Type 1 fibers stain dark with this reagent, and the fast twitch type 2 fibers are light. A, Zygomaticus major; B, gracilis; and C, sternohyoid muscle (original magnification ×20 for all).
A, Subplatysmal flap elevation and exposure. B, Retraction of the omohyoid and exposure of the ansa on the superficial surface of the internal jugular vein. C, Isolation of the omohyoid muscle. D, Divided omohyoid muscle. E, Fully exposed ansa.
A, Great vessel dissection to isolate the superior thyroid artery pedicle. B, Completed pedicle dissection with the branch to the sternohyoid noted on the image. Distal ligation at the superior pole of the thyroid completes this dissection.
A, Medial release of the flap. B, Separation of the sternohyoid from the thyroid cartilage. C, Attachments of the thyrohyoid muscle with the release point noted with the shaded rectangle. The sternothyroid also must be released.
A, The hyoid release. B, Isolation of the hyoid and planned saw cuts around the central sternohyoid attachment. C, The mobilized hyoid-sternohyoid complex. D, The entire harvested flap in situ after the inferior sternohyoid release. STA indicates superior thyroid artery.
A, Harvested flap illustration. B, Harvested flap example.
A, Complete inset. B, Native zygomaticus major and zygoma. C, Sternohyoid flap inset at the oral commissure with the hyoid segment reflected laterally. This is plated into position on the superior surface of the zygoma to establish rigid fixation. D, Complete sternohyoid flap inset.
The common carotid and the internal and external carotid are ligated, isolating the superior thyroid artery for this injection. The injection study reveals the perfusion of the muscle through the descending branch and some superior perfusion through the superior laryngeal branch.
Alam DS, Haffey T, Vakharia K, Rajasekaran K, Chi J, Prayson R, McBride J, McClennan G. Sternohyoid Flap for Facial ReanimationA Comprehensive Preclinical Evaluation of a Novel Technique. JAMA Facial Plast Surg. 2013;15(4):305-313. doi:10.1001/jamafacial.2013.287
Neuromuscular reanimation of the face provides the correct specific neural functional input and thereby prevents synkinesis. Unfortunately, this ideal situation is rarely encountered in the clinical setting.
To assess the technical feasibility of and define the surgical procedure for harvesting the sternohyoid muscle as a novel free flap for use in facial reanimation indications.
Design, Setting, and Participants
Fresh, postmortem, nonfixed cadavers were used to define the anatomy and perform the flap harvest procedures. Twenty-four flap harvests were performed. Angiography was performed on the pedicle of the harvested flaps to assess potential flap perfusion. Adenosine triphosphatase staining was performed on the muscle specimens to establish fiber type.
Main Outcome Measures
The harvest technique, pedicle (arterial or venous), nerve length, and flap geometry parameters were characterized.
The sternohyoid muscle was found to be reliably vascularized by the superior thyroid artery in all cases with an appropriate arterial and venous pedicle for vascular anastomosis. The mean arterial (5.5 cm) and venous (5.9 cm) pedicle lengths are comparable with gracilis flaps. The mean motor nerve length was 10.7 cm. The inclusion of the hyoid bone allows rigid fixation, and the muscle size, fiber type, and volume profiles all compare favorably to the gracilis flap for use in the indication of facial reanimation. Mock surgical procedures were performed to define inset parameters. This flap potentially allows single-stage cross-facial neurorrhaphies to be performed.
Conclusions and Relevance
This is the first article, to our knowledge, of the sternohyoid muscle as a potential donor site for free-tissue transfer. This muscle has a predictable vascular pedicle and neural innervation along with size and fiber type parameters that make it an ideal potential free flap for facial reanimation.
Level of Evidence
The rehabilitation of facial nerve paralysis associated with nerve transection is ideally accomplished through neuromuscular reanimation of the face. When possible, the best-case scenario is reestablishing a neural connection between the native facial nerve and the facial musculature at a distal location in the nerve. This technique provides the correct specific neural functional input and thereby prevents synkinesis. Unfortunately, this ideal situation is rarely encountered in the clinical setting. In most cases, nerve transection or resection occurs proximal to the pes anserinus, which, in turn, limits the functional outcomes of neurorrhaphies to synkinesis. Even this option becomes unavailable in long-standing cases of paralysis. Beyond 18 to 24 months after injury, the motor end plates of the denervated muscles irreversibly atrophy. In this situation, a neuromuscular transfer that incorporates both a new neural input source and muscular mechanism becomes necessary. The 2 most commonly used options in these circumstances are local transfer of the temporalis tendon in an orthodromic manner vs free-muscle transfer using the gracilis muscle.
The gracilis free flap technique offers the advantage of potentially using a cross-facial nerve graft to allow facial nerve input to act as the motor drive for the muscle. In an optimal outcome, this technique allows the rehabilitation of a spontaneous smile. This muscle has become the mainstay of free-tissue facial reanimation. It is used widely in pediatric and adult indications with relatively good outcomes and results.1- 3 The ease of harvest and reliability of the flap have made it an excellent option in patients with long-standing paralysis.
The use of this flap is not, however, without its limitations. The transfer of the flap has definite anatomical constraints, which can make the results suboptimal. The most frequently cited concern is the bulk of the flap, especially in relation to the temporal region and the zygoma. The skin envelope and soft tissue overlying the zygoma in adults is generally less than 8 mm thick. Draping a free-muscle flap over this convex bony landmark often creates a visible facial deformity.4 This approach is less than ideal given that the purpose of facial nerve rehabilitation and reanimation is to reduce physical deformity.
Many authors have reported successful volume reduction of the harvested gracilis flap as a modification to reduce this problem.5 Even with such measures, the muscle bulk is significantly larger than the zygomaticus major or minor that the gracilis flap is intended to functionally replace. Not surprisingly, the flap is more favorable aesthetically in younger patients in whom the facial soft-tissue envelope is proportionately thicker and more robust, which allows for better flap camouflage.
The numerous modifications of the gracilis flap reported in the literature highlight the historical reality that this flap was not originally designed and reported as a means to rehabilitate facial nerve paralysis. Initial reports in the literature and the widespread use of this flap were for indications in the peripheral extremities (eg, wound coverage).6- 8 Over time the flap has been regarded as a well-established muscle flap with a reliable pedicle and nerve input that could be harvested with ease. This clinical history and reliability led to its eventual use in the face, but the choice of this muscle for this particular indication was not based on a directed search for the ideal muscle to replace facial musculature.
Even with the evolution of the gracilis flap, it is still not ideal. When considering several parameters, including muscle size, length-contraction ratio, and intrinsic type 1 or type 2 fiber ratios (although these can shift to some extent with reinnervation in certain muscle transfers), the muscle is quite distinct from the zygomaticus major.9
An anatomical comparison of the 2 muscles is shown in Figure 1. The images illustrate an obvious difference in the size of the muscles. The mean length of the gracilis muscle in an adult male is 41 cm, and only a small portion of the muscle is harvested in a facial reanimation procedure. The contracted length of this muscle in situ is only a few centimeters less than that seen at rest. The length-contraction ratio is unfavorable in contrast to that seen in the zygomaticus major, which only averages 5 cm in length and can have discursion greater than 2 cm in some cases. As a rule, shorter muscles are stronger as well.
Another potential difference is in muscle physiology. A comparison of fast twitch type 2 muscle fiber activity of the gracilis to the zygomaticus major reveals significant differences between these muscles. Our findings corroborate the accepted literature (Figure 2). This is expected when one considers the purpose of these muscles and their function within their native location. The gracilis is a muscle that is involved in chronic low-grade contraction associated with maintaining balance and stability of the torso. The zygomaticus major, in contrast, is involved in rapid facial expression and has almost the highest ratio of fast twitch fibers of any muscle in the body just behind the orbicularis oculi.
Because muscle fiber type is thought to be determined by the neural innervation source, this may not represent a clinical problem, but the limited data available on gracilis flaps suggest this transformation is not seen clinically.10 In many regards, the choice of the gracilis for facial reanimation is an example of using the available flap options instead of developing a specific flap for the indication from the ground up. The advances of microvascular reconstruction have revealed numerous shifts in clinical practice patterns, as new flaps have been developed to replace the available ones because they can address the limitations found in the available ones. Examples include the gradual transition from the iliac crest to the fibula flap for bony reconstruction and more recently the advent of the anterolateral thigh flap as a replacement for suboptimal soft-tissue flaps. In its purest sense, a free flap is simply a reliably harvested angiosome. The vascular territory of any nonessential vessel is potentially a new flap.
This led us to a directed, need-based search for an alternative to the gracilis as a potential free flap for facial reanimation. Ideally, this flap would have the following characteristics: (1) reliable and nonessential vascular pedicle, (2) motor nerve input with a reliable nerve (nonessential function and appropriate length, diameter, and branching), (3) fast twitch muscle fibers with a rapid and brisk contraction profile or clinical evidence of fiber type transition with reinnervation, (4) shorter length from origin to insertion and a superior length-contraction ratio, (5) rigid or semirigid origin and insertion point for better fixation, and (6) smaller muscle mass and size (Figure 3).
Using these criteria, we hypothesized that the sternohyoid muscle potentially could be an ideal replacement. This muscle has a high level of fast twitch fibers.11 It has the appropriate length and volume and serves a noncritical function with minimal morbidity if its native function is lost. The donor nerve (ansa) is long, easily identified, and easily dissected. The vascular supply of the strap muscles has been previously found to be potentially reliable for use as a pedicled flap,12 but its potential use as a free flap has never been studied.
We report the findings of our comprehensive preclinical evaluation of the novel use of the sternohyoid muscle as a potential free flap for facial reanimation. The study has numerous components: (1) a cadaver-based anatomical study of the development of a technique of flap harvest for free-tissue transfer, (2) characterization of pedicle length of the vascular supply along with angiographic analysis of the perfusion (arterial or venous) of the angiosome, (3) comparative physiologic and histochemical analysis of the sternohyoid muscle with the zygomaticus major and the gracilis, and (4) a preclinical study of the use of this muscle as a neuromuscular free-tissue transfer for facial reanimation. We have undertaken a human clinical trial, which will be reported in an accompanying study in the future.
Fresh, nonfixed human cadavers were used for this study. Exclusion criteria included prior neck surgery or tracheostomy. The sternohyoid muscle was harvested using a novel technique and was examined for pedicle length and characteristics and recipient nerve parameters.
Harvested flaps were transferred immediately to the angiography suite, where dye injection studies were performed. The superior thyroid artery (STA; arterial pedicle) was cannulated with a suture-fixed angiocatheter. Then 60% diatrizoate meglumine (Hypaque; GE Healthcare) was injected as a digital subtraction video angiographic study was obtained with real-time fluorography (Siemens USA). The arterial phase tissue, perfusion, and venous phase were examined.
Muscle samples from fresh, postmortem, nonfixed cadavers (<24 hours post mortem) were used for fiber type analysis. Samples were frozen, sectioned, and stained with hematoxylin-eosin and adenosine triphosphatase at the standard 4.6 and 9.8 pH levels to evaluate fiber type.
An evaluation of the technical feasibility of the flap for facial reanimation was performed by conducting mock surgical procedures using the harvested flaps. This was also performed to characterize the pedicle length, size match, and ability of the recipient nerve to reach the contralateral donor nerve to allow for single-stage cross-facial reanimation.
The surgical steps of the harvest are shown in Figures 4, 5, 6, and 7. First, a hemiapron incision is made centered in a neck crease between the hyoid and the sternum just superior to the placement of the typical incision for a thyroidectomy. Subplatysmal flaps are then elevated to expose the underlying strap muscles and the sternocleidomastoid muscle. The sternocleidomastoid muscle is retracted laterally to expose the great vessels and identify the ansa cervicalis. It is then dissected from its origin at the hypoglossal nerve to the point of innervation of the sternohyoid muscle approximately 2 cm above the sternum. The omohyoid muscle is divided to facilitate the completion of this dissection (Figure 4A-E).
Second, the external carotid artery is dissected to identify the origin of the STA. The STA is dissected distally, with care taken to identify and preserve the accompanying pedicle vein (Figure 5A and B).
Third, the STA is suture ligated distal to the medial branches to the sternohyoid at the superior pole of the ipsilateral thyroid gland lobe.
Fourth, a second potential venous outflow through the middle thyroid vein may be isolated at this point.
Fifth, the proximal vein pedicle is then dissected to a draining vein with a luminal diameter of 2 to 3 mm. This circumstance is most often seen with the ranine veins associated with the hypoglossal nerve but in some cases is directly drained into the internal jugular vein.
Sixth, the sternohyoid muscle is reflected posterolaterally as the thyrohyoid and sternothyroid attachments are released from the thyroid cartilage (Figure 6A-C).
Seventh, the superior release is performed next via isolation of the anterior face of the hyoid bone. The suprahyoid muscles are released and the hyoid cut medially and laterally to the attachment to the ipsilateral sternohyoid muscle. Care is taken to avoid injury to the underlying hypoglossal nerve (Figure 7A-D).
Eighth, the inferior muscle cut is made below the level where the ansa fibers innervate the muscle body.
The harvested flap is shown in Figure 8. The inset technique is shown is Figure 9. The muscle is anchored with rigid fixation of the superior hyoid attachment to the superior and inner surface of the zygoma with a 1.0-mm microplate. This technique allows rigid fixation with minimal if any temporal fossa distortion. The pedicle is anastomosed to the superficial temporal artery and the superficial and/or deep temporal vein. Alternatively, the pedicle length allows the use of the facial vessels (artery or vein), but this would have an expected size mismatch. The distal muscle is suture fixed at the modiolus and can be divided to attach around the commissure as is often done with the gracilis. One unique advantage is the graft length of the recipient nerve that can easily be transferred in a single stage across the upper lip in a subcutaneous plane to the contralateral facial nerve due to its long mean length (Figure 10).
The harvested flaps were examined for relevant flap characteristics (Table). A lengthy vascular pedicle was identified in all specimens, with a mean pedicle length of 5.5 cm. There was a small range of variation around the mean, suggesting a reliable harvest pedicle. The venous side had some variability. The draining vein was noted to be either a relatively larger-caliber superior thyroid vein or a ranine vein accompanying the hypoglossal nerve. In all cases, an appropriate vein to perform a microvascular anastomosis (2- to 3-mm diameter) was identified. The venous pedicle lengths were usually slightly longer than the arterial by a mean of 5 mm.
Angiography of the pedicle revealed brisk perfusion of the entire muscle and a venous phase to the superior thyroid vein. An image of the arterial phase of the digital subtraction angiography is shown in Figure 11. This figure illustrates the branching pattern seen in all of the specimens with a reliable, inferiorly directed, intramuscular arterial arcade.
The fiber type match was assessed with adenosine triphosphatase staining (Figure 2). Calculations were performed of the native fiber type ratios by counting the relative percentage of fibers in the 5 high-power cross-sectional fields in each sample. The percentages of fibers were as follows: for adenosine triphosphatase pH 4.6: zygomaticus major, 85%; gracilis, 35%; and sternohyoid muscle, 67%; and for adenosine triphosphatase pH 9.8: zygomaticus major, 83%; gracilis, 36%; and sternohyoid muscle, 66%.
Rehabilitation of long-standing facial paralysis with the gracilis free flap has been widely accepted as the criterion standard for free-tissue transfer for facial reanimation. Although effective, the technique has limitations of excess bulk, poor length-contraction ratios, and a short recipient nerve. These limitations are expected because this flap was not originally designed for this indication and has over time been modified extensively to try to make it adapt better in the role. The gracilis was used primarily for soft-tissue coverage in peripheral extremities and is perfectly suited for this indication.
In the past 20 years, better understanding of flap physiology and design has led to the development of several new flaps, which have replaced predecessors in clinical use because of their better harvest and inset characteristics. A notable example is the increasing use of the anterolateral thigh flap for soft-tissue reconstruction.
On the basis of the criteria for the ideal facial reanimation free flap, we began a systematic search for an alternative muscle to the gracilis for this use. The sternohyoid muscle has previously been reported as a potential pedicled flap, and this muscle had characteristics that are favorable when compared with the gracilis. Better size match, rigid fixation options, longer recipient nerve, and data that reveal higher type 2 fast twitch fiber content, as well as evidence of rapid fiber type transition in animal models, all point to this muscle as a near-perfect alternative.
Our comprehensive evaluation of the harvest, clinical parameters, and preclinical evaluation of the inset for use in facial reanimation is the first article, to our knowledge, of the use of the sternohyoid muscle as a potential new free flap. The muscle was harvested reliably in every cadaveric specimen bilaterally with a reliable vascular pedicle. The mean pedicle length is similar to that seen in the gracilis, and reassuringly there seems to be little clinical variability in pedicle length, which is an important consideration in any free flap choice.
On the basis of our findings, we believe this flap has significant potential advantages over the gracilis. The ansa nerve is 10.7 cm in mean length, which allows the potential for single-stage cross-facial neurorrhaphies. The trend in gracilis surgery has been away from cross-facial procedures to the use of masseter nerve braches. This approach sacrifices one of the fundamental holy grails of facial nerve reanimation in trying to achieve a spontaneous smile. The need for a staging babysitter graft with the gracilis followed by a second cross-facial innervation results in more axonal loss, thereby hindering function. Estimates of 20% distal axonal fiber counts across cross-facials have been reported. The length and orientation of the sternohyoid recipient nerve would allow an easy single-stage cross-graft to be performed.
The sternohyoid muscle has produced favorable neuromuscular characteristics as well. As shown in Figure 2 the neuromotor profile of the sternohyoid muscle has a few axonal fibers that drive larger muscle groups. This fiber type bundling is commonly seen in reinnervation, but clinically this muscle functions well with this as its baseline. Studies have also found that the sternohyoid muscle has reliable fast twitch fiber enrichment where the recurrent laryngeal nerve was sutured to the ansa. These findings support the potential advantages of this muscle.
Finally, the sternohyoid muscle allows for simple fixation and minimal facial disfigurement on inset. As shown in Figure 7, the size match is excellent, and the hyoid allows rigid fixation of muscle that is often difficult to accomplish with suturing techniques alone. This approach allows preservation of the temporal fossa anatomy as well.
Our preclinical findings indicate that the sternohyoid muscle can easily be harvested with a reliable vascular pedicle, and the inset characteristics are favorable. This needs-based method of flap design is a novel approach to a clinical problem. We have traditionally looked at flap reconstruction by asking how we can use existing free flaps for our new clinical problems instead of viewing the issue from the more rational but opposite perspective of how we can design new flaps tailored specifically to solve old clinical problems. With the advances in microsurgery, new flap development can become readily examined and pursued. Any nonessential angiosome in the body is a potential new free flap if it possesses consistent and reliable vascular pedicle anatomy.
The ultimate measure of the success or failure of this particular novel approach will be clinical success in actual patients. On the basis of the strength of the preclinical data presented here and the results of numerous mock surgical procedures, we believe the sternohyoid is an excellent alternative to the gracilis with significant potential advantages. We have begun a clinical trial of this technique. The results of this trial will be presented in an accompanying publication in the future once appropriate follow-up has been obtained.
Corresponding Author: Daniel S. Alam, MD, Desk A71, Head and Neck Institute, Cleveland Clinic, 9500 Euclid Ave, Cleveland, OH 44195 (firstname.lastname@example.org).
Accepted for Publication: December 30, 2012.
Published Online: May 23, 2013. doi:10.1001/jamafacial.2013.287
Author Contributions:Study concept and design: Alam, Haffey, Vakharia, and McClennan.
Acquisition of data: Alam, Haffey, Rajasekaran, Chi, Prayson, McBride, and McClennan.
Analysis and interpretation of data: Alam, Haffey, Vakharia, Chi, Prayson, and McClennan.
Drafting of the manuscript: Alam, Rajasekaran, and Prayson.
Critical revision of the manuscript for important intellectual content: Alam, Haffey, Chi, McBride, and McClennan.
Administrative, technical, and material support: All authors.
Study supervision: Alam and McClennan.
Conflict of Interest Disclosures: None reported.