1.8 Right selective neck dissection (1Y3) 2.5 No
Y
0
8
130,826.32
0
8
114,051.27
1
4
101,439.21
1
3
88,339.90
0
7
99,643.37
4
85,077.15
43
No
Y
SCC right auditory canal and parotid gland SCC of the right ear, right temporal bone, and right parotid SCC of right auricle and right temporal bone BCC of left ear with involvement of the concha SCC of tempora bone (CNVII paralysis) SCC left cheek, auricle, and lateral skull base (CNVII paralysis) SCC of right ear
2.5 2.1 2
Y
SCC of left ear canal
G1
Left modified radical neck dissection No
N
0
5
95,735.82
2 1.4 Unavailable
No No
N N
0 1
1 3
60,570.95 53,887.50
Unavailable
No
N
BCC of left auditory canal Basal cell carcinoma Cholesteatoma right; chronic petrousitis Cholesteatoma
0
0
40,323.05
2
No
Y
SCC
0
9
127,956.97
1 23 Unavailable
No No No
N Y Y
0 0 1
4 6 4
75,912.87 95,299.79 110,044.99
1.4 0.6
No
N
0
1
55,230.80
Unavailable
No
N
BCC of the right ear SCC of left helix SCC of left ear and parotid gland SCC of right auditory canal Osteoradionecrosis of left temporal bone
0
0
46,626.73
2.9 2.4 2.8 Right posterior neck dissection 2.8 2.7 3 Left modified neck dissection
overnight observation in the ICU because of extent of resection, use of multiple drains, medical comorbidities, or poor cardiovascular/respiratory status compromising extubation in the operating room. Average hospital stay was approximately 5 days. There were no flap failures and cosmetic outcomes for the submental island flap were excellent (Fig. 3). Patients were followed after hospitalization by two head and neck microvascular surgeons (A.E./T.J.G.); those who underwent SIF showed a quicker functional recovery and required less outpatient postoperative wound care translating to fewer office visits the first 6 to 8 weeks after surgery. Cost analysis performed using documented facility charges at our institution showed that, on average, patients undergoing submental island flap reconstruction had average cost savings of greater than 40%
Y
N Y Y
compared to the free flap patient. Flap related operative time was less than 50% for all submental island flap elevations/ insets as compared to our free flap patient; however, the primary determinants of cost were length of ICU stay and length of hospitalization. Therefore, we believe that future analysis using larger cohorts of patients and stratifying by presurgical comorbidities could help better define the cost advantages of using the submental island flap.
DISCUSSION This study presents the largest published series of submental island flap reconstructions for patients with temporal bone defects and demonstrates that this flap is Otology & Neurotology, Vol. 36, No. 5, 2015
Copyright © 2015 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
884
C. MILLER ET AL.
FIG. 3. Healed left-sided submental island flap patient 1 year after surgery. Four views demonstrating excellent cosmetic outcome with minimal depression of the pinna.
quick, reliable, and has acceptable cosmetic and functional outcomes. Martin and colleagues first described the flap in 1993. The flap design and technique was investigated on 20 fresh cadavers as well as eight patients undergoing radical neck dissection. With a primary goal of reconstructing orofacial defects, Martin described this flap’s long, reliable pedicle, large cutaneous dimensions, excellent skin color match, and wide arc of rotation. In this investigation, the authors were able to extend the flap to the entire ipsilateral face, with the exception of the superior aspect of the forehead. The original eight patients described utilized this mechanism for reconstruction of both trauma and cancer ablation of the cheek, orbit, auricular region, and temporo-orbitofrontal region. Notably, no patients in this series had defects of the temporal bone nor did the article describe elevation or inset of this flap as an option for these specific defects. However, the authors noted the submental island flap’s overall versatility. More recently, Hayden and colleagues described the use of a variant of the submental flap at their institution between 2002 and 2012. Along with a detailed description of the anatomy and surgical technique, the authors described a variant of the flap, namely the hybrid pedicled flap. This flap is partially pedicled but also utilizes a draining vein close to the recipient site with reanastomosis to the submental or facial vein. In their review, they note successful reanastomosis with the external jugular, retromandibular, common facial, and internal jugular veins. This technique allows for even greater arc or rotation because of the ability to release the tethering by a short outflow pedicle. Their investigation looked at the utility of the flap for a broad range
of facial reconstructions including three cases in the auriculo-temporal region. Unlike microvascular free flaps, the submental island flap is a pedicled flap and decreased flap related operative time by over 50% in our series compared to performing a radial forearm free flap. Furthermore, there was no need for frequent postoperative flap checks and no flap failures during the postoperative period. Also, fewer wound surveillance visits were required immediately after discharge. The submental flap’s skin paddle is large and allows coverage of large parotid-bed defects as well as both partial and total auriculectomy sites. The bulk provided by harvesting the mylohyoid muscle along with this flap allows obliteration of dead space within the largest of temporal bone defects. Our investigation into the cost benefit of utilizing the submental island flap over free flaps found greater than 40% reduction in hospital charges. The free flap patient in this series had the longest ICU stay, providing one explanation for increased costs as validated by studies with published costs for ICU stays (5). The best candidates for this flap are those patients with excess skin laxity along the midline neck and substantial adipose tissue in the submental region. Less than ideal candidates lack neck skin laxity and do not have adequate soft tissue bulk. The latter may result in an inability to close the ablative defect, difficulty with closure of the donor site, and potential dysphagia caused by difficulty in neck extension from soft tissue tethering. Flaps of inadequate bulk can lead to significant depression at the ablative site and poor cosmesis; however, in these patients, abdominal fat grafting to bolster the underside of the flap may be a viable solution. Extremely poor candidates are patients who have undergone previous level
Otology & Neurotology, Vol. 36, No. 5, 2015
Copyright © 2015 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
SUBMENTAL ISLAND FLAP FOR TEMPORAL BONE DEFECTS one neck dissection. In patients with small auricular defects, the submental island flap may be more than what is needed for reconstruction. For these patients with tumors localized to the external auditory canal and ablative sites without large auriculectomy defects, we prefer to use a temporalis muscle rotation flap and abdominal fat placed within the temporal bone defect (deep to the muscle). The posterior aspect of the temporalis muscle is harvested, and these patients recover from surgery without any of the donor-site complications. CONCLUSION Invasive skin cancer requiring temporal bone resection as part of a patient’s surgical management is epidemic in the desert Southwest. As a result, our group has performed an unusually large number of these resections in a relatively short period of time. In appropriately chosen patients with
885
skin and soft tissue laxity in the midline neck along with adequate underlying bulk, the pedicled submental island flap provides a quick, reliable, less costly, and cosmetically acceptable option for reconstruction. REFERENCES 1. Merten SL, Jiang RP, Caminer D. The submental artery island flap for head and neck reconstruction. ANZ J Surg 2002;72:121Y4. 2. Martin D, Pascal JF, Baudet J, et al. The submental island flap: a new donor site. Anatomy and clinical applications as a free or pedicled flap. Plast Reconstr Surg 1993;92:867Y73. 3. Faltaous AA, Yetman RJ. The submental artery flap: an anatomic study. Plast Reconstr Surg 1996;97:56Y60; discussion 61Y2. 4. Hayden RE, Nagel TH, Donald CB. Hybrid submental flaps for reconstruction in the head and neck: part pedicled, part free. Laryngoscope 2014;124:637Y41. 5. Dasta JF, McLaughlin TP, Mody SH, Piech CT. Daily cost of an intensive care unit day: the contribution of mechanical ventilation. Crit Care Med 2005;33:1266Y71.
Otology & Neurotology, Vol. 36, No. 5, 2015
Copyright © 2015 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
