Temporalis Tendon Transfer Versus Gracilis Free Muscle Transfer

Temporalis tendon transfer (T3) and gracilis free muscle transfer (GFMT) are popular techniques in lower facial rehabilitation when reinnervation techniques are unavailable. T3 involves a single-stage outpatient procedure resulting in immediate improvement in resting symmetry and a volitional smile. GFMT allows a spontaneous smile, customized vectors, and increased excursion but requires longer surgical time, a delay before movement, and specialized equipment. Ultimately, shared decision making between the clinician and patient should focus on the patient’s goals and unique medical condition.

Key points

•

Temporalis tendon transfers (T3) involve a single-stage outpatient procedure resulting in immediate improvement in resting tone and a volitional smile. Patients are unable to achieve true spontaneity and surgeons have less control over smile vector and shape.
•

Modifications of the T3 with minimal incisional access, increased tendon harvest along the medial ramus, and intraoperative assessment of muscle tension relationship improves outcomes in this technique.
•

Gracilis free muscle transfer (GFMT) allows a spontaneous smile, customized vectors, and greater excursion but requires longer surgical and recovery time, a delay before movement, and specialized equipment.
•

Degree of resting symmetry is a key factor in surgical decision making between T3 and GFMT.

Introduction

Facial paralysis causes debilitating physical and psychological effects. The restriction of emotional expression, disruption of oral competence during speech and swallow, and static asymmetry cause significant impacts to quality of life and psychological distress. Rehabilitation of lower facial movement, and specifically smile restoration, is a central focus of facial reanimation. Surgical treatments have documented improvements in quality of life, facial attractiveness, and psychosocial function. ^, Reinnervating native facial musculature is the preferred method of treatment when feasible. No single muscle transfer option can duplicate the nuanced movements of mimetic facial musculature. The ideal technique would reestablish symmetry at rest and during smile. In patients with prolonged denervation, absence or myopathy of native facial musculature, abnormality of distal facial nerve branches, and/or incomplete paralysis with insufficient commissure movement, muscle transfer techniques are preferred for static and dynamic smile restoration. Although a myriad of muscle transfer techniques exists, the 2 most widely used techniques are the temporalis tendon transfer (T3) and gracilis free muscle transfer (GFMT). This article discusses surgical decision making between these 2 surgical options.

Temporalis tendon transfer

Many variations using the temporalis muscle for facial reanimation exist. Gillies initially described transposing the temporalis muscle belly over the zygomatic arch to suspend the paralyzed lip with a fascia lata extension in 1934. This technique causes a visible bulge over the zygoma accentuated by an adjacent hollow in the temporal donor site. In addition, folding the muscle belly over the zygomatic arch alters the vector and dynamics of contraction and likely decreases contractile capacity. Improving on the donor site deformity and oral commissure excursion, McLaughlin described an orthodromic use of the temporalis muscle via a coronoidectomy with fascia lata extension in 1949. Labbe and Huault subsequently described the lengthening temporalis myoplasty (LTM), eliminating the need for a fascia graft extension by releasing the periosteal attachments of the posterior third of the temporalis muscle within the temporal fossa to create additional length. Additional modifications of the orthodromic temporalis tendon transfer reduce and refine the use of fascia lata extension and use minimal access incisions.

Orthodromic temporalis tendon transfer techniques and modifications are well described in the literature. Critical to maximizing contractility and commissure excursion, the temporalis muscle tension must be maintained as close to its optimal passive length as possible. Intraoperative muscle stimulation allows surgeons to objectively determine optimal muscle tension length. ^, Although classically described as inserting on the coronoid process, the temporalis tendon extends down the medial aspect of the mandibular ramus along the buccinator ridge as far as the mandibular third molar. Additional length may be gained by dissecting along the medial aspect of the ramus and dividing the tendon as low as possible. In many patients, this modification obviates a fascia lata extension. Eliminating an adynamic segment and additional point of suture fixation reduces the risk of tendon dehiscence and improves the efficiency of commissure excursion. This dissection may be accomplished via a small nasolabial or intraoral incision. Objective outcomes of minimally invasive temporalis tendon transposition (MIT3) reveal distinct improvement in vertical commissure symmetry with smile and expert-graded Terzis score. The authors prefer this MIT3 and comparative discussion in this article is based on this technique.

Key advantages to this technique include:

1.

A single-stage surgery that is routinely performed in the outpatient setting.
2.

There is an immediate and reliable improvement in resting symmetry after surgery.
3.

Volitional movement may begin immediately after surgery, although the authors suggest patients adhere to a soft diet and avoid volitional movement for 6 weeks postoperatively to prevent tendon disinsertion or dehiscence from the commissure.
4.

Minimal access incision via the nasolabial fold or oral mucosa may be used.
5.

Pterygoid muscles compensate for temporalis muscle function, leading to minimal donor site morbidity.
6.

Use of T3 does not preclude patients from receiving subsequent reanimation procedures such as GFMT.

Limitations of this technique include:

1.

There is a fixed, single vector for oral commissure insertion. This limitation may be modified to some degree with variations in fascia lata extension grafts along the upper and lower lip.
2.

Given the trigeminally mediated activation of the temporalis muscle, true spontaneity is unable to be achieved.
3.

Although harvesting temporalis tendon along the medial ramus allows for increased tendon length, the quality of this distal tendon extension varies. Additional length may be required via fascia lata extension grafts or disinsertion of periosteal attachments within the temporal fossa or along the zygomatic arch.
4.

Degree of commissure excursion achieved remains inconsistent. Excellent excursion may be achieved but this is likely to depend on preservation of native muscle tension, degree of insertion site dehiscence, and use of adynamic grafts.
5.

A functioning temporalis muscle unit is required.
6.

Neuromuscular retraining is necessary to optimize results.

Gracilis free muscle transfer

The gracilis flap in smile reanimation was initially described by Harii and colleagues in 1976 and has become the preferred technique in most institutions. Refinement of size of the gracilis muscle based on fascicular units allows increased precision in muscle size and tissue bulk transplanted. Although the gracilis muscle has a single motor nerve, it comprises multiple nerve fascicles producing contraction in different regions of the muscle. Understanding gracilis neural fascicular anatomy has led to innovations such as the multivector gracilis in pursuit of a genuine enjoyment smile or Duchenne smile. Duchenne smiles are characterized by periocular muscle activation with maximal dental display and high cheek elevation naturally associated with positive emotion. In contrast, a social or Mona Lisa smile is a smile elicited mainly from the corners of the mouth with minimal dental display. By dividing the gracilis muscle into 3 functional neuromuscular units, separate muscle segments may be used to imitate levator labii superioris and orbicularis oculi activity in addition to zygomaticus activity. A slip of muscle placed medially and oriented slightly more vertically mimics levator labii superioris and a short slip of muscle placed within the lateral lower eyelid mimics orbicularis oculi activity. A wide variety of choices for neural innervation of GFMT exist. In the original description by Harii and colleagues, the deep temporal nerve was used. Recognizing the need for facial nucleus input to achieve spontaneous movement, O’Brien and colleagues described a 2-stage approach to innervate the gracilis muscle with a cross-facial nerve graft (CFNG). Because of the long distance (and consequently time) required for nerve regrowth, a CFNG is placed in a first-stage surgery and at least 4 months are allowed to elapse before muscle transplant. Although described innervation sources include spinal accessory, phrenic, and hypoglossal nerves, the masseteric nerve has become favored for its high axonal density, improved excursion, minimal donor site morbidity, and favorable proximity. The ipsilateral facial nerve is the ideal donor nerve when feasible but is not available in most scenarios. Recent literature has supported using both a CFNG and the masseteric nerve for dual innervation of GFMT. Although there are inadequate data to compare various methods of innervation, dual innervation is reported to be a safe procedure, movement initiation is similar to single masseteric nerve innervation, and most patients achieve spontaneity.

Nuances of GFMT techniques vary between institution and no ideal technique has been established. Key advantages to GFMT include:

1.

Spontaneous activation may be routinely achieved in GFMT with CFNG input. Spontaneity is reported with trigeminal innervation, although this point is debated.
2.

Unencumbered adjustment of muscle size, position, and number of vectors allows approximation of a more natural smile based on patient characteristics.
3.

Meaningful commissure excursion is reliably achieved in most patients.
4.

Adductor compensation for the gracilis muscle results in minimal morbidity in the donor site.
5.

Multiple sources of neural input are available.

Limitations of this technique include:

1.

Microvascular surgical expertise and equipment are needed to perform GFMT, and patients require inpatient admission with free flap monitoring for 3 to 5 days postoperatively.
2.

Patients require an initial procedure at least 4 months (and on average 6–9 months) before single-innervation GFMT with CFNG. Single-stage dual innervation with CFNG and masseter nerve has been described with satisfactory preliminary results.
3.

Initiation of movement depends on neutral ingrowth. Movement starts around 4 to 6 months postoperatively and progresses over 2 to 3 years after surgery.
4.

The addition of gracilis muscle may result in midfacial bulk requiring secondary revision.
5.

Similar to T3, neuromuscular retraining is necessary to optimize results.

Evidence-based outcome comparison

Few articles directly compare temporalis tendon transfer and GFMT outcomes. Existing studies use widely varied surgical techniques and heterogeneous outcome measures, limiting aggregate comparison.

Erni and colleagues performed a comparison of 10 orthodromic T3s via the McLaughlin technique and a group of 7 heterogeneous muscle transfers (gracilis, latissimus dorsi, pectoralis minor) in 1999. They found no difference in resting symmetry outcomes between the 2 techniques but more excursion with free neuromuscular transfers (mean, 5.5 mm; standard deviation, 1.6 mm) versus T3 (1.7 mm, 0.8 mm). Observers rated overall aesthetic appearance of neuromuscular transfers as markedly impaired by cheek swelling and skin tethering.

Gousheh and Arasteh retrospectively reviewed 505 2-stage, CFNG-innervated GFMTs with 73 antidromic temporalis muscle transpositions and 4 lengthening temporal myoplasties. Results were graded as excellent (≥2 cm lateral commissure movement), good (1.5 to <2 cm), satisfactory (<1.5 cm), and failed (<1 cm). The gracilis group achieved 14% excellent, 76% good, 8% satisfactory, and 2% failed results, whereas the temporals muscle transfer achieved 30% good, 59% satisfactory, and 11% failed. LTM patients were all classified as satisfactory.

A systematic review from 2016 evaluated GFMT versus lengthening temporalis myoplasty and revealed no randomized and controlled clinical trials of these techniques and heterogeneity in objective measurements. GFMT studies reported better commissure excursion compared with LTM studies, although few studies of LTM reported quantifiable data.

A retrospective study from the Netherlands compared 10 masseter-driven gracilis and 12 antidromic temporalis muscle transposition patients. The investigators reviewed excursion measurements, Facial Clinimetric Evaluation (FaCE) scale scores, and casual observer ratings. Smile excursion and symmetry analysis via FACE-gram software showed no statistically significant differences. FaCE scale subscores differed by only 1.7 points in favor of GFMT and were not statistically different. Casual observer aesthetic assessment did not show a preference between the two techniques.

In 2018, Oyer and colleagues performed a retrospective study evaluating patients with unilateral flaccid facial paralysis undergoing MIT3 (14 patients) versus GFMT (14 patients). They evaluated commissure position in vertical, horizontal, and angular dimensions in repose and with smile. Both groups had significant improvement in commissure symmetry in repose as well as with smile. The MIT3 group had increased improvement in angular symmetry compared with GFMT at rest and with smile. The GFMT group had improved vertical symmetry in repose and smile compared with the MIT3 group. Horizontal symmetry was most improved in the GFMT during smile and did not improve significantly in repose in either group. Commissure excursion was present in both groups with a larger improvement in the GFMT group (11.3 mm; 95% confidence interval [CI], 7.0–15.5 mm) versus the MIT3 group (4.8 mm; 95% CI, 0.2–9.3 mm). The GFMT group achieved commissure symmetry to within 1 mm of the contralateral side. Given the retrospective nature of this study, the investigators note a significantly older population receiving the MIT3 compared with the GFMT.

More recently, Nguyen and colleagues retrospectively reviewed LTM (6 patients) with 2-stage CFNG GFMT (10 patients) in the pediatric population. LTM patients showed a significant increase in commissure excursion within 0 to 3 months postoperatively (3.33 ± 1.58 mm) with steady increase in excursion the first year after surgery up to 7.27 +/- 1.96 mm. GFMT patients showed movement at 3 to 6 months postoperatively (4.01 ± 1.77 mm) and, on average, 5.28 mm at more than 12 months postoperatively. Resting symmetry was improved in both groups with superolateral overcorrection noted in LTM patients.

Patient-centered surgical decision making

Temporalis tendon transfers allow reliable, immediate improvement in static symmetry with a single-stage outpatient procedure. The MIT3 technique further reduces surgical dissection and the use of extension grafting. Dynamic outcomes can be excellent but are inconsistent. Based on the limited existing data, temporalis tendon transfers likely result in reduced commissure excursion compared with GFMT. Resting symmetry is at least equivalent and potentially superior to GFMT depending on the gracilis innervation source. GFMT has consistent dynamic excursion results, more tailored smile vectors, and the opportunity for spontaneous movement. These long-term benefits come at the upfront cost of a (likely) multistaged, longer surgery with a protracted period before initiation of perceivable movement and the risk of midfacial bulk. Although morphologic change and asymmetry may occur with T3, this is typically related to nuances in insertion technique rather true facial bulk caused by the additional tissue.

Ultimately, surgeons aim to deliver the maximal psychological and functional benefit to their patients with the simplest, least morbid techniques. The larger change or improvement delivered to the patients, the more benefit they will likely receive. Resting symmetry is one definitive patient characteristic that demarcates the expected improvement achieved between the T3 and the GFMT. Simply put, the most consistent benefit achieved with T3 is improvement in resting symmetry. If a patient is relatively symmetric at rest, there is no opportunity to achieve a beneficial change in this aspect and the result depends on the dynamic movement ( Fig. 1 ). Dynamic movement is more consistently achieved with GFMT. Thus, good resting symmetry diminishes temporalis tendon transfers from consideration. In cases of incomplete paralysis, GFMT is typically preferred to T3 for this reason. In contrast, if a patient has poor resting symmetry, a large beneficial change may be achieved from both gracilis or T3 and consideration turns to secondary factors described previously ( Fig. 2 ). In these situations, the consideration for desired dynamic outcome, ability to tolerate a long anesthetic, life expectancy, and desire for immediate results comes into play.