A, This defect is the result of resection of a recurrent squamous cell carcinoma in a patient who received solid organ transplant following previous surgery and adjuvant radiation. B, One month after surgery, the flap is well healed, with good obliteration of the space. This can also create a platform for future auricle reconstruction.
A, The patient received parotidectomy, neck dissection, and temporal bone resection, preserving the auricle (reflected forward). B, The flap was de-epithelialized and then inset. C and D, Two years following surgery, the flap has provided excellent contouring and support for the auricle despite adjuvant radiation. The frontal image (C) demonstrates that the height and projection of the auricle has been maintained as well as the cheek volume.
Emerick KS, Herr MW, Lin DT, Santos F, Deschler DG. Supraclavicular Artery Island Flap for Reconstruction of Complex Parotidectomy, Lateral Skull Base, and Total Auriculectomy Defects. JAMA Otolaryngol Head Neck Surg. 2014;140(9):861-866. doi:10.1001/jamaoto.2014.1394
There are limited data on the use of the supraclavicular artery island flap (SCAIF) for parotid and lateral skull base (LSB) surgery. This flap can be an important reconstructive tool for these procedures.
To describe the use of the SCAIF for parotid and LSB surgery and its success, as well as important technique modifications for successful use of the flap in this setting.
Design, Setting, and Participants
Retrospective single-institution review from July 1, 2011, to September 30, 2013, of patients in a tertiary care referral center. A prospectively collected institutional database was reviewed to identify patients who received SCAIF reconstruction for parotid and/or LSB surgery. Forty-six SCAIF reconstructions were identified; 16 were performed for the indication of parotidectomy or LSB surgery.
The SCAIF reconstruction for parotid and/or LSB surgery.
Main Outcomes and Measures
Indication for reconstruction, flap viability, flap size, reconstruction site complication, and donor site complication.
Resection was performed for advanced cutaneous malignant tumor in 10 patients, primary salivary gland malignant tumor in 4 patients, and chronic infection and mastoid cutaneous fistula in 2 patients. All defects were complex, involving multiple subsites; 5 patients underwent facial nerve resection and 4 had previous radiation therapy. No complete flap loss occurred. One partial flap loss occurred. The average flap island size was 7 × 10 cm. No major complications occurred. Two minor reconstruction site complications and 3 donor site seromas occurred.
Conclusions and Relevance
The SCAIF can be successfully and reliably used for complex defects following parotid and LSB surgery. There are 3 important technique modifications to help facilitate rotation and coverage of this region.
Parotid and lateral skull base (LSB) surgery defects create a different set of reconstructive challenges compared with typical aerodigestive tract head and neck cancer reconstruction. Typical goals for aerodigestive tract reconstruction focus on separating the aerodigestive tract from the skin and external surface of the body and restoring functions of swallowing and voice production and modulation. Because the glandular function cannot be replaced, parotid defect reconstruction is primarily focused on restoring facial contour. In the case of total parotidectomy and LSB surgery, reconstruction also provides additional support for the auricle to prevent rotation and inferior displacement. When total auriculectomy and skin resection is required, reconstruction must also provide external skin coverage. Last, LSB reconstruction must provide important coverage of bone, the eustachian tube, and in some cases, the great vessels and/or dura mater.
Historical approaches to parotidectomy and LSB reconstruction have evolved over time. The simplest option is fat grafting. This is sufficient for limited parotidectomy defects or simple mastoid obliteration, but it is inadequate for larger skull base defects. Free fat grafts large enough to support and reconstruct extensive LSB defects cannot survive, leading to liquefaction, volume loss, and potential infection.1 Local options, such as the temporoparietal fascia flap and temporalis muscle flap, have been successfully used for mastoid obliteration.2,3 However, the temporoparietal fascia provides a thin flap that cannot provide enough tissue for auricle support for defects beyond a simple mastoid defect. The temporalis muscle flap offers more volume but is limited in its range and rotation while creating a temporal contour defect at the donor site.
More substantial pedicled reconstructive options such as the pectoralis major myocutaneous (PM) flap and latissimus dorsi flap are limited by the length needed to reach the superior aspect of the defect. Given these limitations, free tissue transfer has become the most common type of reconstruction.4- 8 Early experience with the rectus abdominus provided a good and reliable option, but this has been replaced by the anterolateral thigh (ALT) flap owing to less donor site morbidity. The ALT can be used as a fasciocutaneous or myocutaneous flap. Other fasciocutaneous flaps have been used. An excellent experience has been reported with the lateral arm flap.8 However, small vessels and lack of familiarity with the technique have limited its widespread use. The radial forearm free flap (RFFF) is the most widely used flap in head and neck reconstruction and is a reliable option for this defect, but may be limited by insufficient soft-tissue volume. All of these flaps have unique strengths that make them good options for reconstruction. However, all have some limitation such as poor color match, imperfect tissue volume, donor site morbidity, or issues related to complexity introduced by microvascular reconstruction.
One reconstruction that has not been described in great detail for complex parotidectomy and LSB defects is the supraclavicular artery island flap (SCAIF). The SCAIF is not a new flap; however, recent modifications and developments have increased its use in head and neck reconstruction.9,10 The SCAIF is a fasciocutaneous pedicled flap based on the supraclavicular artery and vein. The anatomy was initially described by Toldt in 1903 and further described by Kazanjian in 1949.11 Mathes and Vasconez12 described its clinical use in the 1970s. The PM flap and, subsequently, free tissue transfer became the dominant reconstructive options in the head and neck during the 1980s and 1990s and were preferred to the SCAIF. However, in the late 1990s, Pallua and colleagues9,10 described its extensive use for post-burn contractures in the head and neck. This led to SCAIF application for other head and neck reconstructions.
This series describes our experience to date with SCAIF reconstruction for complex parotidectomy and LSB surgery defects. We have found this flap to be a reliable option that provides excellent skin color match and adequate soft tissue to achieve the goals of recontouring the cheek and providing adequate auricular support, eustachian tube obliteration, and bone coverage.
Institutional review board approval from the Massachusetts Eye and Ear Infirmary Human Studies Committee was obtained. All patients who received SCAIF reconstruction for an indication that included parotidectomy or LSB surgery were identified from a prospectively collected database of SCAIF reconstructions. Once identified, the following data points were recorded: tumor histological features, type of reconstruction performed (fasciocutaneous or fascia and dermis only), extent of surgical resection, flap size, flap viability, reconstruction site complications, and donor site complications. During this period, no other pedicled or free flaps were used for the LSB defect during the study period by the surgeons (K.S.E. and D.G.D.).
Forty-six SCAIF reconstructions were present in the database from July 1, 2011, to September 30, 2013. Sixteen reconstructions were identified as treating parotidectomy or LSB resection. This was further investigated and characterized to understand how the flap was being used; these results are summarized in Table 1. Ten patients had primary advanced cutaneous malignant tumor involving the auricle, external auditory canal, or preauricular skin. Four patients had a primary salivary gland malignant tumor. Two patients received extended mastoidectomy and LSB resection for chronic infection and mastoid cutaneous fistula related to Actinomyces infection.
Table 1 highlights the range of defects reconstructed; 14 involved a parotid defect, 11 a temporal bone defect, and 10 both, including 3 that required total auriculectomy. The SCAIF was harvested as a fasciocutaneous flap in 5 cases and then de-epithelialized for use as a fascial and dermal flap. Ten patients had extensive skin resections and 5 had facial nerve sacrifice, further highlighting the complexity of the tumors and defects. Additional complexity resulted from previous radiation therapy in 4 patients, and 2 flaps were used to salvage failed microvascular free flaps.
Table 2 summarizes the details of the flap sizes and complications. A wide variety of flap sizes were used. Skin island width ranged from 4 to 10 cm, and length ranged from 6 to 15 cm (mean, 7.2 × 10.3 cm). There was no complete flap loss, 1 major partial loss in a fascial flap, and 1 minor distal tip loss in a fasciocutaneous flap. This is consistent with our larger series in which there was only 1 other major partial loss among the 46 patients in the database and no total flap loss. There were 4 other buried fasciocutaneous flaps. There was no evidence of drainage, fluid collection, erythema, or volume loss to suggest flap loss for these buried flaps. Recipient site complications were minimal. The patient with major partial flap loss required operative debridement, but ultimately had no significant complications after further wound care and healing. One patient also developed an incisional dehiscence in a heavily radiated field. The flap was placed for carotid and mandible coverage in the setting of persistent tumor treated with re-irradiation and chemotherapy. Good vessel and bone coverage was achieved and maintained. Donor site complications were limited to seroma formation in 3 patients; 1 subsequently developed incision dehiscence. All seromas were managed with needle aspiration. The patient with seroma and incisional breakdown was treated with negative pressure wound therapy to facilitate closure. This was fully healed in 2 weeks. Significant donor site pain was limited to 1 patient; it eventually resolved and no significant range of motion limitation was reported.
The SCAIF can be reliably used for complex parotid and LSB reconstruction. Parotid and LSB defects can be complex, based on the need to reconstruct different 3-dimensional defects and varying amounts of overlying skin. The tissue chosen for reconstruction must meet these challenges. Levy et al13 described 5 patients with posterior LSB defects reconstructed with the SCAIF. Epps and colleagues14 also used the SCAIF for recontouring the cheek after parotidectomy. The SCAIF reconstructions described in this article add to these initial studies and are unique in a couple of aspects. To our knowledge, this is the largest reported series of parotid or LSB reconstruction with SCAIF. More important, this is the first series to demonstrate the ability to reconstruct a broader range of defects that combine parotid, LSB, skin, and total auriculectomy as well as a greater degree of defect complexity created by previous radiation or failed free flap.
Despite numerous articles, there is limited reporting in the otolaryngology literature of success with this flap. There were no complete flap losses in this series, and in our larger series of 46 patients,15 we have not had a complete flap loss to date. We had 1 major partial loss in this series and a second major partial loss in the larger series as well. The patient with single partial flap loss in this series developed an infection that required operative debridement but did not require further reconstructive procedures after local wound care. This series and our larger experience demonstrate the reliability of this flap.
One of the great advantages of a free flap is that it is not limited by pedicle attachment. Reconstruction of LSB routinely needs to reach as high as the temporalis fascia superior to the auricle. The proximity of multiple potential recipient vessels makes a free flap a great option to ensure the flap can reach high enough. This is one of the potential limitations for any pedicled flap. However, this challenge has been met by other pedicle flap reconstructions. The PM flap historically was considered unable to reach adequate height. Technical modifications and variations in incisions and inset technique described by Resto et al,16 however, have made the PM a reliable option. The occipital flap is another regional pedicle flap option.17 In our experience, the SCAIF has reliably been able to reach the temporalis muscle and the superior-most aspect of parotid and LSB defects. Body habitus can contribute to this. A patient with broad shoulders and a short neck is the ideal candidate. Free tissue may also be a better option in specific cases in which additional refinement is needed that can only be offered by the freedom provided by free tissue transfer.
To reach the superior aspect of the defect, there are a couple of important technical considerations. The classic supraclavicular artery, as described by Mathes and Vasconez12 and further refined by Pallua and colleagues,9,10 arises from the transverse cervical artery and crosses the lateral aspect of the clavicle and then takes a course to the ventral deltoid. Recently, Pallua and Wolter18 also described an anterior perforating supraclavicular artery, which starts from the transverse cervical artery in a more medial and anterior position. This vessel then crosses the medial third of the clavicle and courses toward the deltopectoral fossa.
In this series, the SCAIF skin island was designed over the inferior aspect of the ventral deltoid and encompassed an area of the deltopectoral fossa. Because this series includes patients before and after this description, the initial intent was to use the classic pedicle as described by Pallua and colleagues.9,10 However, this positioning likely allows the flap to incorporate an angiosome from both the classic and anterior vessels. We do not attempt to identify either of these vascular pedicles; rather, our technique, dated prior to the description of this anterior branch, involves creating a wide soft-tissue pedicle that includes the course of both of these vessels.15 To do this, the inferior aspect of the soft-tissue pedicle is designed to remain inferior to the clavicle in a line from the deltopectoral fossa to the clavicular head. At this point, a soft-tissue back cut can be made superiorly over the posterior border of the sternocleidomastoid muscle. This back cut facilitates additional rotation and extension of the flap. This can be safely done because it is medial to the pedicle. A similar approach can be taken superiorly. The soft-tissue pedicle is designed to stay posterior near the trapezius all the way to the point where the external jugular vein crosses the posterior border of the sternocleidomastoid muscle. A similar back cut can be made in an inferior direction over the sternocleidomastoid muscle, providing additional rotation and extension. Surgeons need to be aware of these different vascular pedicles when planning the skin island location if both vessels are not going to be included. In addition, this back-cut technique can be used with either the classic or anterior branch; however, one must be aware of the course of these vessels to appropriately plan the soft-tissue pedicle.
A second key technique is to carefully release the supraclavicular fascia. We elevate the soft-tissue pedicle over the clavicle in the subperiosteal plane. At the medial aspect of the clavicle, the periosteum and fascia are incised. We do this by creating tunnels through the fascia in the direction of the pedicle. A fine-tipped dissecting instrument is used to connect the tunnels, and the see-through fascia is transected. This provides significant release and stretch to reach the defect without tension. These technical modifications have been useful in achieving adequate length for parotid and LSB reconstruction.15 Granzow et al19 recently described their technique for additional length. They use a Doppler to define the lateral extent of the vessel, resulting in a viable flap for 5 cm beyond where the Doppler signal is lost. This may be another consideration for helping to extend the length and potential reach of this flap.
Perhaps the most important functional aspect of parotid and LSB reconstruction is having the appropriate amount of soft tissue to adequately fill the defect. This includes deep soft tissue to help with eustachian tube obliteration, bone coverage, vessel coverage, and dura mater coverage. Figure 1 highlights such a defect with a good outcome. For this reason, potentially large volume flaps, such as the ALT and rectus abdominis, are good options. However, fasciocutaneous flaps can still provide adequate tissue. In some cases, the thinner fasciocutaneous flap may be preferred if there is a deep and narrow area in which it is difficult to place vascularized tissue. We did not have any issues with inadequate coverage or breakdown leading to exposure of any critical structures.
Fasciocutaneous flaps are rarely as thick as myocutaneous flaps. Therefore, additional technique modifications need to be made in most cases to achieve adequate thickness. Lack of thickness is a potential limitation of all fasciocutaneous flaps. The SCAIF is variable in its thickness. Like other flaps, this is body habitus–dependent. We have encountered flaps as thick as 5 cm. The SCAIF is generally thicker than the RFFF but thinner than the ALT and most similar to the lateral arm flap. To generate thickness, these flaps can be designed to be longer and wider than needed. Flaps can be folded in different dimensions to create additional thickness. Most commonly, we found folding the distal tip of the flap into the deep aspect of the defect to be most effective at achieving this additional bulk. As shown in Table 2, we routinely used skin islands of 10 to 13 cm in length and as long as 15 cm to achieve this goal. Ultimately, however, it is important to assess the thickness of the tissue over the deltoid and see if it is appropriate for the defect. Myocutaneous flaps such as the ALT are superior for very large and very deep defects in a thin patient.
The second aspect of soft-tissue reconstruction is related to appearance. In the parotid region, this is simply a matter of restoring the volume that was previously present to maintain cheek symmetry. When the auricle is intact, the LSB surgery and total parotidectomy defects need adequate volume to support the natural position of the ear. If not, inadequate volume will lead to an inferior and medial rotation and displacement of the auricle, yielding a less-than-satisfactory long-term result. Figure 2 shows adequate filling of the parotid soft tissue and supported auricle in a natural position. The duration of follow-up for these patients is a maximum of 2 years; however, we have yet to observe significant volume loss in this period, even after radiation. This will need to be monitored to be sure there is not volume loss with time. This may also provide an area of future investigation to more objectively assess any changes over time.
The SCAIF is a good choice for reconstruction of these complex defects for other reasons as well. Facial reanimation is an important consideration following facial nerve sacrifice. One possibility is microvascular reconstruction with the gracilis free flap. This free flap offers the greatest potential for reanimation by providing volitional facial movement with good resting tone and symmetry. By using a pedicled flap, potential recipient vessels may be preserved for future facial reanimation with a free flap. In our series, 10 of 15 patients also required significant skin reconstruction. This is an area in which the SCAIF is considerably superior to other reconstructive options. In general, the closer the reconstructive skin is to the defect, the more likely it is to provide a good match. This is the case for cheek and auricular defects. The skin match from the SCAIF is quite similar in color, texture, thickness, and hair-bearing status. The PM flap has a much different dermal thickness and has thick hair in many men, preventing a good cosmetic outcome. In women, the breast tissue is often not the right texture and the color match from the chest is poor. The RFFF has a different texture and has a much thinner dermis compared with the cheek. In many men and women, the RFFF is hair-bearing. Last, the ALT is usually hair-bearing and has a much thicker dermis, with a different texture and color. This has not been objectively studied to date but appears intuitive based on our experience.
Of the 15 patients in this series, 10 had advanced cutaneous malignant tumors requiring complex resection. There are a growing number of patients with advanced cutaneous malignant tumors related to immunosuppression for solid organ transplantation, rheumatologic conditions, and hematologic malignancies. These patients have medically complex conditions requiring careful preoperative preparation and postoperative care. In these patients, a pedicled flap may decrease the complexity of surgical care. The rapid harvest (30-60 minutes) and absence of a requirement for microvascular surgical expertise offer the potential to decrease operative time. In addition, there is lower risk of take-back for vascular compromise and no need for specialized monitoring postoperatively. These factors may be important to consider as surgeons choose the best overall reconstruction for an individual patient.
The preceding discussion highlights some of the important considerations for flap selection. Ultimately, the first step in flap selection begins with the defect. We have found that the SCAIF has been successful at reconstructing a broad range of defects. However, defects greater than 10 to 15 cm in width may extend beyond the angiosome of the SCAIF, depending on body habitus, and challenge the ability to achieve primary closure at the donor site. In this case, free tissue transfer may be a better option. In addition, patients with minimal subcutaneous tissue over the ventral deltoid may also be suboptimal for reconstruction of a deep and extensive temporal bone defect. In this case, the muscle offered from an ALT may provide a better option. Last, it is important to remember that cutaneous malignancies of the posterior scalp, and potentially the auricle, carry a risk of lymph node metastasis to the lower level V region. If a comprehensive lymphadenectomy including the transverse cervical vessels is indicated to optimize oncologic outcome, the SCAIF will not be an option because the vascular supply will be disrupted.
The challenge of complex parotid, LSB, and total auriculectomy defect reconstruction can be met in many ways. There is not a single perfect solution. This series demonstrates that the SCAIF can be successfully and reliably used for parotid, LSB, and total auriculectomy defect reconstruction. This includes large defects that span the parotid bed and LSB, including the overlying skin and ear. The flap is excellent when good skin color match is a priority. Last, it is an important consideration in patients with multiple comorbidities and when there is a need to decrease the overall complexity of the surgical management.
Submitted for Publication: November 22, 2013; final revision received May 27, 2014; accepted June 13, 2014.
Corresponding Author: Kevin S. Emerick, MD, Division of Head and Neck Surgical Oncology and Reconstruction, Massachusetts Eye and Ear Infirmary, 243 Charles St, Boston, MA 02114 (email@example.com).
Published Online: August 7, 2014. doi:10.1001/jamaoto.2014.1394.
Author Contributions: Dr Emerick had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Emerick.
Acquisition, analysis, or interpretation of data: Herr, Lin, Santos, Deschler.
Drafting of the manuscript: Emerick.
Critical revision of the manuscript for important intellectual content: Herr, Lin, Santos, Deschler.
Administrative, technical, or material support: Emerick, Herr.
Supervision: Lin, Deschler.
Conflict of Interest Disclosures: None reported.